| 00:08.57 | ``Erik | stupid fucking snow |
| 00:13.15 | Maloeran | So you bought your bicycle already! I know how you feel |
| 00:16.20 | ``Erik | heh |
| 00:16.21 | ``Erik | no |
| 00:16.27 | ``Erik | it went blizzard style while I was driving |
| 00:17.21 | ``Erik | so I was trying to navigate twisty hilly backroads with an inch of fresh powder (unsalted roads), a jackass tailgating me, and summer tires on 1.75 tons of grossly overpowered steel |
| 00:17.40 | ``Erik | oh, and visibility of ~10m at some points |
| 00:18.29 | ``Erik | took me over an hour to get home :/ and the first half our of it was ok, full speed... (usual travel time of 35-40 minutes) |
| 00:18.36 | ``Erik | so that last 5-10 minutes took half an hour or so |
| 00:19.36 | Maloeran | Eheh. That's a long way from work, spending 10% of one's conscious life in a car driving |
| 00:20.03 | ``Erik | yeah... it is |
| 00:20.37 | ``Erik | however; it's pleasantly relaxing, it gives my brain something fairly simple to focus on so it can change gears from the bs of work |
| 00:20.44 | Maloeran | Can you work from home some/most days? |
| 00:20.53 | ``Erik | not unless I filled at forms and got approved |
| 00:21.09 | Maloeran | It could be worth the time investment |
| 00:21.16 | ``Erik | twinky was approved for 3 days a week, I just havne't gotten to the paperwork |
| 00:26.13 | Maloeran | I'm a bit annoyed by how buying 3 dual-dual-opterons is about the same price as one octo-opteron |
| 00:27.42 | Maloeran | The extra circuitry in 8 ways Opterons sure is costly, the motherboards too |
| 00:27.42 | dtidrow_work | octo-opteron meaning eight actual chips, or four dual-core Opterons? |
| 00:27.42 | Maloeran | Four dual-core |
| 00:29.28 | dtidrow_work | depending on the type of workload, having a single quad dual-core system might still be the best way to go |
| 00:29.50 | Maloeran | Yes, clearly. Gigabit will saturate with distributed raytracing on 3 boxes of 4 cores |
| 00:30.12 | dtidrow_work | heh |
| 00:30.51 | dtidrow_work | are there any eight socekt Opteron mobo's out there? |
| 00:31.43 | Maloeran | Yes, haven't really looked into these |
| 00:32.11 | dtidrow_work | http://www.boxxtech.com/Products/APEXX/apexx8_overview.asp - drool.... |
| 00:34.31 | Maloeran | To scale properly on such boxes, we will have to store a copy of the graph in the memory bank of each processor die |
| 00:35.04 | Maloeran | Graph & geometry, of course |
| 00:35.50 | Maloeran | Any recommendation, Erik, (2or3)*dual-dual or quad-dual? |
| 00:36.09 | Twingy | numaPIC |
| 00:36.35 | Maloeran | Yes! Highly scalable processing for PIC |
| 00:38.07 | Maloeran | Perhaps a motherboard with multiple gigabit ports would almost scale, one per processing node |
| 00:39.57 | Maloeran | Infiniband cards are somewhat expensive, but that would be fun to play with |
| 00:47.55 | Twingy | you should see if mellanox has a hardware implementation of tcp/ip to allow for > 1.1Gb on their infiniband |
| 00:48.29 | Twingy | they mentioned they might be working on a next gen card in 2004 |
| 00:49.00 | Twingy | Maloeran, you can always buy a 10Gb PCI Express NIC for alot less |
| 00:49.32 | Maloeran | Can't get a price on these without contacting them, tsk |
| 00:49.54 | Maloeran | That generally means "out of my budget" :) |
| 00:51.00 | Twingy | $750 a card |
| 00:51.11 | Twingy | $3k for an 8 port switch |
| 00:51.19 | Maloeran | That's about what I saw elsewhere |
| 00:52.08 | Twingy | a PCI Express to PCI Express card would be useful |
| 00:52.15 | Twingy | with a ribbon cable |
| 00:52.40 | Twingy | writing a driver for that would be a snap |
| 00:53.19 | dtidrow_work | interesting idea |
| 00:53.53 | Twingy | I could probably make one on my CNC mill |
