| 00:02:02 | SchalaZeal | Plus Google has no release notes about Dart |
| 00:03:04 | SchalaZeal | or Chrome for that matter |
| 00:03:05 | avarus | google is stupid |
| 00:03:16 | avarus | they do have a chrome with dart support |
| 00:03:26 | avarus | but not the public version they promote |
| 00:03:48 | SchalaZeal | after trying to comprehend the absurdity behind Go, I believe that |
| 00:12:49 | * | SchalaZeal quit (Quit: Konversation terminated!) |
| 00:25:17 | reactormonk | Anyone knowledgable about hardware? |
| 00:47:54 | * | q66 quit (Remote host closed the connection) |
| 01:26:43 | * | Trix[a]r_za is now known as Trixar_za |
| 01:54:53 | * | avarus quit (Remote host closed the connection) |
| 05:32:36 | * | Trixar_za is now known as Trix[a]r_za |
| 07:17:51 | * | gour joined #nimrod |
| 11:19:07 | * | avarus joined #nimrod |
| 11:19:19 | avarus | hola cabrones! |
| 11:30:09 | dom96 | hola amigo! |
| 11:37:51 | * | q66 joined #nimrod |
| 11:38:45 | dom96 | avarus: You just called us something insulting didn't you? :P |
| 11:39:21 | avarus | na, it's meant friendly :) |
| 11:39:33 | dom96 | lol ok |
| 12:46:59 | * | avarus quit (Read error: Connection reset by peer) |
| 13:09:24 | * | Zerathul joined #nimrod |
| 13:55:59 | * | exhu joined #nimrod |
| 13:56:05 | exhu | hey |
| 13:56:09 | Araq | hi |
| 13:57:51 | exhu | I have a problem to organize/group a tuple/struct describing a database result, a table name (where this result is from) and the sql column names (which differ from field names of the struct). |
| 13:59:48 | exhu | Right now I have to specify all three in three places in the source (type, const sections) and they seem to far away, something like namespace would be better. |
| 14:01:08 | exhu | Araq, any ideas to solve the problem? |
| 14:01:28 | Araq | use a macro? |
| 14:02:00 | exhu | I wrote a class macro once upon a time, it took several days. |
| 14:02:54 | exhu | And I've forgotten how to use macros again -) |
| 14:03:04 | Araq | well the API will be improved |
| 14:03:19 | Araq | for now it only provides the absolute minimum |
| 14:04:46 | exhu | I've been looking for a proc like indexOf, for taking integer index of the tuple field specifying its name -) Is that possible? |
| 14:06:27 | exhu | i.e. given a T:tuple[a: int, b: int], I can set a = 3 with T.a = 3 or T[0] = 3 |
| 14:06:53 | exhu | but is it possible to do T[indexof(T.a)] = 3 ? |
| 14:08:10 | Araq | I think so |
| 14:08:35 | Araq | proc indexOf(t: typedesc, name: string): int {.compiletime.} = |
| 14:08:49 | Araq | var d: t |
| 14:09:21 | Araq | for n, x in fieldPairs(d): if n == name: return i |
| 14:09:33 | Araq | (I forgot to define i, but it's trivial) |
| 14:09:45 | exhu | wow! |
| 14:22:36 | exhu | why nimrod suddenly stopped reading nimrod.cfg in the current directory? |
| 14:23:06 | Araq | I don't know? |
| 14:23:58 | exhu | it failed to compile the program with that new function, showing error "underfined reference indexof_121212" (some number was here) and another tries stop now that it can't find any modules. |
| 14:24:43 | exhu | some weird magic |
| 14:26:53 | exhu | looks like {.compiletime.} is not working anyway, it can't link the program tests/nimcache/palitsa_db_tests.o: In function `palitsa_db_testsInit': |
| 14:26:54 | exhu | palitsa_db_tests.c:(.text+0x7d83): undefined reference to `indexof_136842' |
| 14:26:54 | exhu | collect2: ld returned 1 exit status |
| 14:27:38 | Araq | yeah hrm |
| 14:29:28 | exhu | http://pastebin.com/4bdRvK16 |
| 14:29:47 | exhu | this function is placed in a separate module and used from another one. |
| 14:36:51 | Araq | it compiles for me |
| 14:37:09 | Araq | if I don't use it anywhere ;-) |
| 14:41:13 | Araq | exhu: you need to do: |
| 14:41:22 | Araq | const arg = TMyTuple.indexOf("a") |
| 14:41:25 | Araq | that works for me |
| 14:41:57 | Araq | but yeah please submit a bug report |
| 14:42:08 | Araq | .compileTime should enforce it already |
