| 00:20.14 | *** join/#gpe Hallski (~micke@h134n1c1o1035.bredband.skanova.com) |
| 00:43.56 | *** part/#gpe RobiWan (~RobiWan@p50824C19.dip0.t-ipconnect.de) |
| 00:46.34 | Debolaz | Grr |
| 00:47.45 | *** join/#gpe Tyg_Wrk (~me@Toronto-HSE-ppp3859672.sympatico.ca) |
| 00:47.54 | Tyg_Wrk | 'lo all |
| 00:49.39 | Tygger_Bob | Florian, you around? |
| 00:50.37 | Debolaz | Err, why can't I change someones password in GPE when logged on as root? |
| 00:50.43 | Debolaz | (Without knowing the original password) |
| 00:53.41 | Tygger_Bob | Debolaz: worked for me.. |
| 00:53.51 | Tygger_Bob | from an terminal window on the ipaq |
| 00:54.03 | Debolaz | It asks me for the original password when trying to change a users password from the GUI logged in as root. |
| 00:54.50 | Debolaz | And as I said, it works from terminal, but it shouldn't ask for this from the GUI. |
| 00:54.52 | florian | Tygger_Bob: yep |
| 01:01.25 | *** join/#gpe TyggrBob (~me@Toronto-HSE-ppp3859672.sympatico.ca) |
| 01:01.44 | TyggrBob | Arrggg |
| 01:02.09 | TyggrBob | So, Debolaz.. where were we? |
| 01:02.14 | Debolaz | Norway. |
| 01:02.29 | TyggrBob | hehehe |
| 01:04.06 | Debolaz | :P |
| 01:04.14 | Debolaz | Seriously though. |
| 01:04.21 | florian | TyggrBob: yep |
| 01:04.22 | Debolaz | Just installed GPE on my iPAQ. |
| 01:06.18 | *** join/#gpe TyggerBob (~me@Toronto-HSE-ppp3859672.sympatico.ca) |
| 01:06.58 | TyggerBob | my internet access is down.. I'm using my neighbour's wireless.. (with permission) and the connection is tenuous at best |
| 01:07.05 | TyggerBob | gnight Florian.. |
| 01:07.08 | TyggerBob | hvae a good one. |
| 01:07.24 | TyggerBob | watching 21 Days later.. |
| 01:07.38 | TyggerBob | looks pretty good so far |
| 01:07.46 | Debolaz | Nah. |
| 01:07.49 | Debolaz | I didn't like it that much. |
| 01:08.13 | TyggerBob | reminds me of Omega Man |
| 01:19.44 | shtrs_ | Debolaz anything new with gpe? |
| 01:19.52 | shtrs_ | Debolaz is it more usable then opie? |
| 01:20.00 | shtrs_ | with h5550 of course |
| 01:20.44 | Debolaz | I can't see any menu options to suspend/resume. |
| 01:20.55 | shtrs_ | the power button works? |
| 01:21.26 | Debolaz | The power button seems to shut it down completely, because when I try to push it again, it starts a fresh boot. |
| 01:25.28 | shtrs_ | hmm |
| 01:25.39 | shtrs_ | better then the always on with opie |
| 01:25.58 | shtrs_ | I think I will try GPE. |
| 01:26.27 | Debolaz | + GPE is probably better for me because then I can run GTK applications without a fuzz. |
| 01:26.53 | shtrs_ | I think the device is at least usable with gpe. |
| 01:27.01 | shtrs_ | because opie has too many issues. |
| 01:27.32 | shtrs_ | well. I will backup everything now and install gpe. |
| 01:27.38 | shtrs_ | thx for the answers |
| 01:28.18 | Debolaz | I'm also going to try to set up an encrypted partition on the device. |
| 01:28.28 | Debolaz | That won't get erased with every reflash. :P |
| 01:29.00 | shtrs_ | one more question |
| 01:29.48 | shtrs_ | when you shutdown and start the device using the power button. everything get back as it was, or you have to re-configure everything? |
| 01:30.35 | Debolaz | With familiar, everything is written to flash (Unlike PocketPC, which stores settings and 3rd party applications in the temporary memory storage) |
