Da
Show enters and exits. Hide enters and exits.
| 00:01:17 | marcandre | OK, it's split... |
| 00:02:06 | marcandre | Ah, missing a nil?. |
| 00:03:16 | marcandre | Cool. That was it. Thanks for the help! |
| 00:11:34 | evan | marcandre: weird. |
| 00:11:48 | evan | ok |
| 01:29:41 | marcandre | Is the current rbx broken or did I do something wrong? mspec core/enumerable/all_spec.rb -t rbx gives 4 errors. |
| 01:29:48 | marcandre | All "should raise_error(...)" seem to fail for me. |
| 01:30:05 | marcandre | Expected ...Error but no exception was raised |
| 01:50:15 | marcandre | Weird. bin/mspec from rbx is ok, but not my (up to date) mspec. |
| 01:50:29 | marcandre | There's some really basic thing that's escaping me. |
| 02:53:41 | marcandre | Can anyone try "mspec core/enumerable/all_spec.rb -t rbx" with the current head of mspec, rubyspecs and rubinius? |
| 03:21:25 | marcandre | evan: ping |
| 04:37:54 | boyscout | Used Undefined instead of nil for optional arguments of Strign#initialize, #split, #gsub, #gsub!, #sub & #gsub - b8d6262 - Marc-Andre Lafortune |
| 04:37:54 | boyscout | 1.8.7: String#gsub, #each, #each_line, #each_byte now have optional blocks - 6f7b355 - Marc-Andre Lafortune |
| 04:37:54 | boyscout | 1.8.7: String#end_with?, #start_with? (implemented new methods) - 6b0ec71 - Marc-Andre Lafortune |
| 04:41:46 | boyscout | CI: 6b0ec71 success. 2682 files, 10332 examples, 32881 expectations, 0 failures, 0 errors |
| 04:46:32 | perlooooooooo | why a timeout could happen when we use agent.get(url) ???? |
| 04:46:35 | perlooooooooo | thnals |
| 14:05:02 | rue | http://murmurhash.googlepages.com/ |
| 14:05:20 | rue | 2.0 version is, apparently, ~52 instructions on x86 :P |
| 14:47:11 | tilman | rue: the simple version isn't portable though (unaligned accesses) |
| 15:38:09 | rue | But, but, ~52 instructions! |
| 16:47:11 | sandal | Anyone want to help with a newbie FFI question? |
| 16:47:36 | sandal | I'm wrapping a function that takes a char* buffer and writes to it |
| 16:48:02 | sandal | I can't figure out the right way to deal with this |
| 16:48:13 | sandal | as just passing in a string doesn't capture the side effect |
| 16:48:29 | sandal | I've been playing around with FFI::Pointer but not sure of the exact setup |
| 16:54:37 | yakischloba | sandal: MemoryPointer? |
| 16:56:31 | sandal | yakischloba: yep, just got it to work thanks! |
| 16:56:42 | yakischloba | cool |
| 16:56:58 | sandal | I just had to tweak the parameter to be of type :pointer instead of :string |
| 16:58:54 | yakischloba | right |
| 18:25:13 | rue | Hm, is it looking for a char** instead, then? |
| 18:25:49 | rue | Or I suppose it could be that the string does not re-evaluate itself |
| 18:26:28 | yakischloba | are you supposed to be able to pass a string in as a char* and have the string be modified if thats what it does? |
| 18:26:54 | yakischloba | i hadn't encountered that situation before, but it sounded like thats what hes talking about |
| 18:27:04 | rue | A :string *is* char* |
| 18:27:17 | yakischloba | right |
| 18:27:42 | rue | So if so desired, it could be modified |
| 18:27:53 | yakischloba | but if my function writes "foo" to a passed pointer, and i attach_function :func, [:string], :void |
| 18:27:56 | yakischloba | can I then do |
| 18:28:00 | rue | (It could be const char*, too, but it is somewhat less general-purpose) |
| 18:28:15 | yakischloba | str = "barbaz"; func(str); str == "foobaz" ? |
| 18:28:55 | rue | It depends on exactly the char* <-> String mapping is done. Your function can certainly alter the string it gets |
| 18:29:42 | yakischloba | So it is dependent on ffi/ruby impl? |
| 18:29:57 | rue | But if the conversion dups the String you give it, then obviously it will not reflect the changes back |
| 18:30:24 | rue | It depends on what the conversion mechanism is |
| 18:31:30 | tilman | at least you cannot have ffi functions fill in buffers you pass though |
| 18:31:47 | tilman | a la readlink(2) |
| 18:35:46 | rue | Well, you can if you give a :pointer to a buffer |
| 18:36:51 | tilman | oh? |
| 18:37:02 | tilman | takes a note |
| 18:37:25 | yakischloba | tilman: otherwise it would be quite useless, no? ;) |
| 18:43:09 | rue | And giving a char* with some memory allocated works too, on the C side...apparently it just does not propagate back right |
| 18:43:52 | rue | Although maybe it would be neater to use char** anyway |
| 18:45:34 | evan | I changed it at one point |
| 18:45:56 | evan | passing a :string in copies the String#data into a local buffer and passes that to the funtction |
| 18:46:10 | evan | if we pass the internal String#data directly, it could easily break |
| 18:46:11 | yakischloba | i see. |
| 18:46:17 | evan | because the GC could move the String object |
| 18:46:35 | evan | if you want it to be mutated, you use a MemoryPointer |
| 18:46:46 | evan | which always points to memory thats unmanaged and nonmoving |
| 18:48:07 | evan | ok, off to breakfast! |
| 18:48:19 | yakischloba | :) |
| 22:07:30 | boyscout | Adding rb_big2dbl to capi, with specs. - 37cf902 - Cezar Sa Espinola |
| 22:07:30 | boyscout | Adding rb_cProc constant to capi, specs updated. - e723e16 - Cezar Sa Espinola |
| 22:07:30 | boyscout | Adding rb_str2cstr to capi and fixing STR2CSTR, specs included. - 363f8fa - Cezar Sa Espinola |
| 22:07:30 | boyscout | Adding rb_range_new to capi, with specs. - 930a31b - Cezar Sa Espinola |
| 22:14:10 | boyscout | CI: 930a31b success. 2683 files, 10344 examples, 32915 expectations, 0 failures, 0 errors |