| 14:42:39 | exhu | ok |
| 14:42:42 | exhu | thanks |
| 14:43:46 | Araq | looks easy to fix; I already have some idea what's wrong here |
| 14:44:52 | exhu | const r = TA.indexof("b") does not work inside a template (unittest) |
| 14:45:08 | exhu | palitsa_db_tests.nim(148, 29) Error: constant expression expected |
| 14:46:53 | Araq | interesting -- add that to the bug report ;-) |
| 14:50:16 | exhu | done. |
| 14:54:57 | exhu | I also can't add documentation comments to enum types. |
| 14:55:42 | exhu | it complains about wrong identation, tried everything and don't get how to document each enum value. |
| 14:56:28 | Araq | well .... I can ... |
| 14:59:03 | exhu | http://pastebin.com/fpPkfvAq |
| 14:59:17 | exhu | only the last line works, it documets the enum as a whole. |
| 14:59:59 | exhu | other combinations produce "Error: identifier expected, but found '## medi'" |
| 15:00:06 | Araq | type x = enum ## document x here |
| 15:00:19 | Araq | valueA = "value A", ## document valueA |
| 15:00:29 | Araq | valueB = "value B" ## \ |
| 15:00:32 | Araq | ## also possible |
| 15:02:32 | exhu | now it works |
| 15:03:02 | Araq | I have to go, see you later |
| 15:03:44 | exhu | bye |
| 15:09:58 | exhu | actually i need that weird stuff like indexOf(tupleType.fieldName) to translate tuple definition into column names to form an sql request. |
| 15:12:34 | exhu | maybe i'm doing it all wrong and there's a simpler solution :) i need it to force compilation error on wrong code (i.e. forming an sql request with wrong column names -- the case that is possible in runtime but is eliminated with nimrod compile-time features). |
| 15:13:58 | exhu | So that a typo in a column name would be prevented or showed up before running unit tests at all. |
| 15:18:26 | exhu | Eventually there's a tuple (name:type definition for "business logic") where field names obey coding convention and an array of strings with native sql names of associated columns, tuple and array of sql column names share the same index value. |
| 15:20:35 | exhu | Index here is a way to attach an "annotation" (like in java) to a field. |
| 15:37:05 | * | gradha joined #nimrod |
| 15:41:46 | gradha | dom96: builder seems to be behaving, maybe I should stop it and recompile with the default timeout to verify the change actually affected the cpu usage |
| 15:47:19 | gradha | Araq: send me your public ssh key to let you access my laptop |
| 15:53:39 | * | gradha quit (Read error: Connection reset by peer) |
| 15:54:09 | * | gradha joined #nimrod |
| 15:57:39 | * | gradha quit (Read error: Connection reset by peer) |
| 15:58:44 | * | gradha joined #nimrod |
| 16:20:39 | * | gradha quit (Read error: Connection reset by peer) |
| 16:21:24 | * | gradha joined #nimrod |
| 16:21:49 | gradha | build.nimrod-code.org seems to be down |
| 16:22:09 | gradha | dom96: builder failed with http://pastebin.com/vfP9EdFT, is that ok or should be the builder resilient to this? |
| 16:25:49 | * | gradha quit (Read error: Connection reset by peer) |
| 16:26:24 | * | gradha joined #nimrod |
| 16:27:49 | * | Zerathul quit (Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]) |
| 16:48:39 | dom96 | gradha: Yeah, sure go ahead and do that. |
| 16:48:54 | dom96 | Nimbuild is back up now. |
| 16:49:37 | dom96 | gradha: The builder should be resilient. |
| 16:49:40 | * | dom96 will fix it |
| 16:50:31 | gradha | before rebuilding the builder I will try at some point to bring the laptop to sleep state, I think it might have had something to do with the cpu usage |
| 16:51:51 | gradha | usually time dependant software goes crazy due to hibernation |
| 17:39:38 | * | gradha quit (Ping timeout: 255 seconds) |
| 17:41:08 | * | gradha joined #nimrod |
| 17:43:19 | * | Zerathul joined #nimrod |
| 17:48:12 | * | gradha quit (Ping timeout: 250 seconds) |
| 18:11:57 | * | gradha joined #nimrod |