| 01:31.25 | Debolaz | So there's no need to reconfigure anything no matter what happends (Except for reinstallation) afaik. |
| 01:31.34 | shtrs_ | hmm |
| 01:31.35 | shtrs_ | good |
| 01:31.50 | shtrs_ | and is it possible to use the 128mb ram available on the device? |
| 01:33.39 | Debolaz | It's not extensively utilized by default, but /mnt/ramfs gives you access to use it. |
| 01:35.33 | *** join/#gpe shtrs (~shooter@cm-netfortaleza-for-C8B1DD90.brdterra.com.br) |
| 01:35.55 | Debolaz | [02:35] <Debolaz> It's not extensively utilized by default, but /mnt/ramfs gives you access to use it. |
| 01:35.58 | Debolaz | In case you didn't see that. |
| 01:39.58 | shtrs | oh |
| 01:39.59 | shtrs | thx! |
| 01:40.22 | shtrs | I'm installing gpe2 now. |
| 01:40.30 | shtrs | Will give you the feedback |
| 02:35.49 | *** join/#gpe jg_ (~jg@h005018015b26.ne.client2.attbi.com) |
| 02:51.34 | *** join/#gpe cmarqu_ (colin@buug.mind.de) |
| 02:51.59 | Debolaz | Interesting, trying to record a memo crashes GPE. |
| 03:00.24 | shtrs | shit |
| 03:00.30 | shtrs | im still installing gpe :) |
| 03:00.48 | shtrs | got the write failed because the flash was protected |
| 03:02.08 | Debolaz | Ouch. |
| 04:16.19 | *** join/#gpe Debolaz (~Debolaz@rgbqi71.cm.chello.no) |
| 04:28.09 | *** join/#gpe Debolaz (debolaz@sheep.hardship.net) |
| 06:02.43 | *** join/#gpe destroyer (~destroyer@029.a.001lg.syd.iprimus.net.au) |
| 06:49.49 | *** join/#gpe destroyer (~destroyer@095.c.001lg.syd.iprimus.net.au) |
| 06:50.33 | Debolaz | My log is slowly filled with Xlib: extension "MIT-SCREEN-SAVER" missing on display ":0.0". |
| 06:50.43 | Debolaz | Is this vengance on me for having so much memory? :P |
| 09:14.40 | *** join/#gpe PaxAnima (~petter@b134b.studby.ntnu.no) |
| 09:36.24 | *** join/#gpe destroyer (~destroyer@095.c.001lg.syd.iprimus.net.au) |
| 09:55.47 | *** join/#gpe __moray (~moray@dsl-217-155-67-241.zen.co.uk) |
| 10:53.50 | *** join/#gpe pb_ (~pb@cpc2-cmbg4-3-0-cust239.cmbg.cable.ntl.com) |
| 11:34.27 | *** join/#gpe BillSteamShovel_ (~BillSteam@CPE-144-132-26-113.vic.bigpond.net.au) |
| 11:34.47 | BillSteamShovel_ | Anyone know why java1.3 on gpe 0.7.2 is making X crash? |
| 11:34.58 | BillSteamShovel_ | .. |
| 11:35.00 | BillSteamShovel_ | [Loaded java.io.PushbackInputStream from /usr/j2re1.3.1/lib/rt.jar] |
| 11:35.00 | BillSteamShovel_ | [Loaded java.util.zip.CRC32 from /usr/j2re1.3.1/lib/rt.jar] |
| 11:35.00 | BillSteamShovel_ | [Loaded java.util.zip.Checksum from /usr/j2re1.3.1/lib/rt.jar] |
| 11:35.00 | BillSteamShovel_ | [Loaded sun.awt.motif.DrawStringOneBitMono from /usr/j2re1.3.1/lib/rt.jar] |
| 11:35.00 | BillSteamShovel_ | [Loaded sun.awt.FontDescriptor from /usr/j2re1.3.1/lib/rt.jar] |
| 11:35.00 | BillSteamShovel_ | SIGTERM |
| 11:36.34 | *** join/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 11:41.22 | *** join/#gpe florian (~fuchs@pD9E5FDA0.dip.t-dialin.net) |
| 11:41.27 | florian | re |
| 11:42.04 | puks | hi florian |
| 11:43.02 | florian | hi puks |
| 11:46.04 | pb_ | florian: good afternoon |
| 11:49.33 | *** part/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 11:50.17 | florian | pb_: hello... oh, yes... it's already afternoon |
