| 00:01:04 | Russ|werk | heh, I think I get alltel directory assistance for free |
| 00:06:29 | mallum | jacques: do you have render extension running with tinyX ? |
| 00:07:11 | andersee | mallum: was it you that was reworking init? |
| 00:07:17 | andersee | forgets |
| 00:07:41 | jacques | number of extensions: 12 |
| 00:07:41 | jacques | BIG-REQUESTS |
| 00:07:41 | jacques | FontCache |
| 00:07:41 | jacques | MIT-SCREEN-SAVER |
| 00:07:41 | jacques | MIT-SHM |
| 00:07:42 | jacques | MIT-SUNDRY-NONSTANDARD |
| 00:07:44 | jacques | RANDR |
| 00:07:46 | jacques | RENDER |
| 00:07:48 | jacques | SHAPE |
| 00:07:50 | jacques | SYNC |
| 00:07:52 | jacques | TOG-CUP |
| 00:07:54 | jacques | XC-MISC |
| 00:07:56 | jacques | XTEST |
| 00:08:40 | mallum | jacques: apparently render steels all color cells |
| 00:09:03 | mallum | jacques: may be worth disabling it and seeing if matchbox get some color |
| 00:09:04 | jacques | mallum: say what? |
| 00:09:09 | mallum | jacques: could be a quick fix |
| 00:09:20 | jacques | why do other apps work then? |
| 00:09:25 | mallum | jacques: render extension makes most of the colormap readonly |
| 00:09:36 | mallum | jacques: good point ... |
| 00:09:44 | mallum | jacques: back to drawing board |
| 00:09:46 | jacques | even aewm works |
| 00:09:51 | mallum | jacques: its late |
| 00:09:54 | jacques | np |
| 00:10:00 | mallum | jacques: aewm uses 2 colors |
| 00:11:46 | ctr | is away: bed |
| 00:11:52 | mallum | jacques: Im checking it ... |
| 00:12:22 | jacques | well when I get home I'll try blackbox or something |
| 00:12:37 | mallum | jacques: aewm might be enough |
| 00:12:41 | jacques | flwm also works but it does not use a lot of colors |
| 00:12:46 | jacques | ad it uses dltk |
| 00:12:46 | andersee | Anybody got some time to help me get busybox 0.60.3 released? |
| 00:12:47 | mallum | aewm is pretty simple |
| 00:12:48 | jacques | fltk |
| 00:14:43 | jacques | andersee: what do you need? |
| 00:15:27 | Verty | is away: Meeeeeeeeerp! |
| 00:15:57 | mallum | jacques: I checked blackbox code, its very complicated |
| 00:16:09 | mallum | jacques: I will check the source of fltk, when I get a mo |
| 00:16:32 | jacques | mallum: dont bother - it's probably even more complicated |
| 00:16:42 | jacques | as is mapped color stuff in x |
| 00:16:44 | jacques | :-( |
| 00:17:49 | file | sighs |
| 00:17:58 | file | sprintf( timestring, "%d : %d", time.hour, time.minute ); |
| 00:18:03 | file | what's wrong with that? |
| 00:19:45 | jacques | file: looks ok to me |
| 00:19:50 | jacques | assuming |
| 00:20:02 | jacques | andersee: ??? |
| 00:20:07 | file | time.hour should return 9 |
| 00:20:11 | andersee | jacques: I need people to pound on the current busybox.stable, and look over the bugs at http://bugs.busybox.net/db/pa/lbusybox.html and test if things are fixed |
| 00:20:12 | file | and time.minute should return 23 |
| 00:20:28 | andersee | jacques: I am updating the daily-snapshot now by manually |
| 00:21:18 | file | writes his GUI |
| 00:21:59 | andersee | jacques: k, snapshot is ready. I am about to test bug #1226 to be sure it is fixed |
| 00:22:21 | jacques | locales arg |
| 00:22:26 | jacques | !1225 |
| 00:23:26 | file | sighs |
| 00:23:27 | file | why isn't this working?!?!?! |
| 00:24:11 | prpplague | howdy howdy |
| 00:24:24 | jacques | is it just me or is the who net screwed u ptoday? |
| 00:24:27 | jacques | whole |
| 00:24:31 | mallum | bedtime for me, goodnight all + have a good weekend |
| 00:24:37 | jacques | later mallum |
| 00:24:43 | mallum | goes off the dream of Xcolors ..... |
| 00:26:39 | Morn | hi prpplague |
| 00:26:52 | Morn | prpplague: I got opie running on the tuxscreen today |
| 00:27:01 | Morn | I'm so excited! |
| 00:27:22 | as_ | Whats opie? |
| 00:27:29 | prpplague | Morn: cool |
| 00:27:33 | Morn | it is the open version of qtopia |
| 00:27:41 | Morn | a qt based GUI for handhelds |
| 00:27:42 | prpplague | Morn: sounds like you had a more rewarding day than i |
| 00:27:50 | as_ | Ahh, cool. |
| 00:28:05 | Morn | prpplague: what's wrong? |
| 00:28:19 | as_ | Do I need unix running on my PC to drop files to the tux? |
| 00:28:26 | prpplague | Morn: spent all day doing sco ksh to linux sh |
| 00:28:48 | Morn | ick |
| 00:29:01 | prpplague | Morn: its kinda like shaving by pull each hair out with twissors |
| 00:29:18 | prpplague | Morn: i.e. slow and painful |
| 00:29:19 | Morn | as_: you can use windows with the hyperterminal |
| 00:29:48 | prpplague | Morn: but i did just get back from this cool beer store i found 45 minutes from my house |
| 00:30:09 | Morn | awesome |
| 00:30:11 | Morn | what do we have? |
| 00:30:35 | as_ | Morn, what page should I be reading the instructions from? |
| 00:31:01 | as_ | Ive been going through and everything seems spread out. |
| 00:31:29 | prpplague | Morn: duvel, fuller's 1845, youngs ram rod, youngs old nick, and altenmuster premium |
| 00:31:31 | Morn | what are you trying to do? |
| 00:32:01 | as_ | Id like to get a linux image bootable on the tux. |
| 00:32:14 | jacques | Morn: you build opie against uclibc ? is it running locally on the tux ? |
| 00:32:18 | jacques | how big is it? |
| 00:32:26 | as_ | Is it possible to dual boot inferno and linux or isnt there enough room? |
| 00:32:41 | Morn | jacques: no it is compiled against the familiar libc |
| 00:32:49 | jacques | oh |
| 00:32:54 | Morn | and I have familiar on a nfs mount |
| 00:33:00 | Morn | and then opie in there |
| 00:33:18 | jacques | ok |
| 00:33:20 | Morn | I just have the barest min of a ramdisk to get the nfs up and pivoted |
| 00:34:04 | Morn | familiar 0.5.2 doesn't seem to boot properly though on the tuxscreen |
| 00:34:15 | Morn | I used 0.4 and then upgraded some of the packages |
| 00:34:28 | Morn | on a 100MB lan nfs is acceptable for speed |
| 00:35:09 | Morn | and thanks to the person that helped me get the nfs working with solaris |
| 00:35:19 | Morn | since that is the server that is up all the time |
| 00:38:35 | jacques | i wonder why familiar doesnt boot |
| 00:44:52 | file | it chooses not to |
| 00:46:48 | Morn | it got to /sbin/init |
| 00:46:49 | Morn | and hung |
| 00:46:59 | Morn | I had no idea how to troubleshoot it |
| 00:47:08 | Morn | so I went back to 0.4 I knew it worked |
| 00:47:18 | file | uh huh |
| 00:47:33 | as2 | Is the tuxscreen serial port DTE or DCE? |
| 00:47:37 | file | Thinkgeek finally has the Sharp Zaurus! |
| 00:48:44 | jacques | as2: you need a null modem cable |
| 00:48:56 | as2 | Crossover or straight? |
| 00:49:20 | Morn | I used a null modem to connect it to the server |
| 00:50:10 | as2 | Crossover or straight? |
| 00:50:17 | jacques | just not one of the microsoft null modem cables - those dont work |
| 00:50:47 | jacques | as2: I dunno - are you saying there are two different types of null modem cable? |
| 00:50:59 | as2 | Yeah. |
| 00:51:10 | jacques | beats me then |
| 00:51:20 | jacques | both of my cables worked |
| 00:51:36 | jacques | actualyl one is a regular modem cable with a null modem adaptor |
| 00:51:50 | as2 | Null modem is the wrong name. Sometimes people mean a model calbe, which is DTE to DCE, and sometimes its a crossover cable DTE to DTE. |
| 00:52:10 | jacques | but one guy lots tons of time by trying to use a microsoft null modem cable |
| 00:52:18 | jacques | as2: did you look at the wiki ? |
| 00:52:32 | Morn | null modem or null mind (microsoft) |
| 00:52:47 | file | lol |
| 00:53:09 | as2 | Ive tried, I keep hitting stuff I dont understand, or doesnt follow from previos stuff. |
| 00:53:31 | as2 | Does your null modem cable work for connecting 2 PCs? |
| 00:53:33 | file | I must go |
| 00:53:37 | file | I'm very tired |
| 00:53:40 | Morn | as2: yes |
| 00:53:48 | Morn | I use it to connect to my solaris box too |
| 00:53:51 | as2 | Ahh, its a crossover then. |
| 00:53:56 | as2 | L8r file. |
| 00:54:08 | file | snores |
| 00:54:24 | jacques | does it even matter if there is no hardware flow control ? |
| 00:55:17 | as2 | I've no idea. |
| 00:56:17 | as2 | It doesnt help I'm almost completely ignorent of linux. |
| 00:56:36 | prpplague | as2: sssshhh, don't say that |
| 00:56:41 | cbiffle | Okay, I've managed to resurrect most of the phone, but does anyone else get loud crackly static on the audio? :-) |
| 00:56:59 | jacques | cbiffle: i would tell you if I actually had audio |
| 00:57:18 | cbiffle | jacques: Well, good to know some people are worse off than I! :-) What're your symptoms? |
| 00:57:31 | jacques | cbiffle: which audio? I thought you meant shanip mod |
| 00:57:35 | jacques | which I have not done |
| 00:57:47 | as2 | Oh dear, am I in danger of being beaten to death? |
| 00:57:57 | as2 | What series farscape? |
| 00:58:08 | prpplague_farsca | as2: no, just given special attention |
| 00:58:22 | cbiffle | jacques: No, I mean phone functions. |
| 00:58:27 | as2 | Will it involve needles :/ |
| 00:58:42 | Morn | cbiffle: I got opie working! |
| 00:58:44 | cbiffle | openly declares to not only be a Linux newbie, but in fact to have BSD experience! *gasp* |
| 00:58:47 | Morn | has to tell everyone |
| 00:58:49 | cbiffle | Morn: Groovy! |
| 00:58:54 | as2 | How did you mange to kill it cb? |
| 00:59:05 | cbiffle | Morn: I'm getting ready to test it with my new kernel/userland to see if my memory errors clear up. |
| 00:59:24 | Morn | cbiffle: I did it different from you though |
| 00:59:33 | Morn | I nsf mounted a familiar bootstrap |
| 00:59:37 | cbiffle | as2: I have no idea! :-) It's just staticy, and...oop, there it goes, ringing eerily again. |
| 00:59:39 | Morn | and then just ipkg installed opie |
| 00:59:50 | cbiffle | Morn: I did that the first time around, but I tried to get away from Familiar. |
| 00:59:52 | as2 | Ghost in the machine. |
| 01:00:07 | cbiffle | as2: Pretty much! :-) |
| 01:00:23 | cbiffle | I'd strip the thing down to the StrongARM board except that I really did buy it for the phone functions. :-) |
| 01:00:24 | Morn | cbiffle: I had to use familiar 0.4 0.5.2 wouldn't boot |
| 01:00:31 | Morn | it hung on /sbin/init |
| 01:00:36 | cbiffle | Morn: I'm not sure which I used; I ripped it off a friend's iPaq. |
| 01:00:41 | cbiffle | Morn: Same symptoms here. |
| 01:01:00 | Morn | I have it working pretty good now |
| 01:01:04 | Morn | most things run |
| 01:01:10 | Morn | except the qipkg |
| 01:01:13 | cbiffle | as2: And now it's sitting there quietly crackling to itself. Even though speaker isn't on. |
| 01:01:16 | Morn | that causes an out of memory error |
| 01:01:31 | Morn | I'm going to write a dialer for address book |
| 01:01:33 | jacques | you bought a tuxscreen for a phone? what a radical concept :-) |
| 01:01:46 | cbiffle | Morn: I was already working on it, until I lost the phone functions. :-) |
| 01:01:56 | cbiffle | jacques: Do you know if the screen is driven by the DSP board entirely? |
| 01:02:01 | Morn | I recompiled tuxphone for libc |
| 01:02:33 | Morn | and created a /etc/init.d file to start and stop it |
| 01:02:46 | prpplague | as2: when did you start using linux? |
| 01:03:05 | as2 | Ive only used it for eggdrop I'm afraid. |
| 01:03:10 | cbiffle | Hm. And now the phone functions have fled completely, leaving me with a glowing message light. *sigh* |
| 01:03:27 | Morn | cbiffle: weird |
| 01:03:30 | Morn | mine works here |
| 01:03:43 | prpplague | as2: you don't have linux box your learning on? |
| 01:03:45 | cbiffle | I wish you guys could hear this, it's tres creepy. :-) From time to time it'll emit bursts of static that sound vaguely like a voice. |
| 01:03:57 | Morn | I had to disable all the debug info in tuxphone though, because it messed up the opie screen |
| 01:04:13 | cbiffle | Morn: Yup, same here. |
| 01:04:28 | Morn | simple turning DEBUG_LEVEL to 0 fixed that |
| 01:04:31 | as2 | prp, I dont have linux on anything yet. |
| 01:04:33 | jacques | cbiffle: I dont know |
| 01:05:29 | prpplague | as2: are you a student? |
| 01:05:43 | as2 | Recently postgrad. |
| 01:06:10 | Russ|werk | cbiffle: does it do it if tuxphone isn't running? |
| 01:06:49 | cbiffle | Russlwerk: Same thing. Even does it when Linux isn't working, and it's only on batteries. I really feel like it's the DSP. |
| 01:06:53 | as2 | The tux interests me becuase I used to code ARM asm. The OS doesnt matter so much as long as I can program for it. |
| 01:07:22 | cbiffle | I have not yet installed Inferno because I'm hesitant to do so without a JTAG dongle. |
| 01:07:45 | prpplague | as2: well if you need some linux cd's or even some hardware don't hesitate to drop me an email, my company has an education foundation |
| 01:08:12 | as2 | Thanks for the offer :] |
| 01:09:00 | Russ|werk | cbiffle: a lot of stuff is soldered to the board with stranded copper |
| 01:09:09 | Russ|werk | might check and see if anything is shorting |
| 01:09:27 | cbiffle | Interestingly, now the phone is sitting there warbling and isn't responding to keystrokes. |
| 01:09:50 | cbiffle | Russlwerk: I've been trying not to take it apart too thoroughly, so that if I was to send it back to Tim it would be more credible. |
| 01:10:31 | Russ|werk | cbiffle: its very disassemblable |
| 01:10:40 | Russ|werk | just don't pull the hotglue off the solder joints |
| 01:10:58 | cbiffle | Russlwerk: I know, I've seen the pictures, but with the problems I'm having I'd prefer that no fingers pointed to me. :-) |
| 01:12:09 | Morn | cbiffle: do you have the work you started on the dialer? |
| 01:12:16 | Morn | Or should I start fresh? |
| 01:12:34 | cbiffle | Morn: Start fresh; I was just doing Qt interface design. |
| 01:12:47 | Morn | Ok |
| 01:12:53 | Morn | I have to learn QT first :-) |
| 01:13:45 | Russ|werk | too bad its not gtk |
| 01:13:51 | Russ|werk | gtk/glib rules |
| 01:14:03 | cbiffle | ... |
| 01:14:08 | Morn | have you seen the opie interface? |
| 01:14:11 | Morn | it is very clean |
| 01:14:12 | cbiffle | Have you played with Qt? |
| 01:14:44 | Russ|werk | not much |
| 01:15:11 | Morn | it will be really nice when the touchscreen works |
| 01:15:15 | cbiffle | From a programming perspective, it's a real joy to work with. |
| 01:15:25 | Morn | but I have figured out many of the keyboard keys |
| 01:15:40 | cbiffle | F12 to switch tabs, F11 for menu, Esc to close are what I found. |
| 01:15:45 | Morn | cbiffle: do you know how to get in the menus with the keyboard? |
| 01:15:53 | Morn | F11 for the bottom menu |
| 01:15:57 | Morn | but what about menus in an app |
| 01:16:13 | cbiffle | If an application is running without the taskbar (if you start it with -qws), f11 gets to its menus. Other than that, no. |
| 01:16:35 | Morn | I wanted the task bar for the applets |
| 01:16:43 | Morn | having the time applet is nice |
| 01:16:48 | cbiffle | That's a good tip, actually; if you need to run an app without running the whole environment, run it with '-qws'. |
| 01:17:00 | Russ|werk | I really can't stand c++ |
| 01:17:02 | Morn | do you know how to switch between running apps? |
| 01:17:25 | cbiffle | Russlwerk: I'm a Java guy myself, so C++ looks like Pascal; Qt uses a preprocessor that removes a lot of the crufty committee-designed language features. |
| 01:17:29 | cbiffle | Morn: No. :-) |
