| 00:06:18 | Araq | I'm not sure {.push cdecl.} works as intended :-/ |
| 00:24:41 | Araq | still around fowl? |
| 00:37:33 | * | XAMPP-8 quit (Read error: Connection reset by peer) |
| 01:20:14 | NimBot | nimrod-code/Aporia 76412ac Dominik Picheta [+0 ±6 -0]: When text is selected, all occurrences of the selected text are now... 4 more lines |
| 01:20:24 | dom96 | Araq: Test it!! ^ |
| 01:21:56 | Araq | dom96: I'm already sleeping |
| 01:22:26 | * | q66 quit (Quit: Quit) |
| 01:24:13 | dom96 | Araq: Aporia has been tested by the Sleep Programmer association, and it fits the criteria for being used while you are sleeping. |
| 01:25:59 | Araq | ugh I don't get this bug |
| 01:26:16 | Araq | it's a trivial anon proc that doesn't even require a closure |
| 01:29:35 | dom96 | You know what will help you? |
| 01:29:38 | dom96 | Trying out the new aporia |
| 01:29:51 | Araq | perhaps you're right |
| 01:37:09 | Araq | dom96: you've implemented one of these "I can't live without anymore" features ... |
| 01:37:10 | Araq | you're doing an impressive job |
| 01:37:27 | dom96 | thanks :D |
| 02:21:56 | Araq | good night |
| 02:22:48 | dom96 | good night |
| 02:35:18 | * | XAMPP-8 joined #nimrod |
| 02:40:22 | * | fowl quit (Quit: Leaving) |
| 04:29:59 | shevy | hehe |
| 04:30:03 | shevy | dom96 is a coding-machine |
| 04:30:19 | shevy | I am stuck right now rewriting a ruby-projects |
| 04:30:23 | shevy | unix-like pipes |
| 04:41:53 | shevy | \o/ |
| 04:42:01 | shevy | can we use aporia to install addons for nimrod now? :) |
| 04:42:17 | shevy | (1) code in nimrod (2) code for nimrod - all through aporia |
| 04:42:32 | shevy | next addon feature would be to directly upload to nimrod repository... distributed coding hehehe |
| 08:30:11 | Araq | hi shevy |
| 08:30:27 | Araq | do you rewrite your ruby-projects in nimrod? |
| 08:32:24 | shevy | Araq hmm possibly not all of them |
| 08:32:38 | shevy | with bioinformatics, it's ok to look at it from different angles |
| 08:32:59 | shevy | other projects, such as compiling/installing something via ruby, I think it would be too much work to rewrite this in another language |
| 08:33:11 | shevy | smaller projects would be ok I guess |
| 08:33:58 | shevy | ruby has bindings to biology-related actions (bioruby), but I feel they are not really that good :( |
| 08:34:07 | Araq | np; yay for some bioinformatics code :D |
| 08:54:49 | * | FreeArtMan joined #nimrod |
| 10:01:42 | * | FreeArtMan quit (Remote host closed the connection) |
| 11:04:27 | * | shevy quit (Quit: "") |
| 11:09:34 | * | q66 joined #nimrod |
| 11:54:05 | * | codeallergy joined #nimrod |
| 12:19:34 | dom96 | hello |
| 12:19:41 | Araq | hi dom96 |
| 12:19:57 | Araq | aporia needs word bounds checking :P |
| 12:20:11 | Araq | in it's search feature |
| 12:20:23 | Araq | forgot to tell you :P |
| 12:20:35 | Araq | but it's an annoying ommission |
| 12:21:00 | dom96 | ok. Should be simple with regex. |
| 12:21:34 | Araq | please don't use a regex for that |
| 12:21:53 | Araq | make a checkbox "whole words" instead please |
| 12:22:02 | dom96 | well yeah. |
| 12:22:10 | dom96 | I mean it will use regex under the hood |
| 12:22:23 | Araq | ugh |
| 12:22:27 | Araq | no need |
| 12:22:34 | dom96 | well how should I do it? |
| 12:23:01 | Araq | just check that foundPos-1 < 0 or buffer[foundpos-1] notin {'a'..'z', ...} |
| 12:23:32 | Araq | and ditto for buffer[foundPos+length] notin {...} |
| 12:24:29 | dom96 | alright |
| 12:24:58 | dom96 | First I will commit my do fix |
| 12:25:11 | Araq | ok good |
| 12:37:55 | dom96 | Araq: I'll add a 'do' test, is that ok? |
| 12:38:02 | Araq | sure |
| 12:48:22 | NimBot | Araq/Nimrod 92c7504 Dominik Picheta [+1 ±1 -0]: Fixed a bug with the 'do notation' consuming statements after it. |
| 12:48:50 | dom96 | Araq: There |
| 12:49:10 | Araq | dom96: lets see what the tester says :P |
| 12:49:20 | dom96 | :P |
| 13:06:12 | * | FreeArtMan joined #nimrod |
| 14:35:15 | * | FreeArtMan quit (Quit: Leaving) |