| 00:54.08 | Twingy | just a double sided PCB really |
| 00:54.50 | Twingy | won't get much faster than that |
| 00:55.12 | Maloeran | I think I'll just start slowly with two dual-dual-opterons on gigabit, I'll grow my home cluster over the months |
| 00:55.48 | Twingy | you should take one class at a university |
| 00:56.00 | Twingy | then you will get university account and access to infinite computing resources |
| 00:56.29 | Twingy | I have a bazillion computers at my finger tips at njit |
| 00:57.16 | Maloeran | I see :) |
| 00:57.33 | Twingy | http://images.google.com/imgres?imgurl=http://www.dansdata.com/images/cmoy/dcin280.jpg&imgrefurl=http://www.dansdata.com/cmoy.htm&h=174&w=280&sz=15&hl=en&start=3&tbnid=U0FzNIGLlOw2EM:&tbnh=71&tbnw=114&prev=/images%3Fq%3DDC%2BInput%2BJack%26svnum%3D10%26hl%3Den%26lr%3D%26sa%3DN |
| 00:59.03 | Twingy | every time I eat macaroni and cheese I get gas, hrmph |
| 01:00.39 | dtidrow_work | Twingy: tinyurl.com - for future reference ;-) |
| 01:01.37 | Twingy | what's a matter your web browser 8-bit or something? |
| 01:02.00 | dtidrow_work | heh |
| 01:02.19 | dtidrow_work | just more convienent |
| 01:06.59 | Twingy | Maloeran, you could install two gigabit cards and expand the socket buffer in the kernel |
| 01:07.10 | Twingy | then plex across the two |
| 01:07.35 | Twingy | that should scale almost linearly |
| 01:08.04 | Maloeran | Right. I don't think the bandwidth would saturate if high-level results ( pixels or so ) are transfered |
| 01:08.21 | Twingy | adrt run smooth on 100Mb |
| 01:08.29 | Twingy | across 8 nodes |
| 01:08.32 | Maloeran | Pixels or raw intersection results? |
| 01:08.40 | Twingy | compressed pixels |
| 01:08.52 | Twingy | on gigabit I don't compress |
| 01:09.24 | Maloeran | Ah. 1024*768*3, 2.5mb per frame, 40 frames per second for 100mb/s. -> Gigabit is almost filled up |
| 01:09.42 | Twingy | tried compression? |
| 01:10.03 | Twingy | I had decent success with libz |
| 01:10.19 | Maloeran | Yes I noticed |
| 01:10.34 | Maloeran | I haven't used compression, I still got stuff to complete in graph state synchronisation between nodes |
| 01:11.07 | Maloeran | I'm just saying that theorically, 3 boxes of 4 cores would saturate the gigabit |
| 01:11.15 | Twingy | I'm so happy, my shrink tube came today |
| 01:11.25 | Maloeran | Shrink tube? |
| 01:11.28 | Twingy | I've been making cables galore, this triton Jr rocks my socks off |
| 01:11.37 | Twingy | yes, it shrinks when you heat it |
| 01:11.47 | Twingy | I have a 1500W heat gun I bought for $19.95 somewhere |
| 01:11.59 | Twingy | the triton outputs banana |
| 01:12.15 | Twingy | so I made a banana to j-type, banana to banana, banana to dc jack, and banana to deans is in the mail |
| 01:12.56 | Maloeran | Is the final objective related to rocketry, robotics? |
| 01:12.58 | Twingy | I can charge lithium polymer, lead acid, nickel metal hydrides, and nickel cadmiums |
| 01:13.07 | Twingy | both actually |
| 01:13.13 | Maloeran | Ah, neat |
| 01:13.25 | Twingy | an aerial platform for testing my rocketry electronics |
| 01:13.42 | Twingy | I just burnt $650 on getting back into r/c planes |
| 01:13.51 | Twingy | I'm all tooled up |
| 01:13.53 | Maloeran | Eheh, I saw the picture yes |
| 01:14.20 | Twingy | I'm going to convert half the basement (the unfinished part) into my rocket and r/c workshop |
| 01:14.36 | Twingy | and leave the garage for making parts and boards |
| 01:15.05 | Twingy | until I build my shed out back |
| 01:15.09 | Maloeran | I see, sounds good. I almost assumed you had enough room in the garage |
| 01:15.12 | Twingy | either this summer or next |
| 01:15.26 | Twingy | I don't have the garage temperature controlled |
| 01:15.31 | Twingy | (yet) |
