| 01:27.47 | *** join/#stepcode hoiji (671b082a@gateway/web/cgi-irc/kiwiirc.com/ip.103.27.8.42) |
| 04:24.33 | hoiji | Hi, I am almost done with making cllazyfile thread safe. (just need to add comments and organize into commits) |
| 04:27.04 | hoiji | can anyone point me to some large step files ~300mb (similar to ferrari sharknose) so that I can see the performance improvements. |
| 04:30.42 | *** join/#stepcode hoiji (671b082a@gateway/web/cgi-irc/kiwiirc.com/ip.103.27.8.42) |
| 04:33.35 | brlcad | hoiji: I'll PM you one model you can test with |
| 05:13.51 | *** join/#stepcode hoiji (671b082a@gateway/web/cgi-irc/kiwiirc.com/ip.103.27.8.42) |
| 06:26.15 | *** join/#stepcode hoiji (671b082a@gateway/web/cgi-irc/kiwiirc.com/ip.103.27.8.42) |
| 06:43.18 | *** join/#stepcode hoiji (671b082a@gateway/web/cgi-irc/kiwiirc.com/ip.103.27.8.42) |
| 13:20.15 | *** join/#stepcode kesha (~kesha@14.139.122.114) |
| 14:37.14 | *** join/#stepcode hoiji (671b082a@gateway/web/cgi-irc/kiwiirc.com/ip.103.27.8.42) |
| 16:04.40 | starseeker | mpictor: BRL-CAD commit r61065 addresses the performance issue I was seeing with some large step files on the header read. |
| 16:05.36 | starseeker | if I get a chance I'll forward port it to the current github tree, but wanted to make sure you were aware of it |
| 16:16.16 | starseeker | suspects that "GenNodeArray" should probably morph into std::vector (or at most a wrapper around std::vector) |
| 16:17.05 | starseeker | actually, a lot of those "custom containers" probably should be garden variety STL these days... |
| 16:17.15 | starseeker | might simplify the code a bit too |
| 16:29.22 | brlcad | you can only do that when the stl container is fully encapsulated, not part of public API |
| 16:35.57 | brlcad | or if it's in an application front-end or only staticly compiled, otherwise it's a world of hurt |
| 16:36.54 | brlcad | (i.e., run-time incompatibilities, crashes, unpredicatable/incorrect container behavior, and more) |
| 16:37.51 | brlcad | easily would make a "top ten list of things to never do with c++" |
| 17:26.38 | *** join/#stepcode cwstirk (~charlie@c-107-2-138-189.hsd1.co.comcast.net) |
| 17:42.03 | *** join/#stepcode github-notify (~github-no@192.30.252.35) |
| 17:42.03 | github-notify | [stepcode] hoiji09 pushed 7 new commits to hj/cllazyfile-thread-safety: https://github.com/stepcode/stepcode/compare/343c9c54e2c5...313e99c7f350 |
| 17:42.03 | github-notify | stepcode/hj/cllazyfile-thread-safety b865666 hoiji09: Added test for opening multiple files in parallel... |
| 17:42.03 | github-notify | stepcode/hj/cllazyfile-thread-safety 22a6b90 hoiji09: Added a compatible wrapper for mutex lock / unlock... |
| 17:42.03 | github-notify | stepcode/hj/cllazyfile-thread-safety 0e2ab60 hoiji09: Removed static variable (string str) in sectionReader.cc... |
| 17:42.03 | *** part/#stepcode github-notify (~github-no@192.30.252.35) |
| 17:43.28 | hoiji | yay! I think I have completed my first milestone: (thread safety in cllazyloading) (albeit I am late). Off to clstepcore now :) |
| 18:28.35 | starseeker | defining what's "part of the public API" could get a bit tricky for stepcode, as matters stand |
| 18:30.12 | brlcad | not from an application / linking perspective, it's very clear cut |
| 18:31.17 | brlcad | basically, you do not want STL containers in something that is ever dynamically loaded |
| 18:37.26 | *** join/#stepcode cstirk (~charlie@c-107-2-138-189.hsd1.co.comcast.net) |
| 18:55.28 | starseeker | brlcad: as API, or used at all? |
| 19:02.14 | starseeker | stepcode shouldn't be implementing generic containers, imho... |
| 21:24.14 | *** join/#stepcode stirk (~charlie@c-107-2-138-189.hsd1.co.comcast.net) |
| 21:27.26 | *** join/#stepcode cwstirk (~charlie@c-24-9-78-79.hsd1.co.comcast.net) |
| 23:47.35 | starseeker | brlcad: I could use a suggestion on how to approach reworking r61065 from an encapsulation standpoint |
| 23:56.46 | *** join/#stepcode cwstirk (~charlie@c-107-2-138-189.hsd1.co.comcast.net) |