| 14:35:24 | * | FreeArtMan joined #nimrod |
| 14:36:39 | reactormonk | Araq, utf8? |
| 14:51:42 | Araq | reactormonk: unicode.nim |
| 14:51:57 | Araq | but for JS I don't know ... |
| 14:52:12 | Araq | a char is a unicode char for the JS target |
| 14:55:07 | dom96 | hrm, maybe reactormonk was referring to your suggested implementation of word bounds checking for aporia? |
| 14:57:03 | Araq | good point |
| 14:57:32 | Araq | but for a first version my approach is good enough |
| 14:57:50 | Araq | and it can easily be fixed later, though it's a bit of work |
| 15:02:51 | reactormonk | dom96, exactly |
| 15:03:11 | reactormonk | Araq, if the regexp-implementation were unicode-aware ;-) |
| 15:05:19 | Araq | hey I think it is :P |
| 17:35:07 | NimBot | Araq/Nimrod b3a416b Dominik Picheta [+0 ±2 -0]: Reverted my 'do notation' fixes. |
| 17:40:58 | * | XAMPP-8 quit (Ping timeout: 244 seconds) |
| 18:41:37 | NimBot | nimrod-code/Aporia 4012e00 Dominik Picheta [+0 ±3 -0]: Added a list of characters which will not be highlighted.... 3 more lines |
| 18:43:46 | apotheon | https://twitter.com/CopyfreeNews/status/293066113833648128 |
| 18:44:04 | dom96 | :O |
| 18:44:12 | dom96 | apotheon: Nice |
| 18:44:48 | apotheon | I like to promote people who do good things (even while recovering from the flu). |
| 18:45:10 | dom96 | apotheon: get well soon! :) |
| 18:52:03 | apotheon | thanks |
| 19:15:11 | * | gradha joined #nimrod |
| 19:16:30 | * | codeallergy quit (Read error: No route to host) |
| 19:20:14 | gradha | Araq: the answer to http://forum.nimrod-code.org/t/118 is missing the gc stop-the-world/incremental discussion from yesterday's irc |
| 19:21:09 | Araq | gradha: why? |
| 19:21:33 | Araq | the answer misses quite a lot as I left out the D bashing :P |
| 19:22:44 | gradha | you don't say anything about the GC for suitability of real time software |
| 19:22:53 | Araq | oh ... |
| 19:23:22 | gradha | a link to http://nimrod-code.org/gc.html would be enough |
| 19:23:59 | gradha | maybe |
| 19:25:59 | * | FreeArtMan quit (Quit: Leaving) |
| 19:27:18 | Araq | gradha: you should answer too then and give that link ;-) |
| 19:48:29 | gradha | you mentioned games as time wasters, and lan cable plugin as solution, yet that leaves you without irc. What to do? |
| 19:48:57 | Araq | I never mentioned games as time wasters, did I? |
| 19:49:14 | gradha | is masters of magic a life enhacer then? |
| 19:49:16 | dom96 | IRC is usually a time waster too :P |
| 19:49:23 | gradha | true |
| 19:56:19 | gradha | dom96: I installed locally with babel and compiled some other project, it found automatically the directory and added to the path, is this correct? |
| 19:57:44 | dom96 | gradha: yep |
| 19:59:30 | gradha | how will a user deal with multiple versions of the same lib? any way to specify that? |
| 20:00:28 | gradha | let's say I have epak-1.0.0 and epak-1.1.3 installed, can I pick which one? |
| 20:00:45 | Araq | yes |
| 20:00:54 | Araq | the compiler searches everything in $babel |
| 20:01:24 | Araq | but you can exclude dirs explicitly with --excludePath:PATH |
| 20:01:58 | Araq | you can put that into your per project configuration file or in the global config file or ... |
| 20:02:30 | Araq | oh and the compiler knows about versions in $babel and picks the latest version per default |
| 20:02:57 | gradha | indeed, it picks 1.1.3 |
| 20:03:04 | dom96 | Also isn't suppose to pick the version before the version you exclude? |
| 20:04:23 | gradha | --excludePath:epak-1.1.3 doesn't work, needs full path? |
| 20:04:39 | Araq | yes I think so |
| 20:05:02 | gradha | no tilde expansion I presume |
| 20:05:08 | Araq | true |
| 20:06:03 | gradha | looks like ancestors don't work either, I used "nimrod c --excludePath:/Users/gradha/.babel/libs/epak-1.1.3 -d:release submarine_hunt_optimus.nim" |
| 20:06:10 | gradha | got "config/nimrod.cfg(36, 11) Hint: added path: '/Users/gradha/.babel/libs/epak-1.1.3/nimrod' [Path]" |
| 20:06:29 | Araq | well hrm that doesn't mean much |