| 11:50.29 | pb_ | :-) |
| 11:50.33 | *** join/#gpe BillSteamShovel (~BillSteam@CPE-144-132-26-113.vic.bigpond.net.au) |
| 11:54.24 | *** join/#gpe destroyer (~destroyer@136.a.002.syd.iprimus.net.au) |
| 12:22.52 | *** part/#gpe BillSteamShovel (~BillSteam@CPE-144-132-26-113.vic.bigpond.net.au) |
| 13:14.25 | *** join/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 13:25.40 | florian | bbl |
| 13:51.29 | *** join/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 14:10.58 | *** join/#gpe florian_kc (~fuchs@kc.celox-online.de) |
| 14:42.44 | *** join/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 14:58.00 | *** join/#gpe shtrs (~shooter@cm-netfortaleza-for-C8B1DD90.brdterra.com.br) |
| 14:58.41 | shtrs | Debolaz Hi |
| 14:58.52 | Debolaz | Hey. |
| 14:58.57 | Debolaz | How did the installation go? |
| 14:59.02 | shtrs | LoL |
| 14:59.23 | shtrs | The main problems of opie i could "turn around" with gpe |
| 14:59.32 | shtrs | but gpe has lots of other problems :~ |
| 14:59.37 | Debolaz | Heh, I know. |
| 14:59.51 | Debolaz | Attempting to record audio or scan for wireless networks will crash X on my iPAQ. |
| 15:00.04 | shtrs | same here |
| 15:02.25 | Debolaz | And I haven't had any success finding a good video player either. |
| 15:02.39 | Debolaz | You'd think a 400MHz CPU would be able to play video at more than 2-3 fps. |
| 15:03.22 | shtrs | wow! sure it does. |
| 15:03.28 | florian_kc | If you have a player that doesn't use floating point stuff :-) |
| 15:03.52 | Debolaz | famplayer and opieplayer2 didn't manage to get any better on "Jennifer Love Hewitt - Cheerleader Tryout.mpg". |
| 15:04.10 | Debolaz | opieplayer2 didn't even manage to get the audio right. |
| 15:04.24 | shtrs | i think depends only how stable the video player is and if it was compiled with otimizations to the 400mhz processor of h5550 |
| 15:05.46 | shtrs | god damn it |
| 15:12.07 | Debolaz | I've got to make an ipk that installs a working ppp setup on the system. |
| 15:16.46 | pb_ | What's the problem with the default ppp setup? |
| 15:17.57 | Debolaz | pb_: Well, there's not a problem with it, but it's not very usefull against Windows. |
| 15:18.06 | pb_ | Ah, I see. |
| 15:18.23 | pb_ | What does Windows need? |
| 15:18.46 | Debolaz | Connecting to the ipaq "just works", however, Windows is unable to share internet with the ipaq then because it considers itself the client and hence deems it uneccesary to share its own net with the server. |
| 15:19.04 | shtrs | Debolaz wich windows? |
| 15:19.09 | Debolaz | Any Windows afaik. |
| 15:19.36 | shtrs | with 2k and xp i got t working enabling the ip forward stuff editing the registry |
| 15:19.51 | Debolaz | There are very detailed wiki pages on how to set up a connection the other way around though. |
| 15:19.54 | shtrs | and masquerading the ipaq IP on my linux gateway |
| 15:20.20 | shtrs | i also had to tell to my linux gateway that my workstion ip is the gateway to the ipaq ip |
| 15:20.21 | shtrs | :~ |
| 15:27.34 | Debolaz | GPE will probably be better for me than OPIE though, since I want to use applications like GPA. |
| 15:28.05 | shtrs | what gpa does? |
| 15:28.54 | florian_kc | eh.. feels |
| 15:33.30 | florian_kc | bbl |
