| --- Log opened Fri Aug 19 00:00:22 2011 |
| 00:00 | < RichardBarrell> | Oh, Python, no wonder. |
| 00:01 | < Derakon> | Yeah, I didn't really bother optimizing it either. |
| 00:01 | < RichardBarrell> | Don't bother multithreading your program, whatever you do. :) |
| 00:01 | < Derakon> | :p |
| 00:01 | < Derakon> | Silly old GIL. |
| 00:01 | < RichardBarrell> | Depending on exactly what it is that you do per-pixel, you might see crazy speedups for modest amounts of effort by using Cython. |
| 00:02 | < RichardBarrell> | (Or less-crazy speedups for much less effort, just by using pysco) |
| 00:02 | < Derakon> | Yeah. Also depending on the image being generated, pre-calculating its angle and distance with respect to the center of the image, since that's re-used a lot. |
| 00:03 | < Derakon> | I doubt I'm going to pursue this much further, though; it was something to keep me busy while traveling, and now I'm back home I have more than enough stuff to do. |
| 00:04 | < RichardBarrell> | Neat, anyway. |
| 00:04 | < Derakon> | Thanks. |
| 00:09 | | shade_of_cpux is now known as cpux |
| 00:44 | | Thaqui [Thaqui@27B34E.D54D49.F53FA1.6A113C] has joined #code |
| 01:33 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has quit [Ping timeout: 121 seconds] |
| 02:03 | | RichardBarrell [mycatverbs@Nightstar-3b2c2db2.bethere.co.uk] has quit [Client closed the connection] |
| 02:17 | | SmithKurosaki [smith@Nightstar-87f33f96.cpe.mountaincable.net] has quit [Ping timeout: 121 seconds] |
| 02:18 | | Reivles [orthia@Nightstar-75c3c27b.lightwire.co.nz] has joined #code |
| 02:29 | | gnolam [lenin@Nightstar-202a5047.priv.bahnhof.se] has quit [[NS] Quit: Z?] |
| 03:33 | | cpux is now known as shade_of_cpux |
| 04:06 | | Kindamoody[zZz] is now known as Kindamoody |
| 04:14 | < Derakon> | Argh, how do I tell git to overwrite my current copy with the remote copy? |
| 04:14 | < Derakon> | I keep running into this problem and the solution is typically horribly obtuse or "blow away your repo and create it anew". |
| 04:16 | < Derakon> | (Also, why the fuck do you use "git checkout foo" to revert an uncommitted change? Instead of, say, "revert" or "undo" or "forget" or something that actually makes some kind of mnemonic sense?) |
| 04:26 | <@Tamber> | 'git checkout -f; git pull'? |
| 04:27 | <@McMartin> | Because you're checking it out of the repository and into your working copy, wiping out whatever was there. |
| 04:27 | <@McMartin> | And because we are total sociopaths immune to context or intent; the mechanics are all that matter~ |
| 04:29 | < Derakon> | Heh. |
| 05:25 | < ToxicFrog> | Derakon: well, "revert" is "create a new commit that reverts the effects of a previous one", and I can't think of a better name for that. |
| 05:25 | < ToxicFrog> | And yeah, "checkout" is consistently "overwrite stuff in your working copy with stuff from the repository" |
| 05:26 | < ToxicFrog> | Also, the answer to this depends on exactly what you mean. |
| 05:26 | < ToxicFrog> | Is it: |
| 05:26 | < ToxicFrog> | - you have local uncommitted changes that you want to discard |
| 05:26 | < Derakon> | That. |
| 05:26 | < Derakon> | "Restore this file to the state it had as of the most recent commit". |
| 05:27 | < ToxicFrog> | git reset --hard HEAD or git checkout HEAD to do the entire repo, then; git checkout HEAD -- foo bar baz to do just files foo, bar and baz. |
| 05:27 | < ToxicFrog> | If you also want to delete files that aren't versioned, check out 'git clean'. |
| 05:28 | < ToxicFrog> | And finally, "git stash" will reset the files, but also store the changes you made in case you want to restore them later. |
| 05:30 | < Derakon> | I think my main beef here is that "checkout" is the wrong verb. |
| 05:30 | < ToxicFrog> | Why? |
