Perl 6 - the future is here, just unevenly distributed

IRC log for #pdl, 2013-11-29

| Channels | #pdl index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
07:28 rindolf joined #pdl
14:05 jberger_ joined #pdl
17:03 chm joined #pdl
18:45 Mithaldu chm: in OpenGL.pm please rename Changes to ChangeLog and Release_Notes to Changes, as that is the convention recognized and adhered to by perl toolchain
18:46 Mithaldu chm: also Release_Notes without dates are very not nice
18:55 Mithaldu chm: also see: http://changes.cpanhq.org/
19:15 sivoais Mithaldu: hey, btw, could you take a look at this pull request? <https://github.com/cparmer/plotlyperl/pull/2> ;-)
19:16 chm ...
19:16 Mithaldu sivoais: oh hi, sure
19:17 Mithaldu sivoais: you sure the html will work?
19:18 Mithaldu oh neat, you made perl examples :D
19:18 sivoais Mithaldu: let me link you how I tested it.
19:18 Mithaldu mhm
19:18 sivoais didn't make any examples, just linked to them :-P
19:18 chm Mithaldu - please open a ticket to this effect.  I plan to be more conformant in the coming OpenGL2 release.
19:18 Mithaldu chm: will do, thanks! :)
19:18 chm The work on POGL right now is mainly life support...
19:19 Mithaldu what order of magnitude is the ETA on OpenGL2 on? :)
19:19 chm BTW, the Changes files I've seen are *far* from Release_Notes.  Is there something better.  The ones I've seen are usually cryptic cxxp.
19:19 Mithaldu (my work here is making me more and more tempted to put out OpenGL::Bleeding)
19:20 Mithaldu cxxp?
19:20 chm cr.p
19:20 Mithaldu :D
19:20 sivoais Mithaldu: I used <http://search.cpan.org/pod2html> and <http://podwebview.odyniec.net/>
19:20 Mithaldu chm: we have a spec for Changes files: https://metacpan.org/pod/CPAN::Changes::Spec
19:20 chm I'm aiming for the end of the year----mostly slowed down by the 64bit index support in PDL
19:21 Mithaldu and yes, only 44% of CPAN confom to the spec
19:21 Mithaldu otoh
19:21 chm which I also need to support modern OpenGL stuff without 32bit limitations.
19:21 Mithaldu 44% of CPAN DO manage to conform to the spec :)
19:21 * sivoais & # back in ~1 hour
19:22 Mithaldu sivoais: sounds good, i'll merge and push it out, see how metacpan likes it
19:22 chm I'm hoping to get the double/longlong cast bug resolved this weekend which would
19:22 Mithaldu chm: huh, 64 bit calculations in 32 bit perl?
19:22 chm leave december to finish the POGL2.
19:22 Mithaldu oh wow, end of this year
19:22 Mithaldu that sounds good
19:22 chm 64bit calculations in perls with 64bit indexing.
19:22 chm The problem is understood but just no cycles.
19:23 Mithaldu ok, i'm not all too sure what that means, but i trust it's something useful :)
19:23 chm Are you able or willing to help out?  Things usually go *much* faster with teamwork.
19:23 Mithaldu depends
19:23 Mithaldu do you have some defined tasks?
19:23 chm I mean, this is what needs to be done:
19:24 chm - implement Alien::OpenGL to find and determine the local OpenGL libraries and parameters
19:24 chm - Implement OpenGL bindings via GLEW
19:24 chm -this is pretty mechanical but there are a lot of routines
19:24 chm - the first cut is planned to be very close to the C API
19:25 chm - the final stage would be to add perl-ish bindings that are simple and easy to use
19:25 chm including transparent interoperation with PDL ...
19:25 chm - of course testing and feedback are very helpful.  What platforms do you usually work with?
19:26 Mithaldu windows 7, 32 bit activeperl
19:26 Mithaldu i might be able to do some work on the alien stuff if you explain the scope of the task better
19:26 chm - as I'm not up to speed on the OpenGL 2.x+ coding (have the book), I could use a suite of
19:26 Mithaldu the glew stuff, maybe, i'm not sure how much it goes beyond my capabilities
19:26 chm test programs for the various features that are simple, self contained and could also be
19:26 Mithaldu ahahaha
19:26 chm useful as tutorial intro...
19:26 Mithaldu give me a second please :)
19:27 Mithaldu chm: https://github.com/wchristian/ogldev/commits/perl
19:27 Mithaldu tutorial 17 is in progress like, right now
19:28 chm Looks like a lot of prgress since the last we spoke...
19:29 Mithaldu basically, i got this thing going right now: http://ogldev.atspace.co.uk/www/tutorial16/txt_example.png
19:29 Mithaldu with lighting next
19:29 chm nice.
