--- Log opened Mon Jun 19 00:00:05 2017
20170619 00:00:26< celticminstrel> But maybe it's not a problem; executing arbitrary code on the GPU might be less of a security risk than doing so on the CPU. I dunno. It should be carefully considered, in any case. Don't just do it because you can.
20170619 00:01:36< celticminstrel> I guess there are probably some things you could do with shaders but not with ~CHAN() - distortions and stuff.
20170619 00:01:57< vultraz_iOS> chan is literally color changing :|
20170619 00:02:19< celticminstrel> For example, in shaders it should be possible to distort an image into the shape of a circle, I think. ~CHAN() certainly can't do this.
20170619 00:02:27< vultraz_iOS> don't we want UMC authors to be able to create complex effects
20170619 00:02:50< celticminstrel> Sure, but that doesn't necessarily mean allowing them to execute arbitrary code.
20170619 00:03:17< celticminstrel> For example, you could offer a ~DISTORT() IPF which uses a WFL formula to transform the texture coordinates however you want.
20170619 00:03:31< vultraz_iOS> There's this this weird thing in wesnoth's design where UMC authors are just given a set set of actions and told "use these"
20170619 00:03:51< vultraz_iOS> Instead of developing an API and giving them the API
20170619 00:03:55< vultraz_iOS> For example
20170619 00:03:58< celticminstrel> Um, vultraz_iOS 
20170619 00:04:04< celticminstrel> That set of actions is an API.
20170619 00:04:20< celticminstrel> It may or may not be a good one, but it's an API.
20170619 00:04:21< vultraz_iOS> it's the difference between a [fade_out_text] action and giving them access to alpha controls of various areas
20170619 00:04:29< vultraz_iOS> including text
20170619 00:04:49< celticminstrel> The API can be improved, sure.
20170619 00:05:22< celticminstrel> Anyway, I'm not saying don't allow UMC authors to write their own shaders. I'm saying consider the consequences carefully.
20170619 00:06:56< vultraz_iOS> Our philosophy has always been to expose a narrow feature to creators instead of providing an environment where they can craft that feature.
20170619 00:07:01< vultraz_iOS> I think that should change.
20170619 00:07:38< celticminstrel> I don't think allowing authors to write their own shaders is particularly useful, TBH.
20170619 00:08:05< vultraz_iOS> It's part of the reason Anura's object system is so robust
20170619 00:08:37< vultraz_iOS> It provides a framework for interacting with objects and lets module creators work within that given some core utility functions
20170619 00:08:47< celticminstrel> Or at least, their own shaders in shader language. If we let them write shaders in WFL, that might be another story.
20170619 00:08:58< vultraz_iOS> why not?
20170619 00:09:35< celticminstrel> Because that means they need to learn yet another language in order to get everything they can out of Wesnoth? And a rather low-level one to boot (at least compared to Lua, WFL, or WML).
20170619 00:11:20-!- travis-ci [~travis-ci@ec2-54-224-155-18.compute-1.amazonaws.com] has joined #wesnoth-dev
20170619 00:11:21< travis-ci> gfgtdf/wesnoth#882 (fix_1436 - 4f44c87 : gfgtdf): The build is still failing.
20170619 00:11:21< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/244344020
20170619 00:11:21-!- travis-ci [~travis-ci@ec2-54-224-155-18.compute-1.amazonaws.com] has left #wesnoth-dev []
20170619 00:11:22< vultraz_iOS> There is no way we can provide a wrapper that encapsulates the functionality of OGL and shaders without severely limiting options.
20170619 00:11:42< celticminstrel> That's probably false.
20170619 00:14:58< celticminstrel> WFL is probably Turing-complete. It could probably fully substitute for the shader language without reducing capabilities.
20170619 00:16:00< celticminstrel> And if I understand shaders correctly, ultimately all they do is take in some input and put out some output, making them sort of pure-functional and thus WFL particularly suitable.
20170619 00:16:24< celticminstrel> A shader has no side-effects, right?
20170619 00:16:37< celticminstrel> It's just a calculation. Potentially a very complex calculation, but still just a calculation.
20170619 00:18:00< celticminstrel> WFL for shaders would probably need its own function library, (amost?) completely separate from the built-on function library, and I guess it'd use floating-point instead of fixed-point.
20170619 00:18:04< vultraz_iOS> https://open.gl/drawing
20170619 00:18:07< celticminstrel> ^built-in
20170619 00:18:48< celticminstrel> IIRC geometry shaders are post-3.2? Not entirely sure.
20170619 00:19:32< celticminstrel> So I don't think we can use them.
20170619 00:19:40< vultraz_iOS> they're 3.2
20170619 00:19:50< vultraz_iOS> "Since OpenGL 3.2 there is a third optional type of shader that sits between the vertex and fragment shaders, known as the geometry shader. This shader has the unique ability to create new geometry on the fly using the output of the vertex shader as input."
20170619 00:19:56< celticminstrel> We could use vertex shaders. ~ROTATE() and ~FL() at least are best implemented as vertex shaders.