| 05:30 | < Derakon> | Mercurial uses "revert" for "forget uncommitted changes, or, with a -r flag, go back to a previous commit". |
| 05:30 | < Derakon> | (i.e. there's an implicit "-r HEAD") |
| 05:31 | < ToxicFrog> | What does it use for what git calls revert, then? |
| 05:31 | | * Derakon hits "git help revert". |
| 05:31 | < Derakon> | AFAICT this undoes a commit? |
| 05:31 | < ToxicFrog> | <ToxicFrog> Derakon: well, "revert" is "create a new commit that reverts the effects of a previous one", and I can't think of a better name for that. |
| 05:32 | < Derakon> | I don't know off the top of my head if there's a Mercurial verb that does that in one step; I assume so but it's simply not a functionality that I ever need. |
| 05:32 | < Derakon> | Well, need all that often. |
| 05:32 | < Derakon> | When I do need it, "hg revert -r commitBeforeBadCommit; hg commit" does the trick. |
| 05:33 | < ToxicFrog> | ...wouldn't that also revert all subsequent commits? |
| 05:33 | < Derakon> | Ah, I see. |
| 05:33 | < Derakon> | Again, this is not functionality that I need. |
| 05:33 | < Derakon> | Whereas I want to forget pending changes, or take a look at a previous version, all the time. |
| 05:33 | < ToxicFrog> | I can see that; my point is that since it is functionality git has, it needs a name. |
| 05:33 | < Derakon> | Thus it has much higher namespace priority than the current "revert" verb does, in my mind. |
| 05:33 | | Alek is now known as Cloud |
| 05:34 | < Derakon> | I would call git's current "revert" "undo". |
| 05:34 | | Cloud is now known as Alek |
| 05:34 | < ToxicFrog> | I also don't really see why calling it checkout is "wrong" |
| 05:34 | < Derakon> | Note also that "revert" is commonly used in other programs for "revert to most recently saved version". |
| 05:34 | < Derakon> | Checkout to me means "get a new copy of X that you didn't have before". |
| 05:34 | < ToxicFrog> | Why? |
| 05:35 | < Derakon> | Because that's how it's used in most contexts? |
| 05:35 | < Derakon> | Just like "revert" is usually used to mean "restore to the state as of most recent save". |
| 05:36 | < ToxicFrog> | Ok, part of the disconnect here may be that git is the first tool I've used extensively that has a "checkout" command by that name |
| 05:36 | < ToxicFrog> | So I have no preconceptions there |
| 05:36 | < Derakon> | I'm not just thinking about checkout "commands", but also of the word "checkout" as a general verb. |
| 05:36 | < Derakon> | E.g. checking out a book from a library. |
| 05:36 | < Derakon> | Or clicking "checkout" at an online store. |
| 05:37 | < Derakon> | In my mind, using "checkout" to mean "overwrite this file with a fresh copy from the repository" is stupid because I already have the file. |
| 05:37 | < Derakon> | "checkout" would work better as "grab the remote version of this file from a different repository". |
| 05:37 | < ToxicFrog> | ..but you don't. You have a different file that happens to have the same name. Otherwise checkout would be a no-op. |
| 05:38 | < ToxicFrog> | And, uh, doesn't that have the same issue if you already have a modified local copy? I don't see why it matters where the source repo is stored. |
| 05:38 | < Derakon> | It matters because of my expectations based on the use of the word "checkout". |
| 05:38 | < Derakon> | Git does not exist in a void. It can't just blindly choose words and assign them meanings. |
| 05:38 | < Derakon> | (Which I'm not saying it's done, just that it's done an astoundingly bad job of it in this specific case) |
| 05:39 | | Stalker [Z@Nightstar-3602cf5a.cust.comxnet.dk] has quit [Ping timeout: 121 seconds] |
