tynet-lichat/shirakumo

10:06:19 Grolter I think I'll write here as well... I wrote a draft version of the PLN standard (weirdly there hasn't been one for over 10 years), and it would be nice to get your opinion on the issues presented there.

10:06:37 shinmera sure

10:06:43 Grolter https://github.com/Gleefre/cdr-package-local-nicknames [ link ]

10:07:19 Grolter Thanks :)

11:17:46 shinmera Grolter: I think it should be *disallowed* to make a local nickname the same as the package name or one of its nicknames, as that behaviour is extremely confusing.

11:19:31 Grolter Well, I tend to agree about the name, but maybe not one of nicknames?

11:19:52 shinmera I find it extremely hard to justify either case

11:20:01 Grolter Well, consider cl-mixed

11:20:06 shinmera Like, I cannot think of an example where it would be reasonable to allow

11:20:49 Grolter You have a package org.<...>.mixed that has a local nickname mixed. Say I use some renaming to add nicknames and I add a global nickname mixed to that package (seems reasonable)

11:21:04 Grolter Then this package has local nickname shadowing its global nickname

11:21:43 shinmera hrm

11:22:34 Grolter I mean, it is confusing, but (IMHO) not enough confusing to signal a full error.

11:23:26 shinmera I tend to hope people don't add global nicknames for packages :)

11:23:36 shinmera especially when said package is not theirs to control

11:24:19 shinmera arguably rename-package should fail, but that would be confusing too, since then internal information of the package would leak

11:24:32 shinmera I guess your proposal makes sense in that case

11:24:58 shinmera As for read consistency, I don't think your proposal for ::: and :::: will work, I seem to remember some implementations already using them.

11:25:22 Grolter Yeah, I think so too, just was trying to find different ways to deal with it

11:25:42 Grolter Thanks for feedback on "self-shadowing" :)

11:26:18 shinmera I'm also not a fan of using #`, since it isn't self-explanatory either

11:26:36 shinmera The #., perhaps with another internal operator for a "global-find-symbol" would be best.

11:26:47 shinmera SBCL already uses #. for certain types of arrays, for instance.

11:27:03 Grolter Hm, really?

11:27:05 shinmera yes

11:27:18 Grolter #. has a problem with safe read...

11:27:59 Grolter Even though it probably won't be of concern in any case

11:28:07 shinmera Hmm, or maybe they changed it

11:28:23 shinmera Istr that they used it for array literals with specific element types

11:28:32 shinmera But now it's an extension to #A

11:29:00 shinmera Maybe it could be done as an extension to #:

11:29:05 shinmera like #:(package name)

11:29:24 shinmera though I don't know if that would be compliant

11:30:25 Grolter I think it would not...

11:30:41 Grolter I choosed #` because it was (1) free and (2) resembles quote

11:30:54 Grolter Though quote doesn't have anything to do with it

11:31:14 Grolter Maybe there is some other name that will make sense, hm...

11:31:48 Grolter I kind of like #:(package name) anyway, will include it into the list of possible solutions

11:31:50 shinmera Either way, your write-up is good and I see no problems with it

11:32:10 shinmera Though I also never use the programmatical interface for PLNs so this has no real impact on me :)

11:32:21 Grolter Thanks :)

11:33:00 Grolter I hope to get some more feedback in near future and then also try asking in different implementation-specific channels

11:33:14 shinmera Thanks for putting in the work! :)

11:35:26 Grolter Hm, about #:(package name), how one would denote whether the symbol is external or internal?

11:37:17 Grolter #:(package name :external) and #:(package name) probably, but it kind of complicates things...

11:38:09 shinmera #::(package name) #:(package name) ;)

11:38:35 Grolter Ah, that looks good :)

12:36:02 Grolter FYI binpack package on shirakumo dist (quicklisp) is broken - a lot of files are missing :/

12:37:10 shinmera like what?

12:37:33 shinmera oh

12:37:35 shinmera what in the world

12:37:42 Grolter It contains like only one .asd file and 2 .lisp files

12:37:58 shinmera that is extremely weird, considering the online file listing shows all the files

12:38:40 shinmera well, don't have the energy to look at that at the moment.

12:39:03 Grolter It doesn't matter to me much anyway

12:39:20 Grolter But I though it might be important for you to know

12:39:20 shinmera it matters a lot to me

12:40:04 Grolter By the way, do you know anything about quicklisp becoming more and more stale? Or it is no becoming more stale?

12:40:22 shinmera Xach has had some serious real life things pop up in recent years

12:40:27 shinmera and I think he's stepping back a bit.

12:40:35 shinmera Which is why I started the redist project

12:40:39 Grolter I see

12:40:44 Grolter Redist?

12:40:44 shinmera But, as you can see, it still has some pretty glaring issues

12:40:51 shinmera https://github.com/shirakumo/redist

12:40:55 shinmera which is what generates my dist

12:41:31 Grolter I see

12:41:48 Grolter Seems nice

12:43:46 Grolter Is your dist definition open source btw?

12:44:41 shinmera I'm not sure what you mean?

12:44:46 shinmera like, what packages it contains, or?

12:44:58 Grolter I mean the definition with redist?

12:45:07 shinmera I still don't know what exactly you're referring to

12:45:25 Grolter I guess I mean something similar to quicklisp-projects

12:45:29 shinmera Ah. No.

12:45:40 shinmera You can see what packages are in it http://dist.shirakumo.org

12:45:54 shinmera but adding ones is on a as-needed basis for shirakumo releases.

12:46:38 Grolter I see, just thought it might be a good example for redist to look at :/

12:46:58 shinmera You can just use the command line interface to add stuff

12:47:21 Grolter I see

12:47:33 Grolter (I'm repeating it too much, :/ )

15:17:17 shinmera More in-progress work https://github.com/Shirakumo/trial/pull/45

15:20:22 shinmera Once I have that I should probably extend the docs and readme and all with some cool screenshots or whatever

15:20:31 shinmera you know, to make people think this is as good as unreal or what

16:55:23 Colleen Shinmera: do you do a render pass per dynamic light or multiple lights at once?

16:55:23 dthompson Shinmera: do you do a render pass per dynamic light or multiple lights at once?

17:00:22 Colleen quite some time ago I learned a bit about IBL but it was just used for ambient lighting so you could get cool skybox reflections off that pbr helmet model

17:00:22 dthompson quite some time ago I learned a bit about IBL but it was just used for ambient lighting so you could get cool skybox reflections off that pbr helmet model

17:14:49 shinmera yes, that is what IBL is.

17:14:57 shinmera right now I've only implemented a forward renderer.

17:15:16 shinmera though I would like to implement a more advanced scheme at some later point

17:16:40 shinmera the new method combination scheme should allow me to re-use pretty much everything from the forward scheme and just implement the necessary logic in a subclass.

17:19:29 shinmera |3b|: is there an easy way in cl-opengl to enable the debug message callback stuff?