| <<<Back 1 day (to 2018/04/04) | 20180405 |
tor8 | Robin_Watts: ping. | 11:00.11 |
| Robin_Watts: ugh. I hate harfbuzz commit cbec0cd65e1787b20ea55dd5583a7444938bd381 | 15:02.24 |
| it no longer builds on vs2005 | 15:02.28 |
Robin_Watts | oh, anything in particular ? | 15:03.22 |
tor8 | it removes the typedefs for uint8_t and friends | 15:03.38 |
| assuming that stdint.h will always be available | 15:03.44 |
| I *hate* stdint.h and workarounds | 15:03.50 |
| it's headache #2 in our build system | 15:03.59 |
Robin_Watts | So do we have stdint_.h which either #defines or #includes "stdint.h" ? | 15:04.50 |
| s/have/add/ ? | 15:04.56 |
tor8 | Robin_Watts: doesn't really help the situation with thirdparty libraries | 15:05.10 |
Robin_Watts | tor8: Well, it's a single small commit we have to keep merging on a branch in thirdparty libs. | 15:05.36 |
tor8 | what we *could* do is add a 'include/vs2005-workaround/stdint.h' and put stuff in there and -I that directory | 15:05.38 |
Robin_Watts | tor8: We could. | 15:05.45 |
tor8 | it would be nice if we didn't have to keep our own branches in thirdparty libs though... | 15:05.58 |
Robin_Watts | It would. | 15:06.03 |
tor8 | I wonder if he could be persuaded to revert that commit | 15:06.08 |
Robin_Watts | The workaround include seems sensible. | 15:06.13 |
| tor8: I've spoken to him in the past, and he was... inflexible... in terms of wanting to pass a context through. | 15:06.58 |
| Apparently "using custom allocators is a bad idea". | 15:07.19 |
| "An inability of a library to use custom allocators is a huge shortcoming" would be my view, but... | 15:07.57 |
tor8 | yeah. stubborn is my impression last time I talked to the maintainers too. | 15:08.24 |
| anyway, gotta go. back in a couple of hours. | 15:09.18 |
Robin_Watts | tor8, sebras, fred: http://git.ghostscript.com/?p=user/robin/mupdf.git;a=commitdiff;h=46407fb9c49a6c4a30dbdf50347bbe60c3f598fb | 18:58.23 |
| That's a fix for a multithreaded store problem that fred ran into. | 18:58.35 |
| It seems to have cured his problem. | 18:58.47 |
fredross-perry | I'm testing and it looks good so far. | 18:59.00 |
| Forward 1 day (to 2018/04/06)>>> | |