20170619 00:20:23< celticminstrel> The only problem is that if they appear later in a chain, you can't do it as a vertex shader without transforming stuff that came later in the chain.
20170619 00:20:35< celticminstrel> (The second later means "after the ROTATE or FL")
20170619 00:20:43< celticminstrel> So like...
20170619 00:20:53< celticminstrel> Can I think of an example...
20170619 00:21:38< vultraz_iOS> perhaps it's not just shaders authors need access to if they want to do complex effects
20170619 00:22:26< celticminstrel> In any case, I don't think letting authors use shaders should be a priority. Get the stuff we already support working first.
20170619 00:22:55< vultraz_iOS> sur
20170619 00:22:55< vultraz_iOS> e
20170619 00:23:42< vultraz_iOS> like, honestly, it would be easy for me to just keep the surface handling of IPFs, finish up a few things, do some refactoring and done
20170619 00:23:56< vultraz_iOS> embarking on this OGL road is complex..
20170619 00:23:59< celticminstrel> Do you know what GLSL version is used in OpenGL 3.2?
20170619 00:24:07< vultraz_iOS> I have no idea of anything
20170619 00:24:13< celticminstrel> Heh, okay.
20170619 00:24:32< vultraz_iOS> " OpenGL 3.2 also included an upgrade to GLSL version 1.50."
20170619 00:24:57< celticminstrel> So we should put "#version 150" at the top of all our shaders.
20170619 00:25:11< celticminstrel> Pretty sure that won't work on my computer, sadly.
20170619 00:25:21< celticminstrel> My Mac that is, should be no problem on the Windows.
20170619 00:25:33< vultraz_iOS> no idea why we can't target 3.3
20170619 00:25:49< celticminstrel> I think we should target the lowest reasonable version.
20170619 00:26:07< celticminstrel> ^the lowest reasonable version possible
20170619 00:26:44< celticminstrel> IOW, unless you absolutely need something new in 3.3, stick to 3.2. 
20170619 00:27:07< vultraz_iOS> well we don't really need D3D10 stuff anyway
20170619 00:27:53< celticminstrel> "The glShaderSource function can take multiple source strings in an array, but you'll usually have your source code in one char array. "
20170619 00:28:00< celticminstrel> That seems potentially useful.
20170619 00:28:26< celticminstrel> Might mean it's possible to construct an array of shader fragments and just pass that in - no need to actually concatenate the strings.
20170619 00:28:44< vultraz_iOS> what's the 4.0 hardware support..
20170619 00:28:47< vultraz_iOS> "Hardware support: Nvidia GeForce 400 series and newer, AMD Radeon HD 5000 Series and newer, Intel HD Graphics in Intel Ivy Bridge processors and newer.[36]"
20170619 00:29:18< celticminstrel> Again, unless you absolutely need something new in 4.0 (which I seriously doubt), stick to 3.2.
20170619 00:29:30< vultraz_iOS> I want to use the newest version possible
20170619 00:29:37< celticminstrel> Don't.
20170619 00:30:08< vultraz_iOS> eh whatever
20170619 00:30:09< celticminstrel> Use the oldest version possible.
20170619 00:30:10< vultraz_iOS> 3.2 is good
20170619 00:30:16< celticminstrel> Yeah, 3.2 is probably fine.
20170619 00:30:36< vultraz_iOS> now if someone could tell me how to actually get a goddamn thing drawing :|
20170619 00:30:56< celticminstrel> Is the screen pitch black?
20170619 00:31:01< vultraz_iOS> yup
20170619 00:31:18< vultraz_iOS> I need to rewrite the tetxure wrapper class
20170619 00:31:19< vultraz_iOS> texture*
20170619 00:31:20< vultraz_iOS> since it
20170619 00:31:23< celticminstrel> I dunno then. IIRC that happened to me too when I attempted to use a 3.2 profile via SDL.
20170619 00:31:42< vultraz_iOS> well of course nothing is drawing because nothing is using OGL drawing calls!
20170619 00:31:53< celticminstrel> But I suppose the chances are that the problem in your case is totally different.
20170619 00:32:22< celticminstrel> Maybe you can just replace some CreateRenderer calls with a different type and some things would magically work?
20170619 00:32:38< vultraz_iOS> No no 
20170619 00:32:48< vultraz_iOS> I need to write custom texture loading code
20170619 00:32:57< vultraz_iOS> remember I can't use SDL_Texture anymore
20170619 00:33:09< vultraz_iOS> SDL_Texture requires a renderer
20170619 00:33:17< vultraz_iOS> and you can't use a renderer with a custom OGL context
20170619 00:33:26< celticminstrel> Oh. That's annoying.
20170619 00:33:38< vultraz_iOS> you can do the opposite, I think - OGL with a renderer context - but all you get is OGL 1.1
20170619 00:33:54< celticminstrel> What version introduced shaders again?