| 01:17:48 | Russ|werk | mainly its the idea of classes that bother me |
| 01:17:55 | cbiffle | Really. So you're not much of an OO person. |
| 01:17:58 | Morn | is a java programmer |
| 01:18:01 | Russ|werk | no, I'm very oo |
| 01:18:08 | cbiffle | ... But not class-based OO? |
| 01:18:16 | Russ|werk | classes is one way of doing oo, but its very restrictive |
| 01:18:22 | cbiffle | You prefer prototypes, or...? |
| 01:18:39 | cbiffle | Interface-based abstraction? |
| 01:18:54 | Russ|werk | structs |
| 01:19:05 | Russ|werk | every source file does one type of thing |
| 01:19:15 | Russ|werk | I wouldn't mind operater overloading though |
| 01:19:27 | cbiffle | structs as in C-style structs? |
| 01:19:34 | Russ|werk | ya |
| 01:19:40 | Morn | any language can be well or badly written |
| 01:19:40 | Russ|werk | structs are very powerfull |
| 01:19:45 | Morn | it is a personal preference |
| 01:19:45 | cbiffle | I don't consider those to be OO, because they can't have attached code. |
| 01:20:12 | cbiffle | What differentiates OO and procedural languages, imho, is the ability to attach an abstract interface to zero or more routines to the data structure itself. |
| 01:20:15 | Russ|werk | cbiffle: what do you mean by attached? |
| 01:21:21 | prpplague_away | Morn: btw you still need those scsi drives? |
| 01:21:31 | Morn | prpplague_away: I thought you would never ask :-) |
| 01:21:32 | Morn | yes |
| 01:21:56 | cbiffle | Such that, rather than mysql_query(name_of_connection), you're instead doing connection.query. It (1) organizes functions related to the data structure within the data structure itself, and (2) provides an abstractable interface under which multiple types of implementations can be made. Using the example earlier, connection.query() might be a MySQL database, or might be Access, and you need not know, because the objects are interchangeabl |
| 01:22:29 | cbiffle | But more importantly, there doesn't need to be one massive central query function with a switch() statement that can distinguish between every possible type of connection; the connection-specific functions are 'attached' to the connection data structures. |
| 01:22:55 | Russ|werk | ah, thats where function pointer come it |
| 01:22:56 | Russ|werk | er in |
| 01:23:09 | Morn | asks her laptop to compile faster |
| 01:23:10 | cbiffle | Sort of, yes. You can simulate OO using function pointers. |
| 01:23:37 | cbiffle | But the function you're calling has no concept of 'this', of what data structure it's applying to; you generally have to pass it a pointer to the structure that it's a member of. |
| 01:23:38 | levi__ | That's not 'simulating' OO. OO is a programming paradigm, not a syntax. |
| 01:24:08 | Russ|werk | cbiffle: but you are already doing that |
| 01:24:09 | cbiffle | levi__: Agreed; I say 'simulating' because the syntax is similar but the implementation is very different. |
| 01:24:10 | Morn | I'll give you a pair a dime for a nickle |
| 01:24:14 | cbiffle | Russ---: sort of. |
| 01:24:29 | Russ|werk | object->function is the same thing as function(object) |
| 01:25:03 | Morn | Can I add one thing? |
| 01:25:13 | cbiffle | Except that with function(object), 'function' is defined in a global namespace, while 'object' may either contain its own implementation of 'function' or inherit it from somewhere else, even though the interface and name is identical. |
| 01:25:15 | cbiffle | Morn: of course! |
| 01:25:18 | Morn | if the program written accompilishes the goal does it matter how it does it? |
| 01:25:25 | cbiffle | Morn: Of course not. :-) |
| 01:25:28 | Russ|werk | cbiffle: write you own namespcae |
| 01:25:37 | cbiffle | Russ---: I don't follow. |
| 01:25:52 | cbiffle | And I'm not completely clear on why my IRC client is attaching dashes to peoples' names. |
| 01:25:52 | Russ|werk | cbiffle: the main problem I have with classes, is that its oriented at working with only one datatype at a time |
| 01:25:57 | cbiffle | How so? |
| 01:26:33 | cbiffle | (Disclaimer: I'm not really an OO bigot. I spend most of my time in C. I'm just interested in other viewpoints, so I'm taking the OO one for now.) |
| 01:26:52 | as2 | I hate the OO idea. ;] |
| 01:27:03 | Morn | I like OO, but then I'm weird |
| 01:27:16 | cbiffle | Russ---: This whole discussion would be simpler for me if we were using code examples. |
| 01:27:20 | cbiffle | as2: Why? |
| 01:27:43 | as2 | How long have you got ;] |
| 01:28:01 | as2 | Probably for the same reason I like assembler. |
| 01:28:10 | cbiffle | as2: I don't see why they can't coexist. |
| 01:28:29 | as2 | They are the oppasit ways of thinking. |
| 01:28:30 | prpplague_away | Morn: remind me tomorrow and i'll get them boxed up |
| 01:28:38 | Morn | I haven't done any assembler coding in ages |
| 01:28:44 | Morn | there is a reason I don't any more |
| 01:28:54 | Morn | prpplague_away: great thanks |
| 01:28:57 | cbiffle | as2: I'd say that assembler and prolog are opposites; in one, you control everything at a low level, while in the other, you define high-order rules without regard to procedure. |
| 01:29:02 | as2 | It is diffuclt to do large projects in asm. |
| 01:29:11 | prpplague_away | Morn: why because you lose too many phone when you code assembler? |
| 01:29:23 | as2 | Did you know prolog went OO. |
| 01:29:25 | Morn | assembler for a large project is unmanageable |
| 01:29:40 | Morn | too difficult to maintain |
| 01:29:41 | as2 | I agree Morn, I avoid large projects. |
| 01:29:45 | cbiffle | as2: It's not the data organization I'm concerned with in prolog; it's that assembler is imperative while prolog is declarative. |
| 01:30:08 | Russ|werk | cbiffle: the kernel is a very good example of oo c, and namespaces |
| 01:30:09 | cbiffle | My theory on assembler in large projects is that a programmer can produce a fixed number of debugged lines of code per hour (plus or minus small language variances). |
| 01:30:13 | as2 | I know. Goal oriented programming. |
| 01:30:49 | as2 | Put in all the staements without any regard to speed of calculation, or what is really going on under the hood. |
| 01:31:03 | cbiffle | as2: Right. That's why I look at prolog as the polar opposite to assembler. :-) |
| 01:31:14 | as2 | OO to me seems to wrap everything up in protective layers that wouldnt be needed if the programming was done properly. |
| 01:31:34 | BZFlag | andersee: modprobe does now? modules.dep? modules.conf? etc? |
| 01:31:35 | cbiffle | as2: Various people have different notions of 'proper'. |
| 01:31:57 | cbiffle | as2: The ability to drop in modules of code with defined interfaces and not have to worry about implementation is one of the main benefits of OO in large team projects. |
| 01:32:04 | as2 | And the result is a massivly bloated program wasting space and computing power. |
| 01:32:19 | andersee | as2: thats like saying that having a police force wouldn't be necessary if everyone would just be nice to eachother. |
| 01:32:20 | cbiffle | as2: *sigh* That's all in the compiler. |
| 01:32:42 | Russ|werk | as2: you are thinking of classes, not oo |
| 01:32:53 | cbiffle | Russ---: I don't think that's inherent in classes, even! |
| 01:33:00 | as2 | I'm mainly thinking of classes yes Russ. |
| 01:33:03 | andersee | BZFlag: it now supports reading and using /lib/modules/<version>/modules.dep |
| 01:33:13 | cbiffle | Russ---: What you're doing in C is much closer to classes than any of the alternative OO models, like prototyping. |
| 01:33:23 | Russ|werk | there is a lot about c++ I would like to see done well, but classes kind of screw it all up |
| 01:33:27 | andersee | BZFlag: no modules.conf -- feel free to open a seperate bug on that |
| 01:33:31 | as2 | andersee, I agree, but programs are not like people, they can be written to behave perfectly. |
| 01:33:39 | Russ|werk | I'd like to see catching done well, as well as operator overloading |
| 01:33:49 | cbiffle | Russ---: Have you used Java much? |
| 01:34:04 | cbiffle | (The language, not the VM.) |
| 01:34:13 | Russ|werk | and more autocasting |
| 01:34:20 | Russ|werk | I've helped people with it quite a bit |
| 01:34:22 | levi__ | Russ---: I still don't see what you've got against classes. |
| 01:34:25 | Russ|werk | but again, its class centric |
| 01:34:25 | cbiffle | as2: But programmers can't. |
| 01:34:40 | cbiffle | Russ---: I'm with levi, I don't see how classes are to blame for what you're describing. |
| 01:34:45 | cbiffle | Russ---: Can you explain? |
| 01:34:49 | Russ|werk | levi__: too restrictive, and the variable->function thing is really inconsistant |
| 01:35:03 | cbiffle | So instead you're using function pointers in structs? :-) |
| 01:35:04 | as2 | Classes are like having one policeman per citizen. |
| 01:35:06 | levi__ | Russ---: What's restrictive about them? |
| 01:35:16 | cbiffle | Russ---: Yes, what is restrictive? |
| 01:35:21 | andersee | thunderstorm rolling in. time to rsync my laptop just in case... |
| 01:35:24 | Russ|werk | they are asking you to organize things in a paticular way |
| 01:35:33 | BZFlag | andersee: sweet |
| 01:35:35 | cbiffle | .....yes... |
| 01:35:43 | cbiffle | Most hardware abstraction models do that. |
| 01:35:51 | Russ|werk | work with one data object at a time |
| 01:36:00 | cbiffle | You mentioned that before, I'm not sure I follow that. |
| 01:36:01 | as2 | Classes are restrictive thats what they are for, to stop a program making mistakes if there is a mistake in the code. |
| 01:36:07 | andersee | BZFlag: I'm trying _very_ _very_ hard to get busybox 0.60.3 out tonight. |
| 01:36:10 | Morn | hi Time |
| 01:36:14 | Morn | er |
| 01:36:16 | andersee | BZFlag: Just going through the bug lists... |
| 01:36:17 | Morn | hi Tim |
| 01:36:23 | Russ|werk | as2: like how qbasic is restrictive? |
| 01:36:40 | cbiffle | as2: I disagree. I don't think classes are there to catch mistakes, I think they're there to make code extensible and modular. |
| 01:36:57 | as2 | qbasic is another way to acomplish the same thing. |
| 01:37:07 | Russ|werk | cbiffle: which can be done just as well, and better, without classes |
| 01:37:21 | cbiffle | Russ---: I'm not sure I agree with 'better', but yes, it can also be done without classes. |
| 01:37:25 | cbiffle | Russ---: I |
| 01:37:28 | cbiffle | Russ---: I |
| 01:37:30 | Russ|werk | I think gtk is a very good example of this |
| 01:37:30 | cbiffle | heh. Right. |
| 01:37:32 | cbiffle | New keyboard. |
| 01:37:45 | Russ|werk | the simplicity with which bindings are added |
| 01:37:56 | Russ|werk | (language bindinds) |
| 01:38:04 | cbiffle | Russ---: I'm just not seeing the -problems- you're seeing with classes. I see that things can be done without them, certainly. |
| 01:38:05 | levi__ | Russ---: How is putting a function pointer in a struct substantially different than defining a class? |
| 01:38:40 | Russ|werk | because you are putting it there for an entirely different purpose |
| 01:38:47 | Russ|werk | functions are put in classes for namespace |
| 01:39:03 | cbiffle | ...and organizational modularity, and ease of reference to the data items in the class. |
| 01:39:12 | as2 | Doing math with pointers is very effitiant code when done properly, and a total disaster when it isnt. |
| 01:39:27 | Russ|werk | cbiffle: why would you want to call a function to return a data element? |
| 01:39:38 | Russ|werk | why would you do math with pointers? |
| 01:39:41 | as2 | I'm not sure the answer is to prevent people from being able to do it though. |
| 01:39:48 | cbiffle | Russ---: ...as opposed to functions that return nothing? |
| 01:40:05 | as2 | Russ, jump tables for example. |
| 01:40:14 | Russ|werk | cbiffle: I mean, for example, if you have a string as part of a class, and a function that returns that string |
| 01:40:23 | cbiffle | Ah, yes. The accessor-method question. |
| 01:40:30 | andersee | Russ|werk: data hiding |
| 01:40:45 | levi__ | Russ---: Whether or not you can access data elements in a class depends on the language. |
| 01:40:54 | cbiffle | Russ---: I agree that it's inconvenient if all you're doing is fetching a single data method. But say in a later revision of the class, that string isn't stored in RAM; it's generated dynamically using a bunch of data sources. Then having it wrapped in a function suddenly comes in quite handy. |
| 01:41:33 | cbiffle | levi__: Also true. |
| 01:41:39 | Russ|werk | cbiffle: rare case |
| 01:42:29 | Morn | I find java to be very convient for web based applications |
| 01:42:30 | cbiffle | Russ---: Not in my experience. |
| 01:42:39 | Russ|werk | then you have very poor planning |
| 01:42:43 | cbiffle | *sigh* |
| 01:42:50 | cbiffle | Russ---: It's not always possible to plan for things like that. |
| 01:43:24 | cbiffle | Russ---: Say later on it becomes evident that having your class able to access remote data sources transparently over a network is necessary. That's the sort of situation where data hiding comes in handy. |
| 01:43:26 | Russ|werk | I've never had a case where I needed that |
| 01:43:34 | cbiffle | How much professional development have you done? |
| 01:43:47 | cbiffle | Where management changes the direction of the project without checking with dev? |
| 01:43:59 | Russ|werk | about 4 years now, but I am my own management |
| 01:44:06 | Russ|werk | I also do a lot of hobby development |
| 01:44:08 | cbiffle | You're very fortunate. |
| 01:45:14 | cbiffle | *shrug* I've seen this happen many, many times. I'm glad you have the sort of foresight to be able to foretell and leave hooks for every possible feature your program might develop; in my case, being able to reimplement functions modularly comes in handy because of my limited scope. |
| 01:45:46 | Russ|werk | with c, its very simple to modify the interfaces though |
| 01:46:02 | cbiffle | Unless you're directly accessing a data member in a struct 400 times throughout 60 modules. |
| 01:46:15 | cbiffle | And then that data member suddenly becomes two, or a function. |
| 01:46:53 | Morn | I think that for certian applications one language is better than another |
| 01:47:02 | Morn | the right tool for the right job |
| 01:47:05 | cbiffle | Morn: Absolutely true. Generally it's C. :-) |
| 01:47:08 | cbiffle | ducks |
| 01:47:12 | Russ|werk | then write it that way in the first place, you have a choice |
| 01:47:24 | as2 | Haha. |
| 01:47:33 | as2 | I like FORTRAN for some jobs. |
| 01:47:34 | Morn | but Russ, his point is managment can change the direction of the code |
| 01:47:36 | cbiffle | Russ---: Unless you have no idea that need is going to arise! Have you had to maintain legacy code left by other teams a decade ago? This sort of thing happens! |
| 01:47:38 | as2 | hides under the table. |
| 01:47:41 | levi__ | Russ---: You may have a choice. Not everyone does. |
| 01:47:42 | Russ|werk | (I mean, write it in the way where its a function that returns data) |
| 01:47:55 | cbiffle | as2: You'd be surprise how much FORTRAN still goes around the psych dept at my university. |
| 01:48:27 | BZFlag | really really hates accessor methods and operator overloading. |
| 01:48:28 | as2 | Actually, I doubt it, at my ex uni everything is still fortran in the chem dept. |