| 15:37.02 | Debolaz | shtrs: It's a frontend for GnuPG. |
| 17:01.00 | *** join/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 17:34.08 | treke|home | ibot, seen mallum? |
| 17:34.12 | | mallum <~mallum@82-35-112-58.cable.ubr07.dals.blueyonder.co.uk> was last seen on IRC in channel #gpe, 18h 38m 30s ago, saying: 'g'night all'. |
| 17:59.29 | *** join/#gpe flypiper (~deb-user@12-222-141-210.client.insightBB.com) |
| 18:08.57 | *** join/#gpe florian (~fuchs@pD955BD45.dip.t-dialin.net) |
| 18:12.43 | florian | re |
| 18:19.58 | puks | hi |
| 18:20.16 | puks | does python run with tkinker as well as GTK? |
| 18:20.47 | *** join/#gpe luc_ (luccio@cs78168215.pp.htv.fi) |
| 18:24.13 | *** join/#gpe ricmm (ricky@200.47.143.122) |
| 18:24.16 | ricmm | hello |
| 18:24.28 | ricmm | florian: hey |
| 18:25.41 | florian | hi ricmm |
| 18:26.22 | ricmm | i need some help in here |
| 18:26.32 | ricmm | how do i set up a new background image for GPE2 desktop |
| 18:26.35 | pb_ | hi luc_ |
| 18:26.59 | pb_ | luis: would you fix the missing .desktop extension on /usr/share/applications/gpe-today? |
| 18:27.11 | luis | pb_: again? |
| 18:27.21 | pb_ | oh, did you fix it already? it's still wrong in 0.7.2. |
| 18:27.39 | luis | pb_: i uploaded a fixed package to unstable |
| 18:27.52 | florian | ricmm: Use gpe-conf... "Settings" and then "Global Look"... |
| 18:27.58 | luis | pb_: 0.8 |
| 18:28.07 | ricmm | nothing happens when i set an image there, ive already done that |
| 18:28.08 | pb_ | ah, yes, I see it. |
| 18:28.39 | pb_ | nelson: would you move gpe-today_0.8 (and the associated locale packages) into 0.7.2? |
| 18:28.52 | ricmm | florian: do i need an specific file type, i am using a jpeg image |
| 18:29.29 | luis | pb_: is there a more automated way of uploading packages? it's a bit boring to upload all the locale packages one by one |
| 18:29.59 | pb_ | luis: if you're using the normal build scripts, you should get a .batch file that you can pass to ipkg-upload. |
| 18:30.23 | florian | ricmm: I'm nut sure... just try if mbdesktop likes it :-) |
| 18:30.28 | luis | pb_: right, thanks |
| 18:30.28 | ricmm | eh |
| 18:30.29 | ricmm | heh |
| 18:30.51 | ricmm | hmm it set a gif image, but the desktop things dissapeared |
| 18:31.02 | ricmm | when i called the desktop again, it is still blue |
| 18:32.06 | luc_ | hi all |
| 18:32.13 | florian | hi luc_! |
| 18:33.00 | florian | ricmm: Plese file a bug against this... mbdesktop is not that stable when it comes to background settings :-/ |
| 18:38.35 | ricmm | is there any problem with shutting down the ipaq with shutdown -h now |
| 18:38.48 | ricmm | it doesnt comes back when i press power button, i have to press reset |
| 18:39.34 | pb_ | why are you shutting it down like that? |
| 18:40.48 | ricmm | trying it |
| 18:41.28 | pb_ | well, I think the behaviour you're seeing is expected. |
| 18:41.43 | ricmm | and it is scary too xD ! |
| 18:42.13 | puks | gpe-calendar2 still pretends me to have locale-base-pt+pt installed... |
| 18:59.41 | *** join/#gpe mallum (~mallum@82-35-112-58.cable.ubr07.dals.blueyonder.co.uk) |
| 19:00.58 | *** join/#gpe RobiWan (~RobiWan@p508251E2.dip0.t-ipconnect.de) |
| 19:12.00 | *** join/#gpe ipaqprobs (ipaqprobs@ip68-228-58-21.tc.ph.cox.net) |
