00:40.24 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1096601470.dsl.bell.ca) |
00:42.00 | IriX64 | http://rafb.net/p/hng0Fb41.html <--- new kids on the block have a problem |
01:05.57 | IriX64 | http://rafb.net/p/HmDN2c78.html <--- found a critter in ltmain.sh |
01:12.55 | ``Erik | heh, EBCDIC, that's funny |
01:13.10 | IriX64 | not if you work for ibm :) |
01:13.23 | ``Erik | erm |
01:13.36 | ``Erik | and the last ibm that used that was what, the as400? |
01:13.47 | IriX64 | mv370 :) |
01:14.49 | IriX64 | doesn't solve *my problem tho, but what the heck |
01:15.25 | ``Erik | or ebcdic->ascii... I don't remember anymore |
01:15.29 | IriX64 | why not 8086 |
01:16.29 | IriX64 | the glory of unix=the iron doesn't matter :) |
01:17.04 | ``Erik | um, it was mostly 8086 crap, but 286 stuff was permitted |
01:17.11 | ``Erik | for some reason, 386 was not |
01:17.17 | ``Erik | and this was, like, '98 or '99 :( |
01:17.23 | ``Erik | (my second run at college) |
01:17.27 | IriX64 | was that around at that time |
01:17.42 | IriX64 | ah it was |
01:17.44 | ``Erik | pentium was around in like '95 |
01:18.10 | IriX64 | my memories like yours too far back to be accurate :) |
01:18.24 | ``Erik | I sure enjoyed doing asm on r2k a lot more |
01:18.31 | ``Erik | even though it was just using 'spim' |
01:18.35 | *** join/#brlcad rpaddock (n=bob_padd@c-24-131-109-50.hsd1.oh.comcast.net) |
01:18.43 | IriX64 | don't know r2k or is that r2d2 :) |
01:18.55 | ``Erik | r2000, early MIPS chip |
01:19.00 | IriX64 | ah |
01:19.05 | ``Erik | um, predecessor to what you may've seen in an sgi |
01:19.20 | ``Erik | the o2 bookend I have sports an r10k |
01:19.29 | IriX64 | im only familiar with pc's and dec schtuff |
01:19.40 | ``Erik | alphas were almost good |
01:19.47 | IriX64 | still are |
01:19.51 | ``Erik | if they woulda used the right endian, I coulda loved 'em :D |
01:19.56 | IriX64 | heh |
01:20.08 | ``Erik | they also did very horrible things |
01:20.09 | IriX64 | hate reversals do you |
01:20.32 | IriX64 | ?? prioris |
01:20.37 | ``Erik | yeah |
01:20.42 | ``Erik | shaped like a coffee table |
01:20.48 | IriX64 | dont know that one |
01:20.54 | ``Erik | cabinet for like 8 scsi drives, big honkin' backplane board |
01:21.02 | IriX64 | ah |
01:21.03 | ``Erik | the cpu and memory boards popped into the backplane |
01:21.12 | ``Erik | I had a dual p133 cpu card |
01:21.20 | IriX64 | unibus machine or omnibus |
01:21.31 | ``Erik | and like an 8 slot 30pin simm card |
01:21.35 | ``Erik | um, eisa |
01:21.41 | ``Erik | it was all eisa |
01:21.45 | IriX64 | ah a dec pc attempt |
01:21.48 | ``Erik | yes |
01:21.50 | ``Erik | ... |
01:21.52 | ``Erik | p133's... |
01:21.55 | ``Erik | simm... |
01:21.58 | IriX64 | i know |
01:22.12 | ``Erik | cheaper than, say, an sbus equiped machine... but... |
01:22.19 | IriX64 | never seen it the rainbow tho i played with |
01:22.31 | ``Erik | amusingly enough, I couldn't get nt3.51 or nt4 to install on it... linux took it though |
01:22.38 | ``Erik | the alphas were a hell of a lot better |
01:22.40 | IriX64 | heh |
01:22.48 | ``Erik | 27xsoemthingsomething... |
01:23.03 | IriX64 | depends who you talk to :) |
01:23.06 | ``Erik | damn fast, both in crunch per clock and in straight clock at the time |
01:23.23 | ``Erik | and I think the nt on alpha had... real smart shit. |