| 18:22:58 | gradha | dom96: I think I found out when the cpu usage spikes, it's when the builder timeouts |
| 18:24:22 | gradha | want me to create some github issues for tracking? |
| 18:25:59 | dom96 | sure |
| 18:27:54 | gradha | the cpu spikes after the first ping timeout, and this is using the 2000 timeout too |
| 18:29:00 | dom96 | Does the CPU usage go down after it reconnects? |
| 18:29:33 | gradha | I don't think so, but it's something to test |
| 18:34:18 | gradha | made d-code/nimbuild/issues/6 |
| 18:34:24 | gradha | err https://github.com/nimrod-code/nimbuild/issues/6 |
| 18:37:35 | gradha | what's the host the builder tries to connect to? I can block it in /etc/hosts and see what it does |
| 18:38:27 | dom96 | whatever hubAddr is set to in your config |
| 18:38:51 | dom96 | I've never seen "nodename nor servname provided, or not known" before |
| 18:38:55 | dom96 | Any ideas what that is? |
| 18:39:10 | gradha | no idea |
| 18:39:59 | dom96 | interesting |
| 18:41:06 | dom96 | This seems like the same reconnection bug NimBot is suffering from :\ |
| 18:41:29 | gradha | bug in the stdlib? |
| 18:41:58 | dom96 | perhaps |
| 18:42:37 | gradha | ah, crap, blocking /etc/hosts doesn't work because the connection is already open |
| 18:42:51 | gradha | the problem happens when the connection breaks, I gues |
| 18:43:25 | dom96 | I will try to reproduce this after dinner. bbl. |
| 18:52:02 | * | Trix[a]r_za is now known as Trixar_za |
| 18:57:47 | dom96 | back |
| 18:58:15 | gradha | just in time for an update of https://github.com/nimrod-code/nimbuild/issues/6 |
| 18:58:44 | dom96 | cool |
| 18:59:52 | gradha | maybe the first step would be to verify if this is a macosx only issue or you can reproduce the problem on your side, since we know what triggers it |
| 19:02:08 | gradha | btw, the unhandled exception seems to be a default message for dns reachability problems, it happens also in another program I wrote that connects to the web http://pastebin.com/6EZRQWVx |
| 19:03:09 | dom96 | ahh, I suppose it makes sense since you basically disconnected from the internet and DNS is the first thing it tries to connect to. |
| 19:03:18 | * | dom96 would have thought this stuff is cached though |
| 19:03:49 | dom96 | gradha: Please try disconnecting your WiFi and then launching the builder. |
| 19:04:05 | gradha | that's https://github.com/nimrod-code/nimbuild/issues/7 |
| 19:04:14 | dom96 | It should keep running and try to connect every 5 seconds |
| 19:04:25 | dom96 | oh |
| 19:05:22 | gradha | in any case I'll revert the timeout value changes I had locally, now that we know they don't help |
| 19:08:45 | gradha | blocking your domain in /etc/hosts doesn't fail with an exception, the builder simply starts and quits, but no error is generated |
| 19:08:55 | gradha | so /etc/hosts is different from actual wifi disconnection |
| 19:10:54 | Araq | exhu: the compiler never uses the cwd for the configuration, it uses the *.nimrod.cfg in the same dir as the project.nim |
| 19:10:58 | dom96 | that's weird, does the builder echo anything in that case? |
| 19:11:02 | dom96 | I can reproduce this btw |
| 19:11:14 | Araq | are you sure it's because of some commit I made? |
| 19:11:28 | reactormonk | Working knowledge of at least one programming language: C#, C++, C, Java, JavaScript, Basic, HTML # Good ol' craigslist :D |
| 19:11:41 | exhu | Araq, http://build.nimrod-code.org/docs/nimrodc.html |
| 19:11:47 | gradha | dom96: when /etc/hosts is blocked it only says "Started builder: built at 2013-03-10 20:05:59" |
| 19:12:31 | dom96 | gradha: hrm, could you debug it using endb? |
| 19:12:34 | exhu | Araq, i've created issue, there's some weird bug with extension, i.e. it works when full filename is specified. |
| 19:12:53 | gradha | what do you want me to look for? |
| 19:12:56 | Araq | ok but that's not a new bug, is it? |
| 19:13:23 | dom96 | gradha: Try to figure out why it exits silently. |