20170619 00:34:22-!- louis94 [~~louis94@109.131.102.51] has joined #wesnoth-dev
20170619 00:35:06< vultraz_iOS> 2.0 I think
20170619 00:35:14< celticminstrel> I guess that's no good then.
20170619 00:35:27< vultraz_iOS> "OpenGL 2.0 was originally conceived by 3Dlabs to address concerns that OpenGL was stagnating and lacked a strong direction.[29] 3Dlabs proposed a number of major additions to the standard. Most of these were, at the time, rejected by the ARB or otherwise never came to fruition in the form that 3Dlabs proposed. However, their proposal for a C-style shading language was eventually completed, resulting in the current formulation of the
20170619 00:35:27< vultraz_iOS> OpenGL Shading Language (GLSL or GLslang)."
20170619 00:35:43< matthiaskrgr> porting wesnoth to OGL :o 
20170619 00:35:46< matthiaskrgr> ?
20170619 00:35:51< celticminstrel> Apparently.
20170619 00:35:52< vultraz_iOS> seriously, 1.1 was from 1997
20170619 00:36:03< vultraz_iOS> Who uses 20 year old APIs
20170619 00:36:15< celticminstrel> Well, if not for the lack of shaders though, I'd probably say 1.1 is good enough.
20170619 00:36:35< vultraz_iOS> By using 3.2 we're only using an 8 year old API
20170619 00:36:37< matthiaskrgr> I tried to compile the branch to figure out what it is but linking failed because of missing glew foo 
20170619 00:37:02< vultraz_iOS> matthiaskrgr: yeah i have done nothing to set up the cmake/scons files with GLEW
20170619 00:37:05< vultraz_iOS> someone else will have to
20170619 00:37:13< matthiaskrgr> oki
20170619 00:37:27< matthiaskrgr> I'll wait until then 
20170619 00:37:28< matthiaskrgr> :>
20170619 00:37:37< vultraz_iOS> you'll be waiting a long time
20170619 00:37:46< matthiaskrgr> :/ 
20170619 00:38:07< vultraz_iOS> I've only just started the OGL work 
20170619 00:38:20< vultraz_iOS> even if you could build you wouldn't see anything
20170619 00:38:53< celticminstrel> vultraz_iOS: TBH, I think it's in your best interested to get the scons/cmake stuff set up sooner rather than later, as it increases the chances of someone else being able to help you with stuff.
20170619 00:38:58< celticminstrel> ^best interests
20170619 00:39:07< vultraz_iOS> *I don't know how*!
20170619 00:39:11< vultraz_iOS> I use neither of those!
20170619 00:40:51< vultraz_iOS> It is still certainly an option to use SDL_GPU but I dont like the API
20170619 00:41:10< vultraz_iOS> He decided to cal his render function GPU_Flip() :|
20170619 00:41:11< matthiaskrgr> I tried installing some glew dev package but it didn't help surprisingly
20170619 00:42:08< celticminstrel> matthiaskrgr: Yeah, because scons/cmake isn't told that it's needed.
20170619 00:42:42< matthiaskrgr> mmh
20170619 00:47:18< matthiaskrgr> hah !   :D 
20170619 00:47:41< matthiaskrgr> 5 minutes of googling => stack overflow => 
20170619 00:47:45< matthiaskrgr> https://stackoverflow.com/questions/23171894/cant-compile-easy-source-in-c-and-opengl-glfw-in-linux-in-netbeans  
20170619 00:48:07< matthiaskrgr> blindly adding       -pthread -lGLEW -lGLU -lGL -lrt -lXrandr -lXxf86vm -lXi -lXinerama -lX11       to the linker invocation makes the linking work :D
20170619 00:48:49< vultraz_iOS> nice
20170619 00:48:55< vultraz_iOS> not enjoy starting at a black screen :P
20170619 00:48:58< vultraz_iOS> now*
20170619 00:49:36< matthiaskrgr> black is the new RGB 
20170619 00:50:01< vultraz_iOS> celticminstrel: btw, a downside of having to use OGL is really do need to load a surface for every texture i create :/
20170619 00:50:19< celticminstrel> ?
20170619 00:50:36< matthiaskrgr> vultraz_iOS: you, wanna know about  heap use after free I found ? :P
20170619 00:50:38< matthiaskrgr> *yo 
20170619 00:50:46< vultraz_iOS> I expect you found much
20170619 00:50:52< vultraz_iOS> but go ahead
20170619 00:50:54< matthiaskrgr> wesnoth -t 
20170619 00:51:12< matthiaskrgr> https://pastebin.com/MP3K3Zt7  
20170619 00:51:17< matthiaskrgr> on the render branch 
20170619 00:51:42< matthiaskrgr> mm
20170619 00:51:51< matthiaskrgr> ==3099==AddressSanitizer: while reporting a bug found another one. Ignoring.