01:23.41 | IriX64 | remember the alpha with nt |
01:23.49 | IriX64 | nice enoufgh |
01:23.50 | ``Erik | I'm under the impression (having never used it) that it had a machine code translator that cached the converted machine code |
01:23.53 | IriX64 | enough |
01:24.12 | ``Erik | so you could slap an i386 program on it, run it, it'd sit and spin for a minute the first time... then run just fine |
01:24.19 | ``Erik | next time ya launched it, it just popped right up and worked |
01:24.26 | IriX64 | i was happy i could do serial port stuff with it |
01:24.57 | ``Erik | oh, yeah, ya got leenewx or fbsd, PIC chips, and a programmer? |
01:25.16 | ``Erik | what's his name is making a new release of picprog and wants people to test the new serial timing code :D |
01:25.38 | IriX64 | seriously? i'm a hobbiest :) |
01:26.15 | ``Erik | this is free software kinda targetting the hobbyist |
01:26.35 | ``Erik | he has a short lift of people he knew were interested (I maintain the fbsd port, so I was on it) |
01:26.53 | ``Erik | but I only have a couple 16f88's and haven't done anything honestly noteworthy with 'em |
01:27.06 | ``Erik | other than making LED's blink in patterns |
01:27.12 | IriX64 | don't have the tools man |
01:27.20 | ``Erik | software or hw? |
01:27.24 | IriX64 | hw |
01:27.26 | ``Erik | ah |
01:27.36 | ``Erik | the programmer I got was like $15, and the chips were like 2.50 a pop |
01:27.47 | ``Erik | the serial cable was probably the most expensive part, I think it was $20 :( |
01:28.02 | IriX64 | got enough to keep me busy but thanks |
01:28.09 | ``Erik | but I already had a breadboard and a huge box o' parts |
01:28.51 | ``Erik | and a 5w cheap power supply I made from the transformer out of a cheap broken alarm clock, 4 diodes, 2 caps and a resistor :D |
01:29.01 | IriX64 | heh |
01:29.09 | ``Erik | oh, and a candle to drop wax on the assembly once I was happy with the output DC signal |
01:29.13 | IriX64 | zeners are cheap |
01:29.31 | ``Erik | yeah, btu I didn't have any |
01:29.44 | IriX64 | :( |
01:29.52 | ``Erik | <-- scratches his head and wonders how zeners would simplify a bridge |
01:30.00 | IriX64 | heh |
01:30.03 | IriX64 | wouldn't |
01:30.17 | IriX64 | just stating that they're cheap :) |
01:30.22 | ``Erik | the kind I had were the cheapest, not even germaniums |
01:30.34 | ``Erik | the ones that you used to be able to get at radioshack in a package of like 100 for $2 |
01:30.41 | ``Erik | back when radioshack did that kinda thing |
01:30.44 | IriX64 | sliconized mutational diodes were they |
01:31.00 | ``Erik | I've have to look... the cheapest on the rac |
01:31.03 | ``Erik | per unit |
01:31.06 | IriX64 | heh |
01:31.21 | ``Erik | I have way more ceramic caps than electrolytics, too |
01:31.38 | IriX64 | teardrops too are cheap |
01:31.41 | ``Erik | my big dollar stuff is mostly the 74xx series ic's I think |
01:32.02 | IriX64 | the hobbisyts friend standard ttl |
01:32.12 | ``Erik | thought I have a few "weird" items, like a 10 bar LED, and a couple 8 piece LED's |
01:32.22 | ``Erik | 555's and 556's, some 741's I think |
01:32.35 | IriX64 | stuff like that gets costly too |
01:32.45 | ``Erik | the LED's? or the ic's? |
01:32.50 | IriX64 | leds |
01:32.52 | ``Erik | the ic's are ass cheap, common as hell |
01:32.57 | IriX64 | true |
01:33.02 | ``Erik | yeah, I think they were a few bucks each |