| 19:13:28 | dom96 | actually |
| 19:13:43 | gradha | the exit status code of the process is 141 if it helps |
| 19:13:46 | dom96 | no, my hunch is wrong |
| 19:14:42 | * | Zerathul quit (Ping timeout: 256 seconds) |
| 19:15:14 | dom96 | gradha: how did you block it in /etc/hosts? |
| 19:15:28 | gradha | You can add "0.0.0.0 domain.name" |
| 19:16:00 | gradha | I have a really long /etc/hosts with ad servers blocked in such a way, works like magic |
| 19:16:27 | exhu | Araq, https://github.com/Araq/Nimrod/issues/353 |
| 19:16:35 | gradha | I've been thinking a whole lot about automatic the process though, now it takes me to open the web inspector, see what domain is serving stupid stuff and manually edit the file |
| 19:16:46 | gradha | it would be so much nicer with some browser plugin |
| 19:17:04 | gradha | but then, have a plugin with admin pass to change /etc/hosts, meh |
| 19:17:58 | gradha | the endb manual doesn't say how to pass the process command line arguments |
| 19:18:01 | dom96 | gradha: hrm, the issue you are having with /etc/hosts and the builder seems to be Mac OS X only. |
| 19:18:04 | exhu | Araq, it's not about commit, it seems i used full filename before it triggered the bug -) |
| 19:18:37 | gradha | nice, segmentation fault: 11 from endb |
| 19:18:37 | Araq | exhu: good :-) |
| 19:18:59 | Araq | hey, endb is pure gold and never crashes :P |
| 19:19:15 | dom96 | I blame Apple! :P |
| 19:19:18 | gradha | Program received signal EXC_BAD_ACCESS, Could not access memory. |
| 19:19:19 | gradha | Reason: 13 at address: 0x0000000000000000 |
| 19:19:19 | gradha | 0x0000000100057855 in endb () |
| 19:19:19 | gradha | (gdb) bt |
| 19:19:19 | gradha | #0 0x0000000100057855 in endb () |
| 19:19:19 | gradha | #1 0x000000010007cd1f in systemInit () |
| 19:19:21 | gradha | #2 0x00000001000910f7 in NimMain () |
| 19:19:23 | gradha | #3 0x0000000100092737 in main () |
| 19:19:25 | gradha | ok, ok, Apple is to blame |
| 19:20:02 | dom96 | Try gdb I guess |
| 19:20:58 | reactormonk | Araq, would you mind making the default target for nimrod js node? |
| 19:21:11 | Araq | reactormonk: yeah I would |
| 19:21:25 | Araq | can't see the point of running nimrod over node except for testing things |
| 19:21:30 | reactormonk | oke |
| 19:21:50 | reactormonk | TypeError: Cannot read property 'prev' of null |
| 19:21:53 | reactormonk | O.o I see |
| 19:22:16 | gradha | got http://pastebin.com/2kAyz6G2 with gdb, what's the switch/option to make it find line numbers? |
| 19:22:38 | dom96 | --debuginfo iirc |
| 19:22:57 | dom96 | oh and --lineDir:on |
| 19:23:33 | * | gour quit (Disconnected by services) |
| 19:23:33 | * | gour_ joined #nimrod |
| 19:23:58 | gradha | ok, much better http://pastebin.com/35M8Tmy2 |
| 19:27:02 | dom96 | weird. |
| 19:27:14 | dom96 | is that right before it exits? |
| 19:27:37 | gradha | Program received signal SIGPIPE, Broken pipe |
| 19:27:44 | gradha | that's the actual gdb error |
| 19:28:17 | exhu | call for plumber -) |
| 19:29:00 | dom96 | ahh |
| 19:29:20 | dom96 | http://stackoverflow.com/a/450130/492186 |
| 19:29:24 | dom96 | That's what we need to do. |
| 19:30:09 | dom96 | or perhaps the equivalent to MSG_NOSIGNAL needs to be used for send() on Mac OS X and Windows. |
| 19:30:18 | dom96 | It seems strange that Mac OS X and Windows share code here |
| 19:30:26 | gradha | I have added some echos to that call and the socket has a file descriptor of 3, and the program is sending |
| 19:30:27 | gradha | sendAsync data 0x10021c050"{ \"name\": \"builder\", \"platform\": \"macosx-x86_64\", \"version\": \"1\", \"pass\": \"nimbuildisawesome\"}\13\10" |
| 19:30:27 | gradha | "" |
| 19:30:27 | gradha | sendAsync len 99 |
| 19:30:34 | Araq | well there is no MSG_NOSIGNAL on mac afaik |
| 19:30:36 | gradha | oh, crap |
| 19:30:59 | gradha | dammit, didn't notice what was being sent |
| 19:31:01 | gradha | sorry dom96 |
| 19:31:13 | dom96 | what are you sorry about? |