| 01:15.33 | Maloeran | Ah yes, good point |
| 01:15.42 | Twingy | soldering in sub zero temperatures is tricky |
| 01:15.47 | Maloeran | ;) |
| 01:15.51 | Twingy | and I'm not heating the garage while it's not insulated |
| 01:16.30 | Twingy | 3 more gcam releases and it's back to rocketry |
| 01:16.46 | Twingy | next one should be tonight or tomorrow |
| 01:17.34 | Maloeran | I'm vaguely planning to build some kind of home cluster over the months/years, I think I'll need that when I'm ready to get back into AI |
| 01:19.15 | Twingy | buying computers sucks you dry >_< |
| 01:19.43 | Twingy | nice thing about the tools I'm buying is they won't be obsolete for a good 15 - 25 years |
| 01:20.20 | Twingy | all I'm missing is the $2500 lathe |
| 01:20.27 | Maloeran | Eheh, indeed. Unfortunately, I have no other need or interest than computer hardware |
| 01:21.01 | dtidrow_work | unless you pick up used computers |
| 01:21.18 | Twingy | hard to build next gen algorithms on obsolete hardware |
| 01:21.20 | Maloeran | I already got my cluster of amd-k6 and Pentium 2, I need an upgrade ;) |
| 01:21.30 | dtidrow_work | heh |
| 01:21.47 | Twingy | especially "timing" related ones |
| 01:22.51 | dtidrow_work | well, for 'home' clusers... |
| 01:23.16 | dtidrow_work | you generally take what you can get :-) |
| 01:23.31 | Twingy | I speant almost two hours writing this yesterday: *(int *) &_f = (t&0xff000000)>>24|(t&0x00ff0000)>>8|(t&0x00007f00)<<8|(t&0x000000ff)<<23|(t&0x00008000)<<16; } |
| 01:23.41 | Maloeran | That's what I did.. but my overclocked desktop beats the other 6 boxes combined |
| 01:23.49 | dtidrow_work | heh |
| 01:24.14 | Twingy | converts big endian PIC floats to little endian IEEE 754 floats |
| 01:24.15 | Maloeran | Twingy, a weird partial byte swapping? |
| 01:24.22 | Maloeran | Oh I see |
| 01:24.30 | dtidrow_work | PIC floats? |
| 01:24.33 | Twingy | for the autopilot |
| 01:24.39 | Twingy | microchip floats |
| 01:24.45 | Twingy | exponential, sign, mantissa |
| 01:24.47 | dtidrow_work | ah |
| 01:24.54 | Twingy | where IEEE754 is mantissa, exponent, sign |
| 01:24.55 | dtidrow_work | that PIC :-) |
| 01:25.45 | dtidrow_work | haven't played with those in years (PICs, that is) |
| 01:25.56 | Twingy | I use them religiously at work now |
| 01:26.58 | dtidrow_work | IIRC, they have an 'interesting' machine architecture |
| 01:27.10 | dtidrow_work | Twingy: for what? |
| 01:28.07 | Twingy | all sorts of stuff |
| 01:28.21 | Twingy | autopilots mainly |
| 02:06.22 | ``Erik | <-- probably woulda loaned ya some of the junk ya may've needed for r/c's... |
| 02:24.11 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/TODO: |
| 02:24.11 | CIA-5 | BRL-CAD: add support for arbitrary matrix transformations to torus primitive .. while |
| 02:24.11 | CIA-5 | BRL-CAD: there's certainly question on what this means to the mathematics of the implicit |
| 02:24.11 | CIA-5 | BRL-CAD: form (they're no longer mathematically a torus), the underlying modeling |
| 02:24.11 | CIA-5 | BRL-CAD: capacity is still needed and storing a transformation matrix or at least scaling |
| 02:24.13 | CIA-5 | BRL-CAD: vectors in the primitive (or above the primitive) would go a long way to |
| 02:24.15 | CIA-5 | BRL-CAD: handling this somewhat unique case. |
| 02:53.27 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168051616.dsl.bell.ca) |
| 03:03.12 | *** join/#brlcad IriX64_ (n=IriX64@bas3-sudbury98-1168051616.dsl.bell.ca) |
| 03:07.26 | ``Erik | wee, squashed and skewed torii |
| 03:09.18 | ``Erik | kith style, crrrushink your torus! crrrushink your torus! |
| 03:11.35 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libbn/ (axis.c font.c symbol.c): simplify, use MAT_DELTAS_VEC when setting translation elements to a given vector |