01:33.03 | IriX64 | not 741's tho |
01:33.20 | ``Erik | hum, I got a box with several 741's for a couple bucks |
01:33.27 | IriX64 | 555 are as common as dirt as far as timer chips go |
01:33.29 | ``Erik | have yet to find a good use for an opamp like that |
01:34.17 | ``Erik | I want to buy an m68020 or something, some crud for memory, mebbe something to do video signals... see if I can make me a little unix machine |
01:34.24 | ``Erik | and code up a unix workalike on it :D |
01:34.27 | IriX64 | the memory data book as usefull as the ttl data book |
01:34.39 | ``Erik | I think the 20 introduced the mmu |
01:34.47 | IriX64 | i have faichild what do you use |
01:34.53 | IriX64 | fairchild |
01:35.00 | ``Erik | huh? |
01:35.03 | IriX64 | for reference |
01:35.09 | IriX64 | books now |
01:35.11 | ``Erik | um, these days? the intarweb |
01:35.22 | IriX64 | i prefer paper in my hand |
01:35.42 | ``Erik | I used to have, um, the little paperbacks with the grid pattern on the cover that you could get at radioshack |
01:35.48 | ``Erik | btu they're at my parents |
01:35.57 | IriX64 | never used those |
01:36.07 | ``Erik | I have a notebook with the pinouts for msot of those chips from a college class |
01:36.15 | ``Erik | hand drawn notes from class |
01:36.15 | IriX64 | ah |
01:36.25 | ``Erik | I was introduced to the 74xx's in a class |
01:36.42 | ``Erik | and the 555 and 556 were just that damn common, that I put those in there, too |
01:36.51 | IriX64 | i was introduced to them in industry |
01:37.04 | ``Erik | the 741, I have a book from one of those ancient "101 projects in electronics" kids |
01:37.04 | ``Erik | kits |
01:37.14 | IriX64 | ah |
01:37.17 | ``Erik | too bad those don't seem to exist anymore |
01:37.37 | IriX64 | i never looked , really they stopped them |
01:37.55 | ``Erik | I know the chemistry kits are gone, I haven't SEEN an electronics kit |
01:38.06 | IriX64 | blargh |
01:38.23 | IriX64 | kids used to eat those up |
01:38.33 | ``Erik | chemistry kits were beaten down by politics, cuz, uh, some of those chemicals might possibly create something dangerous if mixed just right |
01:39.03 | IriX64 | heh and leave it to a bright kid to "mix" them right :) |
01:39.41 | ``Erik | <-- personally kinda thinks parents should be more involved and should be more education *shrug* :/ |
01:40.32 | ``Erik | and if'n ya dick with anything ya don't know, understand it may be a little dangerous and learn appropriate safety policies |
01:40.33 | IriX64 | truth |
01:40.50 | IriX64 | well its making again, bbiab |
01:52.04 | *** join/#brlcad yukonbob (n=yukonbob@CPE001125477e9c-CM0011e6be27b1.cpe.net.cable.rogers.com) |
02:05.55 | *** part/#brlcad rpaddock (n=bob_padd@c-24-131-109-50.hsd1.oh.comcast.net) |
02:35.38 | ``Erik | oh ho ho |
02:35.41 | ``Erik | how ironical |
02:35.48 | ``Erik | http://science.slashdot.org/article.pl?sid=07/10/31/2256206 |
03:03.15 | IriX64 | heheh they heard you |
03:58.06 | ``Erik | starting about 6 years ago. |
05:37.49 | IriX64 | man there's a dnl without a preceding # in configure.in in tk/unix |
05:43.31 | IriX64 | the ' in don't presents a problem |
06:18.26 | IriX64 | http://www3.sympatico.ca/mario.dulisse2/7.8.4doesvista64.png :) |
06:35.26 | IriX64 | nite all |
08:16.27 | *** join/#brlcad Z80-Boy (n=clock@zux221-122-143.adsl.green.ch) |