| 01:48:48 | cbiffle | Russ---: Some of the projects I've been on has been modernizing software that was developed back when the company was saying "Someday, we'll be able to tell our investors with pride that we've never been connected to the Internet." (No joke.) |
| 01:49:04 | Russ|werk | I hate crufty code |
| 01:49:08 | BZFlag | c++ is very good at hiding performance problems. |
| 01:49:10 | cbiffle | Russ---: That's valid. |
| 01:49:14 | Russ|werk | lindent hels a little though though |
| 01:49:15 | as2 | They use 50 year old machines becuase the code is allready written, and FORTRAN for new stuff becuase of the vast libraries of code they have. |
| 01:49:27 | prpplague_away | as2: qbasic!!! |
| 01:49:40 | cbiffle | BZFlag: Operator overloading strikes me as a Bad Thing, but I haven't used it much. |
| 01:49:57 | Russ|werk | I like the idea of operator overloading myself |
| 01:50:04 | Morn | operator overloading is useful, but can hide some bad programming |
| 01:50:24 | cbiffle | I do like function overloading. There are times when I'll compile my C code as C++ just to get function overloading. |
| 01:50:35 | prpplague_away | Russ|werk: opverload hides problems, which i've recently been made aware of in my migration from SCO to linux |
| 01:50:39 | Russ|werk | function overloading? |
| 01:50:55 | cbiffle | Russ---: Having multiple functions, same name, different argument sets point to different routines. |
| 01:50:55 | as2 | qbasic doesnt have the code base they need for chem. If someone needs to find the determinant for a matrix they just look up the library and it takes a few mins. To do it in anything else would need it completely writing from the ground up. |
| 01:50:55 | BZFlag | if you are the only one touching the code, or if the code is supposed to be dog slow, then operator overloading makes for very readable, very slow code. quite easy to reach those goals. |
| 01:50:57 | Morn | taps her fingers waiting for opie to finish compiling |
| 01:51:12 | Russ|werk | gtg |
| 01:51:13 | cbiffle | as2: Are people really using QBasic for something still? |
| 01:51:35 | as2 | Not that I know of. |
| 01:51:52 | as2 | I used to program in Acorn BASIC. |
| 01:51:52 | cbiffle | BZFlag: Does operator overloading really introduce a performance hit? I figured it'd be resolved in the compiler to machine instructions. |
| 01:52:06 | Morn | I wish I still had the source code to the Back propigated neural net I wrote in pascal |
| 01:52:13 | as2 | Its still the only high level lang I'm fluent in, and now the platform that runs it is dead. |
| 01:52:13 | cbiffle | as2: I used to use QuickBasic, which was sort of what QBasic would have grown up to be, back in...88 or so. |
| 01:52:14 | Morn | in 1992 |
| 01:52:23 | BZFlag | cbiffle: operator overloading does not in and of itself cause performance problems. |
| 01:52:33 | cbiffle | BZFlag: compiler issues then? :-) |
| 01:52:55 | BZFlag | the problem is when the time impact of the overloading is not taken into account by the developer using said overloading. |
| 01:53:30 | cbiffle | BZFlag: Hm. As in they'd treat '+' being 'strcat' the same as '+' being an add instruction, or...? |
| 01:53:47 | BZFlag | yes. |
| 01:54:01 | as2 | Makes me glad I went over to electronics though, I get to use software designed in the last 50 years. |
| 01:54:04 | cbiffle | BZFlag: Ah, okay. That makes sense. (Particularly for someone like me who has the 68k instruction timings memorized. :-) |
| 01:54:07 | BZFlag | copying entire structures into locals by using = etc. |
| 01:54:20 | cbiffle | BZFlag: Oooh, that one in particular stands out. I see. |
| 01:54:57 | cbiffle | BZFlag: Are you Tim? |
| 01:55:19 | as2 | BZ is Tim. |
| 01:55:21 | cbiffle | (Which stands out as an obvious example of IHRP, the IRC Handle Resolution Protocol.) |
| 01:55:25 | | hmmm... bzflag is TimRiker or a 3D internet multiplayer multiplatform (linux, irix, win32, etc) opensource opengl Battlezone capture the flag game that you must try at http://BZFlag.org/ debian arm linux contains a version that will run on the ipaq at like 6 fps |
| 01:55:25 | BZFlag | ibot: bzflag |
| 01:55:49 | Morn | ack |
| 01:55:51 | Morn | irix |
| 01:55:53 | Morn | runs |
| 01:55:59 | cbiffle | I've been meaning to try BZFlag. The only thing holding me back was that a friend had had issues on *BSD. |
| 01:56:00 | levi__ | ack, 6fps |
| 01:56:22 | as2 | Noone converting it for the tux, haha. |
| 01:56:35 | cbiffle | BZFlag: Did you hear my bitching about the most recent DSP weirdnesses a few minutes ago? :-) |
| 01:57:01 | Morn | hey, what is this shanip2 I hear about |
| 01:57:19 | BZFlag | cbiffle: nope. SIGDINNER |
| 01:57:36 | cbiffle | BZFlag: *grin* |
| 01:59:56 | | okay, prpplague_away. |
| 01:59:56 | prpplague_away | ibot BZFlag is also CTO of Embedix |
| 02:03:16 | andersee | BZFlag: http://www.embedix.com/ |
| 02:03:33 | andersee | BZFlag: whois embedix.com is also amusing |
| 02:03:42 | as2 | How fast is the ARM in an ipaq? |
| 02:03:58 | cbiffle | 206, iirc |
| 02:04:10 | prpplague_away | as2: there is nothing wrong with legacy code, god knows i've been dealing with alot lately, but moving to an environment like gcc is a rude awakening, i'm finding arounf 20-30 problems a days with legacy code just because of the demands of POSIX compatiblity |
| 02:04:45 | as2 | I think legacy code is good if its well written. |
| 02:04:48 | prpplague_away | andersee: thats odd |
| 02:05:08 | as2 | The problem is someone has to write it well in the first place. |
| 02:05:36 | prpplague_away | as2: amen |
| 02:05:53 | prpplague_away | andersee: i know i've asked, but are you attending OLS? |
| 02:06:58 | as2 | The other problem is people keep changing the specs of the hardware. Which turns well written legacy code into utter rubbish legacy code. |
| 02:07:01 | andersee | prpplague_away: would love to, but probably not... |
| 02:07:10 | andersee | prpplague_away: Sortof a $$$ thing |
| 02:07:25 | prpplague_away | andersee: hmm |
| 02:09:24 | prpplague_away | as2: code is code is code , well writen does interact with hardware but with OS |
| 02:09:28 | Morn | well folks I'm going to get some rest |
| 02:09:29 | BZFlag | andersee: any power trouble up there? |
| 02:09:34 | prpplague_away | s/does/doesn't |
| 02:09:47 | BZFlag | thunderstorm is at full force here now. |
| 02:09:51 | prpplague_away | Morn: later babe |
| 02:09:52 | andersee | BZFlag: not yet |
| 02:09:53 | prpplague_away | lol |
| 02:10:12 | andersee | BZFlag: I've rsynced everything to my laptop though, just in case. |
| 02:10:14 | Morn | is away: while (tired) { sleep(5000); } |
| 02:12:02 | as2 | prp, everything interacts with the hardware at some level, and what I'm refering to is mainly asm optimisation. What runs fast on an 8086 is very substandard for 386, and at every level code should be doing other things to be running quickly. |
| 02:13:51 | prpplague_away | as2: but writing code that doesn't approtately determine the max/mins of short and long ints doesn't make sense |
| 02:13:59 | andersen_ | power died... |
| 02:13:59 | cbiffle | is glad he finally figured out linuxrc |
| 02:15:05 | as2 | Hmm, dont understand that. I dont use variables, I use memory. |
| 02:16:00 | as2 | The main reason I use asm, is becuase I find for what I want to do CPUs are agonisingly slow. |
| 02:16:23 | andersen_ | as2: perhaps you should use DSPs then? |
| 02:16:27 | cbiffle | You know, I really don't understand the antagonism a lot of my fellow BSD hackers have for Linux. I'm not finding it too terribly nasty on the phone... |
| 02:16:45 | cbiffle | is thinking out loud again. |
| 02:17:12 | as2 | DSPs are just as slow in my experience, and compaired to late x86 architecture, much slower. |
| 02:17:31 | cbiffle | BTW, is the LCD controlled by the mainboard or DSP board? i.e. were I to have to amputate the DSP board and build my own around a Ti-5x, would I need an LCD controller? |
| 02:18:04 | cbiffle | as2: What the hell are you -doing-, simulating nuclear explosions? |
| 02:18:25 | as2 | Oddly enough, thats not far away from one of the things Id like to do ;] |
| 02:19:11 | as2 | IT cant even rewrite a complete frame on a byte basis in realtime. |
| 02:19:28 | cbiffle | *nod* I do a lot of my development for Palm, and we have similar constraints. |
| 02:20:21 | as2 | I wanted to do realtime ray tracing, which still isnt possible. |
| 02:21:05 | cbiffle | You need more than one CPU, my friend. |
| 02:21:48 | as2 | I cant afford a connection machine. Even if thinking machines were still solvent. |
| 02:22:18 | cbiffle | BZFlag: If I power off the machine for a few minutes and bring it back up, the phone functions work for several minutes (~10), but gradually get cracklier and cracklier, and the message light starts blinking, first every few minutes, then closer together. Then it starts warbling and talking static. |
| 02:22:29 | cbiffle | BZFlag: If I didn't know better I'd say it was a cooling issue. |
| 02:23:24 | cbiffle | BZFlag: Thoughts? :-) |
| 02:23:31 | as2 | Buying multiple CPU machines isnt a wise move finantially, you pay more than twice the price for a 2CPU machine and get less than x2 the perfornace. And in 18months they release a single CPU machine that does the same. |
| 02:23:35 | BZFlag | back. reading. |
| 02:23:47 | as2 | cb, try freezer spray. |
| 02:23:59 | cbiffle | Oooh, it's warbling now. (Kind of like a ring, except it changes pitch every few seconds. Sounds like those electronic halloween ghosts.) |
| 02:24:24 | as2 | Makes finding thermal problems much easier. |
| 02:24:37 | cbiffle | as2: But it seems important to you to have the level of performance -now-, rather than 18 months from now. :-) |
| 02:25:03 | BZFlag | cbiffle: it's possible. is anything on one of the boards hot? |
| 02:25:19 | as2 | I want the level of performance we should have in 100 years today. |
| 02:25:25 | cbiffle | I popped it open earlier when it was doing it, and while I couldn't reach the mainboard, nothing on the DSP board is. |
| 02:25:28 | BZFlag | yeah, freezing it is a good idea. |
| 02:25:41 | cbiffle | It does this with power off and batteries on, though; is the CPU powered up when it's on batteries? |
| 02:25:48 | BZFlag | the rf shield is cool? |
| 02:25:56 | cbiffle | BZFlag: room temp. |
| 02:26:14 | cbiffle | Tell 'ya what. I'll open it up again. |
| 02:26:16 | as2 | The CPU shouldnt be the problem thermally. |
| 02:26:21 | BZFlag | cbiffle: nope, cpu is only on wall power. |
| 02:26:45 | cbiffle | BZFlag: Right, that's why I suspected the DSP. |
| 02:28:58 | cbiffle | The RF shield, while it's running, is just barely over room temp. |
| 02:29:49 | cbiffle | I don't think it's cooling. |
| 02:29:53 | jacques | andersen_: they have these new things now called UPS |
| 02:30:34 | andersen_ | jacques: got one on my server, switch, and DSL modem. Not my workstation. |
| 02:30:39 | cbiffle | If I reboot it, when tuxphone initializes, it blinks the msg light as usual but emits a burst of static rather than boo-de-beep. |
| 02:30:52 | andersen_ | jacques: my workstation is an power gobbling Athlon... |
| 02:31:31 | cbiffle | BZFlag: How would I go about getting Inferno back on here, then? |
| 02:31:38 | cbiffle | BZFlag: Sounds like it's my last hope. |
| 02:31:46 | jacques | andersen_: you can get 500kVa (units?) now for $39.95 at compusa |
| 02:31:50 | BZFlag | know how to use dd to split files? |
| 02:31:55 | cbiffle | BZFlag: sure. |
| 02:32:47 | BZFlag | dd an inferno image which you can get off http://TuxScreen.net/download/inferno/ into three pieces. 128k, 1M-128k, 3M. |
| 02:33:19 | BZFlag | flash the last over the ramdisk. the second over the kernel. and when you are sure those worked, then flash the first over blob. |
| 02:33:55 | cbiffle | Blob can't boot Inferno? |
| 02:34:12 | andersen_ | jacques: oh? |
| 02:34:20 | cbiffle | *grin* |
| 02:36:18 | jacques | andersen_: http://www.compusa.com/products/product_info.asp?product_code=288428 |
| 02:36:27 | BZFlag | cbiffle: nope, blob can't. If someone can figure ou how, i suppose we could add it. |
| 02:37:18 | andersen_ | jacques: looking... |
| 02:38:02 | jacques | andersen_: oh crud it's a rebate thing - but two qweeks ago they had them for $40 without the rebate :-\ |
| 02:38:12 | cbiffle | BZFlag: When you say 1M-128k, you mean 896k? |
| 02:39:43 | | 917504 |
| 02:39:43 | jacques | ibot: 1024 * (1024 - 128) |
| 02:39:57 | jacques | cbiffle: 917504 bytes |
| 02:40:00 | | 896 |
| 02:40:00 | cbiffle | ibot: 1024 - 128 |
| 02:40:03 | cbiffle | jacques: 896k |
| 02:40:10 | jacques | 917504 bytes |
| 02:40:14 | cbiffle | jacques: I was just verifying that he meant subtraction. |
| 02:40:25 | | 20 |
| 02:40:25 | andersen_ | ibot: (10^10^10^10^10)*2 |
| 02:40:32 | jacques | 128k is used for the bootldr |
| 02:40:40 | jacques | since erase size is 128k... |
| 02:40:45 | cbiffle | jacques: Right. I wanted to make sure he didn't mean 1M based at 128k. |
| 02:40:45 | jacques | cat make it any smaller |
| 02:40:48 | prpplague | later folks |
| 02:40:58 | jacques | cbiffle: ok, nm :-) |
| 02:41:01 | cbiffle | jacques: Just making sure I have everything straight before I brick the tux. |
| 02:41:23 | jacques | cbiffle: it works better if you dont brick it |
| 02:41:30 | cbiffle | jacques: I've heard that. |
| 02:41:37 | as2 | Damn, and I had the brick ready. |
| 02:42:25 | jacques | lol |
| 02:43:12 | cbiffle | If I wind up cannibalizing the phone for its mobo and LCD, I'll name the resulting product the Brick. |
| 02:43:42 | as2 | Because its small dense and you build houses out of it? |
| 02:43:49 | cbiffle | Absolutely. |
| 02:44:29 | as2 | A JTAG interface should make the system work again if the update screws it up. |
| 02:44:38 | cbiffle | as2: I'm out of moneys. :-) |
| 02:44:42 | cbiffle | as2: But yes. |
| 02:44:54 | jacques | when i order my next batch, I am going to get jtag |
| 02:45:04 | as2 | If your into electronics you can make one for peanuts. |
| 02:45:09 | cbiffle | When I can get the damn thing to run for a week straight, I'll order my next batch. :-) |
| 02:45:27 | as2 | But you need a glitch free LPT port. |
| 02:45:41 | cbiffle | as2: I wonder if the software groks BSD-style printer ports. |
| 02:46:34 | as2 | No clue, I dont know BSD or unix well, but I dream in JTAG statemachine currently. |
| 02:46:35 | jacques | i bet once num left < 100 sales / time goes way up |
| 02:46:51 | cbiffle | *nod* |
| 02:47:15 | cbiffle | I'd like to pick up one or two more, but I have to wonder if by the time mine was sold we'd entered into some sort of a 'bad' batch. |
| 02:47:36 | as_ | Re. |
| 02:48:33 | jacques | wb as_ |
| 02:50:14 | cbiffle | The phone is warbling while I'm updating the flash from blob...it's amusing now, it's like it knows I'm erasing its brain. Mwa ha ha ha. |
| 02:51:08 | cbiffle | Hm, fun fact. If you boot the kernel with the Inferno filesystem in place, the jffs2 drivers erase parts of it! Yay. |
| 02:51:10 | as_ | Lol. HAL style. |
| 02:51:49 | as_ | MAke it sing "Daisey Daisey" |
| 02:51:53 | jacques | daisy daisy.... |
| 02:51:58 | jacques | LOL |
| 02:51:59 | cbiffle | A lot of the static almost sounds like a voice, it's rather creepy. |