| 19:31:50 | dom96 | oh I see. |
| 19:31:55 | dom96 | Don't worry about it |
| 19:32:51 | dom96 | Araq: Shouldn't Nimrod print a traceback on SIGPIPE? |
| 19:33:45 | Araq | I never heard of SIGPIPE so system.nim doesn't handle it :P |
| 19:34:18 | dom96 | No wonder those pipes feel neglected and seek attention... |
| 19:34:25 | gradha | never played mario bros? |
| 19:35:29 | dom96 | gradha: Please find the equivalent of MSG_NOSIGNAL for Mac OS X and make a pull request |
| 19:36:29 | gradha | hrm... sure, when I find out what that means |
| 19:37:12 | dom96 | it suppresses the SIGPIPE signal during a send() call. |
| 19:37:30 | gradha | I'll note down the stackoverflow answer and this dialog for later |
| 19:37:58 | dom96 | Araq: Maybe it would be a good idea to find a list of all signals and handle them all? :P |
| 19:45:07 | reactormonk | Araq, any way to make nimrod return a non-null exit code when the code via -r fails? |
| 19:45:51 | Araq | reactormonk: sure, patch it |
| 19:46:39 | Araq | it's in compiler/nimrod.nim, make it inc(gErrors) when 'execProcess' fails |
| 19:46:51 | Araq | (or similar) |
| 19:48:05 | reactormonk | Araq, you use execExternalProgram which does not return anything :-/ |
| 19:48:20 | reactormonk | let's make it. |
| 19:49:51 | reactormonk | Araq, ye sure it's gErrors? |
| 19:50:33 | reactormonk | gErrorCounter more likely. |
| 19:51:32 | Araq | whatever man |
| 19:51:40 | reactormonk | doesn't help. Still returns 0. |
| 19:52:31 | Araq | maybe node is too blame? |
| 19:52:48 | reactormonk | yep. |
| 19:54:00 | Araq | dom96: I used the ansi C list of signals iirc |
| 19:54:20 | Araq | never occured to me that pipes trigger a signal handler |
| 19:54:29 | Araq | braindead if you ask me |
| 19:54:56 | Araq | this "we don't need exception in C let's use signals" is a bad unix joke |
| 19:55:49 | gradha | aren't signals just a "hack" for programs without event loops? |
| 19:58:58 | reactormonk | Araq, nope. Used $! instead of $? |
| 19:59:18 | reactormonk | Araq, remember when it worked? |
| 20:01:50 | * | Zerathul joined #nimrod |
| 20:08:27 | NimBot | nimrod-code/nimbuild b6faf7a Dominik Picheta [+0 ±3 -0]: Fixes #6 and fixes #7 |
| 20:08:44 | dom96 | gradha: ok, try it now. |
| 20:15:20 | * | gradha quit (Read error: Connection reset by peer) |
| 20:15:27 | * | gradha_ joined #nimrod |
| 20:15:48 | gradha_ | this time the CPU stayed 0 while getting messages like Waiting 5 seconds before reconnecting... |
| 20:15:48 | gradha_ | Could not reconnect: nodename nor servname provided, or not known |
| 20:16:02 | gradha_ | however, turning on the wifi made the process exit without message |
| 20:16:06 | gradha_ | maybe another sigpipe? |
| 20:16:25 | gradha_ | I'll try now with connected wifi but disconnected router |
| 20:16:34 | dom96 | it reconnected fine for me |
| 20:18:52 | * | gradha_ quit (Read error: Connection reset by peer) |
| 20:18:54 | * | exhu quit (Quit: Leaving) |
| 20:25:12 | * | Trixar_za is now known as Trix[a]r_za |
| 20:25:56 | reactormonk | Araq, did it _ever_ work? |
| 20:26:50 | Araq | reactormonk: no idea |
| 20:27:33 | * | gradha joined #nimrod |
| 20:28:35 | gradha | dom96: it keeps dying on reconnection here, so maybe it's the SIGPIPE happening, hmm, I'll try once more with gdb |
| 20:28:41 | reactormonk | Araq, apparently nope |
| 20:30:46 | reactormonk | Araq, and it does return a non-zero exit code btw |
| 20:31:35 | Araq | reactormonk: there is a known bug in osproc that I wanted to fix after 0.9.2 because it can break stuff |
| 20:31:57 | * | gradha quit (Read error: No route to host) |
| 20:32:07 | * | gradha joined #nimrod |
| 20:33:00 | gradha | interesting, under gdb it doesn't exit, but I get loops like these: |
| 20:33:00 | gradha | Waiting 5 seconds... |
| 20:33:00 | gradha | Incorrect welcome message from hub |
| 20:33:05 | reactormonk | Araq, relation? |