| 03:22.14 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/librt/dg_obj.c: simplify, use MAT_DELTAS_VEC_NEG when setting translation elements to negative vector values |
| 03:30.14 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/vmath.h: add additional helper macros MAT_DELTAS_(ADD|SUB|MUL)(_VEC)? for modifying the translation elements (3, 7, 11) in a 4x4 transformation matrix |
| 03:30.41 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/ (util/pl-dm.c util/plrot.c rt/do.c rt/read-rtlog.c): simplify, use MAT_DELTAS_VEC_NEG when setting translation elements to negative vector values |
| 03:31.18 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/proc-db/ (common.c kurt.c tube.c): simplify, use MAT_DELTAS_VEC when setting translation elements to a given vector |
| 03:31.55 | brlcad | mm.. pure donuts |
| 03:34.12 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/mged/ (chgmodel.c edsol.c): simplify, use MAT_DELTAS_VEC when setting translation elements to a given vector |
| 03:34.15 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/mged/ (rtif.c usepen.c): simplify, use MAT_DELTAS_VEC_NEG when setting translation elements to negative vector values |
| 03:36.11 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/vmath.h: s/NUL/MUL/ typo |
| 03:44.38 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/vmath.h: add similar MAT_SCALE sibling macros for adding, subtracting, and multiplying the scaling elements just to be consistent |
| 03:48.55 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/vmath.h: elements_per_mat is a square of elements_per_plane |
| 03:50.05 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168051616.dsl.bell.ca) |
| 04:11.24 | *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net) |
| 05:15.17 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/tclscripts/ (14 files in 14 dirs): |
| 05:15.17 | CIA-5 | BRL-CAD: since the pkgIndex.tcl and tclIndex files are now in CVS, disable the CLEANFILES |
| 05:15.17 | CIA-5 | BRL-CAD: directive that deletes them. now that they're sorted.. if they're different, |
| 05:15.17 | CIA-5 | BRL-CAD: the cvs update notice should be significant now that they're sorted |
| 05:15.17 | CIA-5 | BRL-CAD: consistently. |
| 05:29.20 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/librt/ (30 files): allow the import function transformation matrix argument to be NULL, use an identity matrix in that situation. |
| 05:30.21 | *** join/#brlcad brlcad (n=sean@bz.bzflag.bz) |
| 05:30.32 | *** mode/#brlcad [+o brlcad] by ChanServ |
| 07:02.18 | *** join/#brlcad clock_ (i=clock@84-72-63-118.dclient.hispeed.ch) |
| 07:50.11 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/sh/ (copyright.sh indent.sh): sanity checking - don't care if configure.ac is readable, just see if it exists |
| 08:34.00 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 08:49.55 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/sh/indent.sh: skipping |
| 09:01.50 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/sh/indent.sh: sorta support filenames with spaces (the for loop will still choke) |
| 09:39.33 | *** join/#brlcad cad53 (n=c1312aab@bz.bzflag.bz) |
| 09:39.58 | *** part/#brlcad cad53 (n=c1312aab@bz.bzflag.bz) |
| 09:47.32 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 13:23.33 | *** join/#brlcad docelic (n=docelic@212.15.184.11) |
| 13:31.01 | *** join/#brlcad docelic (n=docelic@212.15.172.177) |
| 14:34.05 | ``Erik | brlcad: lunch... japanhouse... the usual time... |
| 15:16.25 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 15:53.03 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 16:03.50 | brlcad | hm, that means I'd need to shower |