09:29.13 | *** join/#brlcad Elperion (n=Bary@p5487576C.dip.t-dialin.net) |
09:39.48 | *** join/#brlcad rpaddock (n=bob_padd@c-24-131-109-50.hsd1.oh.comcast.net) |
10:08.52 | *** part/#brlcad rpaddock (n=bob_padd@c-24-131-109-50.hsd1.oh.comcast.net) |
10:46.02 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-022-084.pools.arcor-ip.net) |
12:50.49 | *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net) |
13:26.08 | *** join/#brlcad bpoole (n=bpoole@UNIX39.andrew.cmu.edu) |
13:26.47 | *** join/#brlcad elite01_ (n=elite01@dslb-088-070-022-084.pools.arcor-ip.net) |
13:31.55 | *** join/#brlcad PrezKennedy (i=Matt@74.86.45.130) |
13:42.30 | *** join/#brlcad bobbmt (n=user@static-71-240-124-130.pitbpa.east.verizon.net) |
14:47.13 | *** join/#brlcad Elperion (n=Bary@p5487576C.dip.t-dialin.net) |
15:02.02 | CIA-27 | BRL-CAD: 03brlcad * 10brlcad/include/conf/Makefile.am: need all-am to be dependent upon built_sources else someone doing a make install without first running make will end up failing to generate the COUNT, DATE, HOST, USER, PATH files |
15:02.22 | CIA-27 | BRL-CAD: 03brlcad 07STABLE * 10brlcad/include/conf/Makefile.am: merge from head, need built_sources for all-am |
15:03.07 | ``Erik | it ist time to do the big honkin' join? |
15:05.38 | brlcad | not yet |
15:05.53 | brlcad | not until the remaining builds are posted |
15:06.03 | brlcad | (still need irix and solaris) |
15:07.43 | brlcad | this will possibly be the last irix release :( |
15:08.42 | ``Erik | heh |
15:25.26 | *** join/#brlcad nita (n=nita@i-83-67-61-194.freedom2surf.net) |
15:25.53 | *** join/#brlcad MinuteElectron (n=MinuteEl@silentflame/member/pdpc.base.minuteelectron) |
15:36.27 | *** join/#brlcad Elperion (n=Bary@p5487576C.dip.t-dialin.net) |
18:18.26 | *** join/#brlcad yukonbob (n=yukonbob@CPE001125477e9c-CM0011e6be27b1.cpe.net.cable.rogers.com) |
18:46.52 | dtidrow_work | brlcad: yep, IRIX is a dying OS :-\ |
18:47.12 | dtidrow_work | the company isn't doing much better, afaik |
18:54.09 | PrezKennedy | people still use IRIX? |
19:20.40 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1096601470.dsl.bell.ca) |
19:21.07 | IriX64 | http://rafb.net/p/I3oL3a29.html <---- have i reported this already? |
19:23.06 | minute | PrezKennedy: HAHA |
19:31.48 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-022-084.pools.arcor-ip.net) |
19:34.19 | IriX64 | pssst, there's 9 dnl's in config.in in src/other/tk/unix. |
19:35.54 | IriX64 | some in tcl too |
19:37.38 | IriX64 | http://rafb.net/p/BEB5Wr15.html <---- rtsrv duplicates bu_bomb too |
19:41.27 | *** join/#brlcad MinuteElectorn (n=MinuteEl@cpc4-flit1-0-0-cust283.lutn.cable.ntl.com) |
19:42.29 | IriX64 | errr +o :) |
19:52.55 | ``Erik | um, ok? |
19:53.07 | IriX64 | so eh :) |
19:53.37 | ``Erik | 'dnl' is m4 for a line comment (do not list or something)... autoconf will remove thos elines when the configure script is generated |
19:53.59 | IriX64 | my build platform is kinda weird ``Erik so i'm not sure.... |
19:54.03 | ``Erik | <-- personally prefers shell comments as he looks through the generated configure script more often than he'd like |
19:55.54 | *** join/#brlcad iMinute (n=MinuteEl@silentflame/member/pdpc.base.minuteelectron) |
19:58.06 | *** join/#brlcad Z80-Boy (i=clock@77-56-72-166.dclient.hispeed.ch) |
20:17.16 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1096601470.dsl.bell.ca) |