20170619 00:51:54< matthiaskrgr> this is kind of bad xD
20170619 00:51:57< vultraz_iOS> celticminstrel: with SDL_Texture, I could just use IMG_LoadTexture. Now with OGL you need to actually create the texture yourself and set its data
20170619 00:52:07< vultraz_iOS> which means loading a surface
20170619 00:52:11< vultraz_iOS> so I can get the pixel data
20170619 00:52:57< celticminstrel> So many reasons to not use SDL for any moderately complicated game...
20170619 00:53:13< vultraz_iOS> what do you mean
20170619 00:53:51< celticminstrel> Nothing important, really.
20170619 00:54:08< celticminstrel> But SDL is better than most at event handling...
20170619 00:54:14< celticminstrel> Particularly keyboard handling...
20170619 00:55:46< vultraz_iOS> "The fifth parameter should always have a value of 0 per the specification." why in hell would you have a parameter that must always be something...
20170619 00:57:39< celticminstrel> For what?
20170619 00:58:05< celticminstrel> gfgtdf: I commented on your PR.
20170619 01:01:57< vultraz_iOS> glTexImage2D
20170619 01:03:21< celticminstrel> The height must be zero? :S
20170619 01:03:38< celticminstrel> Oh, the border. But that's the sixth parameter.
20170619 01:05:02< celticminstrel> Ah, I see. It used to be used in OpenGL2, but is now no longer used. That's why.
20170619 01:21:14-!- louis94 [~~louis94@109.131.102.51] has quit [Ping timeout: 268 seconds]
20170619 01:21:49-!- travis-ci [~travis-ci@ec2-54-226-197-47.compute-1.amazonaws.com] has joined #wesnoth-dev
20170619 01:21:50< travis-ci> gfgtdf/wesnoth#883 (fix_1436 - 0b28f90 : gfgtdf): The build was fixed.
20170619 01:21:50< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/244346579
20170619 01:21:50-!- travis-ci [~travis-ci@ec2-54-226-197-47.compute-1.amazonaws.com] has left #wesnoth-dev []
20170619 01:27:42-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 01:30:20-!- gfgtdf [~chatzilla@x4e36377d.dyn.telefonica.de] has quit [Ping timeout: 260 seconds]
20170619 01:31:07-!- gfgtdf [~chatzilla@x4e36377d.dyn.telefonica.de] has joined #wesnoth-dev
20170619 01:31:57-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Ping timeout: 240 seconds]
20170619 01:36:06-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 255 seconds]
20170619 01:36:14-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 01:38:26< vultraz_iOS> hmmm
20170619 01:38:32< vultraz_iOS> ok so it looks like in OGL..
20170619 01:38:57< vultraz_iOS> you don't keep pointers to textures
20170619 01:39:09< vultraz_iOS> you keep... numerical....names?
20170619 01:39:10< celticminstrel> You kinda do.
20170619 01:39:15< celticminstrel> But yeah, the pointers are integers.
20170619 01:39:24< vultraz_iOS> GLuint
20170619 01:39:28< vultraz_iOS> which I guess is basically uint32_t
20170619 01:39:42< celticminstrel> Maybe, I dunno.
20170619 01:39:54< celticminstrel> But wrap it in a class. Don't be passing around raw texture names everywhere.
20170619 01:39:57< vultraz_iOS> "GLuint	32	Unsigned binary integer"
20170619 01:40:07< celticminstrel> Guess so, then.
20170619 01:40:24< vultraz_iOS> I'm probably going to put this in the texture class
20170619 01:40:27< celticminstrel> Interesting that it feels the need to specify binary, but I guess BCD does exist, so maybe that's why.
20170619 01:40:37< celticminstrel> Yeah, that's exactly where it belongs. In the texture class.
20170619 01:40:47-!- Waste [~Cracker@blk-138-74-119.eastlink.ca] has quit [Remote host closed the connection]
20170619 01:40:55< vultraz_iOS> I guess I don't need the shared_ptr interface anymore
20170619 01:41:02< celticminstrel> ?
20170619 01:41:21< vultraz_iOS> right now textures are managed inside `texture` by `shared_ptr<SDL_Texture>`
20170619 01:41:28< celticminstrel> Ah, then yes.
20170619 01:41:43< vultraz_iOS> so I think they just need to hold a GLuint now 
20170619 01:41:49< celticminstrel> Right.
20170619 01:42:14< celticminstrel> Or uint32_t if you don't want to include GL.h, though I guess it's a pretty lightweight header.
20170619 01:42:15-!- gfgtdf_ [~chatzilla@x4e368ba3.dyn.telefonica.de] has joined #wesnoth-dev
20170619 01:44:03-!- gfgtdf [~chatzilla@x4e36377d.dyn.telefonica.de] has quit [Ping timeout: 240 seconds]
20170619 01:44:06-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection timed out]
20170619 01:44:10-!- gfgtdf_ is now known as gfgtdf
20170619 01:44:39-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 02:03:49< celticminstrel> Oh, look at that. Maybe we need to use OpenGL 2.0 after all.
20170619 02:03:59< celticminstrel> https://forums.wesnoth.org/viewtopic.php?p=614401#p614401
20170619 02:04:19< celticminstrel> Oh, wait. :|
20170619 02:04:33< celticminstrel> That's not 2.0, that's ES 2.0 which is probably completely different.