| 20:06:49 | Araq | it adds it and then discards it again -- perhaps |
| 20:06:49 | gradha | yes, excluding the full path works, but then the other version is not found and can't compile |
| 20:07:04 | Araq | well yes? |
| 20:07:09 | Araq | what's the problem? |
| 20:07:39 | gradha | I was following dom96's suggestion that nimrod may pick the version before the one excluded |
| 20:07:56 | Araq | I don't think that works :P |
| 20:09:00 | Araq | gradha: feel free to come up with a better scheme that works |
| 20:09:09 | Araq | we have been discussing these things to death |
| 20:09:27 | dom96 | I think this scheme may work, if the implementation works :P |
| 20:09:43 | Araq | IMHO it's better to have a braindead simple package manager that requires hand holding when stuff breaks |
| 20:09:50 | dom96 | Also --excludePath shouldn't need a full path IMO |
| 20:10:03 | Araq | as opposed to a complex package manager that also breaks |
| 20:10:13 | Araq | and makes things hard to fix |
| 20:10:26 | gradha | careful there dom96, soon you will be asking the same feature for --path |
| 20:11:05 | Araq | and don't expect me to improve what's already there until we actually have more than 20 packages :P |
| 20:11:20 | gradha | can you disable the compiler's behaviour to search libs in .babel? |
| 20:11:26 | Araq | sure |
| 20:11:42 | dom96 | well, --excludePath has been implemented specifically for babel. |
| 20:11:45 | Araq | should throw out the --babelPath stuff in your config |
| 20:12:01 | dom96 | So it makes sense that it's relative to babel's path I think. |
| 20:12:01 | Araq | dom96: yes but we have 3-4 packages |
| 20:12:19 | Araq | yes it makes sense |
| 20:12:27 | Araq | but I prefer to fix bugs instead |
| 20:12:45 | dom96 | sure, I don't mind. I'm just stating my thoughts as to how it should work :P |
| 20:13:25 | gradha | I've tried --excludePath:/Users/gradha/.babel to disable babel, but ancestors don't work, also tried --babelPath: with a phony directory and still found the packages |
| 20:13:40 | Araq | you have to remove it |
| 20:13:46 | Araq | can't change it from the command line |
| 20:13:58 | Araq | there can be multiple babel paths ... |
| 20:13:58 | gradha | oh, remove the babel directory? |
| 20:14:26 | dom96 | remove it from the compiler's config file |
| 20:14:34 | Araq | I dunno what you're doing with your config/nimrod.cfg, but mine contains: |
| 20:14:39 | Araq | babelpath="$home/.babel/libs/" |
| 20:15:17 | gradha | yes, so a system wide install of nimrod forces you to use babel, since you can't override that |
| 20:15:54 | Araq | it nevers occured to me that there is such a thing as a "system wide install of nimrod" :P |
| 20:16:06 | gradha | true |
| 20:17:01 | Araq | but you can make the compiler not read the global config file |
| 20:17:20 | Araq | --skipCfg |
| 20:17:33 | Araq | --skipCfg do not read the general configuration file |
| 20:17:34 | Araq | --skipUserCfg do not read the user's configuration file |
| 20:17:36 | Araq | --skipParentCfg do not read the parent dirs' configuration files |
| 20:17:38 | Araq | --skipProjCfg do not read the project's configuration file |
| 20:17:49 | gradha | hehe, weird that those are hardcoded and you have none with a parameter to specify the path |
| 20:18:04 | Araq | good point |
| 20:18:56 | Araq | nimrod's config system is excessive and overdone imo |
| 20:19:53 | Araq | but zahary made good arguments for this system |
| 20:20:11 | gradha | looking at the switches now, why is that --parallelBuild:1 shows the C compile errors and not something like --verbose:x? |
| 20:20:17 | gradha | --verbosity:x |
| 20:20:39 | Araq | ever tried to capture all the output from multiple processes running in parallel? |
| 20:20:55 | gradha | not me |
| 20:21:04 | gradha | things in parallel tend to be scary |
| 20:21:11 | gradha | I run away from them instead |
| 20:21:31 | Araq | yeah well ... unix output redirection is a pita |
| 20:21:45 | Araq | so I didn't even try for parallel builds |