| 05:39 | < Derakon> | Going back a bit, I do have that file already. The fact that I've changed it means, in the most pedantic possible sense, that it's now "a different file with the same name", but it is still the same file. |
| 05:39 | < Derakon> | "checkout" should be a no-op once I've acquired the file. |
| 05:39 | < Derakon> | That's why using the word "checkout" as the verb here makes no sense. |
| 05:40 | < ToxicFrog> | Ok, I'm not completely lost by the fact that you think "checkout" is appropriate for "overwrite my local copy of this file with the version from a remote repository", but not if you s/remote/local/ |
| 05:40 | < Derakon> | s/not/now/? |
| 05:40 | < ToxicFrog> | Yes. |
| 05:40 | < Derakon> | Basically I'm just saying that I'd accept "checkout" instead of "pull". |
| 05:41 | < ToxicFrog> | ...but that's not at all what pull does |
| 05:41 | < Derakon> | (But I don't object to the push/pull nomenclature) |
| 05:41 | < Derakon> | From a user's perspective, there are files Over There. "git pull" gets you those files Over Here. |
| 05:42 | < Derakon> | (From a semi-nontechnical and/or uninterested in the nitty-gritty details of source control user's perspective) |
| 05:43 | < ToxicFrog> | I have to say that I think calling something that creates merge commits "checkout" would be even more confusing. |
| 05:44 | < Derakon> | I didn't say that "pull" should be renamed to "checkout", just that if you're going to call something "checkout", it should be the pull mechanism. |
| 05:45 | < ToxicFrog> | What would you call checkout, incidentally? |
| 05:45 | < Derakon> | What I really want is to rename the current "checkout" to "revert" and the current "revert" to "undo" or something. |
| 05:45 | < ToxicFrog> | Aah. |
| 05:45 | < ToxicFrog> | I do agree that "undo" is probably a better name for "revert". |
| 05:45 | < Derakon> | Because my normal use case is Authoritative Remote Repo > Local Working Repo > Modified Copy. |
| 05:46 | < ToxicFrog> | OTOH, calling "checkout" "revert" now means it's seriously dissonant when you use it to resurrect files from older commits or other branches. |
| 05:46 | < Derakon> | ...er, that's not actually helping my case, nemmind. |
| 05:46 | < Derakon> | Why is reverting to an older version confusing? |
| 05:46 | < Derakon> | Other branches I could see being a problem. |
| 05:46 | < ToxicFrog> | And yeah, the assignment of "checkout" to the remote -> local copy but not the local -> local copy seems completely arbitrary to me. |
| 05:46 | < ToxicFrog> | Because "revert" implies that you are returning something that exists to an earlier state. |
| 05:46 | < Derakon> | Forget about "checkout". It is banished from my nomenclature. ?.? |
| 05:47 | < Derakon> | And the earlier state is the state as of a previous commit. |
| 05:47 | < ToxicFrog> | Whereas "checkout" is also used to create files that do not exist in the working copy, but do exist in some commit. |
| 05:47 | < ToxicFrog> | Ie, 'checkout test-config -- foo.cfg bar.cfg' is meaningful even if foo.cfg and bar.cfg exist neither in HEAD or in the working copy, as long as they exist in the test-config branch. |
| 05:47 | < Derakon> | A.k.a. files that were removed at some point? |
| 05:48 | < ToxicFrog> | Yeah. Or files that were committed to a different branch but not this one. |
| 05:48 | < Derakon> | I admit it -- branches still fuck me up royally. |
| 05:48 | < ToxicFrog> | <3 branches <3 |
| 05:49 | < Derakon> | As for files that were removed, I don't have a huge problem imagining ghost files in my repo that could be restored if I reverted their deletion. |
| 05:51 | < ToxicFrog> | Fair enough. |
| 05:52 | < Derakon> | Unrelated, have a larger version of haloswirl: http://derakon.dyndns.org/~chriswei/temp/proceduralImages/haloswirl2.png |
| 05:52 | < Derakon> | (9MB PNG) |