| 19:15.09 | *** join/#gpe TyggerBob (bdavies@CPE00045a265f64-CM.cpe.net.cable.rogers.com) |
| 19:15.18 | TyggerBob | 'lo all |
| 19:21.28 | Debolaz | Yo |
| 19:22.36 | mallum | hey TyggerBob |
| 19:24.26 | TyggerBob | Heya Mallum.. |
| 19:24.29 | TyggerBob | how's it going today? |
| 19:24.39 | TyggerBob | better for me now.. cable's back up. |
| 19:24.44 | mallum | TyggerBob: cool :-) |
| 19:25.45 | TyggerBob | It's amazing how much the internet has become an appliance, in my house. |
| 19:25.53 | TyggerBob | TV Guide, movie listings, news.. |
| 19:26.11 | TyggerBob | I'd go to check something and "oh yeah, the cable's down.. damn!" |
| 19:27.58 | mallum | hehe, yeah |
| 19:29.40 | treke|home | yo mallum |
| 19:29.48 | mallum | hey treke|home |
| 19:30.12 | mallum | treke|home: hows the west coast today ? |
| 19:30.25 | treke|home | it's well |
| 19:30.31 | treke|home | how is the erm, whichever coast london is closest to :) |
| 19:31.06 | mallum | pretty cold today :/ |
| 19:31.07 | TyggerBob | treke|home: where on which west coast? |
| 19:31.17 | treke|home | TyggerBob: Southern california. |
| 19:31.35 | TyggerBob | Ahhh |
| 19:31.50 | TyggerBob | <PROTECTED> |
| 19:32.15 | mallum | TyggerBob: thats like nearly being british I guess |
| 19:32.30 | TyggerBob | Hehehe. |
| 19:32.40 | TyggerBob | And what do you call cold, by the way.. |
| 19:32.46 | treke|home | mallum: What do you think of the performance difference in X? |
| 19:32.47 | TyggerBob | -9C here today |
| 19:33.16 | treke|home | it's pretty cool here |
| 19:33.17 | mallum | treke|home: between rotations ? |
| 19:33.24 | treke|home | mallum: yeah |
| 19:33.44 | treke|home | down to 16C here |
| 19:33.55 | TyggerBob | oooo brrrr |
| 19:34.13 | TyggerBob | ;) |
| 19:34.25 | treke|home | TyggerBob: our record low is -2 :p |
| 19:34.35 | TyggerBob | hehehehehe |
| 19:35.16 | mallum | treke|home: its very annoying. and afaik very hard to fix without hardware rotation |
| 19:35.38 | treke|home | mallum: I know. have you tried the rotated X to see how it performs? |
| 19:35.42 | TyggerBob | -16 here w windchill |
| 19:35.52 | treke|home | It seemed quicker to me, but I don't use it enough to know if it was just my imagination |
| 19:36.07 | treke|home | TyggerBob: The sad thing is, people are wearing big heavy jaggets |
| 19:36.55 | mallum | treke|home: no not yet. I need to flash your kernel to experiment |
| 19:36.59 | treke|home | cool |
| 19:37.05 | mallum | treke|home: hopefully get a chance to do it later |
| 19:37.09 | TyggerBob | Yeah.. I went down to Disneyland one november.. we're in shorts and tshirts and some californians had scarves |
| 19:37.11 | TyggerBob | hehehe |
| 19:37.19 | treke|home | it's just crazy |
| 19:37.44 | TyggerBob | Treke: what's that about a new kernel |
| 19:38.01 | ipaqprobs | hi. can anyone help me to get started installing my micro-innovations wireless keyboard on my ipaq 54xx |
| 19:38.06 | mallum | treke|home: I have a program you can run to test performance |
| 19:38.15 | treke|home | TyggerBob: It's a c760 kernel with some patches applied so that we can experiment with the w100 performance |
| 19:38.18 | treke|home | mallum: Oh? |
| 19:38.20 | mallum | treke|home: it should be approx 8x faster so should be noticeable |