| 16:06.35 | ``Erik | dot... dot... dot... |
| 16:20.45 | dtidrow | heh |
| 16:20.56 | dtidrow | just getting around to too, eh? |
| 18:21.04 | ``Erik | seems he decided against shorting |
| 18:21.06 | ``Erik | showering |
| 18:21.17 | ``Erik | osake was sugoii ii desu ne |
| 18:29.10 | Maloeran | Tout à fait d'accord |
| 18:44.47 | ``Erik | heh, damn I'm lame, I had to babelfish that one |
| 18:45.33 | ``Erik | (but... you're in agreement? I didn't think you drank? so how would you know if sake is very good?) |
| 18:46.55 | Maloeran | Oh :), just provided an unintelligible answer to a question with the same attribute |
| 18:47.22 | ``Erik | les gallons de saké traversent la veine |
| 18:47.23 | Maloeran | Feel free to keep me informed on any interface talk |
| 18:47.43 | ``Erik | lee gave a thumbs up, I told him where to snarf it |
| 18:48.00 | Maloeran | What did he approve exactly? |
| 18:49.10 | ``Erik | full access, and I told him the name of the person he should talk to at the place the cvs repo machine is, as well as your handle and this channel |
| 18:49.18 | ``Erik | ummm, his nick was like "joevalley" or something I think? |
| 18:49.32 | Maloeran | I don't remember seeing that |
| 18:50.10 | ``Erik | ok, the grammar and ordering is all out of whack for sure on this, but think english order/grammar... |
| 18:50.11 | ``Erik | Je vous considère de devoir commettre votre travail plus souvent |
| 18:50.25 | Maloeran | Understood :) |
| 18:50.43 | Maloeran | I don't like commiting half-way done code |
| 18:50.48 | ``Erik | out of curiousity, how would the well formed sentence be structured? :D |
| 18:51.13 | ``Erik | well, 'half done' is a necessicity, just try not to compile in a "broken" state |
| 18:51.58 | Maloeran | But theres's nothing new to see or test if the new code isn't completed |
| 18:52.30 | Maloeran | Though the bit-packed graph cache is a bit better, not a single bit wasted now |
| 18:52.33 | ``Erik | nothing at the user level, of course, but I tend to read your commits and try to understand what and why |
| 18:52.39 | ``Erik | and where you're going |
| 18:53.25 | ``Erik | and I'd rather read a book made of sections, chapters and paragraphs than a single long stream :D |
| 18:53.49 | ``Erik | so my poor feeble brain has time to grasp the meaning before moving on |
| 18:54.03 | Maloeran | Okay :), give me a day or two and I'll commit |
| 18:54.27 | Maloeran | One of the reasons I don't like commiting early is that there are big blocks of personal notes, comments that would make no sense to other people, right in the code |
| 18:54.36 | Maloeran | Sometimes using a mix of french and english |
| 18:54.51 | ``Erik | <-- works frequent commits into his development mentality, so not seeing it is alien |
| 18:54.57 | ``Erik | but I'm a dork like that |
| 18:55.49 | ``Erik | I don't care if it's in french while you're working on it, as long as teh end product is coherent to me... :D if nothing else, when people want to know how progress is going, I can at least say "it's progressing" if I see commit traffic |
| 18:55.52 | Maloeran | It's just that I put my thinking process as raw comments in the code, so it's heratic ; and the comments take the way out when the code is done |
| 18:56.34 | Maloeran | Okay, 1-2 days and I'll commit |
| 18:56.53 | ``Erik | heh, if you codified your thinking process via repository history, then others can look at the history and gain some understanding into your process and why you made your decisions, no? :D |
| 18:57.27 | Maloeran | That would imply taking time to write comments that could make sense to anyone else... :) I don't understand my own raw comments after a week |
| 18:58.07 | ``Erik | hehehe, millions of postit notes with random sentence fragments? :D |