20:21.46 | brlcad | IriX64: it *intentionally* duplicates bu_bomb as a scope override -- there's some compiler flag amiss if it's complaining about it |
20:22.25 | brlcad | and the dnl's are normal, part of m4 syntax |
20:22.36 | brlcad | there are dnl's throughout our own top-level configure.ac |
20:24.36 | IriX64 | i'll check my flags then thanks |
20:25.00 | brlcad | check the gcc documentation |
20:25.11 | IriX64 | way ahead of you :) |
20:26.31 | IriX64 | found --allow-multiple-definition, will that do it |
20:30.01 | brlcad | maybe |
20:30.53 | brlcad | my gcc nor ld has that option, so can't be sure |
20:31.11 | brlcad | if it's a linker option, it'll need to be an LDFLAGS instead of a CFLAGS |
20:34.15 | CIA-27 | BRL-CAD: 03brlcad * 10brlcad/src/util/pix-png.1: clarify that the user must specify the size if the image is not 512x512 |
20:47.49 | *** join/#brlcad minute (n=MinuteEl@silentflame/member/pdpc.base.minuteelectron) |
20:49.29 | *** join/#brlcad MinuteElectron (n=MinuteEl@silentflame/member/pdpc.base.minuteelectron) |
21:06.02 | *** join/#brlcad iMinute (n=MinuteEl@silentflame/member/pdpc.base.minuteelectron) |
21:10.46 | *** join/#brlcad yukonbob (n=yukonbob@CPE001125477e9c-CM0011e6be27b1.cpe.net.cable.rogers.com) |
21:30.04 | iMinute | =auit |
21:30.07 | iMinute | oops |
21:30.09 | iMinute | wrong channel |
21:36.48 | brlcad | damn skippy! |
21:37.42 | iMinute | hmm? |
21:37.45 | iMinute | oh |
21:37.47 | iMinute | starfox |
21:37.47 | brlcad | nvr mind :) |
21:37.49 | iMinute | hehe |
21:38.00 | iMinute | :D |
21:38.13 | brlcad | :D |
21:49.36 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-029-010.pools.arcor-ip.net) |
21:51.29 | CIA-27 | BRL-CAD: 03brlcad * 10brlcad/src/util/pix-png.c: warn if the output PNG image dimensions are smaller than the input PIX image, related to support request from mallory |
21:57.20 | CIA-27 | BRL-CAD: 03brlcad * 10brlcad/NEWS: (log message trimmed) |
21:57.22 | CIA-27 | BRL-CAD: pix-png now warns the user if the PNG being created has image dimenions smaller |
21:57.24 | CIA-27 | BRL-CAD: than the PIX file that was given as input. The tool compares the size of the |
21:57.26 | CIA-27 | BRL-CAD: input and output images, and reports the difference if there is one; it also |
21:57.28 | CIA-27 | BRL-CAD: suggest what flags the user may need to use if the input has recognizable image |
21:57.30 | CIA-27 | BRL-CAD: dimensions. This feature was implemented in response to a support request from |
21:57.32 | CIA-27 | BRL-CAD: Mallory where he was confused by the black PNG image produced (which happened to |
22:01.39 | CIA-27 | BRL-CAD: 03brlcad * 10brlcad/BUGS: mged isn't updating the graphics window with the new 8.5 tcl (sometimes) |
22:05.07 | *** join/#brlcad rpaddock (n=bob_padd@c-24-131-109-50.hsd1.oh.comcast.net) |
22:26.50 | *** join/#brlcad MinuteElectron (n=MinuteEl@silentflame/member/pdpc.base.minuteelectron) |
22:44.29 | brlcad | 95k for just under 4TB with support, without discounts |
22:45.50 | IriX64 | brlcad=deep pockets? :) |
22:47.53 | brlcad | heh, no |
22:48.03 | IriX64 | :) |
22:52.13 | starseeker | The unthinkable has happened - the STIX fonts have reached a Beta release |
22:52.32 | starseeker | http://www.stixfonts.org/ |
22:52.34 | brlcad | mm.. $50k for a 12TB decked out Mac |
22:52.52 | starseeker | Drool... |