| 20:22:46 | gradha | so whenever you say "we should do something about --parallelbuild" you mean "somebody should implement that correctly" or is it just not possible and something else has to be invented? |
| 20:23:36 | Araq | I'm sure its possible |
| 20:23:36 | zahary | it's not that hard - since I'm digging through the code of build systems from time to time, in ninja there is some code that's supposed to be very optimized |
| 20:23:37 | zahary | https://github.com/martine/ninja/tree/master/src |
| 20:24:10 | zahary | the subprocess-* files |
| 20:25:12 | Araq | lol super optimized ... it uses fork() :P |
| 20:25:33 | Araq | nimrod's uses posix_spawn on linux at least |
| 20:27:35 | zahary | hmm, it really should be throughly tested. ninja is the fast build system in the C++ world |
| 20:27:43 | zahary | *the fastest* |
| 20:28:06 | gradha | gour mentioned some time ago "tap", or something like that |
| 20:28:15 | zahary | tup |
| 20:28:20 | gradha | oh, yeah |
| 20:29:02 | zahary | it has more and cooler features than ninja, but the actual command execution code is a bit slower |
| 20:29:07 | gradha | tup looked nice from it's description on carrying a photon |
| 20:29:31 | zahary | the coolest thing about tup is that it can automatically detect the dependencies in any kind of builder/compiler |
| 20:29:46 | Araq | and the windows version uses CreateProcessA ... |
| 20:29:58 | Araq | CreateProcessW should be faster but maybe it's not |
| 20:30:46 | gradha | isn't the difference between A/W versions the string format? how can that matter for process execution? |
| 20:31:03 | Araq | lol good point |
| 20:31:21 | Araq | but windows uses the utf-16 in its core |
| 20:31:34 | Araq | so the A functions wrap the W functions |
| 20:32:19 | gradha | IIRC when you leave the A/W character out the compiler calls the "native" version, so you would be passing utf-16 strings anyway to it if that was default, right? |
| 20:32:25 | gradha | oh, maybe not |
| 20:32:53 | Araq | that only works in C/C++; the preprocessor maps it to either the A or the W version |
| 20:33:39 | Araq | but the native version is the W version anyway |
| 20:34:06 | Araq | unless they changed it to utf-8 for windows phones |
| 20:36:04 | Araq | looking at the code in osproc.nim:170 |
| 20:36:24 | Araq | it may work with -d:noBusyWaiting -d:debugExecProcesses |
| 20:56:20 | * | Roin quit (Quit: brb) |
| 20:56:30 | * | Roin joined #nimrod |
| 20:56:30 | * | Roin quit (Changing host) |
| 20:56:30 | * | Roin joined #nimrod |
| 21:37:10 | gradha | while using docstrigns I find it irritating that ` is a special character and the compiler sort of "jumps" through docstrings pointing at the wrong line |
| 21:37:40 | gradha | I have "blah blah `blah" in docstring for proc1 |
| 21:37:55 | gradha | then I have another proc2, and the compiler points to somewhere there as error? |
| 21:40:13 | gradha | see here https://gist.github.com/4581948 that's really bad |
| 21:41:24 | gradha | the opening of the quote is on line 125, the error is reported for both doc and doc2 |
| 21:41:32 | gradha | both on line 153 |
| 21:42:24 | gradha | hopefully it would be possible to point at the line with the opening `, not where it is expected to be closed |
| 21:44:25 | dom96 | hrm, gtk doesn't like it when a GtkSourceView from another tab is focused. |
| 22:11:46 | Araq | gradha: I can't really follow |
| 22:11:57 | Araq | what's the problem? stupid compiler error message? |
| 22:12:08 | Araq | wrong doc html generated? |
| 22:12:20 | gradha | error points at random unrelated line |
| 22:13:01 | gradha | whatever the "correct" line it expects the closing quote, it could really say where it found the starting one |
| 22:13:26 | Araq | hey often we try to do that :P |
| 22:13:44 | gradha | btw, how do you specify skipCfg in a nimrod.cfg file? |
| 22:13:52 | Araq | you don't |
| 22:13:59 | Araq | you use it on the command line |
| 22:14:07 | gradha | aha |
| 22:14:10 | Araq | (I think) |
| 22:14:23 | gradha | I was trying skipCfg=true, but didn't work |