| 20:34:03 | gradha | ah, no, those messages were because I was running the gdb proc and another one at the same time |
| 20:39:07 | * | XAMPP_ quit (Quit: My code has no bug's, just random features) |
| 20:46:25 | * | XAMPP joined #nimrod |
| 20:51:49 | reactormonk | Araq, argh. reconsidering, the bug is rather easy to spot imo |
| 20:51:59 | reactormonk | http://sprunge.us/LIKO |
| 20:55:22 | Araq | reactormonk: I never said it was hard |
| 20:55:40 | reactormonk | Araq, but I don't know which would be the right way |
| 20:57:10 | reactormonk | Araq, I'd go for using the one in finally, but I'm not sure if your exception handlers are somehow stateful |
| 20:58:33 | Araq | the tests I mentioned will tell you if you got it right :P |
| 20:59:05 | dom96 | gradha: I guess you should try to fix the SIGPIPE problem first. |
| 21:16:16 | gradha | I have type A and subtype B, now I want to assign A to B and theres a type mismatch |
| 21:16:34 | gradha | is there any shortcut to say that I want fields of A copied into B or do I have to write my own assignment proc? |
| 21:18:31 | Araq | the later |
| 21:18:45 | Araq | you can use the new 'fields' iterator to cut the boilerplate though |
| 21:20:02 | gradha | would work casting the B type to A inside the assignment proc? |
| 21:22:06 | Araq | I suppose that could work |
| 21:22:45 | gradha | yay, it does |
| 21:22:56 | gradha | no need to write my own proc then |
| 21:24:46 | * | gour joined #nimrod |
| 21:26:48 | * | gour_ quit (Ping timeout: 245 seconds) |
| 21:48:45 | reactormonk | Araq, the first one runs by now, not sure if the fix crashes other stuff |
| 21:52:37 | Araq | well you know ... the tests contain the desired output |
| 21:53:45 | reactormonk | Araq, yeah, yeah |
| 21:55:08 | gradha | Araq: did you read my message about ssh? |
| 21:57:47 | reactormonk | Araq, how do I watch a variable in endb, aka whenever it gets reassigned? |
| 21:58:41 | Araq | reactormonk: {.watchpoint: myvar.} after it's been declared |
| 21:58:53 | Araq | but endb only works for the C backend of course |
| 21:58:59 | reactormonk | ... meh |
| 21:59:08 | reactormonk | wait. that's as intended. |
| 22:00:15 | reactormonk | Araq, can I do that for fields as well? |
| 22:00:27 | reactormonk | ... aka all fields of any of the objects created |
| 22:01:35 | Araq | reactormonk: it tracks the full object |
| 22:02:15 | * | Zerathul quit (Read error: No route to host) |
| 22:03:20 | Araq | but only the most recently created object I think |
| 22:09:55 | reactormonk | Araq, I want to watch all nestedTryStmts assignments |
| 22:10:22 | Araq | why? |
| 22:10:52 | Araq | the whole codegen is in jsgen.nim? |
| 22:11:32 | Araq | you can easily grep for it in that file |
| 22:11:45 | reactormonk | sure I can. But I want to watch the code run |
| 22:12:10 | Araq | well endb doesn't support that |
| 22:12:23 | reactormonk | :-( |
| 22:12:27 | Araq | or maybe it does |
| 22:12:33 | reactormonk | \o/ |
| 22:15:01 | Araq | try to add {.watchpoint: p.nestedTryStmts.} to after line 1464 then |
| 22:15:13 | Araq | cause that's where a new 'p: TProc' is created |
| 22:16:18 | Araq | but it's only modified in genTryStmt anyway |
| 22:16:34 | Araq | no need to run it through it to a debugger to see that |
| 22:16:53 | reactormonk | oh, there's two nestedTryStmts... |
| 22:20:29 | reactormonk | Araq, now the bootstrapping with koch goes haywire |
| 22:24:48 | Araq | you can't bootstrap with --debugger:on ... |
| 22:24:55 | Araq | for obvious reasons |
| 22:27:23 | reactormonk | aww |
| 22:37:24 | Araq | gradha: is #351 a recent bug? |
| 22:37:33 | Araq | I remember it to work :P |
| 22:37:40 | reactormonk | git bisect! |
| 22:40:36 | gradha | does github show dates? |
| 22:40:58 | gradha | yes, happened yesterday, IIRC |
| 22:41:02 | Araq | you shouldn't date other githubbers |
| 22:41:16 | gradha | crap, now all my intimate life exposed |