19:29 Mithaldu and it involves bunches of shaders
19:29 Mithaldu thus OpenGL::Bleeding
19:29 Mithaldu also OpenGL::Debug
19:29 jberger__ joined #pdl
19:30 Mithaldu see: https://github.com/wchristian/ogldev/blob/05363133de494a3eaa99ca035d5f1ea92fdd0fa2/arcsyn/framework/OpenGL/Debug.pm
19:30 chm I just read the latest book on OpenGL 4.3 and the new shader stuff is really excellent, and finally
19:30 chm explained in a unified and clear way.  It should really simplify the usability of 3d graphics.
19:30 Mithaldu also of interest, note the pp_def: https://github.com/wchristian/ogldev/blob/05363133de494a3eaa99ca035d5f1ea92fdd0fa2/tutorial16/math_3d.pm
19:31 chm Of particular interest is making a piddle out of a buffer object which would allow for shader computation,
19:31 Mithaldu hm
19:31 chm high dynamic range images, and interoperability with all sorts of OpenGL capable toolkits and frameworks.
19:31 Mithaldu yeah, some shortcuts would be helpful :)
19:31 Mithaldu oh, you meant screen buffer
19:32 chm Do you have a specific goal for your modern OpenGL use?
19:32 Mithaldu yes
19:32 chm I meant framebuffer or renderbuffers.
19:32 Mithaldu reimplement this in modern stuff: http://www.youtube.com/watch?v=Amiybcqtqiw
19:33 Mithaldu i'm unsure how aware you are of dwarf fortress
19:33 chm Cool!  It looks relevant to another interest of mine---use OpenGL as the display interface for a PDL workbook (since
19:33 Mithaldu haha, kind of
19:34 Mithaldu in my case i'm reading the memory of a windows process directly
19:34 chm opengl is supported by a number of GUI toolkits, it should be possible to make it fairly GUI agnostic---very portable.
19:34 chm Are you using PDL?
19:34 Mithaldu i.e. i tell the windows api "give me 2 kilobytes of data from this offset in the app's memory space"
19:34 Mithaldu and then i process that into data structures and render it
19:34 Mithaldu not yet
19:34 Mithaldu but with the shader stuff i will
19:35 chm Is this a constant memory mapped region or do you repeated fetch from different addresses?
19:35 Mithaldu repeated fetch
19:35 chm Different addresses?
19:35 Mithaldu i used Coro to make it smoothish
19:35 Mithaldu yes
19:35 Mithaldu the data i fetch contains offsets to more data to fetch, etc.
19:36 chm Hmm, is that something that would map to a memory mapped file region/shared memory region?
19:36 Mithaldu possibly, but i'm unsure of how well windows supports that
19:36 Mithaldu especially as i'm accessing another processes' memory without using any hooks
19:36 Mithaldu oh, also
19:36 Mithaldu i plan to do some demoscene stuff
19:37 chm Just looking to the future.  I would like the core of PDL to be mainly the underlying data access and computation framework
19:37 Mithaldu like this for example: http://www.youtube.com/watch?v=jlRoOqRs0bs
19:37 chm The details of the rest would be in the outer modules.  Keeping the core minimalist, fast, flexible, and portable.
19:38 Mithaldu and eventually i might want to clone a python 3d engine called beige
19:38 chm I would like to have improved 3d visualization of large data sets, including 3d printer objects...
19:38 Mithaldu http://www.youtube.com/watch?v=1N21V-WZFRo
19:39 Mithaldu that video is recorded from a python game, it should be possible to implement this in perl
19:39 Mithaldu oh, 3d printer visualization sounds useful :D
19:39 chm Is that using pygame?  Now that SDL2 is out, I think that would be a natural vehicle for a toolkit.
19:40 Mithaldu i'm unsure, the repository is here: https://bitbucket.org/duangle/beige/src/55bcb5cfa6fe/beige/?at=default#
19:41 Mithaldu i don't think it does
19:41 chm interesting.