22:53.15 | starseeker | (not that BRL-CAD is likely to need the STIX fonts, I suppose...) |
22:53.27 | brlcad | starseeker: hmm, interesting |
22:54.02 | *** part/#brlcad ewilhelm (n=ewilhelm@pool-71-111-49-155.ptldor.dsl-w.verizon.net) |
22:54.11 | starseeker | That's roughly the scientific equivalent of Duke Nukem Forever being released ;-) |
22:54.12 | brlcad | that actually is pretty interesting |
22:54.43 | brlcad | new modeler is going to ship with fonts, sounds like stix might be what we'd want to go with |
22:54.46 | starseeker | If you download 'em, watch out for the zip file - it dumps everything into the same directory as the zip file itself is in |
22:54.52 | starseeker | cool :-) |
22:54.54 | brlcad | presuming freetype can read them |
22:55.04 | starseeker | one would hope :-) |
22:55.22 | brlcad | looks like at least their "goal" is TTF's |
22:55.40 | starseeker | Yes - if it doesn't do what you need it's probably a bug they want to hear about |
22:56.17 | brlcad | bah, they seriously need a "screenshots" page ... |
22:56.38 | brlcad | not a glyphs page that requires me to install them |
22:57.39 | starseeker | agreed |
22:58.36 | brlcad | I have a sampling of decent free fonts that I was going to use, but a project dedicated to making good scientific fonts would be even better |
22:59.33 | brlcad | mm.. fun, 7.10.4 Mac install of BRL-CAD is nearly 500MB |
22:59.36 | ``Erik | hum |
22:59.39 | brlcad | granted, that's with debugging symbols |
22:59.57 | ``Erik | "t2" is of the t1000 family? |
23:00.19 | ``Erik | this time, we dont' have to crush it, it's here to save us from the goopy one? O.o :D |
23:01.11 | brlcad | we didn't have to crush it the first time .. someone and someone else just thought it'd be fun to set up and play with |
23:01.26 | louipc | wow my install is just 96M |
23:01.40 | brlcad | louipc: is that stripped? |
23:01.53 | louipc | well it doesn't have tcl... |
23:01.53 | ``Erik | stripped, space optimized and 32 bit? |
23:02.11 | louipc | not stripped but 32 bit yep |
23:02.16 | ``Erik | tcl is a big space consumer, all those extra man pages, tz files, tcl files, ... |
23:02.25 | brlcad | oooh, good point, I was going to make these universal binaries... |
23:02.29 | brlcad | that'll make them 2x |
23:02.34 | ``Erik | um, libtool is fond of stripping on install |
23:02.46 | brlcad | it's not supposed to |
23:02.50 | brlcad | there's a flag for stripping |
23:02.51 | ``Erik | go into like src/fb, type "make install" and see if it's getting the -s |
23:03.02 | ``Erik | I've seen it 'just happen' before... unless I'm completely outta it :D |
23:03.36 | ``Erik | the t5220 would be better on infrastructure than the apple, I think... |
23:03.40 | brlcad | http://bzflag.bz/~sean/tmp/quote.pdf is kinda what I'm thinking of |
23:05.13 | brlcad | it would be better of course, just not sure about it for the price point given with that price diff there could even be <1hr on-site support |
23:05.28 | ``Erik | oh yeah, they saddle up for that support |
23:05.59 | brlcad | i mean the apple could be configured for <1hr for less than the sun |
23:06.09 | ``Erik | but, um, is it to be a file server or a compute server? |
23:06.17 | *** join/#brlcad rpaddock (n=bob_padd@c-24-131-109-50.hsd1.oh.comcast.net) |
23:06.18 | brlcad | the sun is configured for same day iirc, woulda cost more for less |