| 22:14:24 | Araq | you could try it though maybe it works |
| 22:14:34 | Araq | well try this syntax: --skipCfg |
| 22:15:05 | gradha | no, I guess by the time the project cfg file is read the global one has been read too |
| 22:15:21 | Araq | yeah |
| 22:15:40 | gradha | back to good old shell scripts for invokation |
| 22:15:52 | Araq | we could get rid of the babel line for the default config |
| 22:16:05 | Araq | and hell no, please no shell scripts |
| 22:16:16 | Araq | they don't run on windows |
| 22:17:03 | gradha | the situation is: I have this epak lib I'm trying to make open, so I have installed it, but it has test cases. |
| 22:17:18 | gradha | compiling the test case now picks the babel library, so I need to avoid that |
| 22:17:33 | gradha | since .cfg can't override that, I will wrap the tests in a shell script |
| 22:17:39 | gradha | don't worry, end users aren't supposed to run the tests |
| 22:18:33 | gradha | for unit testing avoiding the global cfg is sane anyway |
| 22:18:33 | Araq | but still a good use case that we should consider |
| 22:18:43 | Araq | not really |
| 22:18:58 | gradha | well, you would like the unit test environment to be as reproducible as possible |
| 22:18:59 | Araq | what if your global cfg says to avoid GCC because it's broken on the mac? |
| 22:19:06 | gradha | if you have a global cfg, that could change |
| 22:20:03 | gradha | good point, there's no solution |
| 22:21:27 | gradha | short of packing a virtual machine with full compiler setup I don't see what could be done |
| 22:21:34 | Araq | the global cfg also contains things like where to find the stdlib ... |
| 22:21:38 | Araq | including system.nim |
| 22:21:50 | Araq | good luck without that :P |
| 22:22:44 | gradha | you could add the hack of priorities with paths on top of babel |
| 22:23:02 | dom96 | You could just 'babel install' and then run your tests :P |
| 22:23:32 | dom96 | s/run/compile/ |
| 22:23:47 | gradha | then uninstall? otherwise runnina test will break other compilations |
| 22:25:28 | gradha | oh, no babel uninstall yet |
| 22:25:40 | Araq | I guess babelpath shouldn't be in the global config |
| 22:25:49 | Araq | it does more harm than good |
| 22:25:58 | * | XAMPP-8 joined #nimrod |
| 22:26:04 | dom96 | huh, what do you mean it will break other compilations? |
| 22:26:11 | gradha | that's because babel has few users yet |
| 22:26:20 | dom96 | Araq: Please don't remove it. |
| 22:26:32 | gradha | dom96: if I install a development version of my lib with another version number, other compilations expecting a stable version might not work |
| 22:27:08 | gradha | say I want to run the test suite of epak-devel branch, if it is left installed, normal epak-stable compilation could not work |
| 22:27:27 | gradha | depending on if I changed interfaces, backwards compatible, or simply buggy |
| 22:27:54 | dom96 | hrm, I think we do need some sort of path priorities here. |
| 22:28:13 | dom96 | Your tests can have a config file specifying "--path:../epak" |
| 22:28:17 | dom96 | if they are in a tests dir |
| 22:28:26 | gradha | nah, it's easier if developers of modules don't have babel in the global cfg, I'll do that |
| 22:29:58 | Araq | dom96: yeah we need path priorities, ways to override the default, setting the default, having different configs depending on whether you like to test it or not |
| 22:30:04 | Araq | whether you're the author or not |
| 22:30:07 | Araq | etc. etc. |
| 22:30:13 | Araq | this stuff never ends |
| 22:31:02 | Araq | so sorry, babelpath won't be in the global config |
| 22:31:25 | Araq | it already starts to cause trouble |
| 22:31:37 | gradha | don't be so hasty, users will like it |
| 22:31:52 | gradha | users aka non module developers |
| 22:32:10 | Araq | I think compiling with --babel if you want that feature is tolarable |
| 22:34:54 | gradha | sounds good |
| 22:46:20 | Araq | on the other hand, gradha, babelpath uses the latest version of your library that you want to test |