19:42 chm Well, I have to go right now.  Maybe I'll see you on later.  o/
19:42 Mithaldu have a good evening :)
19:48 Mithaldu t sivoais [@MITHALDU/UploadToCPAN] POSTing upload for WebService-Plotly-1.133330.tar.gz to https://pause.perl.org/pause/authenquery
20:34 jberger_ joined #pdl
21:10 sivoais Mithaldu: excellent! The images seem to work on MetaCPAN, but not search.cpan.org. The HTML is probably being scrubbed? <http://search.cpan.org/~tokuhirom/Test-Pretty-0.27/lib/Test/Pretty.pm> works fine on both
21:11 Mithaldu maybe because of the style
21:12 Mithaldu or maybe because of the link
21:12 Mithaldu you should ask in #toolchain :)
21:12 sivoais ok, will do
21:13 Mithaldu sivoais: also: http://grep.cpan.me/?q=%3Dbegin+html
21:14 sivoais I keep forgetting about that service! I always end up doing a site:cpan.org type of search
21:20 Mithaldu yeah, it's really really useful
21:39 sivoais It's definitely the <a><img/><a/> tags that are the problem, see: <https://metacpan.org/pod/Acme::CPANAuthors::Nonhuman> versus <http://search.cpan.org/perldoc?Acme::CPANAuthors::Nonhuman>
21:43 sivoais ok, letting #toolchain know :-)
23:11 jberger_ joined #pdl
23:12 jberger_ Mithaldu, sivoais: see also https://metacpan.org/pod/Acme::XSS
23:14 Mithaldu yeah, the whole matter simply isn't documented well
23:15 Mithaldu sivoais: couldn't you make separate links under the images by mixing html and simple pod?
23:15 jberger_ I use a travis linked image in Galileo: https://metacpan.org/pod/Galileo
23:15 jberger_ at the bottom
23:15 jberger_ (I missed what you guys are doing, so I'm just spouting the things I have seen work)
23:16 Mithaldu jberger: that doesn't work on sco
23:16 jberger_ and yes, it shows on metacpan but not on sco
23:16 Mithaldu heh
23:16 jberger_ just checked
23:17 jberger_ that's too bad
23:18 Mithaldu i think sivoais's pull request will be fine if he also thinks sco simply can't link pics
23:18 * jberger_ never uses sco anymore
23:18 jberger_ do you guys know about mcpan.org?
23:18 sivoais Mithaldu: mmm, yeah, adding links in plain POD is a possibility. Let me add a commit to the PR
23:19 Mithaldu sivoais: the question is if you can make the pics show up in sco
23:19 sivoais I did not... I usually go for p3rl.org
23:19 jberger_ any time you accidentally land on sco
23:19 Mithaldu jberger: we're trying to make Webservice::Plotly look right on both
23:19 jberger_ enter an extra m before the cpan
23:19 sivoais Mithaldu: I can if I don't make them links. I'm pretty sure of it
23:19 * jberger_ looks
23:20 sivoais In fact, I'm going to drop the <ul> since it renders a bit ugly
23:20 Mithaldu sivoais: cool, just force-push and poke me here when i should merge :)
23:21 Mithaldu jberger: i never end up there on accident ;)
23:21 Mithaldu i have dedicated search shortcuts
23:42 jberger_ when I search via google the first result is often sco
23:42 sivoais Mithaldu: ok, commit is up! <https://github.com/cparmer/plotlyperl/pull/3>
23:43 Mithaldu jberger: that's true, but i never google for perl modules :)
23:43 Mithaldu sivoais: thanks!
23:43 sivoais yeah, that's a huge drawback of ignoring s.c.o. s.c.o often has a high ranking
23:46 Mithaldu [@MITHALDU/UploadToCPAN] POSTing upload for WebService-Plotly-1.133331.tar.gz to https://pause.perl.org/pause/authenquery
23:46 Mithaldu dzil <3
23:46 jberger_ mojo cpanify ...
23:46 jberger_ :-)
23:47 Mithaldu just as soon as that software loses the NIH attitude
23:47 Mithaldu oh, it's just an uploader anyhow
23:48 jberger_ Mithaldu, we agree to disagree on that point
23:48 Mithaldu well, people keep bringing it up :v
23:48 jberger_ it was just a playful barb, knowing that we disagree
23:49 Mithaldu look i'm german, my humor involves wood and rough sea
23:49 jberger_ hahaha
23:49 Mithaldu http://www.youtube.com/watch?v=fhkRM1Pa1HA
23:55 jberger_ "local *PDL::TO_JSON = sub { $_[0]->unpdl };"
23:55 jberger_ ++
23:56 Mithaldu :D
23:57 Mithaldu i think that's the only time i'll monkeypatch in production and not have it be a temporary hack
23:57 jberger_ seems like a perfectly fine monkey_patch
23:57 jberger_ <3 local
23:57 jberger_ did I tell you about unpdl or did you find it?
23:58 jberger_ it was my only real contribution to the PDL core
23:58 Mithaldu i knew about it because i needed it for my opengl stuff
23:58 Mithaldu but you might've mentioned it
23:58 jberger_ its the joy of open source, seeing your work get used for real things
23:59 Mithaldu aye :)

| Channels | #pdl index | Today | | Search | Google Search | Plain-Text | summary