| 18:58.21 | Maloeran | Yes, that's exactly what it is :), except that it's in the code |
| 18:58.37 | ``Erik | <-- still interested in the deltas, and it does no harm to 'back up' your work frequently, no? |
| 18:58.43 | Maloeran | All right though, I'll try to do more than a large monthly commit |
| 18:59.13 | ``Erik | I mean, hell, you see the cia messages here, we commit stuff to brl-cad all the time... |
| 18:59.24 | ``Erik | and in my case, a good portion of it is only half thought out ;) |
| 18:59.36 | Maloeran | Very small commits, yes, but there are a lot of people working on it |
| 18:59.42 | Maloeran | Ah :) |
| 19:00.08 | ``Erik | <-- commits just as frequently on his personal projects where he's the only developer |
| 19:00.09 | Maloeran | Do I also commit temporary code for personal testing purposes? |
| 19:00.16 | ``Erik | often in smaller commits, heh |
| 19:00.22 | ``Erik | sure, why not? |
| 19:00.33 | Maloeran | *shakes head* Okay |
| 19:00.47 | ``Erik | it's a way of recording history, communicating, and backing up your data |
| 19:01.25 | Maloeran | Stuff will be commited for Monday when the guy looks at it |
| 19:01.55 | ``Erik | maybe you'll be knee deep in something you're trying to figure out and a commit would let someone see where your head is and maybe ask golden questions that help you, or offer possible alternatives? |
| 19:02.36 | ``Erik | it's your code and you get to decide, but I'm interested and want to keep up to speed on it :D (and if I'm being unreasonable, someone please, ffs, tell me) |
| 19:02.37 | Maloeran | So you want to participate in the code? :) I hadn't assumed so, I'll change my approach then |
| 19:03.02 | Maloeran | Oh, you are not being unreasonable at all. Look Erik, I have always coded alone |
| 19:03.03 | ``Erik | I'm already half participating out on a fringe, but in using an old snapshot |
| 19:03.36 | ``Erik | binding rayforce and adrt into a benchmark program is my current top priority non-emergency task |
| 19:03.51 | Maloeran | Great |
| 19:03.59 | ``Erik | (now I'm focusing very much on tie at the moment, as I know my copy of rayforce is outdated) |
| 19:05.48 | Maloeran | I'll shower and eat breakfast before it closes at 15h, thanks for your guidance |
| 19:06.10 | ``Erik | later, dude :D |
| 19:12.01 | ``Erik | nice http://home.earthlink.net/~krautj/sassy/sassy.html |
| 19:16.02 | archivist | bah too many brackets |
| 19:20.36 | ``Erik | func1(func2(func3())) vs (func1 (func2 (func3))), the second has too many brackets but the first is ok? :D |
| 19:20.52 | ``Erik | or, say, (func1 {func2 [funk3]}) ? |
| 19:21.32 | archivist | I dont expect any brackets in plain assembler |
| 19:24.32 | ``Erik | (lisp or scheme style) macros would be insanely awesome in an asm type thing |
| 19:25.38 | ``Erik | makes me wanna do os work that much more :D |
| 19:26.42 | archivist | ew |
| 19:27.13 | ``Erik | <-- wants a modern lispos |
| 19:27.44 | ``Erik | perhaps because I ate too much wallcandy as a child :D |
| 19:28.26 | brlcad | ~seen joevalleyfield |
| 19:29.05 | ibot | joevalleyfield <n=joevalle@bz.bzflag.bz> was last seen on IRC in channel #brlcad, 371d 1h 7m 36s ago, saying: 'may i buy a vowel?'. |
| 19:31.22 | brlcad | frequent small commits are a "good thing" |
| 20:19.33 | Maloeran | I think Lisp-style macros would be really bad in assembly |
| 20:20.03 | Maloeran | Even if you would repeat the same macro twice in high level code, the assembly code should probably be different for each instance |
| 20:38.51 | ``Erik | erm... yeah... uh... that's why I said lisp style instead of c style... :D |
| 20:45.59 | ``Erik | qjkl |
| 21:30.28 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168051616.dsl.bell.ca) |