| 22:46:58 | gradha | yes, the problem is not for the test, but if I happen to run the tests, forget to manually uninstall it, then try to compile other software expecting a non-developer version |
| 22:47:09 | gradha | that's why I was looking for a babel uninstall |
| 22:47:19 | Araq | but is that really a problem? |
| 22:47:29 | gradha | not really, I've already disabled babel in the global cfg |
| 22:47:34 | Araq | if you break things you'll be reminded to uninstall your devel version |
| 22:50:54 | dom96 | I can (and will) add 'babel uninstall' |
| 22:52:27 | gradha | don't worry about this now, I actually prefer to disable babel anyway and specify my own paths |
| 22:52:39 | gradha | you know, I'm the weird guy asking for static linking and stuff, don't mind me |
| 22:53:14 | Araq | gradha: that's exactly my preference too |
| 22:53:44 | Araq | I think it should be this way: |
| 22:54:06 | Araq | babel can install a thing and its dependencies recursively |
| 22:54:18 | Araq | babel can uninstall a single package |
| 22:54:44 | Araq | and there is --babelpath for "convenient" path choosing but it's not the default |
| 22:54:57 | Araq | and then I can have as many conflicts as I like |
| 22:55:12 | Araq | I can always use project specific path settings to control things |
| 22:55:36 | Araq | so that project A uses pack-0.8.0 and project B can use pack-0.9.0 |
| 22:55:58 | gradha | ok, here goes an idea expanding on https://github.com/nimrod-code/babel/issues/3 |
| 22:56:09 | gradha | how about babel being a package installer and also a .cfg file configurator? |
| 22:56:26 | Araq | too complex |
| 22:56:29 | gradha | the idea being that babelPath would not be installed, and babel would actually create/change the .cfg for the project |
| 22:56:40 | gradha | then you could use babel to "switch" to a specific installed version if prefer so |
| 22:57:05 | Araq | nah if I don't like adding ever more paths manually, I'll use --babelPath |
| 22:57:08 | gradha | basically, it's what you said but using babel as an interface to manually updating the cfg files |
| 22:58:34 | gradha | in a way this would be the equivalent of using `pkg-config blah` for C/C++ compilation |
| 22:58:54 | Araq | perhaps but files like that tend to become a mess anyway |
| 22:59:18 | Araq | with sections like ##begin_autoconf ... ##end_autoconf |
| 22:59:39 | Araq | it's tolerable though |
| 23:01:13 | gradha | still, back to the idea that 2013 is the year of Nimrod, we are at a weird situation now were we have practically the same number of develoeprs as end users |
| 23:01:28 | gradha | the current setup works nicely for end users, and developers can just comment a single line in their global .cfg |
| 23:01:41 | gradha | also sounds fair to me, plus requires the least of development time and works now |
| 23:01:50 | dom96 | Araq: The thing that I don't like about your solution is that dependency information will be duplicated across the .babel and the .cfg files. |
| 23:02:18 | dom96 | But that's also an issue with how it works currently... |
| 23:04:39 | Araq | gradha: I can't really follow |
| 23:04:49 | dom96 | But I myself can't come up with a better solution. |
| 23:04:51 | Araq | we have no "end users" |
| 23:04:58 | gradha | precisely |
| 23:05:02 | Araq | everybody who uses nimrod is a programmer |
| 23:05:10 | gradha | ah, not really |
| 23:05:12 | Araq | because it's a programming language |
| 23:05:17 | gradha | think of nimrod being the default programming language |
| 23:05:23 | gradha | now you need to compile progX |
| 23:05:31 | gradha | progX is in nimrod and requires babel libs a, b, c |
| 23:05:45 | gradha | you are a "more advanced user" than most, but still a user, not a nimrod developer |
| 23:05:48 | Araq | progX should have used niminst and call it a day :P |
| 23:06:26 | gradha | ok, I'll try to stop foreseeing the future and call it a day too |
| 23:07:02 | gradha | speaking of preferences, I dislike package managers and use everything as git submodules (the horror!) |