20170619 02:04:35< vultraz_iOS> he has 1.4
20170619 02:04:42< celticminstrel> Yeah,.
20170619 02:04:44< vultraz_iOS> ES is different
20170619 02:04:47< celticminstrel> Yeah.
20170619 02:04:54< vultraz_iOS> mobile
20170619 02:04:58< celticminstrel> Yeah.
20170619 02:06:06< vultraz_iOS> "The pixels in the texture will be addressed using texture coordinates during drawing operations. These coordinates range from 0.0 to 1.0 where (0,0) is conventionally the bottom-left corner and (1,1) is the top-right corner of the texture image. "
20170619 02:06:10< vultraz_iOS>  don't understand this
20170619 02:06:15< vultraz_iOS> does this mean we can't use pixel coords?
20170619 02:06:21< celticminstrel> We can.
20170619 02:06:26< celticminstrel> It just requires extra work.
20170619 02:13:16-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170619 02:13:42< vultraz_iOS>     glGenTextures(1, &texture_id_);
20170619 02:13:42< vultraz_iOS>     glBindTexture(GL_TEXTURE_2D, texture_id_);
20170619 02:13:43< vultraz_iOS>     glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, surf->w, surf->h, 0, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8_REV, surf->pixels);
20170619 02:14:54< vultraz_iOS> look alright?
20170619 02:17:13-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20170619 02:17:47< celmin> Probably? I think I'm seeing that out of order, mind you.
20170619 02:19:01< celmin> And I'm not sure off the top of my head how to make it use pixel coordinates for textures.
20170619 02:19:54< celmin> Oh, we also probably want to transform the coordinates so that (0,0) is in the top left, rather than the bottom left.
20170619 02:19:59< vultraz_iOS> indeed
20170619 02:20:07< celmin> SFML does both these things.
20170619 02:22:38< celmin> What does the _REV bit mean?
20170619 02:22:45< vultraz_iOS> not sure
20170619 02:22:49< vultraz_iOS> maybe i don't want that
20170619 02:24:17< vultraz_iOS> hm..
20170619 02:24:29< vultraz_iOS> so glTexParameteri doesn't bind to individual textures...
20170619 02:24:51< celmin> ?
20170619 02:25:25< vultraz_iOS> It doesn't take a texture ID..
20170619 02:26:51< celmin> What does it do, again?
20170619 02:27:10< vultraz_iOS> Sets some texture parameters
20170619 02:27:14< celmin> Keep in mind that everything in OpenGL operates on global state.
20170619 02:27:32< celmin> Just because a function doesn't take a texture ID, that does not mean that it doesn't operate on individual textures.
20170619 02:27:46< celmin> More likely, it operates on whatever texture is bound as the "current texture".
20170619 02:28:15-!- gfgtdf [~chatzilla@x4e368ba3.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.93 [Firefox 53.0.3/20170518000419]]
20170619 02:28:20< vultraz_iOS> Blahhhf
20170619 02:28:24< celmin> glBindTexture is the call that sets the current texture, so before calling glTexParameteri, you need to make sure the correct texture is bound with glBindTexture.
20170619 02:28:54< celmin> You absolutely need to learn how to use glPush* and glPop* functions.
20170619 02:28:59< vultraz_iOS> Am starting to consider if this is all worth it...
20170619 02:29:14< celmin> As these enable a pseudo-local state.
20170619 02:29:27< celmin> (In much the same way that Wesnoth auto-stored variables do.)
20170619 02:41:52< celmin> gfgtdf just brought up an interesting point.
20170619 02:42:35< celmin> vultraz_iOS: Apparently typed_formula<int>, typed_formula<float>, etc needlessly converts the number to a string and back.
20170619 02:43:01< celmin> I wonder if that's having performance impacts.
20170619 02:43:17< celmin> We definitely need to profile...
20170619 02:47:30< vultraz_iOS> why would it do that?
20170619 02:48:01< vultraz_iOS> ah, yes...
20170619 02:48:21< vultraz_iOS> well, it receives the formula as a string
20170619 02:49:02< celmin> Sooo… to fix it you'd probably need a typed_formula constructor that takes a config_attribute_value?
20170619 02:49:29< celmin> I might look into this...
20170619 02:49:38< vultraz_iOS> why would you need that?
20170619 02:49:48< vultraz_iOS> the value isn't int there either
20170619 02:49:54< celmin> Not sure. I'm not currently looking at the code.
20170619 02:50:06< celmin> It would be int if it's not a formula.
20170619 02:50:34< vultraz_iOS> glGenTextures
20170619 02:50:36< vultraz_iOS> oops
20170619 02:50:40< vultraz_iOS> meant that for google
20170619 02:50:40< celmin> Heh
20170619 02:57:46-!- Bonobo [~Bonobo@220-244-64-153.tpgi.com.au] has joined #wesnoth-dev
20170619 03:06:37-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20170619 03:14:52< celticminstrel> Wow, stack overflow when calling a non-existent WFL function.