| 21:32.13 | *** part/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168051616.dsl.bell.ca) |
| 21:55.34 | brlcad | curious, how do most feel about embedded tabs? e.g., int foo;[tab][tab]/* some comment */ |
| 21:58.03 | brlcad | at least doesn't answer the question |
| 21:59.10 | Maloeran | If not for comments besides code, I don't see the point of embedded tabs :). In other words, I wouldn't be too fond of these |
| 22:00.08 | brlcad | might also see them in comments for example to line something up perhaps |
| 22:00.25 | brlcad | <PROTECTED> |
| 22:00.39 | brlcad | <PROTECTED> |
| 22:00.44 | brlcad | <PROTECTED> |
| 22:00.47 | brlcad | <PROTECTED> |
| 22:02.03 | brlcad | personally, I've found them to be more annoying than helpful simply because some editors are thrown off by them |
| 22:02.34 | Maloeran | Yes, I'm not too fond of tabs generally. If you have seen my code, I use space everywhere |
| 22:03.51 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168051616.dsl.bell.ca) |
| 22:05.11 | brlcad | it's not so much preference as it is collaborative impact that I'm concerned with .. gut feeling is that it's probably not a big deal either way and the few bytes that are saved probably don't matter much |
| 22:06.51 | brlcad | Maloeran: I believe you perhaps mean to say you can't stand indentation levels of 8 .. rather different statement |
| 22:07.27 | brlcad | the only portable tabwidth is 8, indentation and use of tab characters at intervals of 8 are usually the question |
| 22:07.28 | ``Erik | your embedding is just indenting after the code... I might have a line that says \t\t\tint i;\t\t/* i is the index. nehe taught me commenting */\n |
| 22:08.44 | ``Erik | <-- tabhappy mofo :D but ain't religious on it |
| 22:10.09 | brlcad | use of tabs or not isn't really a concern, it's whether this mass update I'm about to commit that turns all embedded ones into tabs or not is what's at hand |
| 22:10.50 | brlcad | whitespace at the beginning of lines is already commented on in hacking, we have standard practice on that and footers to encourage/enforce it |
| 22:11.05 | brlcad | but embedded is another question |
| 22:11.12 | ``Erik | as long as they linux up with the ex: command in a comment or 8 if none exists, I'm all down with it :D |
| 22:11.18 | ``Erik | erm |
| 22:11.21 | ``Erik | linux??? line |
| 22:12.50 | brlcad | linux? |
| 22:13.06 | brlcad | ahh |
| 22:13.31 | brlcad | yeah, embedded is pretty much orthogonal to ex: |
| 22:14.55 | ``Erik | wait, just -*- option -*- ? :D |
| 22:15.34 | brlcad | we don't use emacs header lines |
| 22:15.48 | brlcad | footer variable blocks |
| 22:15.49 | ``Erik | I thought I saw 'em somewhere in the src |
| 22:15.54 | ``Erik | oh |
| 22:16.04 | ``Erik | <-- not an emacs guy... needs to learn |
| 22:16.35 | brlcad | header lines have funky rules about having to be within the first few lines of the file (similar to vi's need to be near the end) |
| 22:17.02 | brlcad | the emacs variable blocks can be anywhere, so it's more convenient to bundle them with the vi line |
| 22:17.31 | brlcad | then all your formatting junk is out of sight most of the time and respected by the two main editors |
| 22:19.50 | ``Erik | aaanyhoo, was just throwing out that it'd be interesting to see how many files are delinquent wrt |
| 22:21.50 | brlcad | oh, most are atm |
| 22:22.02 | brlcad | i've only run the indent script on a few directories |
| 22:22.32 | brlcad | there's preprocessor gunk that screws things up that has to be manually/slowly fixed, otherwise I'd let it run on all files |
| 22:29.55 | louipc | making the code more readable? |
| 22:30.22 | brlcad | readable depends on the reader |
| 22:30.29 | brlcad | but more consistent, yes ;) |