23:06.31 | ``Erik | <-- not keen on mixing the two, kinda likes the seperation of services |
23:07.12 | brlcad | it's neither and both, needs to be big enough to fit all the data and fast enough to not be annoying |
23:08.01 | ``Erik | that'll give us, what, ~3tb live at raid5? |
23:08.06 | brlcad | which is probably a few TB, really fast disks, not sucky cpus .. bus bandwidths, wide pipes |
23:08.22 | ``Erik | or almost 2 at 0+1 ? |
23:08.28 | brlcad | yeah, about that much |
23:08.44 | ``Erik | are they still windowed big endian chips? |
23:08.47 | ``Erik | or didja look? |
23:08.56 | brlcad | yeah, that was something else I liked |
23:09.08 | brlcad | one of the few big end servers remaining |
23:09.14 | ``Erik | we'd be able to retire the e420 when it goes out of service |
23:09.39 | brlcad | the new "super" threading servers, able to run 64 threads simultaneously without skipping a beat |
23:09.52 | brlcad | that would be *really interesting to test out |
23:10.23 | ``Erik | <-- likes suns cuz he did sysadmin at a place where downtime was measured in millions an hour :) the price was never an issue, so'z is not used to really judging that angle |
23:10.42 | ``Erik | we can always apply for a 60 day test ride |
23:10.56 | ``Erik | heh |
23:11.18 | brlcad | they also undoubtedly paid 7 figures for their infrastructure :P |
23:11.38 | ``Erik | when I lived in a place where the entire mgmt chain cared about getting shit done instead of politics, we had so many $'s in sun, they'd loan us high end (like 4800, 6800) boxes for a year at a whack |
23:12.01 | ``Erik | uhm, more than 7... :) |
23:12.25 | brlcad | meant at least 7 |
23:13.03 | ``Erik | the, uh, maintenance, staff, infrastructure, etc costs for the admin/web/it/softwaredev was in the rough neighborhood of a billion a year |
23:13.37 | brlcad | yeah, that's counting too much (for what is being compared) |
23:13.40 | ``Erik | and it wasn't seen as "productive" from the other companies, so under constant attack to defend those $'s |
23:13.44 | brlcad | the orgs are massively different in size too |
23:13.45 | ``Erik | yeah, tha'ts the only number I know |
23:14.07 | ``Erik | I showed up when I wanted, left when I wanted, more or less did what I wanted, and got paid a lot more than I do here :) |
23:14.38 | brlcad | for any reasonably small dev team, I think it's reasonable to try to stay under 6 figures |
23:14.46 | ``Erik | yeah |
23:14.54 | brlcad | so then .. uh, why'd you leave? :P |
23:15.18 | brlcad | those $$'s were too much apparently ;) |
23:15.38 | ``Erik | was a limited contract to perm position, and bullshit politics lost me on the pickup |
23:16.15 | ``Erik | a cohort who changed teams got the inside scoop from some people, the "you didn't hear this, but" word was that I was seen talking to the wrong people (the guys on my team who quit for google right after I was jobless) |
23:17.17 | brlcad | no strip on install |
23:17.25 | ``Erik | I was in one of those pesky rogue groups that doesn't play buy the corporate rules and outshines everyone with success, and useless politicos really hate that :D |
23:18.07 | brlcad | sounds like how the cad team used to be |
23:18.31 | ``Erik | hrm, should be less than a 4x size bump for just 64b/uni over 32/PF... must be the tickle |
23:18.52 | ``Erik | there were traces of that bubble when I got here :( |