| 23:07:26 | gradha | but again, that happens only because I'm always wrong |
| 23:07:33 | Araq | hey no worries I dislike package managers too |
| 23:07:43 | Araq | never used one that doesn't break |
| 23:08:01 | gradha | we must be dom96's dream of an audience then |
| 23:08:51 | Araq | true ;-) |
| 23:09:43 | Araq | seriously I consider the idea of using babel to install end user software misguided |
| 23:10:25 | dom96 | This babel stuff is really annoying me... if anyone wants to take over its development they are welcome to do so :P |
| 23:10:32 | gradha | you can probably say the same about perl's CPAN |
| 23:10:40 | gradha | yet it is really popular |
| 23:10:49 | dom96 | look at ruby's gems |
| 23:11:00 | Araq | I don't use CPAN to install perl based software on Linux |
| 23:11:03 | gradha | also, python's easy_install is very nice, doesn't work 100% of the times, but nicely for 99% of them |
| 23:11:13 | Araq | I use the distro's package manager |
| 23:11:16 | gradha | so I understand babel from the ease of use POV |
| 23:11:25 | dom96 | No one submits ruby packages to distro repos, instead they just tell you to use gems. |
| 23:11:37 | Araq | interesting |
| 23:11:45 | Araq | I wasn't aware of that, dom96 |
| 23:11:50 | gradha | I sidestep that problem by avoiding to use ruby |
| 23:12:00 | dom96 | Araq: really? |
| 23:12:04 | Araq | really |
| 23:12:56 | Araq | I don't use 'python setup.py install' on Linux either |
| 23:13:28 | Araq | I either use the package manager or compile from source ... and the later is usually a pain ... |
| 23:13:39 | dom96 | But then again do non-ruby developers actually install ruby software through gems? |
| 23:13:40 | Araq | oh I see ... :P |
| 23:14:34 | Araq | well babel already does the recursive deps thing (right?) and that's where the pain is anyway |
| 23:14:46 | dom96 | nope |
| 23:15:30 | gradha | the pain is usually "use progX, there is no binary for your platform, so build it from source!" |
| 23:15:41 | gradha | and the libs are not in the distro, or too old |
| 23:16:47 | Araq | dom96: so now you know what to do with babel :P |
| 23:16:55 | Araq | implement recursive dependencies |
| 23:17:06 | dom96 | Yeah... let me get right on that... |
| 23:17:40 | Araq | "Python 3 solves many problems with packaging. There aren't any packages for Python 3." XD |
| 23:17:46 | Araq | http://stackoverflow.com/questions/2436731/does-python-have-a-package-module-management-system |
| 23:18:31 | gradha | aren't dependencies only required for binary distro programs? I just don't see that as a developer packaging system where you can just list all you need straight away when writing your babel file |
| 23:19:11 | gradha | Araq: you tease, I went in just to upvote "that answer" |
| 23:19:57 | gradha | oh wait, I skipepd it! |
| 23:20:15 | gradha | muahaha, Colonel Panic is my hero |
| 23:21:29 | gradha | but the problem it outlines is true: "Python's devs are ignorant of all this frustration because they are seasoned programmers comfortable building from source, and they use Linux distributions with packaged Python modules" |
| 23:21:43 | gradha | this just happens everywhere where a successful dev package manager rises |
| 23:22:04 | gradha | whenever I've had to install perl/ruby using cpan/gems I've wanted to stab my eyes |
| 23:22:26 | gradha | because "you can just use it!" yet nobody bothers documenting the newbie steps, so you end up having to google for a simple install |
| 23:22:51 | gradha | and when I've submited doc patches to correct this they have not been accepted because "uh, this is so basic we don't think it will help" |
| 23:23:29 | Araq | thinking about it ... I don't install much software these days |
| 23:23:43 | Araq | I download source code to skim over it |
| 23:23:57 | Araq | and that's it |
| 23:26:13 | NimBot | nimrod-code/Aporia 9af24a0 Dominik Picheta [+0 ±4 -0]: The tab which is selected is now saved with the session and reselected... 2 more lines |