| 22:33.51 | louipc | good stuff. Do you need help? |
| 22:35.11 | brlcad | heh, always |
| 22:35.25 | ``Erik | hey |
| 22:35.28 | ``Erik | lets kill vdeck |
| 22:35.28 | louipc | because that sounds like something I could do :D |
| 22:35.41 | brlcad | with over a million lines of code, there's always "something" inconsistent that can be cleaned up |
| 22:35.48 | brlcad | ``Erik: what for? |
| 22:36.12 | ``Erik | why not? I'd hope no one uses gift anymore :D |
| 22:36.16 | brlcad | if it's causing major problems sure, but if it's not.. leave it alone :) |
| 22:36.25 | ``Erik | aw, c'mon, I wanna kill it |
| 22:36.28 | ``Erik | hehehe |
| 22:36.36 | brlcad | actually, I had someone ask for gift stuff just a few weeks ago |
| 22:36.41 | ``Erik | crazy |
| 22:36.46 | brlcad | it is |
| 22:36.48 | Twingy | go shoot them then |
| 22:36.56 | ``Erik | brl-cad is pretty big, a little pruning might be beneficial |
| 22:37.15 | Twingy | ``Erik, not until it gets a flight simulator built in |
| 22:37.26 | ``Erik | heh |
| 22:37.30 | ``Erik | um... hmmm |
| 22:37.33 | louipc | hehe |
| 22:38.08 | brlcad | getting rid of existing functionality that has no/little overhead is of minimal value, and more than likely just would irritate the handful of people that would miss it |
| 22:38.25 | brlcad | more benficial would be to fix the bigger problems.. |
| 22:38.40 | dtidrow_work | somebody would notice and whine bitterly, eh? |
| 22:38.49 | ``Erik | meh *shrug* |
| 22:39.36 | brlcad | just about every time |
| 22:39.38 | brlcad | within six months almost guaranteed |
| 22:40.00 | brlcad | and it because fuel to a fire |
| 22:40.05 | brlcad | s/because/becomes/ |
| 22:41.37 | louipc | is there a guide to how the code should be formatted? |
| 22:42.42 | brlcad | louipc: HACKING file has the details |
| 22:44.00 | brlcad | pretty standard stuff, should resemble linux kernel with 4 char indents mostly |
| 22:44.41 | brlcad | i.e. GNU coding standards with basic K&R indentation style |
| 22:45.40 | louipc | hmm I'll have to read up on the specifics of those |
| 22:46.00 | brlcad | there are examples in the hacking file, wikipedia has a simple coverage as well |
| 22:48.25 | brlcad | there is a script in cvs that will actually utilize emacs to automatically format up sources for you according to the style, and it does a great job |
| 22:48.53 | brlcad | but it does get confused by some preprocessor logic that changes curly brace indentation |
| 22:49.05 | brlcad | e.g. #ifdef __whatever__ |
| 22:49.15 | brlcad | <PROTECTED> |
| 22:49.18 | brlcad | #else |
| 22:49.26 | brlcad | <PROTECTED> |
| 22:49.27 | louipc | yeah hehe |
| 22:49.28 | brlcad | #endif |
| 22:49.36 | brlcad | messes up on the double indent |
| 22:50.01 | brlcad | simple to accommodate, just remove the indent or remove the ifdef even better if possible |
| 23:20.54 | *** join/#brlcad ibot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 23:32.57 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/sh/ (19 files): ws |
| 23:38.33 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/sh/ws.sh: (log message trimmed) |
| 23:38.33 | CIA-5 | BRL-CAD: This script updates a given set of files (or most all files in this |
| 23:38.33 | CIA-5 | BRL-CAD: project) with consistent whitespace formatting. The script can be run |
| 23:38.33 | CIA-5 | BRL-CAD: on individual files or in batch mode, and is intended to be run |
| 23:38.34 | CIA-5 | BRL-CAD: periodically to help ensure consistent formatting. Currently, the |
| 23:38.34 | CIA-5 | BRL-CAD: following actions are performed by default (but can be selected |
| 23:38.38 | CIA-5 | BRL-CAD: individually at run-time): |
| 23:46.53 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/sh/Makefile.am: include ws.sh |