23:18.58 | brlcad | i haven't made the uni yet, just going by estimates I have from other builds |
23:19.22 | brlcad | er, and that should be about 2x .. not 4x |
23:19.30 | brlcad | i expect about a 1GB install |
23:20.03 | ``Erik | hrm, uni should about be 2x... does... wait, you're on an i386 apple, not a g5? |
23:20.13 | ``Erik | does g5 have variable width opcodes? |
23:20.16 | brlcad | 2GB if I make all four platforms (32 ppc, 64 ppc, 32 x86, 64 x86) |
23:20.44 | ``Erik | is there 64b x86 on apple? O.o :D |
23:20.53 | brlcad | oh, you got 4x off what it'd be stripped |
23:21.13 | brlcad | the strip savings is more in tune with just how much symbolage there is |
23:21.19 | ``Erik | unix (2x) * 64b (2x) is what I was thinking |
23:21.33 | ``Erik | er |
23:21.35 | ``Erik | uni |
23:21.50 | ``Erik | heh, damn muscle memory :D I talk about unix way too much, apparently |
23:22.12 | brlcad | yeah, 10.4 has 64bit support and 10.5 has it through and through |
23:22.29 | ``Erik | I didn't think the core duos had that support |
23:22.50 | brlcad | the workstations are quad-core xeons |
23:23.04 | ``Erik | at most, the EMT64 memory hack, not 64b arithmatic and stuff |
23:23.06 | brlcad | laptpos are core 2 duos |
23:23.22 | ``Erik | okie, I haven't looked at asm notions on the footrest yet |
23:23.31 | brlcad | either way, it works -- i've compiled the 4-way uni binaries :) |
23:24.03 | brlcad | and they do kick in on the new workstations if you go 64bit |
23:24.06 | ``Erik | the 4 core opteron fbsd box seems to compile faster, so'z I use it |
23:24.41 | brlcad | yeah, apple has their own optimization loop that does "a little bit better" than gcc's default, but takes way longer |
23:24.57 | brlcad | used to be massive difference |
23:25.16 | ``Erik | that, the short&quick bsd land file ops seem significantly slower (perhaps the mach layer) |
23:25.46 | brlcad | a compile on my old laptop took an hour under mac os x, and 20 minutes on the same hardware running yellow dog linux, same version of gcc |
23:25.53 | ``Erik | and I haven't gotten the tuning to allow bunches of processes to run in parallel, so'z I keep running into fork resource unavaiable crap when I try to od things fast :D |
23:26.16 | brlcad | ah, you have to bump up your limit |
23:26.20 | brlcad | it's not set for devs by default |
23:26.31 | ``Erik | now, if you took, say, a huge single .c file and compiled it with -pipe, what was the difference? :D |
23:26.43 | ``Erik | no, at least, not devs who work in terminal. |
23:27.08 | brlcad | it was way slower even on single file -- that was one of the tests I ran |
23:27.27 | brlcad | i don't remember those exact numbers, but it was more than 2x at O3 |
23:27.48 | ``Erik | and O0? :D |
23:27.50 | brlcad | and after talking to some of the apple devs, they commented on their custom optimization loop |
23:28.39 | brlcad | they got a couple of the old mipspro compiler guys as well, put them to work |
23:29.35 | brlcad | you should check your sysctl kern.maxproc -- it should be > 512 |
23:33.31 | ``Erik | hum |
23:33.41 | ``Erik | I think I made it 4x, and that alleviated it a LITTLE |
23:33.50 | ``Erik | but I think that was still fairly shy of 512 |
23:42.43 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
23:52.02 | ``Erik | mebbe it's maxprocperuid that's got me down |