| 22:42:31 | gradha | hmm... the git commit I'm using is f25b85a7341682024f12ffc524e657ae599e587a from 4th of March, want me to update? |
| 22:42:54 | Araq | hrm now that I'm misusing gradha's machine I notice the repo really is too big ... |
| 22:43:22 | gradha | nah, that's likely I'm torrenting too much porn |
| 22:43:55 | dom96 | Araq: |
| 22:44:04 | dom96 | Told you :P |
| 22:45:01 | gradha | it's really easy to explain, but difficult to understand: I download videos from internet, reencode them, and send them to friends |
| 22:45:05 | dom96 | It's a lot more prominent when I am testing nimbuild for example. |
| 22:46:04 | Araq | yeah we should prune it |
| 22:46:54 | Araq | 267MB is too large |
| 22:47:33 | gradha | will you split the repo from the binary sources? |
| 22:47:36 | gradha | 267MB is not even half the size of any video I'm downloading right now |
| 22:48:09 | gradha | more likely the problem is small files which are slower to create/manage in the fs |
| 22:53:28 | dom96 | Still think the Aporia repo should include all those gtk DLLs? :P |
| 22:53:29 | Araq | yes because 20MB ain't 200 |
| 22:53:29 | Araq | and they are static anyway |
| 22:53:29 | dom96 | what if a new version of GTK comes out? |
| 22:53:30 | Araq | I wouldn't touch a working GTK on windows |
| 22:53:32 | Araq | so I wouldn't update them but tell people how to do it |
| 22:53:32 | dom96 | What if Aporia requires the new version? :P |
| 22:53:32 | Araq | it's likely the changes won't affect aporia anyway |
| 22:53:32 | Araq | aporia shouldn't do that :P |
| 22:53:32 | dom96 | what if it's forced to? |
| 22:53:37 | Araq | what if git becomes cool one day and lets you delete things without breaking forks? |
| 22:53:39 | dom96 | in that case I will consider it. |
| 22:53:39 | Araq | yeah and let your users suffer in the meantime |
| 22:53:39 | Araq | :P |
| 22:53:39 | Araq | but it's fine anyway |
| 22:53:39 | Araq | we have a working gtk.zip on the server now |
| 22:53:59 | Araq | the problem is the generated C code gradha |
| 22:54:13 | Araq | it's too big so we zip it |
| 22:54:35 | Araq | we also don't want github to make it browsable |
| 22:55:13 | gradha | you mean the reason my machine is slow? |
| 22:55:19 | gradha | ah, right, the language issue |
| 22:56:39 | gradha | deletion of history implies different commit checksums, so you won't see that feature in this universe |
| 22:57:29 | dom96 | git should simply learn how to manage binaries properly |
| 22:57:55 | Araq | or be able to deal with files that shouldn't get any history |
| 22:58:40 | gour | </q |
| 22:58:44 | * | gour quit (Quit: WeeChat 0.4.0) |
| 22:59:05 | dom96 | it would be nice if we could say "This file should not be stored in history, only latest version of it should be stored in the git repo |
| 22:59:05 | Araq | it doesn't matter if it's "binary" or not (what does that mean anyway?) |
| 22:59:05 | Araq | dom96: exactly |
| 22:59:06 | gradha | I wonder if there's any versioning system doing that |
| 22:59:22 | gradha | why do you store it in the repo and not download it from 3rd party host? |
| 23:00:22 | gradha | also, wouldn't that be a problem if you needed to go back to a previous version? |
| 23:00:33 | dom96 | Because it's convenient to do so. Also nimbuild detects when it is changed in the repo and automatically builds the csources first. |
| 23:01:07 | dom96 | hrm, not sure if that would ever be required. |
| 23:02:01 | gradha | can't see the problem of having an extra_sources.ini file with the URL for nimbuild to download, changing the .ini file contents would trigger recompilation too |
| 23:03:00 | Araq | dom96: actually build.sh does change too then or at least can easily be made to do so |
| 23:07:16 | dom96 | in that case we can do that |
| 23:10:50 | gradha | there you go http://forum.nimrod-code.org/t/135 |
| 23:10:58 | gradha | just in time before the zerg invasion |
| 23:11:07 | gradha | now humanity at least has a chance |
