Da
Show enters and exits. Hide enters and exits.
| 00:36:21 | boyscout | CI: Built b85e5cf and failed. http://elle.fallingsnow.net:9292/rubinius-jit/builds/2 |
| 00:36:31 | evan | hm, ok. |
| 00:36:45 | evan | oh actually, thats right. |
| 00:37:14 | slava | hi evan |
| 00:37:18 | evan | hi slava! |
| 00:39:48 | evan | slava: when are you headed to NZ? |
| 00:40:18 | slava | 16th |
| 00:41:37 | evan | pretty soon now! |
| 00:42:09 | brixen | slava: great time to be going to the other seasonal calendar |
| 00:42:11 | evan | hm, no announcement that it finished... |
| 00:42:35 | slava | yep |
| 00:50:01 | boyscout | CI: rubinius - jit b85e5cf success. 3018 files, 11557 examples, 35632 expectations, 0 failures, 0 errors |
| 00:50:12 | evan | sweet. |
| 00:50:28 | evan | hm, probably need to set off the thing that built |
| 00:50:35 | evan | perhaps another : in there |
| 00:50:46 | evan | before the hash |
| 00:53:04 | brixen | sure |
| 01:04:59 | boyscout | CI: Built b85e5cf and failed. http://elle.fallingsnow.net:9292/rubinius-no-jit/builds/5 |
| 01:05:14 | evan | THWLP |
| 01:05:28 | evan | fuck |
| 01:05:31 | evan | integrity FAIL. |
| 01:05:36 | evan | you can't have the same repo. |
| 01:05:42 | evan | and the same branch |
| 01:05:51 | evan | (-.-) |
| 01:08:22 | dwaite | thinks that should be failingsnow |
| 01:09:36 | dwaite | (just for the CI failures) |
| 01:09:48 | evan | it will be ci.rubini.us eventually. |
| 01:09:55 | boyscout | CI: rubinius - jit: b85e5cf success. 3018 files, 11557 examples, 35632 expectations, 0 failures, 0 errors |
| 01:09:55 | evan | i'm just setting up the new one. |
| 01:19:13 | dwaite | failings now, the new continuous integration service in the cloud |
| 01:19:57 | evan | hah |
| 01:37:44 | brianmario | yo |
| 01:38:39 | brianmario | first, why did you guys choose c++? (honestly just curious, I'm working on a c++ project right now to force myself to relearn it) |
| 01:39:22 | brixen | it made a lot of code cleaner |
| 01:39:36 | brixen | you can check out the old C code |
| 01:40:13 | brianmario | yeah I'm liking that aspect of it as well |
| 01:40:35 | brianmario | makes it easier to design (IMO) |
| 01:40:51 | brianmario | so second, have you guys started working on the unicode stuff yet? for 1.8 and/or 1.9 ? |
| 01:41:58 | brixen | the m17n stuff will be done in one_nine branch |
| 01:42:03 | brixen | no, we haven't started |
| 01:42:11 | brianmario | k |
| 01:42:14 | brixen | other than the branch is there |
| 01:42:27 | brixen | and marcandre has already added a bunch of 1.9 core methods |
| 01:42:28 | brianmario | what do you guys use for string representation in c++? std::string ? or... |
| 01:42:38 | brixen | what do you mean? |
| 01:42:43 | evan | brianmario_: no no |
| 01:42:44 | evan | never. |
| 01:42:57 | evan | rubinius uses no standard library C++ things to implement Ruby things |
| 01:43:12 | evan | nor will it. |
| 01:43:54 | evan | brianmario_: we have a ByteArray class |
| 01:44:05 | evan | that is just a fixed size array of bytes you can read and write |
| 01:44:09 | evan | and String is built on top of that. |
| 01:44:33 | brianmario | I just spent all day trying to figure out some stupid unicode bug in my project, only to find that std::string uses a <char> - which may or may not be signed/unsigned depending on the compiler/platform right? |
| 01:45:19 | evan | no |
| 01:45:21 | evan | it's always signed. |
| 01:45:35 | brianmario | ok |
| 01:45:41 | brianmario | so I needed it to be unsigned |
| 01:45:51 | brianmario | I've thought about making my own class |
| 01:46:22 | evan | no, you don't have to do that |
| 01:46:27 | evan | string is a template class |
| 01:47:05 | brianmario | just noticed that too (back when I used c++ last, it didn't even have namespaces) |
| 01:47:11 | brianmario | is there a way to declare a specific instance to use an unsigned char instead? |
| 01:48:03 | evan | typedef basic_string<unsigned char> ustring |
| 01:48:13 | evan | then you have to use ustring everywhere. |
| 01:48:17 | brianmario | right |
| 01:49:06 | brianmario | hm |
| 01:49:50 | brianmario | so if I have "void someFunction(ustring buffer)" declared |
| 01:49:56 | brianmario | and they pass a std::string, what happens? |
| 01:50:03 | brianmario | compiler barfs? or will it cast |
| 01:50:42 | evan | barf. |
| 01:51:00 | brianmario | balls |
| 01:51:35 | brianmario | well anyway |
| 01:52:12 | brianmario | I have some code that'll turn 浅 into \u6d45 |
| 01:52:19 | brianmario | also supports surrogates |
| 01:52:36 | evan | neat |
| 01:52:52 | evan | this is C++ |
| 01:52:55 | brianmario | yeah |
| 01:52:57 | evan | so you can always just call .c_str() |
| 01:53:01 | evan | cast that to a unsigned char* |
| 01:53:03 | brianmario | that's what i'm actually doing |
| 01:53:21 | evan | signed char is a clusterfuck imho. |
| 01:53:25 | brianmario | but I spent almost all day finding out I had to :P |
| 01:53:27 | brianmario | no shit |
| 01:53:37 | evan | char should have been unsigned always |
| 01:53:44 | evan | and then they had a 8 bit integer type too |
| 01:53:53 | evan | that could be signed or unsigned |
| 01:54:03 | evan | double duty for that datatype has set the world back years |
| 01:54:40 | brianmario | yeah wasn't there some 15 year old linux kernel bug that was just found, that had to do with signed/unsigned chars? |
| 01:54:55 | evan | probably. |
| 01:55:00 | evan | well shit |
| 01:55:02 | evan | i'm outta here |
| 01:55:03 | evan | later. |
| 01:55:05 | brianmario | later |
| 01:55:10 | yakischloba | brianmario_: strcmp/strncmp |
| 01:55:27 | brianmario | yakischloba: ? |
| 01:55:39 | yakischloba | brianmario_: that was the kernel bug i told u about |
| 01:55:44 | brianmario | oooh |
| 01:55:46 | yakischloba | while we were at the bart |
| 01:55:49 | yakischloba | bar* |
| 01:55:54 | brianmario | yeah |
| 01:57:32 | yakischloba | u could ask joe for more info |
| 13:14:47 | rue | Wish people trying to bolt whatever onto MRI and failing would notice its limitations and move to something better :P |
| 17:20:53 | Zoxc | evan: why did you add an opcode for zsuper? |
| 23:42:07 | botanicus | Hey guys, how can I run rubinius in 1.9-compatible mode? (If it already exists?) |
| 23:51:12 | rue | It does not |