| 00:19.12 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 00:35.18 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 00:42.56 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 01:24.40 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 01:46.39 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 03:16.20 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 04:13.17 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 04:26.25 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 10:41.53 | *** join/#brlcad teepee_ (bc5c2134@unaffiliated/teepee) |
| 11:34.04 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) |
| 12:03.35 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) |
| 12:27.14 | *** join/#brlcad TheJollyRoger (~Peter@gateway/tor-sasl/thejollyroger) |
| 12:28.36 | *** join/#brlcad TheJollyRoger2 (~Peter@gateway/tor-sasl/thejollyroger) |
| 12:30.51 | *** join/#brlcad TheJollyRoger2 (~Peter@gateway/tor-sasl/thejollyroger) |
| 12:37.40 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) |
| 12:40.35 | *** join/#brlcad TheJollyRoger2 (~Peter@gateway/tor-sasl/thejollyroger) |
| 13:03.30 | *** join/#brlcad yorik (~yorik@2804:431:cfc7:b3d9:290:f5ff:fedc:3bb2) |
| 13:38.41 | *** join/#brlcad yorik (~yorik@2804:431:cfc7:d9f2:290:f5ff:fedc:3bb2) |
| 13:40.31 | cad_zulip | @**Sean**: Hi @**jaideep seth**, welcome and thank you for your interest! What brings you to BRL-CAD? |
| 13:41.25 | cad_zulip | @**Sean**: For what it's worth, to be an effective mentor, you'll need to be an active contributor. Youâre more than welcome to be a mentor, but youâll need to demonstrate you can be a capable contributor over the next couple weeks by submitting work, show you can be independently productive, show youâre familiar with the code, show you are or will be effective at giving advice to others, etc. |
| 13:42.31 | cad_zulip | @**Sean**: If you have a reference to your GSoC work (who was your mentor), that would be good to know. |
| 13:48.41 | TheJollyRoger | Yay, the bridge seems to be working ^_^ |
| 13:48.54 | cad_zulip | @**Sean**: :) |
| 13:49.01 | TheJollyRoger | Ahoy Cap'n! ^_^ |
| 13:49.24 | cad_zulip | @**Sean**: greetings! happy friday! |
| 13:49.34 | TheJollyRoger | Yarr! |
| 13:52.43 | TheJollyRoger | I gotta fix up some stuff but I should be around more later today! |
| 13:54.09 | brlcad | sounds good |
| 13:55.58 | teepee_ | brlcad is there someone who would look at the DXF code? I can help, but I don't think I can manage the time to be actual mentor |
| 13:56.36 | teepee_ | I just saw it mentioned somewhere (on the mailing list I think?) |
| 13:56.53 | brlcad | teepee_: absolutely -- that's one I already have a task chain started for since it's more of a priority |
| 13:58.11 | brlcad | would xuwei be interested in mentoring it? |
| 13:58.41 | teepee_ | at some point he said he'd like to continue working on it but I have not heard back |
| 13:58.51 | teepee_ | I can try to catch him via mail |
| 13:58.59 | brlcad | if not mentor it, at least maybe help define a list of the things it needs next |
| 13:59.32 | brlcad | he would get a free t-shirt out of it ;) |
| 14:00.09 | teepee_ | an overview is in https://github.com/openscad/openscad/issues/3044 but I guess it's too high level to directly map to gci tasks |
| 14:03.15 | brlcad | it's better than nothing, but yeah. gci also isn't good for architecture or design (figuring things out), which is a lot of what's in the "What's left" section. |
| 14:04.52 | teepee_ | if there's someone with some basic c++ knowledge there could be tasks to move more c style code to c++ style |
| 14:04.53 | brlcad | what it would be great for is things like creating test data, unit tests, implementing a spline fitting function, creating a summary table of what entities are supported on import/export, etc :) |
| 14:05.41 | teepee_ | nothing done for export yet |
| 14:10.01 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 14:38.22 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) |
| 14:40.02 | brlcad | so that's an easy table to fill out ;) |
| 15:05.07 | teepee_ | is the list of tasks already online somewhere? |
| 15:06.43 | brlcad | not yet, but this was the starter list from last time and several of them are keepers: https://brlcad.org/wiki/Google_Code_In/2017/Tasks |
| 15:08.53 | cad_zulip | @**jaideep seth**: ```quote |
| 15:08.53 | cad_zulip | <PROTECTED> |
| 15:08.53 | cad_zulip | ``` |
| 15:08.53 | cad_zulip | Project 16.1 of INCF in GSoC 19 https://summerofcode.withgoogle.com/projects/#5177840062955520 |
| 15:10.41 | brlcad | teepee_: how possible would it be to extract either .scad or .csg as a (non-gpl) library? |
| 15:12.09 | teepee_ | just the parser should be reasonable |
| 15:12.17 | brlcad | @**jaideep seth**: can you ask your mentor to send me an e-mail or send me their e-mail/irc so I can talk with them? |
| 15:12.58 | teepee_ | including also the processing step to generate actual geometry is probably a big pile of work |
| 15:13.22 | brlcad | no geometry generation, it can stay in expression/implicit form |
| 15:13.33 | brlcad | don't *want* triangles ;) |
| 15:13.55 | teepee_ | you'll get minkowski then :P |
| 15:14.07 | brlcad | it's harkening back to what we talked about a couple years ago |
| 15:14.20 | teepee_ | but yes, that would basically stop at the AST |
| 15:14.33 | brlcad | possibility of full round trip mapping between (or outright replacement) of .g to .csg |