| 02:55:17 | as_ | Have you built your house over an indian burial ground by any chance? |
| 02:55:38 | cbiffle | Starting to wonder. |
| 02:55:56 | as_ | You might find your insurence doesnt cover that. |
| 02:57:58 | as_ | "House sucked through the screen of a linux phone into another dimention". |
| 02:58:15 | as_ | Get it added, you cant be too careful. |
| 03:01:58 | jacques | but you can make it up on the movie rights |
| 03:02:15 | as_ | Now there is a thought. |
| 03:02:50 | jacques | but you would have to dumb down all the technical stuff |
| 03:04:26 | BZFlag | yeah, use macs with macromedia displays. |
| 03:04:40 | as_ | Replace the technical stuff with digging up the bodies and burying them upside down. Dont want to burden the avarage cinima goer with complex stuff like linux. Link it to indian heratage and theyll accept anything. |
| 03:04:54 | jacques | exactly |
| 03:05:24 | as_ | Besides, the microsoft court case would never end. |
| 03:06:25 | as_ | Alians use a mac compatable operating system, makes sense indians use linux. |
| 03:08:29 | as_ | Turn torvalds into a medium channeling spirits and youve got a sequal. |
| 03:08:40 | jacques | if you watch hollyweird movies, you would think everyone in the world uses macs |
| 03:08:54 | as_ | Dont they? |
| 03:09:08 | as_ | looks slightly paniced. |
| 03:09:27 | cbiffle | Hm. Bricked the phone. |
| 03:09:33 | cbiffle | Something about the bootloader image it didn't like. |
| 03:09:45 | cbiffle | Though at least it's a singing brick. |
| 03:09:57 | as_ | Doh. |
| 03:10:08 | jacques | they cut the comm lines |
| 03:11:25 | as_ | The magic talisman will bring it back to life. |
| 03:12:32 | as_ | waves a JTAG interface over his tux to keep the power flowing. |
| 03:14:11 | as_ | Theyr advertising amstrad internet phones here. |
| 03:14:37 | as_ | Mono LCD, runs spectrum games. |
| 03:14:56 | jacques | mono ugh |
| 03:15:45 | as_ | Looks like it was designed in the 19th century when compaired to a tux. |
| 03:16:02 | jacques | heh |
| 03:29:03 | as_ | 4:30am, time to get some sleep. |
| 03:29:06 | as_ | Nite all. |
| 03:35:46 | BZFlag | http://www.amstrad.com/ams121101/default.shtml ? |
| 03:36:07 | scanline | is away: fun, games, and explosions. 100% bannana free |
| 04:05:28 | andersen_ | any last minute patches for busybox? |
| 04:12:18 | andersen_ | who was it that was working on the busybox /dev/initctl patch for busybox? |
| 04:20:53 | | c is for cookie |
| 04:20:53 | jacques | ibot: C |
| 04:20:57 | jacques | arg |
| 04:21:10 | | jacques: i haven't a clue |
| 04:21:10 | jacques | ibot: what is the speed of light in a vacuum? |
| 04:41:28 | BZFlag | andersee: pmisfit |
| 05:27:00 | jacques | is the who damn internet down today??? |
| 05:29:25 | | pmisfit was last seen on #tuxscreen 1 days, 7 minutes and 12 seconds ago, saying: is away: snooooze... [Fri Apr 26 06:22:13 2002] |
| 05:29:25 | andersee | ibot: seen pmisfit? |
| 05:33:43 | jacques | ==============installing ncurses into root dir.... |
| 05:33:43 | jacques | cp -a /home/packages/tuxscreen/buildroot-tux/build/stage/lib/libncurses.so* /home/packages/tuxscreen/buildroot-tux/root/lib/ |
| 05:33:44 | jacques | cp: cannot stat `/home/packages/tuxscreen/buildroot-tux/build/stage/lib/libncurses.so*': No such file or directory |
| 05:33:47 | jacques | ??? |
| 05:38:28 | jacques | grap |
| 05:38:30 | jacques | im 1 for 4 |
| 05:38:54 | jacques | 3 out of 4 things i tried to dl in make dir will not download for various reasons |
| 05:40:14 | jacques | make that 1 for 5 |
| 05:59:17 | Russ--- | ack! |
| 05:59:27 | Russ--- | imap is marking all my messages as new! |
| 05:59:27 | andersee | ping |
| 05:59:37 | Russ--- | pong |
| 05:59:44 | Russ--- | the inbox number won't stop going up |
| 05:59:48 | Russ--- | heeeeelllllp |
| 06:00:11 | andersee | /etc/init.d/imap stop |
| 06:20:13 | fontenot_ | what is wrong with this damn network? |
| 06:20:30 | fontenot_ | us is refusing connections |
| 06:20:34 | fontenot_ | irc is timing out |
| 06:20:58 | | Test Failed! |
| 06:20:58 | fontenot_ | ibot: test |
| 06:21:39 | fontenot_ | at least Im getting my standard 3kB/s from www.uclibc.org |
| 06:28:40 | andersee | fontenot: a test... |
| 06:28:55 | andersee | fontenot: Try grabbing from www.busybox.ne |
| 06:29:01 | andersee | fontenot: Try grabbing from www.busybox.net/downloads/uClibc |
| 06:29:31 | andersee | I think the folk hosting uClibc have throttled things pretty seriously |
| 06:30:03 | andersee | grumbles |
| 06:30:35 | fontenot | The requested URL /downloads/uClibc/ was not found on this server. |
| 06:30:56 | andersee | Ok, try via ftp... |
| 06:31:05 | andersee | I don't think they throttle ftp |
| 06:31:24 | andersee | Try ftp to uclibc.org |
| 06:31:39 | fontenot | what kind of fucked up ftp server is that? |
| 06:31:47 | andersee | dunno |
| 06:31:51 | andersee | what did it do? |
| 06:32:07 | andersee | I think it is proftp |
| 06:32:09 | fontenot | it keeps popping upt "document contains no sdata" |
| 06:32:43 | andersee | for what |
| 06:32:48 | andersee | ? |
| 06:32:49 | fontenot | for everything |
| 06:33:04 | fontenot | ok using wget I am getting full speed at ftp://www.busybox.net/uClibc/snapshots/ |
| 06:33:20 | fontenot | about 10 seconds instead of 5 minutes |
| 06:33:36 | andersee | $ lftp ftp://ftp.uclibc.org/uClibc/ |
| 06:33:44 | andersee | works for me... |
| 06:33:52 | fontenot | yes yes and wget works for me |
| 06:33:59 | fontenot | but my browser pops up those windows |
| 06:34:07 | fontenot | so the server is doing something bad |
| 06:34:09 | andersee | perhaps we need to change the url to bypass stupid apache throttling... |
| 06:34:16 | fontenot | because i dont have that problem on any other ftp servers |
| 06:34:46 | fontenot | 3kB/s - that tells me not many ppl are using the snapshots |
| 06:34:56 | fontenot | because you would probably have heard from more ppl than just me |
| 06:34:57 | andersee | The folks hosting www.uclibc.org for me stuck in an apache throttle. A pretty evil one I guess. |
| 06:35:18 | fontenot | who are those folks ? |
| 06:35:25 | fontenot | i guess they are doing it for free ? |
| 06:35:32 | andersee | An isp run by a friend of mine. yup. |
| 06:36:26 | fontenot | i could not download ntp from work because of theur stupid reverse dns requirement |
| 06:36:52 | fontenot | the gnu ncurses site is just plain off-net |
| 06:36:53 | andersee | Just checked -- 7 GB of uClibc downloads this month (via apache at least). dunno about cvs or ftp. |
| 06:37:23 | fontenot | microcom site was off-net but is back now I think |
| 06:39:26 | fontenot | wow someone put a fltk.mk ?? |
| 06:39:29 | fontenot | do you knwo who ?? |
| 06:39:37 | fontenot | wait - how will that work - stlport? |
| 06:39:52 | andersee | guess so |
| 06:40:06 | fontenot | ARG |
| 06:40:14 | fontenot | # FLTK/FLNX - a low-footprint GUI toolkit compatible with Microwindows |
| 06:40:24 | fontenot | that should be flnx.mk or something |
| 06:40:28 | fontenot | not fltk.mk |
| 06:40:43 | fontenot | FLTK_SOURCE:=flnx-0.16.tar.gz |
| 06:41:23 | fontenot | ah, ftp.gnu.org is back up |
| 06:44:47 | fontenot | and slang.mk is really minislang ? |
| 06:44:54 | andersee | yup |
| 06:45:02 | andersee | Only has the screen handling stuff. |
| 06:45:18 | andersee | Slang provides a tons of other (useless) stuff |
| 06:45:38 | andersee | If you look, slang.mk pulls from busybox.net |
| 06:48:04 | fontenot | yeah i noticed that |
| 06:48:28 | fontenot | i just think ppl might get confised if they try to build stuff that requires fltk or slang |
| 06:49:06 | andersee | fontenot: I don't know of _any_ slang using packages that use it for anything beyond screen handling... |
| 06:49:42 | andersee | fontenot: But clearly if fltk is not the Fast Light Toolkit, it needs a name change |
| 06:50:15 | fontenot | ok, i accept that about slang, but fltk.mk needs to become flnx.mk |
| 06:50:29 | fontenot | cool, strace.mk |
| 06:51:05 | fontenot | new zlib version - also cool |
| 06:51:20 | andersee | Yeah, I did that |
| 06:52:20 | fontenot | :-) |
| 06:54:19 | fontenot | im going to try to do a glib.mk and gtk+.mk |
| 06:55:27 | andersee | :) |
| 06:56:27 | fontenot | i can build them manually but it's a pain |
| 07:00:25 | andersee | I build libglib last night vs uClibc. Worked fine |
| 07:00:43 | fontenot | do you have notes? |
| 07:01:10 | fontenot | i had to edit libtool |
| 07:01:13 | fontenot | did you? |
| 07:01:20 | fontenot | and if not, how did you do it? |
| 07:02:57 | andersee | I just did something like 'CC=foo ./configure --prefix=$(STAGING_DIR); make CC=foo' and it worked. I was compiling for x86 though, not cross copiling for arm |
| 07:03:33 | fontenot | ooooooh |
| 07:03:45 | fontenot | well that's the difference |
| 07:06:49 | fontenot | ARG |
| 07:07:03 | fontenot | ntpdate screws up because timezone is ??? |
| 07:07:11 | fontenot | how do I fix that timezone ??? thing? |
| 07:07:25 | fontenot | # date |
| 07:07:25 | fontenot | Sat Apr 27 07:09:58 ??? 2002 |
| 07:08:04 | andersee | fontenot: uClibc timezone support is stubbed out |
| 07:08:27 | andersee | fontenot: Its on the todo list for the next release, but I gladly accept patches. :) |
| 07:09:17 | fontenot | so in other words it does not work |
| 07:09:29 | pmisfit | is back (gone 25:47:17) |
| 07:09:35 | pmisfit | is reading... |
| 07:09:42 | fontenot | i.e. there is currently no workaround |
| 07:09:50 | fontenot | other than writing timezone support for uclibc |
| 07:10:15 | fontenot | do setuid executables work with uclibc ? |
| 07:10:16 | andersee | fontenot: rdate |
| 07:10:26 | andersee | fontenot: sure |
| 07:10:53 | fontenot | yep they seem to |
| 07:10:59 | pmisfit | damn. i take a day off irc... two people ask about stuff i'm working on that's late/missing/broken. |
| 07:11:19 | pmisfit | andersee: saw tom oehlert's post, and replied. |
| 07:11:26 | fontenot | we need to make /dev/null world writable |
| 07:11:30 | andersee | pmisfit: saw that,,, |
| 07:11:33 | fontenot | well, 666 in general |
| 07:11:36 | pmisfit | fontenot: i'm to blame for ncurses too. |
| 07:11:51 | andersee | pmisfit: Sounds like init will be getting a facelift soon |
| 07:11:51 | | 666 is microsoft |
| 07:11:51 | pmisfit | 666? |
| 07:11:59 | pmisfit | starts singing bad ozzy osbourne tunes... |
| 07:12:39 | pmisfit | andersee: is that ok? i hope i didn't throw gasoline on anything... |
| 07:12:59 | andersee | nah |
| 07:13:32 | pmisfit | cool. |
| 07:13:42 | fontenot | heh I can run X as non-root user |
| 07:13:57 | fontenot | just had to setuid Xfbdev and make /dev/null 666 |
| 07:14:02 | andersee | I'm happy to have things get into shape. I _really_ dislike Tom's idea of hacking the first field to add a tty.... |
| 07:14:15 | andersee | much better to just add a new field if needed |
| 07:14:18 | pmisfit | i raised an eyebrow at that too. |
| 07:14:30 | andersee | s/tty/id/ |
| 07:14:51 | fontenot | hmmm, my load average is 1.0 |
| 07:15:46 | pmisfit | tom's got way more stripes than me, so i probably won't be able to stop the tty field hack. |
| 07:16:08 | fontenot | arg something is running away wioththe cpu and I cant tell what |
| 07:16:13 | fontenot | how can I tell??? |
| 07:16:30 | pmisfit | wasn't someone working on a "top" port for busybox? |
| 07:16:39 | fontenot | yes but we dont have that so how? |
| 07:16:43 | | cbiffle was last seen on #tuxscreen 4 hours, 6 minutes and 58 seconds ago, saying: Though at least it's a singing brick. [Sat Apr 27 04:09:45 2002] |
| 07:16:43 | pmisfit | ibot: seen cbiffle? |
| 07:16:51 | andersee | pmisfit: ps now has memory use, but not yet cpu use |
| 07:17:30 | fontenot | it does? how do I get that? |
| 07:17:55 | fontenot | oh i gues sit's not in 60.2 anyway |
| 07:17:59 | andersee | fontenot: its in CVS now. I'm trying to get 0.60.3 released, which will have it |
| 07:18:14 | fontenot | so right now, there is no way for me to tell which process is taking all my cpu ? |
| 07:18:55 | andersee | not easily... |
| 07:19:15 | fontenot | i am feeling somewhat stifled today |
| 07:19:15 | andersee | you can grovel about in /proc/[0-9]* |
| 07:19:51 | fontenot | here i go killing processes until load average goes down |
| 07:20:12 | andersee | fontenot: run ps |
| 07:20:24 | fontenot | and ? |
| 07:20:32 | andersee | Look for processes doing 'R' (running) |
| 07:20:39 | andersee | As opposed to 'S' sleeping |
| 07:20:44 | fontenot | first process i killed was it |
| 07:20:48 | fontenot | /sbin/udhcpc -i eth0 |
| 07:21:05 | pmisfit | !!? |
| 07:21:07 | fontenot | killed that and now load is down to 0.34 and falling |
| 07:21:15 | fontenot | killed that and now load is down to 0.34 and falling |
| 07:21:29 | fontenot | and heading for 0 |
| 07:21:31 | pmisfit | runaway dhcp? |
| 07:21:35 | fontenot | apparently |
| 07:21:45 | fontenot | it wasnt X - i killed that and also the telnet session and no effect |
| 07:21:55 | fontenot | actually load went up a bit when i killed those |
| 07:22:25 | fontenot | i guess next time I can send over strace and attach to the running process |
| 07:22:36 | fontenot | see what dhcp is doing |
| 07:22:40 | pmisfit | what was in its /proc/xxx/fd directory? |
| 07:22:46 | fontenot | it's happened before - it will happen again |
| 07:22:53 | fontenot | pmisfit: it's dead |
| 07:23:05 | fontenot | but i will look next time this happens, ok? |
| 07:23:14 | fontenot | load is now 0.03 |
| 07:24:27 | pmisfit | fontenot: did you resolve the curses issues you were asking about? |
| 07:24:50 | pmisfit | looks through his notes to remember where he left off with ncurses... |
| 07:25:22 | fontenot | pmisfit: it looks like lib is actually built but for some reason does not get installed into staging and thus cannot be copied to root |
| 07:26:12 | pmisfit | i see. i had a nasty recursive path problem recursive path problem. |
| 07:26:28 | pmisfit | it was getting added twice added twice and i couldn't figure out why. |
| 07:26:45 | fontenot | heh |
| 07:27:16 | pmisfit | if you need it i'll poke at it a bit. |
| 07:27:41 | fontenot | pmisfit: i dont need it right now - I was just trying all the new .mks out |
| 07:28:02 | fontenot | looks like libncuses.so is 260kB stripped |
| 07:28:11 | pmisfit | it's a big boy. |
| 07:28:25 | pmisfit | not as big as openssl tho ;-) |
| 07:28:50 | fontenot | ack |
| 07:29:01 | pmisfit | i still have to yank out all the nonessential ciphers, and i think strip it too. |
| 07:29:12 | fontenot | there really needs to be a tinyssl that leaves out unneeded functioality |
| 07:29:39 | pmisfit | haven't found one yet. they have to exist tho.. wap apps on cellphones, etc... |
| 07:30:20 | pmisfit | tries ncurses again... like a scab he can't stop picking... |
| 07:31:00 | fontenot | anyone here played with links on the tuxscreen? it's in buildroot |
| 07:31:20 | pmisfit | does it build? |
| 07:31:47 | fontenot | yep |
| 07:31:56 | pmisfit | hmm. looks like it requires neither slang or ncurses.. (!) |
| 07:32:10 | pmisfit | s/or/nor/ |
| 07:33:00 | fontenot | nope |
| 07:33:05 | fontenot | 688272 bytes stripped |
| 07:33:14 | pmisfit | :-/ |
| 07:33:17 | fontenot | ARM |
| 07:33:18 | fontenot | ARG |
| 07:33:22 | pmisfit | heh! |