20170619 03:18:30-!- irker312 [~irker@109.237.213.40] has quit [Quit: transmission timeout]
20170619 03:38:22-!- shiki [~Shiki@p54854F0A.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170619 03:38:22-!- sevu [~Shiki@p548577C2.dip0.t-ipconnect.de] has quit [Disconnected by services]
20170619 03:52:38-!- Kwandulin [~Kwandulin@p200300760F7CBAEB51C7F92BAFB0C43D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170619 04:12:42-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20170619 04:58:02-!- Kwandulin [~Kwandulin@p200300760F7CBAEB51C7F92BAFB0C43D.dip0.t-ipconnect.de] has quit [Ping timeout: 255 seconds]
20170619 05:25:32-!- Appleman1234 [~quassel@z190230.ppp.asahi-net.or.jp] has quit [Ping timeout: 260 seconds]
20170619 05:40:34-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Ping timeout: 246 seconds]
20170619 05:51:35-!- zookeeper [zookeeper@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20170619 05:54:02-!- JyrkiVesterinen [~jyrki@87-92-55-234.bb.dnainternet.fi] has joined #wesnoth-dev
20170619 06:28:18-!- JyrkiVesterinen [~jyrki@87-92-55-234.bb.dnainternet.fi] has quit [Quit: .]
20170619 06:35:13< vultraz_iOS> ya know it's rather damn frustrating that I can't work on anything until I get the basic GL stuff running :|
20170619 06:35:45-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20170619 07:13:19-!- atarocch [~atarocch@93.56.160.37] has joined #wesnoth-dev
20170619 07:18:13-!- zookeeper [zookeeper@wesnoth/developer/zookeeper] has quit [Ping timeout: 258 seconds]
20170619 08:04:10-!- Appleman1234 [~quassel@z190230.ppp.asahi-net.or.jp] has joined #wesnoth-dev
20170619 09:00:28-!- louis94 [~~louis94@109.131.102.51] has joined #wesnoth-dev
20170619 09:04:44-!- louis94 [~~louis94@109.131.102.51] has quit [Ping timeout: 246 seconds]
20170619 09:43:44-!- JyrkiVesterinen [~JyrkiVest@85-76-74-185-nat.elisa-mobile.fi] has joined #wesnoth-dev
20170619 10:37:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 11:36:35-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Ping timeout: 258 seconds]
20170619 11:42:00-!- shiki [~Shiki@p54854F0A.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20170619 12:08:12-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20170619 12:20:39-!- Kwandulin [~Kwandulin@p200300760F7CBAF451C7F92BAFB0C43D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170619 12:21:21-!- zookeeper [zookeeper@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20170619 12:26:15-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20170619 12:51:21-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20170619 13:21:28< zookeeper> doing things like turning IPF's into shaders or whatever is something that would come _after_ the actual accelerated rendering. you can't mix the former in with the latter if you wish the actually get the accelerated rendering working ASAP.
20170619 13:21:40< zookeeper> s/the/to
20170619 13:27:35-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20170619 14:33:25-!- JyrkiVesterinen [~JyrkiVest@85-76-74-185-nat.elisa-mobile.fi] has quit [Quit: .]
20170619 14:47:35-!- Kwandulin [~Kwandulin@p200300760F7CBAF451C7F92BAFB0C43D.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20170619 14:51:31-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 15:04:36-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Remote host closed the connection]
20170619 15:05:11-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:d0ba:fb91:d5c3:f816] has joined #wesnoth-dev
20170619 15:05:56-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 246 seconds]
20170619 15:06:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 15:09:26-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:d0ba:fb91:d5c3:f816] has quit [Ping timeout: 246 seconds]
20170619 15:37:12-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:d0ba:fb91:d5c3:f816] has joined #wesnoth-dev
20170619 15:56:35-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20170619 15:57:00-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:d0ba:fb91:d5c3:f816] has quit [Remote host closed the connection]
20170619 15:57:35-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:d0ba:fb91:d5c3:f816] has joined #wesnoth-dev
20170619 16:01:30-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:d0ba:fb91:d5c3:f816] has quit [Ping timeout: 240 seconds]
20170619 16:04:37-!- Kwandulin [~Kwandulin@p200300760F7CBAF451C7F92BAFB0C43D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170619 16:24:05-!- JyrkiVesterinen [~jyrki@89-166-101-10.bb.dnainternet.fi] has joined #wesnoth-dev
20170619 16:33:27-!- clavi [~clavi@v22017034422546657.goodsrv.de] has quit [Quit: ZNC - http://znc.in]
20170619 16:36:16-!- clavi [~clavi@v22017034422546657.goodsrv.de] has joined #wesnoth-dev
20170619 16:38:51-!- Kwandulin [~Kwandulin@p200300760F7CBAF451C7F92BAFB0C43D.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20170619 16:41:16-!- Bonobo [~Bonobo@220-244-64-153.tpgi.com.au] has quit [Ping timeout: 260 seconds]
20170619 16:48:12< vultraz_iOS> zookeeper: it's simply hard to do it properly without the shaders
20170619 16:56:04< vultraz_iOS> zookeeper: unless you advocate sticking to the current method, finishing things, and *then* deploying OGL 
20170619 16:57:53< zookeeper> umm, how is "deploying OGL" and "finishing things" not referring to the same thing?