| 19:38.26 | TyggerBob | <PROTECTED> |
| 19:38.51 | treke|home | TyggerBob: X's rotation on the c760 is pitiful in what most people consider the normal rotation |
| 19:39.05 | TyggerBob | ipaqprobs: external keyboards are not yet supported in Linux on pxa based ipaqs :( |
| 19:39.15 | TyggerBob | Ahhhh |
| 19:39.21 | TyggerBob | I see |
| 19:39.31 | treke|home | the ipaq I had was always a bit sluggish, but it always seemed more like a flicker, not a visible clearing of the screen :) |
| 19:39.33 | mallum | treke|home: http://handhelds.org/~mallum/fullscreen/sources/test-x.c |
| 19:40.01 | mallum | treke|home: the ipaq runs by defualt rotated |
| 19:40.08 | treke|home | mallum: I know. |
| 19:40.18 | TyggerBob | ipaqprobs: I have a snapntype and have the same problem.. |
| 19:40.26 | treke|home | It's probably the size of the screen that makes it more noticable |
| 19:40.40 | TyggerBob | someone needs to write a device driver |
| 19:40.51 | TyggerBob | unfortunately, that ain't me |
| 19:41.30 | mallum | treke|home: see http://handhelds.org/~mallum/fullscreen/x88.html for an idea in performance differences for different orientations |
| 19:42.45 | treke|home | ouch |
| 19:43.56 | ipaqprobs | tyggerbob: ugh. not supported, eh. that sucks. |
| 19:44.44 | mallum | treke|home: I think there are so many factors though. For instance Xvesa on my desktop feels so slow. But when you run it with the composite extension it feels like 10 times faster - but its the same speed |
| 19:44.45 | ipaqprobs | would anyone know if gpe is what causes my /etc/modules file to look different than every other one i see |
| 19:44.54 | ipaqprobs | it says not to modify it, it was generated by update-modules. |
| 19:45.09 | pb_ | mallum: that's probably just an artifact of the shadow framebuffer. |
| 19:45.36 | pb_ | shadowfb will make X feel faster on most PC-class hardware, because reading from AGP memory is very slow. |
| 19:46.08 | *** join/#gpe jg_ (~jg@24.218.46.90) |
| 19:46.24 | treke|home | mallum: reflashing X now to test |
| 19:47.43 | *** part/#gpe puks (~puccio@62-101-126-233.fastres.net) |
| 19:49.59 | mallum | pb_: right. it feels about the same on the ipaq |
| 20:04.21 | treke|home | pb_: Do you build your familiar images on an arm machine? |
| 20:04.32 | pb_ | yeah |
| 20:04.36 | treke|home | ah |
| 20:05.31 | pb_ | The build process needs to run the postinsts, so it can't really be done cross. |
| 20:05.48 | treke|home | yeah. Thats what I was wondering about |
| 20:06.02 | pb_ | I guess we could make like OE and run the postinsts on the target during first boot. |
| 20:09.52 | treke|home | mallum: should I be pressing any buttons? |
| 20:10.41 | *** join/#gpe destroyer (~destroyer@210.50.72.194) |
| 20:10.56 | mallum | treke|home: ? |
| 20:11.18 | treke|home | the test-x app was just sitting there. Wasnt sure if it was supposed to do anything after drawing the colors |
| 20:11.21 | treke|home | It finally quit |
| 20:11.45 | mallum | yeah just run it, it should give you an output |
| 20:13.02 | treke|home | ok. Portrait no w100 usage is 6 times faster. Wow |
| 20:16.02 | treke|home | very noticable |
| 20:16.07 | treke|home | Portrait No w100 |
| 20:16.07 | treke|home | <PROTECTED> |