| 05:53 | < Derakon> | I wouldn't mind having a shirt with that pattern on it all round. Sort of a neo-tyedie type of thing. |
| 05:54 | < ToxicFrog> | MY EYES |
| 05:54 | < Derakon> | Had you seen the other images I made? |
| 05:54 | < Derakon> | http://derakon.dyndns.org/~chriswei/temp/proceduralImages/ |
| 05:54 | < Derakon> | Made, rather slowly, with Python while in airports and airplanes. |
| 06:05 | | Kindamoody [Kindamoody@Nightstar-5507a6b5.tbcn.telia.com] has quit [Client exited] |
| 06:12 | | Kindamoody [Kindamoody@Nightstar-5507a6b5.tbcn.telia.com] has joined #code |
| 06:12 | | KiMo [Kindamoody@Nightstar-5507a6b5.tbcn.telia.com] has joined #code |
| 06:13 | | KiMo [Kindamoody@Nightstar-5507a6b5.tbcn.telia.com] has quit [Client exited] |
| 06:18 | | * Derakon listens to his hard drive swap as he tries to manipulate a 24000x18000x4 array of floats. |
| 06:19 | < Derakon> | (Trying to make a supersized version of the haloswirl image, just for fun) |
| 06:19 | | Derakon is now known as Derakon[AFK] |
| 06:20 | | Kindamoody [Kindamoody@Nightstar-5507a6b5.tbcn.telia.com] has quit [Client exited] |
| 06:20 | | Kindamoody [Kindamoody@Nightstar-5507a6b5.tbcn.telia.com] has joined #code |
| 06:21 | | Rhamphoryncus [rhamph@Nightstar-14eb6405.abhsia.telus.net] has joined #code |
| 09:01 | | AnnoDomini [annodomini@Nightstar-ec6f3e0f.adsl.tpnet.pl] has joined #code |
| 09:51 | | * TheWatcher is attempting to explain to $coworker why an introductory course that teaches php, perl, and python in parallel is a really fucking stupid idea |
| 09:51 | < TheWatcher> | Isn't helped by the fact that $coworker isn't a programmer, and has never taught a programming course |
| 09:58 | < AnnoDomini> | Yeah, seriously. Let's start with C and see what sticks. Those who can master making something like Guess The Number in C are allowed to learn more, uh, modern languages. The rest likely will never learn programming at all, according to research. |
| 10:01 | < TheWatcher> | I can see what she's trying to do, unfortunately she doesn't have any experience of how challenging some people find it just to learn one langauge well enough to use it properly. |
| 10:01 | <@McMartin> | They won't learn it in anything else either; you may as will pick something where the errors are more useful~ |
| 10:01 | | * McMartin is also a fan of starting at one level of abstraction and moving systematically to the other one, so one's first language should either be a LISP family, a scripting family, or actual assembler. |
| 10:02 | < TheWatcher> | Agreed |
| 10:08 | < AnnoDomini> | Assembly would really root out the lazy and the unable. |
| 10:47 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has joined #code |
| 11:45 | | Stalker [Z@Nightstar-3602cf5a.cust.comxnet.dk] has joined #code |
| --- Log closed Fri Aug 19 11:51:31 2011 |
| --- Log opened Fri Aug 19 11:51:35 2011 |
| 11:51 | | TheWatcher [chris@Nightstar-3762b576.co.uk] has joined #code |
| 11:51 | | Irssi: #code: Total of 23 nicks [10 ops, 0 halfops, 0 voices, 13 normal] |
| 11:52 | | Irssi: Join to #code was synced in 36 secs |
| 12:04 | <@froztbyte> | ah, the old "teach them syntax" approach |
| 12:07 | | gnolam [lenin@Nightstar-202a5047.priv.bahnhof.se] has joined #code |
| 12:28 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has quit [Ping timeout: 121 seconds] |
| 12:28 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has joined #code |
| 12:32 | | Attilla_ [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has joined #code |
| 12:33 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has quit [Ping timeout: 121 seconds] |
| 12:39 | | Attilla_ [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has quit [Ping timeout: 121 seconds] |