| 23:26:13 | NimBot | nimrod-code/Aporia 539d565 Dominik Picheta [+0 ±2 -0]: Comment feature now adds spaces between comment chars. |
| 23:27:03 | gradha | my favourite part of software is pain, so I always install software without admin priviledges in non standard paths, then this software misteriously breaks as it expects to be in certain locations or have certain superpowers |
| 23:27:13 | Araq | probably because installing software on linux tends to be a pain |
| 23:28:28 | Araq | which one is the cdrom btw? /media/cdrom or /media/cdrom0 ? |
| 23:29:39 | gradha | usb in linux is fun, every time you plug in something you have a full discovery journey in front of you |
| 23:29:59 | Araq | same for CDs |
| 23:30:22 | Araq | it's no wonder nobody uses linux |
| 23:30:51 | gradha | DOS was so much better, nobody needs more than 26 drive letters anyway |
| 23:31:05 | Araq | true |
| 23:31:22 | Araq | and you can easily have more anyway |
| 23:31:36 | Araq | after Z: AA: could follow |
| 23:31:52 | gradha | I'd run to patent that |
| 23:33:42 | Araq | "This is the wine64-bin helper package, which does not provide wine itself, but instead exists solely to provide the following information about enabling multiarch on your system in order to be able to install and reun the 32-bit wine packages." ... |
| 23:34:43 | gradha | that's really desktop ready, what a coincidence 2013 will be the year of nimrod AND linux |
| 23:35:28 | Araq | and I thought 2008 was the year of Linux for the desktop |
| 23:36:14 | gradha | sorry, I meant to say 2013 is the year of nimrod AND steam |
| 23:36:22 | gradha | or steambox, or linuxsteambox |
| 23:36:41 | gradha | or what-the-hell-I-have-12-ports-for-if-people-only-have-one-or-two-controllers wonderful console |
| 23:37:08 | gradha | hmm... 12 ports, that leaves you room for 14 more drive letters |
| 23:39:23 | gradha | talking about ports, I think nimrod has too few game libraries, so I might wrap cocos2d-x in the future |
| 23:39:38 | Araq | really? |
| 23:39:44 | Araq | I think it has too few math libraries |
| 23:39:51 | Araq | and too few plotting libraries |
| 23:40:10 | gradha | but for math you need a brain, and for plotting you need a plotter? |
| 23:40:15 | Araq | and too few webkit wrappers |
| 23:40:22 | gradha | oh yeah, webkit |
| 23:40:32 | gradha | I'm starting to like the idea of wrapping that |
| 23:43:40 | Araq | stemmers, scientific computing libs, sound libraries, voice recognition, OCR |
| 23:44:02 | gradha | I'm intrigued by your plotting libraries, give me an example |
| 23:44:47 | Araq | matplotlib? |
| 23:44:56 | gradha | oh, so it's math related, boo |
| 23:45:06 | Araq | yeah well |
| 23:45:17 | Araq | a pie chart doesn't qualify as math :P |
| 23:45:57 | gradha | I'm still of the mental age of ponies and unicorns and counting up to 1, 2, 3... er 5 |
| 23:46:55 | gradha | which reminds me how one of my previous employers wanted be to use a specific ios plotting library and it took me longer than to write myself the damn plotting code |
| 23:47:29 | Araq | that's exactly what I did too :D |
| 23:47:44 | Araq | used matplotlib and it wouldn't do what I needed |
| 23:48:02 | Araq | so I wrote code to draw every single pixel the way I wanted it |
| 23:49:26 | gradha | I believe they previously used baboons for coding, so they didn't believe me when I told them it would take me 12 hours |
| 23:49:51 | gradha | not that I cared too much, as they paid by the hour |
| 23:55:28 | Araq | gradha: I fixed the ` error message |
| 23:55:32 | Araq | can you test it? |
| 23:55:47 | gradha | will do |
| 23:57:00 | gradha | the last commits are from dom96 |
| 23:57:50 | Araq | arg! |
| 23:57:58 | Araq | aporia crashed! |
| 23:58:18 | Araq | thanks for the highlighting dom96 :P |
| 23:58:27 | dom96 | I did warn you :P |
| 23:58:34 | Araq | fixed yet? |
| 23:58:40 | dom96 | no. |
| 23:58:44 | dom96 | I can't reproduce it |
| 23:58:51 | dom96 | Can you? |
| 23:59:07 | Araq | highlight a space at the end of the line |