| 07:02.59 | walle | morrn |
| 14:40.38 | *** join/#pegasos.org lisardman (n=mikael@lisardman.org) |
| 16:44.38 | *** join/#pegasos.org TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
| 16:45.08 | TimRiker | where/how can I get a MorphOS boot CD for my Pegasos II? |
| 16:49.34 | Frek | I don't know to be honest |
| 16:49.50 | Frek | I haven't bothered to check either as virtually nothing happened to morphos the past 3 years anyway |
| 16:53.40 | TimRiker | yeah, I just want to re-arrange my partitions a bit. I could blast morphos and use the space, but I'd like a cd in case it want it back. |
| 16:54.06 | TimRiker | the docs online mention a beta test ftp account. I don't think I ever had such a thing. |
| 16:54.42 | TimRiker | does the firmware know about other partitions styles? or is amiga partitioning the only way to fly? |
| 16:55.54 | TimRiker | as a side note: my firewire ports seem to have gone awol recently. The drivers load, but nothing that I plug in gets recognized. |
| 16:57.23 | Frek | no idea |
| 16:57.28 | Frek | I've just got a peg 1 |
| 16:57.50 | TimRiker | er 2.6.18... |
| 16:58.02 | TimRiker | yeah, that new 2.6.28 is great.... heh |
| 17:08.55 | dholm | TimRiker: only amiga partition tables are understood |
| 17:12.01 | TimRiker | dholm: k. thanx. |
| 17:32.45 | kozz | TimRiker: it's download.morphos-team.net, but that is all I know |
| 17:32.53 | kozz | have lost the login :) |
| 17:46.34 | TimRiker | kozz: thx. looks like they do registration on the web site. I need my serial number. there is a S/N on a bar code that looks to be just for the case, and a hand written number on the bottom twice, and then there is the mac address. I don't see one that looks convincingly like a serial number. Thoughts? |
| 17:47.29 | TimRiker | I don't see any mention of a serial number in the firmware, but it may be there someplace. |
| 17:48.10 | TimRiker | also, though some web pages say it's the MAC addres, there are 2 MACs. :( |
| 17:52.58 | TimRiker | hmm. found a forum saying that the serial number is "on the back of the cpu card". hmm. |
| 17:57.17 | TimRiker | anyone know what a valid Peg II serial number should look like? numbers? letters? how many? etc. |
| 17:59.29 | ironfist | TR |
| 17:59.32 | ironfist | hi Tim |
| 18:02.49 | *** join/#pegasos.org cofyr (n=daniel_g@bullerovre127.bitnet.nu) |
| 18:04.31 | TimRiker | hi ironfist |
| 18:46.25 | TimRiker | any idea if morphos from cd would run under MOL? :) |
| 18:55.44 | Frek | TimRiker, yes. it won't |
| 18:55.52 | TimRiker | the number on the back of the cpu card matches the hand written number on the bottom of the case. :) looks like a peg2 serial is 6 digits. |
| 18:56.04 | Frek | MOL simulates a macintosh hw setup |
| 18:56.14 | Frek | the Pegasos is just a pc with a ppc cpu |
| 18:56.27 | TimRiker | morph won't run on a mac then? |
| 18:56.46 | Frek | no |
| 18:56.54 | TimRiker | k |
| 18:56.54 | Frek | and to be honest is there any point? |
| 18:57.25 | TimRiker | nah. just playing. like I was running DOS on my xbox yesterday. just cause I can, not cause it's useful. |
| 18:58.15 | Frek | better install contiki if you want a fast system with resonable modern features :p |
| 18:59.20 | Frek | ( http://www.sics.se/contiki/ ) |
| 19:00.14 | dholm | just as an SGI machine is a PC with a MIPS CPU? :P |
| 19:01.15 | Frek | dholm, everything but machines with ppc are pcs :p |
| 19:01.35 | dholm | but you just called the pegasos a pc O.O |
| 19:02.05 | Frek | the pegasos is an abnomination |
| 19:02.35 | dholm | shiva pees on them every morning before breakfast o.O |