20170619 16:59:58< vultraz_iOS> I have two options here:
20170619 17:00:39< vultraz_iOS> A: I work on deploying OGL now, including moving IPFs to shaders which will be a difficult task
20170619 17:02:39< vultraz_iOS> B: I keep the current method of IPF handling w/ surfaces, convert a few like ~O() to render-time ops, cache as much as I can, and then get everything working, and THEN embark on deploying OGL and "properly" converting IPFs to shaders
20170619 17:03:40< zookeeper> what's the difference between deploying OGL and whatever it is you've been doing in your branch?
20170619 17:04:08< vultraz_iOS> I've been using SDL'2 Renderer API
20170619 17:04:41< vultraz_iOS> it provides basic hw acceleration functionality with in a simple API, but no support for more complex ops, like shaders
20170619 17:05:12< zookeeper> okay. in that case surely B is the sane choice right now, considering how horribly complicated and difficult A would be.
20170619 17:05:56< vultraz_iOS> I'm starting to agree.
20170619 17:06:38< vultraz_iOS> It might be easier to work on deploying OGL too if everything else is working
20170619 17:07:30< zookeeper> after all, the main problem to solve is performance, not lack of shaders. shaders are nice for some fancy future stuff, but lack of them isn't exactly a pressing concern.
20170619 17:08:13< vultraz_iOS> Well, using the surface manipulation method results in some reduced performance.
20170619 17:08:50< vultraz_iOS> That's the downside
20170619 17:09:06< vultraz_iOS> as I said, I can do my best to make some use render-time ops
20170619 17:11:02< vultraz_iOS> If you think time to completion is worth more more than the performance hit (which, I think, will still be better than with plain SW rendering), I can go that route
20170619 17:12:00< zookeeper> yeah the surface manipulation method isn't ideal, but the main performance problem most people seem to have is poor fps, not less than ideal load times or a very short pause when something very IPF-heavy happens for the first time.
20170619 17:12:48< zookeeper> and exactly, it's not going to be _slower_ than now if you keep IPF's working as they do now.
20170619 17:12:57< vultraz_iOS> No, it won't be
20170619 17:13:03< vultraz_iOS> especially since some will be render-time ops
20170619 17:15:18< vultraz_iOS> ~FLIP(), ~ROTATE(), ~O(), ~SCALE(), ~SCALE_INTO(), ~SCALE_SHARP(), and ~SCALE_INTO_SHARP() can all be render-time ops instead of surface ops.
20170619 17:15:19< zookeeper> if we get good/decent/passable render-time performance without IPF processing being any faster than it is now, then that's still a remarkably wonderful result and no one will be complaining
20170619 17:15:59< vultraz_iOS> alright, I'll cordon off my OGL with an include guard and get to work
20170619 17:16:23< vultraz_iOS> OGL work*
20170619 17:16:58< zookeeper> how do you turn some particular IPF into a render-time op if/when that IPF gets used in the middle of an IPF chain, surrounded by non-render-time IPF's?
20170619 17:17:42< vultraz_iOS> *blinks*
20170619 17:18:16< vultraz_iOS> ......yes, that is a good point :|
20170619 17:18:30< zookeeper> maybe hold off on those too then? :]
20170619 17:18:48< vultraz_iOS> yes, perhaps best...
20170619 17:18:56< vultraz_iOS> blehh
20170619 17:20:36< zookeeper> don't worry about it. if you can make gamemap rendering accelerated so that it actually performs notably better than, say, 1.12, then you'll already be a great hero :J
20170619 17:20:57-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 17:24:57-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Ping timeout: 240 seconds]
20170619 17:49:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 246 seconds]
20170619 17:54:24-!- sevu [~Shiki@141.57.60.71] has joined #wesnoth-dev
20170619 18:01:19-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 18:06:22-!- mjs-de [~mjs-de@dorf.cc] has joined #wesnoth-dev
20170619 18:12:21-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Disconnected by services]
20170619 18:12:27-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20170619 18:12:50-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20170619 18:19:45-!- Netsplit *.net <-> *.split quits: Ivanovic, iwaim__
20170619 18:20:24-!- Ivanovic_ is now known as Ivanovic
20170619 18:21:03-!- clavi [~clavi@v22017034422546657.goodsrv.de] has quit [Ping timeout: 240 seconds]
20170619 18:23:40-!- iwaim__ [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20170619 18:25:29-!- clavi [~clavi@v22017034422546657.goodsrv.de] has joined #wesnoth-dev
20170619 18:29:28< vultraz_iOS> zookeeper: warning for travis spam if travis shows up
20170619 18:29:54-!- irker410 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20170619 18:29:54< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 586643470e96 / src/sdl/window.cpp: SDL/Window: set render scaling method to linear https://github.com/wesnoth/wesnoth/commit/586643470e9669f1ff84a8dbd050c60214d0b938
20170619 18:29:55< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 9e34dddc7b74 / src/video.cpp: Video: enabled accelerated rendering with texture target access https://github.com/wesnoth/wesnoth/commit/9e34dddc7b7474a5f620905f761029cbe63c0745
20170619 18:29:56< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering de48ebb05e83 / / (7 files in 4 dirs): GUI2/Canvas: refactor surface blitting out of drawing routines https://github.com/wesnoth/wesnoth/commit/de48ebb05e83b5019a9a6f90dba4e22c784a099f
20170619 18:29:58< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 8b014c0a51b3 / src/gui/widgets/window.cpp: GUI2/Window: properly re-implement draw and restore code https://github.com/wesnoth/wesnoth/commit/8b014c0a51b3ff0d434013e53419bab8e4c7bbc8
20170619 18:29:59< zookeeper> irker you mean?