| 07:33:27 | fontenot | internal compiler error when trying to build openssl |
| 07:33:38 | fontenot | /home/packages/tuxscreen/buildroot-tux/build/stage/bin/gcc -I.. -I../../include -fPIC -DDSO_DLFCN -DHAVE_DLFCN_H -I/home/packages/tuxscreen/buildroot-tux/build/stage/include -DNO_ASM -DL_ENDIAN -DTERMIO -O3 -fomit-frame-pointer -Wall -c -o pem_lib.o pem_lib.c |
| 07:33:38 | fontenot | pem_lib.c: In function `PEM_ASN1_write_bio': |
| 07:33:38 | fontenot | pem_lib.c:410: internal error--unrecognizable insn: |
| 07:33:38 | fontenot | (insn 1093 1090 1089 (set (reg:QI 12 ip) |
| 07:33:39 | fontenot | (mem:QI (plus:SI (reg:SI 5 r5) |
| 07:33:40 | fontenot | (const_int 1264 [0x4f0])) 0)) -1 (nil) |
| 07:33:42 | pmisfit | ARM-ARG. the cross-grumble. |
| 07:34:00 | fontenot | that might be my toolchain :-( |
| 07:34:04 | fontenot | other ppl might not see that |
| 07:34:05 | pmisfit | dude you have the wittiest typos i've ever seen. |
| 07:34:10 | fontenot | lol thanks |
| 07:34:47 | fontenot | maybe i should try building it with gcc3 |
| 07:35:01 | pmisfit | i use 2.95.3 |
| 07:35:08 | pmisfit | builds, runs, no prob. |
| 07:36:10 | fontenot | must be my toolchain then |
| 07:36:13 | fontenot | i built it myself |
| 07:36:38 | fontenot | usually reducing the optim level for the problem file fixes it |
| 07:36:49 | pmisfit | i think yours was 2.95.2, and that was the killer. |
| 07:36:57 | pmisfit | stuff got fixed in .3 |
| 07:37:00 | fontenot | nope mine is 2.95.4 |
| 07:37:14 | pmisfit | hmm.... |
| 07:37:23 | fontenot | with the phil patch |
| 07:37:31 | pmisfit | phil? |
| 07:37:51 | pmisfit | (wow, i asked a question and ibot didn't try to answer it!) |
| 07:37:55 | | no idea, pmisfit |
| 07:37:55 | pmisfit | ibot: you awake? |
| 07:38:01 | pmisfit | just checking. |
| 07:39:20 | fontenot | # Phil Blundell |
| 07:39:29 | fontenot | http://www.armlinux.org/~philb/ |
| 07:39:39 | pmisfit | ah. thanks. |
| 07:40:46 | andersee | fontenot: change the optimization from -O3 to say -O1 |
| 07:40:46 | | andersee: that doesn't look right |
| 07:40:55 | | Sorry, andersee, I'll keep my mouth shut. |
| 07:40:55 | andersee | ibot: shut up |
| 07:41:07 | | ibot is nothing but a fulminating half-mouthful of beslubbering waffle-house grits. |
| 07:41:07 | andersee | ibot: insult ibot |
| 07:42:14 | fontenot | http://intimate.handhelds.org/jacques/matchbox16.png |
| 07:43:05 | andersee | fontenot: is that lynx? |
| 07:43:38 | pmisfit | fontenot: hmm. looks like "phil" recommends 2.95.3 too... |
| 07:44:20 | fontenot | phil recommends 2.95 cvs branch which is what i use |
| 07:44:24 | fontenot | plus his patch |
| 07:44:32 | fontenot | that's links |
| 07:44:41 | fontenot | Out of Memory: Killed process 397 (links). |
| 07:44:43 | fontenot | GRRR |
| 07:44:47 | andersee | err. links. Thats what I meant. |
| 07:44:52 | fontenot | how much memory does it need??? |
| 07:44:57 | fontenot | i had 6MB free |
| 07:45:03 | andersee | added links when lynx wouldn't cross compile |
| 07:45:20 | andersee | hasn't tried it yet though |
| 07:45:51 | fontenot | ok i started it with slashdot.org and i have 5mb free |
| 07:45:59 | fontenot | but now i bet when i go to read more it will die again |
| 07:46:12 | andersee | fontenot: what kernel? |
| 07:46:58 | fontenot | current buildroot kernel |
| 07:47:06 | fontenot | Linux tuxscreen 2.4.18-rmk2-tux1 #1 Tue Mar 26 01:00:57 PST 2002 armv4l unknown |
| 07:47:09 | fontenot | it did it again |
| 07:48:06 | pmisfit | fontenot: sorry, man. looks like it's just not a good evening... |
| 07:48:22 | fontenot | links looks cool |
| 07:48:34 | fontenot | but you;re right - i might have to go watch tivo |
| 07:48:48 | pmisfit | heh. |
| 07:49:53 | pmisfit | clear the head, wake up the next morning, and all of a sudden it makes perfect sense.... |
| 07:50:47 | fontenot | maybe links just has a problem with slashdot |
| 07:50:49 | fontenot | dunno |
| 07:51:03 | fontenot | i am browking another site without it getting killed yet |
| 07:51:39 | fontenot | oh never mind it did get killed |
| 07:52:44 | fontenot | OOM killer seems to be too brutal - it's killing things with 4MB left |
| 07:52:48 | fontenot | that's total crap |
| 07:53:01 | andersee | fontenot: Needs the -aa vm patches |
| 07:53:14 | fontenot | what will that do? |
| 07:53:29 | andersee | fontenot: teach the kernel vm subsystem to behave itself |
| 07:54:00 | fontenot | hm |
| 07:54:05 | fontenot | we need something |
| 07:54:22 | fontenot | 4MB is 25% of my RAM and greater than that if you dont count the kernel |
| 07:56:01 | fontenot | will those patches even apply to the arm kernel ? |
| 07:56:15 | andersee | should |
| 07:57:57 | pmisfit | been running links for the past few minutes... surfing /. , no probs |
| 07:58:10 | pmisfit | fontenot: toolchain again? |
| 07:58:31 | fontenot | pmisfit: on your tux with X? |
| 07:58:42 | fontenot | pmisfit: how much free RAM you have ? |
| 07:58:58 | fontenot | OOM seems to strike around 4.25MB free |
| 07:59:04 | pmisfit | ark.. with x running, but in a telnet window from my desktop... not the same.. |
| 07:59:56 | fontenot | pmisfit: how much free ram ? |
| 08:00:21 | fontenot | cat /proc/meminfo |
| 08:00:33 | pmisfit | total: used: free: shared: buffers: cached: |
| 08:00:33 | pmisfit | Mem: 14528512 8421376 6107136 0 4096 2678784 |
| 08:01:16 | fontenot | if you get that down to 4MB or less I will start to wonder abot myself |
| 08:01:25 | fontenot | what does lowfree say ? |
| 08:01:39 | fontenot | LowFree: 4596 kB |
| 08:01:41 | pmisfit | reading harlan ellison's rant about intellectual "property"... interesting... |
| 08:01:43 | pmisfit | LowTotal: 14188 kB |
| 08:01:43 | pmisfit | LowFree: 5964 kB |
| 08:01:51 | fontenot | that seems to be the point at which it strikes |
| 08:02:11 | pmisfit | strikes! ... from out of the shadows.... |
| 08:02:34 | pmisfit | how can i poke it? hmmm.. what shall i open... |
| 08:03:05 | fontenot | do what i did - start a bunch of rxvt's |
| 08:05:10 | pmisfit | here goes... |
| 08:06:40 | pmisfit | hm |
| 08:06:44 | fontenot | hm? |
| 08:06:52 | pmisfit | DISPLAY issues.. |
| 08:07:03 | fontenot | export DISPLAY=:0 |
| 08:07:27 | pmisfit | rxvt: can't open display :0.0 |
| 08:07:37 | fontenot | maybe OOM killed your x server |
| 08:07:42 | fontenot | tail dmesg |
| 08:07:55 | fontenot | dmesg|tail |
| 08:07:59 | pmisfit | you're right. |
| 08:08:09 | pmisfit | same bug? |
| 08:08:12 | fontenot | this is what mine looke like: |
| 08:08:15 | fontenot | # dmesg|tail |
| 08:08:15 | fontenot | Out of Memory: Killed process 403 (links). |
| 08:08:15 | fontenot | Out of Memory: Killed process 414 (links). |
| 08:08:16 | fontenot | Out of Memory: Killed process 376 (minitime). |
| 08:08:16 | fontenot | Out of Memory: Killed process 377 (minitime). |
| 08:08:16 | fontenot | Out of Memory: Killed process 373 (minisys). |
| 08:08:18 | fontenot | Out of Memory: Killed process 434 (links). |
| 08:08:20 | fontenot | Out of Memory: Killed process 367 (rxvt). |
| 08:08:22 | fontenot | Out of Memory: Killed process 372 (matchbox). |
| 08:08:23 | pmisfit | NWFPE: more[246] takes exception 00000002 at c017a270 from 00037210 |
| 08:08:24 | fontenot | Out of Memory: Killed process 442 (rxvt). |
| 08:08:26 | fontenot | Out of Memory: Killed process 447 (rxvt). |
| 08:08:28 | fontenot | it's the darn kernel |
| 08:08:36 | pmisfit | not on mine |
| 08:08:36 | fontenot | i heard OOM hwas bad |
| 08:08:54 | fontenot | pmisfit: so your x server died for other reasons? |
| 08:09:04 | fontenot | start it and try again please |
| 08:09:04 | pmisfit | looks like. |
| 08:09:07 | pmisfit | sure |
| 08:10:05 | pmisfit | hmm |
| 08:10:16 | pmisfit | my telnet session died too... |
| 08:10:23 | pmisfit | something's flaky... |
| 08:10:27 | fontenot | but not because of OOM ? |
| 08:10:35 | fontenot | maybe reboot if it's not too much of a pain |
| 08:10:38 | pmisfit | what loglevel? |
| 08:10:45 | pmisfit | reboot's easy |
| 08:10:53 | fontenot | what about loglevel? |
| 08:11:21 | pmisfit | maybe you're getting more verbose kernel log messages than i am |
| 08:11:41 | fontenot | you mean when you type dmesg you dont get much ? |
| 08:12:09 | fontenot | found this too (OT): Apr 26 15:57:46 (none) user.debug udhcpc: Sending renew... |
| 08:12:12 | fontenot | several of those |
| 08:13:38 | pmisfit | ya. oh, WHY did you point me at /. |
| 08:14:00 | pmisfit | i can't read /. , gets me too pissed off... i'm too tired to be this angry... |
| 08:15:19 | pmisfit | i avoid /. , politechbot, eff, and of course all the major media too. gets me too upset... too distracting... can't function... |
| 08:15:39 | pmisfit | anyway, back on topic. starting up many rxvt's again. |
| 08:17:09 | fontenot | god damn it |
| 08:17:19 | pmisfit | still holding... |
| 08:17:30 | fontenot | freshmeat has 138 projects marked as browsers and only about 4 of them actually are |
| 08:17:37 | pmisfit | heh |
| 08:17:37 | fontenot | holding? |
| 08:17:43 | pmisfit | i.e. not crashing. |
| 08:18:01 | pmisfit | whoops... links died |
| 08:18:05 | fontenot | this is a browser: |
| 08:18:07 | fontenot | 7. Apache Server Tic Tac Toe |
| 08:18:07 | fontenot | Allows you to play Tic Tac Toe with your Apache server. |
| 08:18:13 | fontenot | according to freshmeat |
| 08:18:24 | fontenot | dis or was killed ? |
| 08:18:53 | pmisfit | Out of Memory: Killed process 200 (links). |
| 08:18:53 | pmisfit | # /usr/X*/bin/rxvt & |
| 08:18:53 | pmisfit | # rxvt: can't load font "7x14" |
| 08:18:53 | pmisfit | Out of Memory: Killed process 215 (links). |
| 08:19:29 | pmisfit | you're right. trouble in river city, and that starts with t and that rhymes with memory. |
| 08:20:20 | pmisfit | exhausted... gotta sleep, man. |
| 08:20:41 | fontenot | ok sleep well |
| 08:20:50 | fontenot | im gonna go watch tivo soon anyway |
| 08:20:59 | pmisfit | thanks for staying on top of these x issues... |
| 08:21:18 | BZFlag | hmmm.. does tinyx include virtual console support? I think that's off on the ipaq, but enabled on the tux |
| 08:21:31 | pmisfit | is away: passing out.. |
| 08:21:32 | fontenot | BZFlag: what's that ? |
| 08:21:37 | fontenot | later pmisfit |
| 08:21:51 | BZFlag | fontenot: ctrl-alt-f* to switch consoles. |
| 08:22:13 | fontenot | BZFlag: that might have something to do with why x likes to die when i hit ctrl-C and why we have the "blue screen" problem |
| 08:22:56 | fontenot | well ctrl-alt-f[1234] has no effect |
| 08:23:32 | BZFlag | fontenot: that's my theory |
| 08:24:55 | fontenot | i can see why that would be off on the ipaq :-) |
| 08:29:02 | fontenot | BZFlag: you know anything we can do about this OOM Killer problem ? |
| 08:30:04 | BZFlag | nope. have not had it happen here. |
| 08:30:52 | fontenot | how much ram you have which kernel you using what's the LowFree: value in /proc/meminfo ? |
| 08:40:48 | BZFlag | total: used: free: shared: buffers: cached: |
| 08:40:48 | BZFlag | Mem: 14528512 4198400 10330112 0 4096 974848 |
| 08:40:56 | BZFlag | LowFree: 10088 kB |
| 08:41:14 | BZFlag | Linux tuxscreen 2.4.18-rmk2-tux1 #6 Fri Apr 12 01:12:34 MDT 2002 armv4l unknown |
| 08:42:31 | fontenot | BZFlag: if you get that lowfree down below 4MB without things getting killed, let me know |
| 08:42:53 | fontenot | killer seems to strike at around 4.5MB |
| 08:47:51 | andersee | prepares to tag busybox 0.60.3 |
| 08:51:51 | andersee | finds a bug |
| 09:14:35 | BZFlag | eep. dead blob. jtag recovery... |
| 09:16:38 | andersee | eep! |
| 09:22:14 | BZFlag | woohoo! just booted with an empty kernel partition but loading the kernel from jffs2. ;-) |
| 09:22:23 | BZFlag | now to make it automatic... |
| 09:22:37 | BZFlag | s/but/by/ |
| 09:23:49 | andersee | :) |
| 09:23:57 | andersee | good news indeed |
| 09:25:44 | andersee | Amazing the amount of development that has centered on these things... |
| 09:27:21 | BZFlag | agreed. handy to have cheap hardware around. ;-) |
| 09:43:39 | BZFlag | Tangent: hey. fixed another trivial ibot bug. |
| 09:43:44 | | zero |
| 09:43:44 | BZFlag | ibot 1-1 |
| 09:44:03 | BZFlag | when math results in a '0' he used to not reply. |
| 09:44:32 | | zero |
| 09:44:32 | andersee | ibot: (10^10^10^10^10^10)*2 |
| 09:44:58 | | 54 |
| 09:44:58 | andersee | ibot: 6*9 |
| 09:45:11 | | andersee: huh? |
| 09:45:11 | andersee | ibot: sqrt(2) |
| 09:47:03 | | 10077696 |
| 09:47:03 | BZFlag | ibot 6**9 |
| 09:51:28 | | fontenot: what? |
| 09:51:28 | fontenot | ibot: infinity / 0 |
| 09:51:51 | | andersee: what? |
| 09:51:51 | andersee | ibot: 9/0 |
| 09:51:57 | | andersee: i'm not following you... |
| 09:51:57 | andersee | ibot: 9 / 0 |
| 09:56:02 | BZFlag | the boot hacking I'm doing is for the vercel, I just don't have one home at the moment, and even if I did, I don't have jtag for it. |
| 09:56:13 | BZFlag | have to reburn the flash in the office. |
| 10:01:36 | BZFlag | cool. flashing a ramdisk at 0x20000 with a kernel that has matching mtd layout and boot args. |
| 10:04:38 | BZFlag | suspense is killin ya, huh? |
| 10:05:03 | fontenot | arg that was an accident |
| 10:06:05 | fontenot | who is in charge of tuxscreen kernel? Russ? |
| 10:07:50 | BZFlag | Creating 2 MTD partitions on "SA1100 flash": |
| 10:07:50 | BZFlag | 0x00000000-0x00020000 : "BLOB boot loader" |
| 10:07:50 | BZFlag | 0x00020000-0x00400000 : "rootfs" |
| 10:08:01 | BZFlag | but then it did not finish booting. =( |
| 10:08:48 | BZFlag | fontenot: Russ has it now, but I could do a release if needed. he's been doing great, so I have just let him run. ;-) |
| 10:13:00 | andersee | just tagged busybox 0.60.3 |
| 10:13:21 | BZFlag | woohoo! |
| 10:13:36 | BZFlag | now update buildroot. ;-) |
| 10:14:18 | andersee | the busybox patch is broken, of course... |
| 10:14:46 | andersee | When we switch to 0.61 then we can use the new config system.... no more patches |
| 10:14:55 | BZFlag | that will be nice. |
| 10:15:21 | andersee | hopes this is the end of 0.60.x... |
| 10:15:59 | fontenot | BZFlag: I just want to ask him about the OOM killer thing |
| 10:16:12 | fontenot | i need to talk to somebody |
| 10:16:24 | fontenot | i just searched the arm-kernel list and no luck |
| 10:18:40 | fontenot | why not switch to 0.61 now? |
| 10:19:21 | andersee | 0.60.3 is stable, and I need to spend some time stabilizing 0.61.... |
| 10:19:34 | andersee | It still has a few build problems for starters |
| 10:19:50 | fontenot | ah ok |
| 10:30:46 | BZFlag | hmm... can't get a prompt. |
| 10:31:19 | andersee | ?? |
| 10:33:56 | BZFlag | wierd jffs2 messages. |
| 10:34:14 | andersee | I just commited the buildroot changes for 0.60.3 |
| 11:02:48 | BZFlag | found some md5 bugs in blob... |
| 11:13:51 | ctr | oooh |
| 13:02:31 | ctr | is away: shower |
| 13:10:02 | Morn | is away: showering, don't peek! |
| 13:16:26 | ctr | is back (gone 00:13:45) |
| 13:29:07 | file | awakens |
| 15:37:18 | prpplague | howdy howdy |
| 15:37:33 | file | prpplague : ever get your thing from Cybiko made? |
| 15:37:48 | prpplague | BZFlag: gosh, i figured we'd less of you with your new job |
| 15:38:10 | prpplague | file: ya we've been doing some looking, but i think the project has been shelved |