| 20:16.07 | treke|home | Portrait w/ w100: |
| 20:16.07 | treke|home | <PROTECTED> |
| 20:16.19 | treke|home | wait. the last one is Landscape w/ w100 |
| 20:18.21 | treke|home | and if I rotate left again it goes to portrait mode and I get 2553 KB/s |
| 20:18.55 | mallum | thats cool, so much faster |
| 20:19.02 | TyggerBob | WOW! |
| 20:21.12 | treke|home | I'll install sash's image and see what I get there |
| 20:21.26 | treke|home | should be an easy way of seeing if he uses the hardware :) |
| 20:23.44 | mallum | yeah, good idea |
| 20:23.58 | treke|home | he only uses gcc2 though, which is odd |
| 20:24.22 | mallum | its still pretty slow though, the ipaq can do twice as fast as that |
| 20:24.50 | pb_ | treke|home: if you're in the mood for benchmarking, I'd find it amusing if you could grab http://handhelds.org/~pb/bw_mem and see what results you get from "bw_mem 10000000 rd" on the Z. |
| 20:24.58 | treke|home | sure. why not. |
| 20:25.19 | mallum | pb_: cool, what does that do ? |
| 20:25.31 | pb_ | mallum: it's part of lmbench; just measures memory read bandwidth. |
| 20:25.43 | treke|home | 10.00 45.12 |
| 20:25.47 | pb_ | for some reason the pxa25x-based ipaqs seem to be very bad at that (worse than the strongarm, in fact). |
| 20:26.06 | pb_ | treke|home: thanks. could you also try with wr in place of rd? |
| 20:26.28 | treke|home | 10.00 45.04 |
| 20:26.46 | pb_ | freaky. I wonder what's going on there. |
| 20:26.58 | treke|home | Is that good or bad? |
| 20:27.12 | pb_ | bad. I was expecting about 80 for the first one, and about 200 for the second one. |
| 20:27.19 | treke|home | fun. |
| 20:27.41 | treke|home | bad as in worse than ipaq-pxa bad? |
| 20:28.33 | pb_ | yeah. if you look at http://handhelds.org/~pb/lmbench.txt, the corresponding numbers are "Mem read" and "Mem write" (last two columns of the last-but-one paragraph). |
| 20:29.07 | treke|home | hmm. |
| 20:29.09 | treke|home | So that's not good :) |
| 20:29.30 | treke|home | could the fact that this is using a gcc2 based kernel be a problem? |
| 20:29.39 | pb_ | No, it shouldn't make any difference. |
| 20:30.03 | treke|home | Anything you can think of other than shit hardware that I might have messed up? |
| 20:30.24 | pb_ | Nothing comes to mind. Unless you have other processes running that are sapping your CPU bandwidth. |
| 20:30.48 | pb_ | I don't think the hardware itself is likely to be bad, though it's possible that the kernel or bootldr is setting it up poorly. |
| 20:32.00 | pb_ | do you have devmem2 on there? if so, you could peek at the sdram controller registers and find out what's up. |
| 20:32.58 | treke|home | nope. You have a copy on hand? |
| 20:34.33 | pb_ | it's in familiar. look for devmem2*.ipk in handhelds.org/feeds/unstable/ |
| 20:34.34 | treke|home | got it |
| 20:34.40 | treke|home | ipkgfind is my friend |
| 20:34.51 | pb_ | aha :-) |
| 20:35.08 | treke|home | ok. I have it, where am I peeking? |
| 20:35.19 | pb_ | the interesting registers are mdcnfg and mdrefr, which you'd read with "devmem2 0x48000000 w" and "devmem2 0x48000004 w". |
| 20:35.33 | treke|home | Value at address 0x48000000 (0x40016000): 0x1A81AC9 |
| 20:35.41 | treke|home | Value at address 0x48000004 (0x40016004): 0x9C017 |