| 23:13:36 | dom96 | oh and with a pull request to babel too. Very nice. |
| 23:14:55 | NimBot | nimrod-code/packages dcdfcae Grzegorz Adam Hankiewicz [+0 ±1 -0]: Adds argument_parser module. |
| 23:14:55 | NimBot | nimrod-code/packages 50ca806 Dominik Picheta [+0 ±1 -0]: Merge pull request #7 from gradha/pr_adds_argument_parser_module... 3 more lines |
| 23:15:26 | gradha | oh man, and here I was expecting weeks for approval in the babel appstore |
| 23:16:25 | dom96 | nah, here at the babel appstore we believe in killing anyone who tries to abuse our appstore. |
| 23:18:48 | dom96 | As soon as you submit a package to babel, you accept this liability. |
| 23:19:16 | gradha | sounds good to me |
| 23:23:56 | gradha | isn't the gem like feature request at https://github.com/Araq/Nimrod/wiki/Feature-suggestions done? |
| 23:24:37 | dom96 | true. I'm not sure the wiki is the best place for feature requests. |
| 23:25:56 | gradha | at some point there will also be an overlap between https://github.com/Araq/Nimrod/wiki/Community-Projects and babel-packages |
| 23:32:33 | Araq | dom96: does nimbuild run the tests with a debug version of the compiler or with a release version? |
| 23:32:59 | dom96 | Araq: release IIRC |
| 23:34:58 | dom96 | gradha: make sure to git tag when you release a new version |
| 23:35:22 | dom96 | gradha: so that babel can install a specific version by simply using the git tags. |
| 23:37:25 | dom96 | Araq: does that matter much? |
| 23:37:54 | gradha | what fomat do you suggest, v0.1.0? |
| 23:38:32 | dom96 | yeah |
| 23:40:55 | gradha | Araq: look, my repo is 100% nimrod |
| 23:42:39 | dom96 | huh, https://github.com/languages/Nimrod |
| 23:43:00 | dom96 | why is nimbuild before gradha's repo in recently updated repositories? |
| 23:43:38 | gradha | maybe they update these stats every n hours |
| 23:43:51 | Araq | yay 104th out of 110 languages! |
| 23:45:44 | Araq | so ... the good news is: the m&s gc fails for linux_x86 too |
| 23:45:51 | dom96 | They bumped the version quite a lot since reactormonk's pull request was merged: https://github.com/github/linguist/commits/master |
| 23:46:10 | dom96 | I guess that doesn't mean they updated it on their servers. |
| 23:46:10 | reactormonk | dom96, maybe my fix didn't work as intended? |
| 23:46:29 | dom96 | reactormonk: I tested it a bit. |
| 23:46:42 | dom96 | A couple of days ago |
| 23:46:44 | dom96 | It seems to work |
| 23:47:08 | dom96 | Although to be honest I couldn't tell properly, because linguist seems to crash when there are any binaries at all in the folder. |
| 23:47:16 | reactormonk | WTF? |
| 23:47:20 | reactormonk | dom96, crash message? |
| 23:48:25 | dom96 | https://gist.github.com/dom96/5b2d25c7f4ec5e624990 |
| 23:49:18 | Araq | the bad news titer8 fails for reasons that puzzle me ... |
| 23:50:10 | reactormonk | oh, they switched to charlock holmes |
| 23:50:17 | reactormonk | which means they likely kicked my code |
| 23:50:37 | dom96 | omg, so it's broken again? |
| 23:50:50 | reactormonk | dom96, That means charlock holmes did not install correctly. |
| 23:51:48 | reactormonk | dom96, got me some example code? |
| 23:51:56 | dom96 | reactormonk: how do I install it correctly? |
| 23:52:15 | dom96 | I simply did: bundle exec linguist samples/ |
| 23:52:19 | reactormonk | hmm |
| 23:53:36 | reactormonk | bundle exec linguist samples/Nimrod |
| 23:53:39 | reactormonk | 100% Nimrod |
| 23:53:56 | dom96 | yeah, that works for me. |
| 23:54:54 | reactormonk | http://sprunge.us/BIdF |
| 23:55:27 | dom96 | I suppose my installation is just broken. |
| 23:55:38 | dom96 | oh the joys of package managers. |
| 23:55:45 | reactormonk | Araq, wanna know why it displays as C? |
| 23:56:02 | Araq | why? |
| 23:56:24 | reactormonk | hmm, wait. I unpacked csources there |
| 23:56:59 | reactormonk | ... still WTF. |
| 23:57:58 | * | gradha quit (Quit: bbl, have youtube videos to watch) |