| 15:38:23 | file | prpplague : hehe |
| 15:38:26 | prpplague | file: we're busy with a complete SCO to linux migration |
| 15:39:04 | file | I see... |
| 15:39:21 | prpplague | file: we might be contracting a sa-1110 board from seletz |
| 15:40:35 | file | could use another Cybiko |
| 15:40:46 | file | tomorrow I take my current one to get air brushed :) |
| 15:40:54 | file | it's getting a coat of black, red, and white |
| 15:40:58 | file | with red maple leafs |
| 15:41:09 | prpplague | file: hmmm interesting |
| 15:41:32 | prpplague | file: if the project gets cancelled i'll see what we are going to do with the dev units |
| 15:42:22 | file | you have dev units? |
| 15:42:46 | file | neat |
| 15:43:26 | prpplague | file: we have 4 or 5 extreme and 2 or 3 classics |
| 15:49:07 | | TomW: i'm not following you... |
| 15:49:07 | TomW | ibot cybiko |
| 15:49:23 | | OK, TomW. |
| 15:49:23 | TomW | ibot, you are dumb |
| 15:50:24 | | OK, prpplague. |
| 15:50:24 | prpplague | ibot: cybiko is a small SH processor based handheld with rf communications at http://www.cybiko.com |
| 15:50:35 | | somebody said cybiko was a small SH processor based handheld with rf communications at http://www.cybiko.com |
| 15:50:35 | prpplague | ibot: cybiko |
| 15:51:59 | prpplague | TomW: morning |
| 15:52:07 | prpplague | TomW: whatcha working on today? |
| 15:52:48 | TomW | Hi prpplague, just fine tuning some of the matchbox stuff. getting rid of manpages, stripping files. |
| 15:53:16 | TomW | andersee: hey! Erik has a colorized ls now! |
| 15:53:43 | TomW | cool! |
| 15:56:22 | TomW | prpplague: working today, or just lonely? |
| 15:56:22 | prpplague | argh, i've lost the din connect for my tux ps |
| 15:56:33 | TomW | working. heh |
| 15:57:25 | TomW | prpplague: tinyx takes up a lot of space, I'll need to get the hard drive working pretty soon. |
| 15:57:54 | prpplague | TomW: that it does |
| 15:58:23 | prpplague | TomW: hey have you considered any cases for the tux dev board? |
| 15:58:25 | TomW | prpplague: trying to get rid of some stuff: xeyes, etc. |
| 15:58:33 | TomW | prpplague: no. |
| 15:59:03 | TomW | prpplague: I am having too much fun playing with the code, and such, right now. |
| 16:06:22 | prpplague | TomW: correct me if i'm wrong, the main tux board only has inputs for 3.3v right? |
| 16:06:45 | prpplague | TomW: i'm looking at schmatics, and don't see a 5v on j600 or j601 |
| 16:09:00 | TomW | prpplague: correct, the Tux ARM CPU board only runs on +3.3v, there is no need for a +5 supply there, unless you are using a +5v pcmcia card. |
| 16:09:29 | TomW | prpplague: all the chips on the PCB are LV (low voltage), +3.3v. |
| 16:10:00 | prpplague | TomW: gotcha, so when i need the 5v card i can do the pcmcia hack and route 5v through an un-used pin on j600 or j601 |
| 16:11:13 | prpplague | TomW: damn this makes me want to go back to school..... |
| 16:11:59 | TomW | prpplague: heh, you are in school. Most knowledge comes later on, school teaches you theory, real life teaches you what works. |
| 16:13:06 | prpplague | TomW: ya, but there are alot basics i'm really missing, any suggestions on a got book on learn digital circuit designs? |
| 16:15:58 | TomW | prpplague: I get asked that a lot. I never have an answer. I only went to a Community College for a math course. I learned by doing, I studied other engineers' designs and studied the datasheets. |
| 16:16:47 | prpplague | TomW: must have a natural gift |
| 16:17:35 | TomW | prpplague: I have always seemed to be able to visualize things well, build large constructs in my head. Gift? Yes. |
| 16:20:15 | TomW | prpplague: I have found that a lot of "engineers" are good studys in school, get good grades and have impressive resume's. But, when it comes down to the real thing, most of them don't do well as they have no passion for what it is that they are doing: it is a job. When I was doing hiring for engineering, etc., I would look for those that had a passion for the work. e.g. "it is fun". Not many of those people. |
| 16:20:33 | TomW | prpplague: lots of mediocre engineers, few good ones. |
| 16:20:45 | prpplague | TomW: which uart is the ttyS on that tux uses for serial console? |
| 16:21:13 | prpplague | TomW: agreed, i do the same on the software side |
| 16:21:36 | prpplague | TomW: i'm more of an emergency problem solver here at my company |
| 16:21:48 | prpplague | TomW: i don't get to stay in one area too long |
| 16:21:48 | TomW | prpplague: I'm not knocking school, I wonder sometimes what it would have been like to have had the advantage of getting a degree when I was very much younger.. :-( |
| 16:24:34 | TomW | ttySA0 |
| 16:25:27 | prpplague | TomW: ya but is that mapped to uart1 or uart2? i know the irkbd is on uart0 |
| 16:27:47 | file | prpplague? |
| 16:27:51 | prpplague | file: yes? |
| 16:28:05 | file | only the classics are of use to me |
| 16:28:36 | prpplague | ok |
| 16:28:42 | file | so if your able to sell them if/when it's cancelled, drop me a line |
| 16:29:26 | prpplague | file: our company just started an education foundation, so they'll probably get drop into that, which means they be free |
| 16:29:44 | file | neat |
| 16:29:59 | file | I think. |
| 16:30:45 | file | basically all I need is a Cybiko Classic Model 2, to see what boot technique it uses |
| 16:31:20 | file | Model 1's use some kind of compressed boot image |
| 16:33:11 | file | rambles on and on |
| 16:34:07 | file | so what's new? |
| 16:34:42 | prpplague | ssdd |
| 16:35:10 | file | nods |
| 16:35:42 | file | curses |
| 16:35:49 | file | I wish icewm would start on turtle |
| 17:13:40 | TomW | back coding... |
| 17:25:00 | david | hello ? |
| 17:28:43 | spd_racer911 | Hi TomW |
| 17:28:58 | TomW | hey speedy! |
| 17:29:19 | spd_racer911 | TomW: have you tested out the shanip2 mod yet ? |
| 17:29:24 | TomW | no |
| 17:29:33 | TomW | not sure what it is for. |
| 17:30:01 | spd_racer911 | TomW: damn! I am trying to get the ucb1x00 audio to work...the shanip2 to mod if for audio output to a headset or through the dsp |
| 17:31:06 | TomW | audio as in what? ARM generated audio, such as stored voice / tones? |
| 17:31:42 | spd_racer911 | TomW: I guess almost anything...I am just try to cat sound_sample.raw > /dev/audio |
| 17:32:31 | spd_racer911 | TomW: basically the Philips UCB1200 chip has a DAC and a speaker output |
| 17:33:19 | TomW | Yes. I would like to play sound files through the speaker (phone hung up). |
| 17:34:14 | spd_racer911 | yeah, it's only mono with the UCB1200 device, but I thought if I could get it working I would hang a stereo DAC onto the Tux |
| 17:34:56 | spd_racer911 | I think scanline has mp3s streaming to his Tux |
| 17:36:19 | prpplague | TomW: wahoo, got a login on my board! |
| 17:36:49 | prpplague | TomW: nothing hot or smoking! |
| 17:42:27 | TomW | prpplague: on the dev board? |
| 17:43:02 | TomW | spd_racer911: yes, scanline is into the MP3 thingy |
| 17:47:24 | prpplague | TomW: ya all i have is the max3232 and power right now, but atleast i got a login |
| 17:57:55 | TomW | prpplague: heh. now you have the Tux down to it's most essential element: the brain! |
| 17:59:01 | prpplague | TomW: tux on life support |
| 17:59:06 | TomW | heh |
| 18:06:08 | gilbertbsd | what is tuxcscreen? |
| 18:06:14 | gilbertbsd | an expensive phone? |
| 18:06:27 | spd_racer911 | http://www.tuxscreen.net |
| 18:06:51 | gilbertbsd | I have seen it. |
| 18:06:59 | gilbertbsd | it looks like an expensive phone... |
| 18:07:04 | gilbertbsd | what is it supposed to do? |
| 18:07:21 | TomW | gilbertbsd: it is for developers to get a cheap ARM system running. |
| 18:07:34 | gilbertbsd | what is ARM? |
| 18:07:43 | gilbertbsd | and why is it expensive? |
| 18:07:50 | TomW | gilbertbsd: do you write code? (software)? |
| 18:08:07 | gilbertbsd | I am learning forth. |
| 18:08:28 | TomW | gilbertbsd: then you probably don't need a Tux. |
| 18:08:49 | gilbertbsd | but why do you need a TUX when you can easily install linux on a cheap computer. |
| 18:09:00 | gilbertbsd | Heck someone donated a P90 to me that works just fine! |
| 18:09:28 | TomW | gilbertbsd: if you must ask, then you don't need to know. ;-) We install linux on the Tux (ARM) because it is a cool thing to do. |
| 18:10:00 | gilbertbsd | Indeed I do not know :D |
| 18:10:16 | gilbertbsd | is that not the reason anyone installs linux? |
| 18:10:19 | gilbertbsd | what is the ARM anyway? |
| 18:11:54 | prpplague | gilbertbsd: ARM is a type of processor |
| 18:12:13 | gilbertbsd | really? |
| 18:12:44 | gilbertbsd | I have heard of Motorala, AMD, Intel and Alpha (ibm does not count because of affordability) |
| 18:12:50 | gilbertbsd | and there is ARM as well? |
| 18:13:32 | prpplague | gilbertbsd: say motorola amd and intel is much like say ford, chyrsler and pontiac, those are manufactures not processors |
| 18:14:05 | gilbertbsd | AMD k6 motorla 680000000000000000000 Intel P4 blah blah |
| 18:14:19 | prpplague | gilbertbsd: http://www.arm.com |
| 18:14:46 | prpplague | gilbertbsd: intel makes an arm core, and so does motorola |
| 18:15:18 | prpplague | gilbertbsd: amd k6 and intel p4 are refered to as x86 architecture |
| 18:15:36 | gilbertbsd | yes I know that. |
| 18:15:55 | gilbertbsd | what is arm referred to as? |
| 18:17:19 | prpplague | ARM |
| 18:17:38 | file | need a hand? |
| 18:17:42 | prpplague | the arm core uses a intruction set specific to ARM |
| 18:17:47 | file | haha - bad joke |
| 18:17:50 | prpplague | file: doh |
| 18:18:28 | gilbertbsd | are ARMS what they use in the tiny hand held devices etc? |
| 18:18:42 | gilbertbsd | also why does the tuxscreen have a phone attached to it? |
| 18:19:26 | TomW | gilbertbsd: http://www.arm.linux.org.uk/ |
| 18:21:10 | gilbertbsd | why is inferno or plan9 not on the arm? |
| 18:21:12 | prpplague | gilbertbsd: many handheld device have arm cores, most ppl don't know it, but a lot of home routers, cable modems, and dsl modems have arm cores |
| 18:21:13 | TomW | gilbertbsd: http://www.arm.com/ |
| 18:21:33 | gilbertbsd | yeah I got that but they don't exactly make it very obvious what they do. |
| 18:21:51 | file | NetWinder! |
| 18:22:13 | file | it's ARM I believe |
| 18:22:27 | TomW | gilbertbsd: This list has to do with Electronics Engineering, Linux Kernel, OpenSouce software and Embedded Systems. |
| 18:22:48 | gilbertbsd | so are you guys forth programmers too? |
| 18:23:01 | TomW | forth? |
| 18:23:11 | TomW | heh |
| 18:23:12 | gilbertbsd | yes forth. |
| 18:23:23 | gilbertbsd | EE, OpenSource, Embedded. |
| 18:23:28 | gilbertbsd | right up your alley this forth. |
| 18:24:49 | file | anybody have a NetWinder? |
| 18:25:14 | gilbertbsd | file: do you have a netwinder? |
| 18:25:24 | file | no |
| 18:25:56 | gilbertbsd | what do you use ARM for file? |
| 18:26:12 | file | nothing that I know of |
| 18:26:17 | file | I use MIPS and Hitachi myself... |
| 18:26:33 | gilbertbsd | for the fun of it? or is profit lurking somewhere? |
| 18:26:42 | file | fun of it |
| 18:26:49 | file | my PDA's run MIPS and Hitachi |
| 18:26:55 | prpplague | slowly fades into the background |
| 18:27:05 | file | prpplague : Cybiko Classic :p |
| 18:27:31 | file | the NetWinder runs StrongARM |
| 18:27:39 | file | runs... I should say uses |
| 18:27:54 | gilbertbsd | so do any of these use the ARMandHammer? |
| 18:28:08 | file | shrugs |
| 18:28:30 | gilbertbsd | lame joke. |
| 18:29:22 | file | asks everyone if they have computer parts for sell |
| 18:29:35 | gilbertbsd | eghey |
| 18:30:04 | gilbertbsd | emm ebay |
| 18:30:17 | file | yeah but EBay can get expensive |
| 18:30:26 | gilbertbsd | _can_ |
| 18:30:30 | file | I just need a laptop floppy drive ( if possible ), and some networking stuff |
| 18:30:40 | file | perhaps a null modem cable? parallel -> parallel cable? |
| 18:30:48 | gilbertbsd | sure go for it. |
| 18:30:58 | file | nah |
| 18:31:02 | gilbertbsd | goodluck buying a cheap dongle. |
| 18:31:10 | file | dongle for what? |
| 18:31:26 | gilbertbsd | modem to rj11 |
| 18:31:32 | gilbertbsd | I tried hard. |
| 18:31:36 | gilbertbsd | nothing came of it :( |
| 18:31:50 | file | lol |
| 18:32:06 | file | do you know what a null modem cable is? |
| 18:32:15 | gilbertbsd | yes I do. |
| 18:32:19 | file | k |
| 18:32:23 | gilbertbsd | its for plip / slip connections. |
| 18:32:43 | file | or direct cable connection in Windows |
| 18:32:45 | gilbertbsd | I only don't know wtf tuxscreen is or is for. |
| 18:32:58 | file | have you gone to the URL in the topic? |
| 18:33:03 | gilbertbsd | yes I have. |
| 18:33:09 | gilbertbsd | but it told me nothing! |
| 18:33:15 | file | and have you read the Wiki? |
| 18:33:20 | file | about how it can be turned into many things? |
| 18:33:22 | gilbertbsd | nope |
| 18:33:26 | gilbertbsd | wait yes. |
| 18:33:35 | gilbertbsd | I am on the communications board part. |
| 18:34:08 | gilbertbsd | but wtf is there no brochure! I thought that was the point of most web pages? |
| 18:34:09 | file | uh huh... |
| 18:34:45 | file | brochure? |
| 18:34:54 | file | tuxscreen.net is a development site for the Tuxscreen |
| 18:35:16 | gilbertbsd | The TuxScreen Phone is a limited edition StrongARM Linux project platform available for purchase from TimRiker |
| 18:35:25 | file | yup |
| 18:35:36 | gilbertbsd | what does the TuxScreen Phone do in the end? |
| 18:35:42 | gilbertbsd | what is its goal! |
| 18:35:46 | gilbertbsd | thats all I wann know. |
| 18:36:03 | gilbertbsd | sure its a development enviroment ... but what does it develop? |
| 18:36:06 | file | why the heck does a Linux-based phone need a goal for? |
| 18:36:20 | file | it's a device that runs Linux, that can run X Windows |
| 18:36:33 | file | it can be used as a web browser, and more |
| 18:36:56 | gilbertbsd | bso the whole point of the Linux Phone IS the linux phone! |
| 18:37:12 | file | pokes other people |
| 18:37:14 | gilbertbsd | thats what I was trying to find out. |
| 18:37:27 | gilbertbsd | so the Linux Phone has ARM in its heart no? |
| 18:37:38 | file | yes it does |
| 18:37:54 | gilbertbsd | who makes the original linux Phone? |
| 18:38:28 | file | it never originally mruns Linux |
| 18:38:36 | gilbertbsd | yes I know that. |
| 18:38:43 | gilbertbsd | what is the original product called? |
| 18:38:58 | file | Shannon... I forget who makes it |
| 18:39:17 | gilbertbsd | Hmmm. |
| 18:39:24 | file | Sharp |
| 18:39:43 | gilbertbsd | Shannon by sharp? |
| 18:39:44 | file | maybe... |
| 18:39:53 | gilbertbsd | I thought they made the LCD? |
| 18:39:56 | file | one moment and I'll tell you |
| 18:40:36 | file | WHO RE-ORGANIZED THE WIKI! |
| 18:40:49 | andersen_ | TomW: yup. Color ls. :) |
| 18:41:07 | file | it's a Philips IS2630 |
| 18:41:27 | TomW | andersen_: tnx, I have been wanting that (and more hours in a day...). |
| 18:41:45 | pmisfit | is back (gone 10:20:14) |
| 18:41:48 | file | gilbertbsd : ok? |
| 18:42:13 | gilbertbsd | file: where did you find that info at? |
| 18:42:33 | pmisfit | andersen_: latest buildroot-tux cvs, with the busybox tarball that came across, fails? |
| 18:42:42 | file | http://www.tuxscreen.net/wiki/view/Main/TuxScreen |
| 18:42:49 | pmisfit | final link of busybox: nslookup.o: In function `nslookup_main': |