| 13:26 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has joined #code |
| 14:25 | | Kindamoody is now known as Kindamoody|out |
| 15:13 | | Thaqui [Thaqui@27B34E.D54D49.F53FA1.6A113C] has quit [Connection closed] |
| --- Log closed Fri Aug 19 15:25:23 2011 |
| --- Log opened Fri Aug 19 15:25:38 2011 |
| 15:25 | | TheWatcher [chris@Nightstar-3762b576.co.uk] has joined #code |
| 15:25 | | Irssi: #code: Total of 23 nicks [10 ops, 0 halfops, 0 voices, 13 normal] |
| 15:26 | | Irssi: Join to #code was synced in 41 secs |
| 17:13 | < Reivles> | aw, shit |
| 17:13 | <@Tamber> | ? |
| 17:14 | < Reivles> | Error 1935, an error occoured during the installation of assembly 'Microsoft.VC80.OpenMP, type="win32",version="8.0.50726.762",publicKeyToken=yaddayaddayadda... |
| 17:14 | < Reivles> | It is apparently something to do with windows update? |
| 17:14 | < Reivles> | Bonus points: It's on a machine I'm trying to get software onto... that currently has no internets. |
| 17:15 | <@Tamber> | Ah. |
| 17:16 | | Phox [Phox@Nightstar-3e5deec3.gv.shawcable.net] has quit [NickServ (GHOST command used by Neophox)] |
| 17:17 | | Phox [Phox@Nightstar-3e5deec3.gv.shawcable.net] has joined #code |
| 17:18 | | * Reivles rages at it, vaugely wonders what to do |
| 17:31 | < Stalker> | Google it. |
| 17:32 | < Stalker> | Then give machine internet, and fix everything. |
| 17:32 | | Stalker [Z@Nightstar-3602cf5a.cust.comxnet.dk] has quit [[NS] Quit: If the world didn't suck, we'd all fall off.] |
| 18:12 | | AnnoDomini is now known as Zon |
| 18:16 | | Reivles [orthia@Nightstar-75c3c27b.lightwire.co.nz] has quit [Ping timeout: 121 seconds] |
| 18:47 | | SmithKurosaki [smith@Nightstar-87f33f96.cpe.mountaincable.net] has joined #code |
| 18:53 | | Stalker [Z@Nightstar-3602cf5a.cust.comxnet.dk] has joined #code |
| 19:27 | | Kindamoody|out is now known as Kindamoody |
| 19:50 | <@Namegduf> | Hmm. |
| 19:51 | <@Namegduf> | If I had a set of points, and I wanted to calculate their concave hull in order to delinate a boundary for the set, and the points were moving |
| 19:51 | <@Namegduf> | What kind of approaches would people here use? |
| 19:55 | < gnolam> | Does it have to be a concave hull? That inherently means "guessing". |
| 20:12 | <@Namegduf> | Yes. |
| 20:12 | <@Namegduf> | And yeah, I know there's multiple solutions. |
| 20:12 | <@Namegduf> | Well, any number of solutions. |
| 20:14 | < gnolam> | Anyway. My instinct is successive agglomeration. Cluster, build convex hulls around each cluster, then agglomerate. |
| 20:18 | <@Namegduf> | Hmm. Sounds like a good idea. |
| 20:31 | | Kindamoody is now known as Kindamoody[zZz] |
| 21:18 | | Zon is now known as AnnoDomini |
| 23:09 | | * McMartin ...s at this caveat |
| 23:09 | <@McMartin> | "Note: Your cups.org username and password are not used by the CUPS web interface. Only accounts on your system can be used." |
| 23:20 | <@Tamber> | *chuckle* |
| 23:20 | <@Tamber> | One would think anyone who's at the point of knowing what CUPS is, shouldn't need that pointing out to them. |
| 23:21 | <@McMartin> | This is actually erring in the opposite direction to most OSS documentation, though, so I have to give it credit for that. |
| 23:21 | <@McMartin> | ... oh. It's also off a potholed link to localhost, and here it is again on a non-potholed link to localhost. |
| 23:22 | <@McMartin> | That does make a little more sense because there the workflow is "go to cups.org, click on link, get auth dialog" |
| 23:42 | | AnnoDomini [annodomini@Nightstar-ec6f3e0f.adsl.tpnet.pl] has quit [[NS] Quit: leaving] |
| 23:51 | | shade_of_cpux is now known as cpux |
| 23:56 | | You're now known as TheWatcher[T-2] |
| 23:58 | | You're now known as TheWatcher[zZzZ] |
| --- Log closed Sat Aug 20 00:00:36 2011 |