| 19:02.45 | TimRiker | contiki... wild. wonder if it would fit on my old HPs. :) ut seriously, I'm running debian unstable on my peg2 and quite happy with that. it does not keep up real well with the desktop at 1920x1200 but it does ok. |
| 19:03.40 | Frek | TimRiker, my point is just contiki does basically all morphos has to offer and more, only it's 30 times smaller and open source |
| 19:03.51 | dholm | debian never was much for speed anyway |
| 19:03.58 | TimRiker | Frek: nod |
| 19:04.44 | TimRiker | dholm: the os it fine it's just the display that dogs at that res. I have 1G ram in the peg2 |
| 19:04.45 | Frek | TimRiker, yes, but I don't think the majority of the pegasos customers bought it to run linux, it's more of something they got forced to as nothing happened to the "lightning os" |
| 19:05.00 | dholm | TimRiker: their binaries have only ppc-generic optimizations |
| 19:05.06 | TimRiker | Frek: probably true. |
| 19:05.24 | TimRiker | dholm: true |
| 19:05.58 | Frek | dholm, but then again how much do you expect processor specific optimizations adds to it? 3-4% ? |
| 19:06.32 | dholm | Frek: hmm, probably a bit more.. ppc-generic is quite generic |
| 19:06.38 | dholm | ;) |
| 19:07.14 | Frek | dholm, well still the ppc is quite similar in all versions, most recent cpus has about the same instructionset / features |
| 19:07.51 | Frek | but anyway, nothing matches morphos when it comes to speed anyway ;) |
| 19:08.39 | Frek | Something I've been pondering about though... is the context switch timer in linux static on the powerpc or is it calculated from the bogomips value ? |
| 19:08.45 | dholm | usually there are differences in instruction scheduling.. yes, the cpu has out-of-order execution but there are still good and bad ways to feed the instructions |
| 19:10.24 | Frek | dholm, that's true, but 70% of all instructions are integer instructions, and if I don't remember all incorrectly the latest generation had pretty much similar integer/branch pipeline setups but anyway, you're right there are some difference I just doubt it adds all that much in non looped code |
| 19:11.54 | Frek | ie. G3 / G4 and variations of them, (G5 breaking that though, but it's not really a "true" powerpc anyway) |
| 19:12.33 | dholm | ppc has to work with everything from the early SoC to the latest G5 (in 32-bit mode) |
| 19:12.52 | Frek | I know |
| 19:13.05 | dholm | and if I wrote the compiler I would make sure the code executes reasonably fast on the old models and not worry too much about newer models |
| 19:13.42 | Frek | ironically I'm pretty certain the G5 would execute unoptimized powerpc code faster than g3/g4 optimized speaking of that btw, but yes you got a point |
| 19:14.20 | Frek | but then again there is nothing that says that generic powerpc is biased towards older powerpcs |
| 19:15.46 | Frek | then you have ppcs that are extremely pedantic about instruction scheduling, such as the 970 cores used in the xbox 2 |
| 19:16.37 | Frek | either way, afk - continuing playing FF XII now |
| 19:22.58 | dholm | me too, the G5 has much longer pipelines.. the G4 only has a 7-stage pipeline so you don't have much time to make scheduling decisions |
| 20:26.43 | Frek | dholm, yes but the G5 chokes on instructions the g3/g4 likes too |
| 20:27.15 | Frek | rlwimini and such that are common when optizming ppc |
| 20:27.40 | Frek | rlwimi I mean |
| 20:29.59 | Frek | (the g5 doesn't really have those instructions in it's "true" instruction set, so instead they're cracked down to regular instructions internally, this causes both pipeline boubles and latency in the execution due to the extra decoding step it has to do for "multi instructions" |
| 20:43.46 | lisardman | metal! |
| 21:58.42 | *** part/#pegasos.org cofyr (n=daniel_g@bullerovre127.bitnet.nu) |