| 18:42:50 | pmisfit | nslookup.o(.text+0x284): undefined reference to `__res_state' |
| 18:42:53 | file | gilbertbsd : the Hardware link gives you specific hardware details |
| 18:43:00 | gilbertbsd | when i went there it said Main Document not found. |
| 18:43:10 | gilbertbsd | *shrugs* I love rtfm IF there is one to read :D |
| 18:43:17 | file | gilbertbsd : it works for me |
| 18:43:33 | file | The TuxScreen Phone is a limited edition StrongARM Linux project platform. |
| 18:43:34 | file | It was previously available as the Philips IS2630 Screen Phone for $650. TimRiker has some in inventory at $99 if you would like to get a hold of one or more. |
| 18:43:49 | pmisfit | checks if __res_state is a libgcc thang.. |
| 18:44:20 | gilbertbsd | is he the only one selling WebPhones? |
| 18:44:34 | file | he is the only one selling the Tuxscreen |
| 18:44:50 | gilbertbsd | no I am a BSDer :D |
| 18:44:59 | gilbertbsd | we need BSDscreen already. |
| 18:45:06 | gilbertbsd | DameonScren perhaps? |
| 18:45:08 | file | too bad :p |
| 18:45:13 | file | NOW GO AWAY! |
| 18:45:16 | file | :p |
| 18:45:22 | gilbertbsd | no I just want a webphone. |
| 18:45:28 | gilbertbsd | I really don't need to have anything on it. |
| 18:45:40 | file | get a Tuxscreen and be quiet :p |
| 18:45:57 | gilbertbsd | but you don't have a TuxScreen either! |
| 18:46:05 | file | because I'm poor and only 15 |
| 18:46:11 | file | so I don't exactly have a source of income |
| 18:46:17 | gilbertbsd | I rest my case. |
| 18:46:19 | file | plus I'm Canadian |
| 18:46:24 | file | and a Tuxscreen would cost $250 my money |
| 18:46:29 | gilbertbsd | I already rested my case. |
| 18:46:42 | file | what case??? |
| 18:46:51 | file | threatens gilbertbsd with Mork and Mindy |
| 18:47:21 | gilbertbsd | The case that 'since neither you nor I has a WebPhone, we can both not be quiet' |
| 18:47:21 | pmisfit | dives for cover and lets file and gilbertbsd duke it out |
| 18:47:29 | pmisfit | is away: back to xcircuit and pcb |
| 18:47:50 | gilbertbsd | 2 Type II PCMCIA slots (exposed) |
| 18:47:50 | gilbertbsd | v.34 modem |
| 18:47:50 | gilbertbsd | Comes running with Inferno |
| 18:47:54 | gilbertbsd | !WOW! |
| 18:47:59 | gilbertbsd | now that I gotta see :D |
| 18:48:48 | gilbertbsd | see? I _run_ inferno! |
| 18:49:36 | file | thwaps gilbertbsd |
| 18:49:49 | file | Any BSD user that pokes fun at the Tuxscreen should be kicked out of here! |
| 18:49:52 | gilbertbsd | I bet you don't run inferno. |
| 18:50:03 | gilbertbsd | hey I did not make fun of the TuxScreen. |
| 18:50:18 | file | That's just a general rule |
| 18:50:27 | spd_racer911 | help /kick |
| 18:50:45 | gilbertbsd | its hte other way around spd_racer911 |
| 18:50:53 | file | gilbertbsd : either buy a Tuxscreen, ask development questions, or get out |
| 18:50:57 | file | :p |
| 18:50:58 | andersen_ | pmisfit: you need a newer uClibc.... |
| 18:51:14 | gilbertbsd | I want a WebPhone _without_ tuxscreen. |
| 18:51:27 | gilbertbsd | whats the fun in buying one already modified? |
| 18:51:31 | file | gilbertbsd : well then go elsewhere!!!!! |
| 18:51:39 | file | gilbertbsd : it isn't modified |
| 18:51:48 | file | gilbertbsd : it comes running Inferno with no mods |
| 18:52:04 | file | Inferno is the thing they decided to put on it at Philips |
| 18:52:44 | gilbertbsd | yes for a reason. |
| 18:52:55 | gilbertbsd | I have inferno on my machine too :D |
| 18:53:09 | file | so why are you going on about modifications? |
| 18:53:30 | gilbertbsd | I thought the whole point of TuxScreen was to buy one of these already modified? |
| 18:53:34 | file | pmisfit: this is fun |
| 18:53:51 | file | gilbertbsd : it doesn't come modified |
| 18:53:51 | spd_racer911 | file: pmisfit: this is ludicrous |
| 18:54:13 | file | gilbertbsd : it's the same thing you would have gotten if you bought from Philips |
| 18:54:25 | gilbertbsd | I see. |
| 18:54:51 | spd_racer911 | operator: /oper |
| 18:54:57 | gilbertbsd | how large is the harddrive? |
| 18:55:15 | file | it has no hardrive |
| 18:55:17 | file | it has flash |
| 18:55:31 | file | 4MB flash |
| 18:55:50 | gilbertbsd | thats all? |
| 18:55:59 | gilbertbsd | and Linux fits in that crawl space? |
| 18:56:06 | file | yes, why not? |
| 18:56:13 | file | Linux fits in that easily |
| 18:56:29 | gilbertbsd | how does one talk to the memory? |
| 18:56:35 | gilbertbsd | does it have floppies? |
| 18:56:47 | file | no... |
| 18:56:51 | file | why don't you read the Wiki? |
| 18:56:53 | file | everything is there |
| 18:57:56 | gilbertbsd | if WIKI work for me I read WiKi WiKi but WiKi WiKi go Wako Wako on me. |
| 18:58:10 | gilbertbsd | Topic "MainWebHome" was not found in web "Main". |
| 18:58:11 | gilbertbsd | You can create the document. |
| 18:59:02 | file | your clicking on links that don't work :p |
| 18:59:08 | file | go to http://www.tuxscreen.net/wiki/view/Main/Hardware |
| 18:59:20 | gilbertbsd | Is it my fault that they don't work? |
| 18:59:30 | gilbertbsd | thats where I am now . |
| 18:59:45 | file | what web browser? |
| 18:59:51 | gilbertbsd | f_ck a PDF. |
| 18:59:58 | gilbertbsd | mozilla |
| 19:00:02 | file | it's not a PDF |
| 19:00:13 | file | gilbertbsd : your quite annoying |
| 19:00:21 | gilbertbsd | what is this then? http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/22358B.pdf |
| 19:00:28 | file | brb |
| 19:00:31 | gilbertbsd | I have a slow modem so I have to wait for ever! |
| 19:02:17 | file | back |
| 19:02:24 | file | you followed a link to specifics on a chip probably |
| 19:02:30 | gilbertbsd | yes I did. |
| 19:02:32 | file | I'm connected at 34.6 kbps :p |
| 19:02:44 | gilbertbsd | I am connected at 1.2 kbps |
| 19:02:49 | gilbertbsd | see the difference? |
| 19:02:54 | file | we aren't responsible for the companies that release chip information in PDF files, so stop complaining! |
| 19:02:59 | file | 1.2 kbps my @$$ |
| 19:03:12 | gilbertbsd | I have a modem |
| 19:03:30 | gilbertbsd | I can see the lights blinking ... now |
| 19:03:33 | gilbertbsd | a |
| 19:04:03 | file | is on a modem |
| 19:04:38 | file | now stop complaining and go away :p |
| 19:04:43 | gilbertbsd | why? |
| 19:04:55 | gilbertbsd | can one not have a healthy interest in new things? |
| 19:05:24 | prpplague | gilbertbsd: interest is one thing, etiquette is another |
| 19:05:42 | gilbertbsd | Did I say something wrong? |
| 19:06:25 | file | prpplague : so what's up? |
| 19:08:13 | andersee | pmisfit: struct _res and co are in the latest uClibc |
| 19:08:53 | gilbertbsd | andersee: have you ever used curfloo ? |
| 19:11:17 | andersee | gilbertbsd: no |
| 19:11:29 | gilbertbsd | ack. |
| 19:12:42 | TomW | is away: mowing the lawn. |
| 19:12:55 | file | bye Tom |
| 19:16:26 | fvgestel | Haha, I just read last conversation LOL... |
| 19:16:32 | as_ | Evnin all. |
| 19:19:45 | as_ | Can someone tell me why sboot isnt at 8000? |
| 19:20:31 | file | bbl |
| 19:23:30 | jgorkos | help, I just bricked my TuxScreen... :( |
| 19:23:37 | as_ | Not fun. |
| 19:23:43 | as_ | What happened? |
| 19:23:56 | jgorkos | tried to install a self-compiled BLOB. |
| 19:24:15 | as_ | A JTAG interface should be able to bring it back to life. |
| 19:24:26 | jgorkos | Does Tim have any for sale? |
| 19:24:48 | as_ | Unsure. If your into electronics you can make one for very little. |
| 19:25:03 | jgorkos | do I need a flash burner? |
| 19:25:14 | as_ | No, just JTAG. |
| 19:25:53 | prpplague | jgorkos: i think BZFlag still has some jtag dongles |
| 19:26:12 | as_ | Hiya prp. |
| 19:26:15 | jgorkos | I'll need to look at the wiki for that. How should I check for "Bricked-ness" before I look into re-flashing it? |
| 19:26:48 | as_ | What happenes when you turn it on? |
| 19:27:05 | jgorkos | backlight comes on. nothing else. Nothing on the serial port (at 115200) |
| 19:28:15 | as_ | All JTAG does is alow you to rewrite the flash without having to boot the machine. |
| 19:29:39 | as_ | Its a 'hardware' rewrite. |
| 19:30:55 | jgorkos | sounds like what i need. I tried to flash a corrupted BLOB and it went away. |
| 19:31:46 | jgorkos | about 90% of the kernels I try to load report errors erasing flash memory. Is that a common problem? |
| 19:32:13 | as_ | Not a clue I'm new at this. |
| 19:33:36 | jgorkos | ok, thanks anyway. I'm going to write BZflag an email about getting a JTAG. Not interested in building my own: more money than time. |
| 19:33:52 | pmisfit | is back (gone 00:46:22) |
| 19:34:22 | as_ | Fair enough. |
| 19:35:26 | pmisfit | andersee: hmmm. buildroot fetches uClibc.0.9.11, which is missing __res, which blows up busybox? |
| 19:37:33 | pmisfit | just wanted to see what happens by typing "make" with all the defaults |
| 19:38:56 | pmisfit | not a problem for me, but someone doing a clean buildroot checkout probably will get stuck. |
| 19:39:10 | pmisfit | fyi |
| 19:45:58 | pmisfit | andersee: i was able to hack around it by removing NSLOOKUP from busybox config. |
| 19:46:12 | pmisfit | should i commit that so others don't get tripped up? |
| 19:47:35 | pmisfit | figures forgiveness is easier to seek than permission. |
| 19:52:18 | pmisfit | can't bring himself to commit such a heinous act. |
| 19:52:21 | pmisfit | nevermind. |
| 19:56:08 | as_ | Bit early for them to be awake maybe. |
| 20:13:41 | mallum | fontenot: are you about ? |
| 20:15:34 | fontenot | kinda |
| 20:16:26 | mallum | fontenot: are you an emacs user ? |
| 20:16:55 | fontenot | mallum: i use emacs but am not an expert |
| 20:17:50 | mallum | fontenot: this -> http://fabrice.bellard.free.fr/qemacs/ would be cool on the tuxscreen |
| 20:19:53 | TomW | mallum: never mind emacs, I vote for vim |
| 20:22:08 | mallum | TomW: Vile ? |
| 20:22:15 | fontenot | ugh |
| 20:22:20 | fontenot | emacs |
| 20:22:26 | mallum | TomW: so un gentlemanly |
| 20:22:55 | fontenot | yes mallum that does look cool |
| 20:23:06 | fontenot | did you guys see my latest problem? |
| 20:23:24 | fontenot | the OOM killer is killing stuff when free RAM get's to 4.5MB |
| 20:23:25 | TomW | mallum: heh, never mind elegant (emacs), need something for doing work with. |
| 20:23:57 | fontenot | so it's like having 12MB instead of 16 |
| 20:24:05 | mallum | TomW: jove and qemacs are both tiny |
| 20:24:20 | mallum | fontenot: hmmm, maybe that was what was causing opie problems |
| 20:24:42 | fontenot | mallum: dmesg should tell you if things were killed by OOM Killer |
| 20:25:23 | fontenot | dmesg|tail for me looks something like thisL |
| 20:25:26 | fontenot | Out of Memory: Killed process 270 (sh). |
| 20:25:26 | fontenot | Out of Memory: Killed process 397 (links). |
| 20:25:26 | fontenot | Out of Memory: Killed process 397 (links). |
| 20:25:26 | fontenot | Out of Memory: Killed process 403 (links). |
| 20:25:26 | fontenot | Out of Memory: Killed process 403 (links). |
| 20:25:27 | fontenot | Out of Memory: Killed process 414 (links). |
| 20:25:30 | fontenot | Out of Memory: Killed process 376 (minitime). |
| 20:25:31 | fontenot | Out of Memory: Killed process 377 (minitime). |
| 20:26:00 | fontenot | and I see this in /proc/meminfo: LowFree: 4452 kB |
| 20:26:10 | fontenot | and i cant get that to go any lower |
| 20:26:21 | BZFlag | mallum: vi in busybox is 21k are hove or qemacs that small? |
| 20:26:36 | BZFlag | er jove |
| 20:27:35 | fontenot | qemacs does a lot |
| 20:29:16 | fontenot | mallum: TomW: have you played with links? it's pretty cool |
| 20:29:21 | BZFlag | so perhaps "jove and qemacs are both tiny" should be restated? ;-) |
| 20:29:30 | mallum | fontenot: yeah a while ago |
| 20:29:36 | fontenot | Previous tiny version (23KB): qemacs-0.2.tar.gz |
| 20:29:42 | prpplague | TomW: hey you alive? |
| 20:29:46 | mallum | BZFlag: bb vi dont do AA text in X |
| 20:30:08 | fontenot | nor kanji |
| 20:30:20 | mallum | nor docbook mode |
| 20:30:24 | fontenot | nor web browsing with frames |
| 20:30:32 | BZFlag | mallum: how large is the qemacs? and it is X with i18 support? |
| 20:30:47 | mallum | nor having the press the damn esacape key all the bloody time |
| 20:30:51 | mallum | ;) |
| 20:31:08 | BZFlag | bb vi does not run X apps either. that was not the point. the point it that it is small. |
| 20:31:50 | mallum | one day I'll hack a bb emacs that will be half the size of bb vi |
| 20:31:51 | BZFlag | feel free to compile and install emacs20 on your CF if you like. |
| 20:32:04 | BZFlag | mallum: and the patch will be welcome! |
| 20:32:17 | mallum | BZFlag: one day, I've though about it a lot |
| 20:32:56 | fontenot | god what a stupid ass makefile it has |
| 20:32:58 | BZFlag | bb has wget, tftp, etc. it would be nice to add those features to the editor. ;-) |
| 20:33:38 | mallum | yeah and a few games |
| 20:33:41 | mallum | its own shell |
| 20:33:57 | mallum | kitchen sink .. etc etc ... |
| 20:34:46 | mallum | fontenot: maybe post to arm list about that mem limit ? |
| 20:36:16 | TomW | mallum: should share/matchbox/defaults be an empty file? |
| 20:36:37 | fontenot | mallum: I guess I will have to |
| 20:36:48 | mallum | TomW: yes |
| 20:37:13 | mallum | TomW: it wasn't planned to be |
| 20:37:25 | mallum | TomW: its read as a base for XResource defaults |
| 20:37:47 | mallum | TomW: it should really be an app-defaults file |
| 20:37:51 | TomW | mallum: that is probably why some of the matchbox apps were not running then. |
| 20:38:04 | mallum | TomW: shouldn't affect that |
| 20:38:20 | BZFlag | matchbox should be in /usr/X11/bin not /usr/local/bin |
| 20:38:23 | mallum | TomW: some were not running due to xpm problems ... fixed now |
| 20:38:34 | BZFlag | share files in /usr/share/ not /usr/local/* |
| 20:38:37 | BZFlag | etc. |
| 20:38:38 | mallum | TomW: ./configure --prefix=/usr/x11 |
| 20:39:28 | mallum | BZFlag: share files are in PREFIX/share/ ? |
| 20:40:04 | mallum | BZFlag: its all autoconfed which makes it v.difficult to put files in non standard places |
| 20:40:10 | file | twitches |
| 20:40:29 | mallum | fontenot: http://savannah.gnu.org/projects/xbubble/ could be nice on the tuxscreen too |
| 20:40:40 | fontenot | TomW: minisys and minitime are running fine now |
| 20:40:56 | TomW | mallum: share/matchbox/themes/default_theme_rc looks ok, it has data in it. |
| 20:40:59 | mallum | fontenot: I remember it was pretty small |
| 20:41:09 | mallum | TomW: yes its only that defualts file |
| 20:41:30 | mallum | TomW: I think that was left over from when I was hacking the X resource stuff |
| 20:41:36 | mallum | TomW: its pretty harmless |
| 20:41:58 | mallum | TomW: Im about to rehack all the conf / themes stuff ... |
| 20:42:45 | mallum | fontenot: actually xbubble is bigger than I remembered ... |
| 20:42:55 | mallum | fontenot: its a nice little game though |
| 20:42:56 | fontenot | TomW: see http://intimate.handhelds.org/jacques/matchbox10.png through matchbox14.png |
| 20:43:45 | TomW | mallum: Ok. so having an empty share/matchbox/defaults is all right. Good, lots of cleaning up here on the makefiles & such, trying to get some more freespace in the flash (like who needs manpages on a Tux? :) |
| 20:44:36 | fontenot | TomW: get rid of that 900kB libXt.so |