| 15:14.54 | teepee_ | which I would like to split out better too, even for just internal use |
| 15:16.30 | brlcad | minkowski would become a new object type in brl-cad's taxonomy, either than or an editing command that evaluates (which is how hull is implemented) |
| 15:16.32 | teepee_ | right now the AST nodes know how to evaluate themselves but it would be much cleaner to use handle that via visitors |
| 15:16.40 | brlcad | s/than/that/ |
| 15:17.31 | brlcad | can you send me a source code link? I can take a look and see if it's at all something GCI students could start to tackle |
| 15:18.27 | brlcad | they can't do the architecture, but they could take care of a lot of other details and make serious progress |
| 15:18.31 | teepee_ | entry point is the bison parser : https://github.com/openscad/openscad/blob/master/src/parser.y |
| 15:18.42 | brlcad | oh no.. yaaaacc |
| 15:18.45 | brlcad | :) |
| 15:19.19 | brlcad | how do you manage than on windows? bison for windows? |
| 15:20.37 | teepee_ | MXE as build environment - that's running on Linux |
| 15:21.43 | teepee_ | there's a mostly working VS build though using vcpkg - that is getting some bison implementation from somewhere else |
| 15:22.02 | teepee_ | command line call is : https://github.com/openscad/openscad/blob/master/src/openscad.cc#L337 which puts the AST into 'root_module' |
| 15:22.27 | teepee_ | and evaluation should be then : https://github.com/openscad/openscad/blob/master/src/openscad.cc#L362 |
| 15:23.51 | brlcad | if it were extracted, what license do you think you'd use? able to do a permissive? |
| 15:29.44 | teepee_ | I suppose that would be possible, there was some discussion about that already - we'd need to check with kintel about the details though |
| 15:30.25 | teepee_ | I know he would not have a problem with making it MIT but I don't know who else has worked on that part and would have a say in that |
| 15:31.43 | kintel | I started working on extracting the file I/O and AST into a library a while ago |
| 15:31.55 | cad_zulip | @**jaideep seth**: ```quote |
| 15:31.55 | cad_zulip | **brlcad@irc.freenode.net**: @**jaideep seth**: can you ask your mentor to send me an e-mail or send me their e-mail/irc so I can talk with them? |
| 15:31.55 | cad_zulip | ``` |
| 15:31.55 | cad_zulip | My main mentor was Mark Watts, Ph.D student with University of Texas at Dallas. |
| 15:31.55 | cad_zulip | Contact Details: |
| 15:31.55 | cad_zulip | watts.mark2015@gmail.com |
| 15:31.56 | cad_zulip | mark@openworm.org |
| 15:31.57 | cad_zulip | There are many tasks ( 100s with many instances ) for each org in GCI 2019, I can help with making some ( while I try to get a commit in the BRL-CAD codbase) |
| 15:32.04 | kintel | And the intent was a BSD-style interop library |
| 15:32.30 | kintel | https://github.com/openscad/openscad/pull/1743 |
| 15:33.04 | kintel | There's a bit of architecture and refactoring work needed. Not quite student material right now |
| 15:33.08 | teepee_ | Ah, right, that's what I remembered :) |
| 15:33.40 | kintel | Once we can preserve use/include as AST nodes, that would greatly simplify the next steps |
| 15:33.43 | teepee_ | maybe some tasks for moving into the direction and re-architecture later? |
| 15:33.57 | cad_zulip | @**Sean**: We converted all of our flex and bison in just a few days to a compatible system that is cross platform |
| 15:34.26 | kintel | The coupling is still a bit high to just separate I think, but it's easier to discover if taking a stab at it :) |
| 15:35.22 | teepee_ | on my list would be some cleanup anyway - like getting all the code out of the parser.y |
| 15:35.23 | kintel | But yeah, reorganizing files and build systems could prob. be done separately |
| 15:35.44 | kintel | Perhaps motivation get the coding style update merged first :) |
| 15:36.12 | teepee_ | right, I guess that's probably a very good point now anyway |
| 15:36.29 | teepee_ | getting out a patch release is pretty much prepared in the branch already |
| 15:36.48 | teepee_ | I think there's only 2 small patches I need to cherry pick |
| 15:37.14 | teepee_ | so otherwise we are up for next release but not very soon realistically :D |
| 15:37.38 | cad_zulip | @**Sean**: Yeah I think thereâs probably a few things GCI could do here |
| 15:38.14 | teepee_ | I agree with kintel we should not try to go for full separation just yet |
| 15:38.48 | teepee_ | but preparing the move into that direction should still be able to produce a number of smaller topics that could fit gci |
| 15:39.02 | teepee_ | those are supposed to be like 2-4h or something like that? |
| 16:07.15 | *** join/#brlcad Macinlew (~MacInLew@2001:1948:212:6010:f4fa:57b7:cfb9:91fd) |
| 17:29.18 | cad_zulip | @**Sean**: More like 1-2 hours for a core dev, someone fully experienced and set up. Target is for it to be something an average student could do in less than a day even if not set up, ideally in the 1-4 hr timeframe depending on their experience. |
| 17:31.01 | cad_zulip | @**Sean**: So if theyâre really good, itâs still at least an hour of work, and if theyâre not, it can still figure it out without getting frustrated or being impossible |
| 18:12.16 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 18:52.27 | *** join/#brlcad kintel (~textual@unaffiliated/kintel) |
| 20:07.03 | *** join/#brlcad yorik (~yorik@2804:431:cfc7:ddfe:290:f5ff:fedc:3bb2) |