| 20:36.03 | pb_ | right, let me look up what those mean. |
| 20:38.10 | pb_ | one bank of sdram, 32 bits, 4 banks, 8 column bits, 13 row bits. tRP=3, CL=3, iRCD=3, tRAS=7, tRC=10 |
| 20:39.47 | pb_ | ah, and SDCLK=50MHz. that'd be the reason. |
| 20:40.12 | pb_ | ipaqpxa runs with 100MHz sdclk. |
| 20:42.00 | mallum | treke: I think theres stuff in that jap patch that allows you to change clock speeds ? |
| 20:44.46 | mallum | this is cool : http://www.cs.umd.edu/~kmconroy/ |
| 20:44.48 | treke|home | Is that something that you can change without breaking things? |
| 20:46.06 | treke|home | mallum: I thought the clock was a w100 thing |
| 20:46.39 | treke|home | mallum: slick |
| 20:51.22 | mallum | arg, turns out tabiword is a win32 only thing :/ |
| 20:55.51 | treke|home | interesting |
| 20:55.57 | treke|home | now I'm getting 10.00 92.09 |
| 20:56.54 | pb_ | for read or write? |
| 20:57.02 | treke|home | both |
| 20:57.22 | mallum | treke|home: what did you change ? |
| 20:57.45 | treke|home | rebooted |
| 20:57.52 | pb_ | weird |
| 20:58.12 | treke|home | mallum: And someone ran your test on sash's image. He just changed the default rotation. You get a huge performance hit for portrait mode |
| 20:58.16 | treke|home | pb_: I have a theory |
| 20:58.26 | pb_ | treke|home: what's your theory? |
| 20:59.37 | treke|home | pb_: I'm thinking it's related to the w100 patch. Running the ioctl changing app becomes signficicantly slower after using it and using fbset. I wonder if it's writing somewhere it shouldnt |
| 21:00.11 | mallum | treke|home: what was there fastest speed ? |
| 21:02.18 | treke|home | 9891KB/se |
| 21:02.21 | treke|home | in landscape |
| 21:04.46 | mallum | and alot worse in portrait ? |
| 21:05.13 | treke|home | 3012 |
| 21:06.13 | treke|home | weird. That didnt change my numbers |
| 21:07.06 | treke|home | devmem2 still returns the same things |
| 21:08.46 | mallum | treke|home: it sound like he's also rotating the display then. but probably just using a non magical kdrive as well |
| 21:09.05 | treke|home | yup. |
| 21:10.25 | treke|home | I just wish I had some clue what was going on |
| 21:10.31 | mallum | would be cool to find out how he;s rotating it |
| 21:10.32 | treke|home | with bw_mem |
| 21:32.28 | treke|home | what was sash thinking |
| 21:32.42 | treke|home | he enabled opaque window dragging by default |
| 21:43.01 | mallum | treke|home: slow ? |
| 21:43.30 | treke|home | yup |
| 21:43.51 | mallum | treke|home: I was experimented with translucent shadowed dragging windows , no chance |
| 21:44.50 | treke|home | just not enough horsepower |
| 21:46.12 | mallum | no looked nice though |
| 21:49.17 | treke|home | I bet |
| 21:56.52 | treke|home | mallum: email kergoth an ssh v1 public key |
| 22:00.27 | mallum | treke|home: what for ? |
| 22:00.41 | treke|home | mallum: So you can commit your fontconfig stuff |
| 22:00.45 | mallum | treke|home: I have oz + oe access I think |
| 22:00.51 | treke|home | ok |
| 22:01.11 | treke|home | I know you have oz, it's a different repository for oe though |
| 22:01.25 | mallum | yeah Im pretty sure he has a key for me on oe |
| 22:05.48 | *** join/#gpe destroyer (~destroyer@046.b.002.syd.iprimus.net.au) |