| 20:44:42 | TomW | BZFlag: your addition of --host forced matchbox ./configigure to ignore the value for --prefix. Remove --host and retry. |
| 20:45:16 | mallum | TomW: look at Makefile.am in packaging/ipkg |
| 20:45:24 | mallum | TomW: that makes a smaller matchbox |
| 20:45:27 | TomW | fontenot: it is now a selection. Edit tinyx.mk and take Xt out of the TINY_LIBS list. |
| 20:45:51 | mallum | fontenot: Xt is like so 1985 ... |
| 20:45:55 | TomW | mallum: k,, I have to run, look later. |
| 20:46:18 | TomW | first to commit what I have so jacques can test this... |
| 20:46:58 | TomW | fontenot: the makefiles are probably going to be broken some how, you can holler at me in a few hours when I get back, k? |
| 20:47:10 | fontenot | TomW: ok I will |
| 20:48:02 | BZFlag | TomW: it fails to build here without --host. |
| 20:49:13 | mallum | are you all building from cvs ? |
| 20:49:45 | TomW | BZFlag: not on my system, you have to tell it what --host=<something>, otherwise --host gobbles the next setting in the configure line. |
| 20:49:56 | TomW | mallum: yes |
| 20:50:01 | BZFlag | TomW: without --host it tries to execute the test binaries the compiler builds and cannot so it fails. |
| 20:50:03 | TomW | fontenot: commited |
| 20:50:58 | TomW | fontenot: you having problems building matchbox from my matchbox.mk file? |
| 20:51:33 | TomW | BZFlag: you have to tell it --host=<something>, otherwise --host gobbles the next setting in the configure line. |
| 20:51:44 | TomW | BZFlag: you left the --host line blank. |
| 20:51:58 | TomW | got to run, bbl |
| 20:53:14 | BZFlag | TomW: ok, --host=foo should work for me here. |
| 20:53:38 | TomW | BZFlag: k |
| 20:53:42 | mallum | TomW: build from a dist tarball, it will be less painfull. There is one ( 0.2.0 ) in my downloadables directory - it is the same as whats in cvs now |
| 20:53:49 | | well, matchbox is a X11 window manager for handheld devices. Info, screenshots & downloadable at http://handhelds.org/~mallum/matchbox . Ipkg is in the familiar unstable feed. Development versions and other goodies at http://handhelds.org/~mallum/downloadables/ |
| 20:53:49 | mallum | ibot: matchbox |
| 20:54:10 | TomW | mallum: no pain right now, everything works fine. |
| 20:54:38 | mallum | TomW: you still have to link the configure.in file though ? |
| 20:54:39 | TomW | mallum: ... except BZFlag keeps bitchin' at me. ;-) |
| 20:54:48 | mallum | TomW: ;) |
| 20:55:39 | mallum | TomW: I dont know why it is, but auto(conf|make) cause a load of problems if you run autoconfigure.sh and your versions of auto(conf|make) are even slightly different to mine ... |
| 20:55:52 | TomW | mallum: I wish I had more time to discuss this, really. I am newbie at matchbox and trying to package it in to buildroot-tux. I need to talk with you about this later. I have a meeting to goto. cul |
| 20:56:00 | TomW | is back (gone 01:43:19) |
| 20:56:03 | mallum | TomW: np |
| 20:56:06 | TomW | is away: I'm busy |
| 21:15:06 | BZFlag | TomW: I committed the --host=arm-linux patch. seems to build fine here. |
| 21:46:40 | Morn | hi everyone |
| 21:46:42 | Morn | :-) |
| 22:02:39 | pmisfit | Morn: howdy. |
| 22:20:13 | Russ | fontenot: yes? |
| 22:30:34 | Russ | BZFlag: should probably change the load |
| 22:30:35 | Russ | er |
| 22:30:44 | Russ | load_kernel stuff to a more appropriate name |
| 22:46:40 | pmisfit | is away: received interrupt... running ISR... |
| 22:50:05 | BZFlag | open to suggestions. ;-) |
| 22:50:24 | Russ | something about file or file system, or fs |
| 22:50:40 | Russ | or maybe if you want to share the code with other things, like network |
| 22:52:27 | Russ | just before its gets too set |
| 23:06:29 | TomW | is back (gone 02:10:24) |
| 23:07:07 | TomW | fontenot: how badly did I break things? |
| 23:11:03 | TomW | mallum: hi |
| 23:11:26 | mallum | TomW: hi |
| 23:11:44 | mallum | anyone played with this -> http://www.emdebian.org/emdebsys.html btw ? |
| 23:12:27 | TomW | I sidestepped the issue of the automake giving errors: |
| 23:12:30 | TomW | ln -sf configure.ac configure.in ; \ |
| 23:12:30 | TomW | echo -e "#!/bin/sh\n./autoconfigure.sh\nexit 0\n" > doauto.sh ; \ |
| 23:12:30 | TomW | sh doauto.sh ; \ |
| 23:12:30 | TomW | rm -f doauto.sh \ |
| 23:12:34 | TomW | ; |
| 23:12:45 | TomW | works for me! :-D |
| 23:13:43 | TomW | mallum: the "exit 0" keeps those nasty old errors from clobbering the toplevel make. |
| 23:15:43 | TomW | I've seen some of these things (menu systems for linux builds), IMO, they create more problems for a serious developer (me) than what they offer to resolve. |
| 23:16:07 | prpplague | mallum: that sounds amazingly like buildroot |
| 23:16:23 | mallum | TomW: build from the tarball at http://www.handhelds.org/~mallum/downloadables/matchbox-0.2.0.tar.gz not CVS, please ! |
| 23:16:31 | TomW | Those things might be a good idea, once linux development stabilizes. Heh, whenever that will happen.. |
| 23:16:31 | prpplague | TomW: hey got a sec for cap question? |
| 23:16:47 | mallum | TomW: if you build from cvs every will need the auto* tools installed on there systems |
| 23:16:55 | mallum | TomW: that tarball is the same as CVS |
| 23:17:13 | mallum | TomW: but no need to run ./autoconfigure.sh |
| 23:17:38 | TomW | mallum: ok, I will rewrite the makefile then. I still will need to run configure, right? but not automake, correct? |
| 23:17:44 | mallum | TomW: yes |
| 23:17:56 | TomW | meekly complies. |
| 23:18:11 | prpplague | TomW: some of my schematics show .1uf caps that are polarized, is it ok to use ceramics there, or should those be tants? |
| 23:18:39 | TomW | prpplague: what schematcis? Tux, which ones? |
| 23:18:44 | mallum | TomW: I think need to put a link in that downloads directory that points to the latest matchbox |
| 23:19:07 | mallum | TomW: but Im about to butcher things so CVS will be unstable for a while any way |
| 23:19:19 | prpplague | TomW: tux and sample schematics for the max3232 chip |
| 23:19:37 | TomW | mallum: if you want to do a "matchbox-release-latest", or something like that, we can implement that now. Then no loopback later. |
| 23:19:59 | TomW | prpplague: max232 / max3232? |
| 23:20:23 | prpplague | TomW: max3232 ( 3.3v version of max232 ) |
| 23:20:38 | TomW | prpplague: damned hard to find polarize 0.1uf caps. |
| 23:20:48 | prpplague | TomW: thats what i thought |
| 23:21:10 | prpplague | TomW: just seemed odd that the schematics were labeled that way |
| 23:22:47 | prpplague | TomW: http://dbserv.maxim-ic.com/quick_view2.cfm?qv_pk=1068 |
| 23:22:57 | TomW | prpplague: well, the max232 requires 1 uf caps. since most values above 0.47 uf are polarized (common), you will find a polarized called for. BUT, the max232cpa device only needs some 0.1uf caps in place of the 1uf caps (improved vesrion of the Max232). |
| 23:23:35 | prpplague | TomW: if you look at the sample it shows the caps with + sides |
| 23:23:36 | TomW | prpplague: probably someone saw a cap on the board, and assumed it had to be a polarized? |
| 23:24:08 | prpplague | TomW: ok thanks, just had me worried |
| 23:25:46 | TomW | prpplague: look on page 8 of the data sheet, under "capacitor selection" |
| 23:26:01 | mallum | TomW: sure, one sec I'll give it ago. Im not sure if apache on hh.org is set to follow softlinks ... I guess we'll find out |
| 23:26:23 | TomW | prpplague: the polarity is shown as a reference in the event you use a polarized cap. |
| 23:27:18 | _as | Yeah, if you use a non polerised it makes no difference, but if you get a polar one the wrong way round... |
| 23:27:51 | mallum | TomW: appears to work think I need to name it .tar.gz though |
| 23:28:13 | _as | You can probably substitute any value over 0.1uf. |
| 23:28:17 | prpplague | TomW: doh, i missed that section all together |
| 23:28:35 | TomW | prpplague: one trick a lot of engineers use: if you need a low current bias (few milliamps) of -5 volts, add more capacitance to a MAX232 chip and steal your bias voltage off it. |
| 23:28:44 | mallum | TomW: http://www.handhelds.org/~mallum/downloadables/matchbox-release-latest.tar.gz |
| 23:28:53 | TomW | mallum: k |
| 23:29:12 | TomW | coding |
| 23:29:17 | prpplague | TomW: thanks |
| 23:29:39 | TomW | not coding! |
| 23:30:15 | TomW | mallum: will take a while, I have full build running out. It will be about 10 minutes before I will have it tested. |
| 23:30:37 | mallum | TomW: np |
| 23:31:02 | TomW | mallum: I like monolaunch, has possibilies. |
| 23:31:15 | mallum | TomW: thanks |
| 23:31:33 | mallum | TomW: look at packaging/ipkg/Makefile.am |
| 23:31:39 | TomW | k |
| 23:31:50 | mallum | TomW: you probably want your make install line like; |
| 23:32:12 | mallum | DESTDIR=/build fakeroot make install-strip |
| 23:32:32 | mallum | and then a |
| 23:32:43 | mallum | rm -fr ./build/usr/man |
| 23:33:00 | mallum | to get things tiny |
| 23:33:17 | TomW | mallum: pretty small right now. tinyx is the hog. |
| 23:33:50 | prpplague | TomW: how do you handle voltage and ground distribution on your dev board? do create a long strip down the board for ground and voltage, or do you just run traces to where they should go?// |
| 23:34:35 | mallum | TomW: the above is assuming you've set your --prefix=/usr |
| 23:36:25 | mallum | TomW: freshmeat page for matchbox is nice -> http://freshmeat.net/projects/matchbox/?topic_id=56 |
| 23:36:29 | TomW | prpplague: I like to use 16ga bus wire for my power grig. the operative word here is 'grid'. think of power distribution as being constructed like the mesh in a window screen. It would be like putting resistors in parallel, lower resistance to the flow of current to the devices. |
| 23:37:21 | TomW | prpplague: one side of the board could be instantly demanding high current, then the current would tend to flow from another direction withing the grid for chips that are on the other side of the board demanding current. |
| 23:38:09 | prpplague | TomW: gotcha |
| 23:38:22 | TomW | prpplague: you can only "shove" so much current down a wire. Then the resistance of the wire starts causing a voltage drop. |
| 23:38:51 | TomW | L + R + C, the lumped impedance of the traces. |
| 23:38:55 | prpplague | TomW: so when possible create a mesh of power connections with the devices |
| 23:40:36 | _as | I think pulse current is normally a bigger problem. |
| 23:40:40 | TomW | prpplague: yeah. You'll find that PCB designers are like witch doctors. Each has their "ceramonies" they do, and incantations they invoke while laying out the PCB. Then, when it don't work right, we get out the high-speed 'scope and take a look around. ;-) |
| 23:41:19 | TomW | _as: true, the frequency can be very high: several hundred MHz. |
| 23:41:23 | _as | CMOS circitry is always dotted with caps next to chips. |
| 23:41:38 | _as | I want to try designing with ECL. |
| 23:41:59 | TomW | _as: good luck |
| 23:42:19 | _as | ECL seems a much cleverer idea on the whole. |
| 23:42:39 | TomW | _as: that stuff is real "magic". Talk about a board that runs almost as an entire electronic wave, that is ECL. |
| 23:43:18 | _as | Transmission line electronics. |
| 23:43:30 | mallum | BZFlag: http://fabrice.bellard.free.fr/tcc/ |
| 23:43:33 | TomW | _as: dunno about modern ECL, but at one time the current consumption of ECL was horrendous! But it would run a 1 GHz. |
| 23:43:34 | mallum | BZFlag: you asked the other day for a small c compiler IIRC ? |
| 23:43:48 | mallum | BZFlag: is x86 only though |
| 23:44:13 | _as | ECL is very wasteful at low frequencies, since current consumption is independent of toggle rate. |
| 23:44:28 | TomW | mallum: I have taken the small-c compiler and rebuilt it to generate 6502 code. that was many years ago. worked good. |
| 23:44:58 | TomW | mallum: alas, the source has been lost. :-( |
| 23:45:22 | _as | CMOS current/power consumption is proportional to toggle rate. But ECL gets you a much faster system for a given amount of power at the high end. Crossover is about 40MHz for 5v ECL versus 3.3v CMOS. |
| 23:46:10 | _as | ECL has a lot of clever tweaks though, and you get no power glitches in it. |
| 23:47:09 | TomW | _as: yeah, but unless you are going to interface to RF, ECL makes no real sense in designing it into todays processor support circuits. |
| 23:47:42 | _as | ITs a real shame they dont make CPUs in it, but I guess I just like speed. ;] |
| 23:47:47 | TomW | _as: the MOS stuff we have now is more than adequate to meet the needs of the processors we employ. |
| 23:47:47 | prpplague | BZFlag: http://www.linuxdevices.com/news/NS2059195371.html i'm not sure i need to take his advice, lol |
| 23:48:41 | TomW | _as: it is the power requirements.. Look at what ECL consumes vs. the complexity of the gate arrays. ECL is fairly simple gates. no? |
| 23:49:15 | _as | CMOS is outdated, but bipolar transistors dont fabricate well with planer processing. |
| 23:49:28 | TomW | _as: if you made a processor of several thousand gates (tens of thousands?), the power requirements would be enourmous! |
| 23:49:40 | _as | Above 40MHz ECL consumes less power gate for gate than CMOS. |
| 23:49:55 | TomW | _as: yes, relatively. |
| 23:50:27 | _as | So an ECL CPU would consume less power in ECL than CMOS, and now were into 2 GHz territory... |
| 23:51:17 | TomW | _as: but, you forget. when ECL is "idle" (holding state), it is consuming a relatively large current. While CMOS has a high current when being switched, the holding current is negligable. |
| 23:52:10 | _as | In a 2GHz CPU Id bet most of the CPU core is toggling over 40MHz. |
| 23:52:42 | _as | And constant current demands vastly reduces noise in the circuits, no power spikes. |
| 23:53:44 | TomW | _as: so, if your logic is idling, holding state (no drive access, no interrupts), those ECL sections are consuming power. An ECL processor would make sense if you were doing continual FFT transformations: but then, we have the AMD2900 series, right? (dating myself, AMD2900 is ancient. |
| 23:55:51 | _as | I dont know what the 2900 series is, but unpredictability in CMOS cPUs is a problem rather than an advantage. A CPU you think works fine can make mistakes during certain bits of code. 3D games were the old benchmark. |
| 23:56:09 | TomW | prpplague: I don't know hos they can make money on Embedded Linux. Too many people like us are willing to share that knowledge for free! |
| 23:56:36 | TomW | s/hos/how/ |
| 23:57:38 | prpplague | TomW: i know, my company is one of the few that has made embedded linux make sense in a business model |
| 23:58:16 | TomW | prpplague: what is interesting. The software is relatively easy to do, the hardware is what trips most people up. IMO, if you can bang-bits into registers, they you have a skill most professionals don't have. |
| 23:58:20 | prpplague | TomW: we sell a proprietary accounting system, out whole embedded linux development is to make products that help sell our accounting software |
| 23:58:34 | prpplague | TomW: exactly |
| 23:58:39 | TomW | prpplague: loss-leader? |
| 23:59:07 | TomW | mallum: finally into compile of tinyx. |
| 23:59:09 | prpplague | TomW: whats that? |
| 23:59:44 | TomW | prpplague: something you provide at / below cost to enhance the value of another product (more expesive: the cash cow). |
