[00:45:01] --- Simon Wilkinson has become available [01:23:07] --- Simon Wilkinson has left [01:34:42] --- Simon Wilkinson has become available [02:22:46] --- Simon Wilkinson has left [02:58:24] --- Simon Wilkinson has become available [04:11:38] --- Marc Dionne has become available [06:08:01] --- mvita has become available [09:17:30] --- deason/gmail has become available [10:46:30] OpenAFS 1.6.x overwrites the standard assert macro in the public header afs_assert.h with a macro that can’t be used in expression context, which makes python-afs (Cython) sad. This was fixed in master commit cac7424 by removing afs_assert.h. Is that a reasonable change for 1.6.x, or do we want a smaller fix? [10:47:52] I thought our code was supposed to use opr_Assert (or osi_Assert) [10:48:06] assertion everywhere other than on master is a mess [10:50:12] I just want it to become a mess that doesn’t infect other applications via the public headers. [10:51:43] If nothing else, we should -#define assert(ex) do{if (!(ex)) AssertionFailed(__FILE__, __LINE__);}while(0) +#define assert(ex) ((ex) ? (void)0 : AssertionFailed(__FILE__, __LINE__)) though I think I’d prefer not exposing our assert() at all? [10:52:36] Not exposing our assert() at all would be good. But I think any patch to fix this that's designed for 1.6 would have to do so in a way that guaranteed that all of our binaries that use assert() and expect special behaviour still receive it. [11:25:53] --- Simon Wilkinson has left [11:56:15] --- ballbery has left [11:56:47] --- ballbery has become available [12:06:21] Okay, 10096 is my proposed solution for 1.6.x (depends on 10095). [12:10:16] --- ballbery has left [12:11:17] --- ballbery has become available [12:16:54] --- Simon Wilkinson has become available [12:19:56] --- Ken Dreyer has left [12:31:42] --- Ken Dreyer has become available [12:32:39] --- Ken Dreyer has left [12:33:31] --- Ken Dreyer has become available [12:33:41] --- Ken Dreyer has left [12:45:13] --- Simon Wilkinson has left [12:45:13] --- Simon Wilkinson has become available [12:45:13] --- Simon Wilkinson has left: Lost connection [12:50:40] --- Simon Wilkinson has become available [13:19:07] --- meffie has become available [14:42:57] --- meffie has left [14:58:58] --- mvita has left [15:53:40] --- deason/gmail has left [20:17:15] --- ballbery has left [20:23:27] --- ballbery has become available [20:37:34] --- Marc Dionne has left [21:45:39] --- ballbery has left [22:05:29] --- ballbery has become available [22:58:08] --- Simon Wilkinson has left [22:58:24] --- Simon Wilkinson has become available [23:10:31] --- Simon Wilkinson has left