20170619 18:30:00< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering ce04dad7abeb / src/ (47 files in 5 dirs): GUI2: removed dirty widget interface https://github.com/wesnoth/wesnoth/commit/ce04dad7abeb372868852446abaf3e5ff7fe373a
20170619 18:30:03< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 327eebd364e5 / src/gui/widgets/ (window.cpp window.hpp): GUI2/Window: removed restore flag and restorer surface https://github.com/wesnoth/wesnoth/commit/327eebd364e59a987ca924939a09e2d2a4a7d381
20170619 18:30:05< vultraz_iOS> zookeeper: right
20170619 18:30:06< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 5ee54c4e1717 / src/gui/ (core/canvas.cpp core/canvas.hpp widgets/panel.cpp widgets/styled_widget.cpp): GUI2/Canvas: renamed blit function and removed surface argument https://github.com/wesnoth/wesnoth/commit/5ee54c4e17175f9c057287c0a0e6117724334faa
20170619 18:30:09< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering e9c795ffac38 / src/gui/widgets/ (31 files): GUI2: removed framebuffer code from draw impl functions https://github.com/wesnoth/wesnoth/commit/e9c795ffac3876d7859f38fb493657218504c8bc
20170619 18:30:12< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 51dcfb75c752 / src/font/text.cpp: Hack to get text to render right https://github.com/wesnoth/wesnoth/commit/51dcfb75c7528a7a480be3cc5bc45ca69559878e
20170619 18:30:15< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 62f2b4e5174d / src/gui/ (6 files in 2 dirs): GUI2: refactored viewport setting and re-added clip rect setting https://github.com/wesnoth/wesnoth/commit/62f2b4e5174dce0cfaa20605c6d4f01cd5f8418b
20170619 18:30:18< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering f9a53c95987c / src/gui/ (core/event/handler.cpp dialogs/lua_interpreter.cpp): Fixed some unused parameter warnings https://github.com/wesnoth/wesnoth/commit/f9a53c95987c180606cd17aee78ff5212c2bb995
20170619 18:30:21< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering bbfa66ebb60e / src/gui/widgets/window.cpp: Removed some debug output https://github.com/wesnoth/wesnoth/commit/bbfa66ebb60ec3092c9928f2a7aed30df45f9582
20170619 18:30:24< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 10362f2a4b4e / src/ (gui/core/canvas.cpp gui/widgets/widget.cpp video.cpp): Added some code to make my recent changes work with a software renderer https://github.com/wesnoth/wesnoth/commit/10362f2a4b4eb39ff6d5e2ecccf6026b863b5100
20170619 18:30:24< zookeeper> well 30 seconds of warning isn't quite enough most of the time...
20170619 18:30:27< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering f18d50aa011c / src/gui/core/canvas.cpp: GUI2/Canvas: removed surface_lock objects https://github.com/wesnoth/wesnoth/commit/f18d50aa011c344b336942d70b89271c6fefcdc6
20170619 18:30:30< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 0b6c258b7327 / src/gui/core/event/handler.cpp: Small code simplification https://github.com/wesnoth/wesnoth/commit/0b6c258b73274be6fd97c70d90d2111f55976ec1
20170619 18:30:33< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 48082701fbc6 / src/gui/core/canvas.cpp: GUI2/Canvas: add error logging if canvas texture creation fails https://github.com/wesnoth/wesnoth/commit/48082701fbc68546b263691bae0576a8d83b692f
20170619 18:30:36< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering b203f5111484 / / (11 files in 7 dirs): Added SDL_Texture wrapper class and some utility functions https://github.com/wesnoth/wesnoth/commit/b203f5111484a6c089f4148d69fed28283a51435
20170619 18:30:39< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering b0ec0b97062b / src/gui/widgets/minimap.cpp: GUI2/Minimap: quick fix to get the minimap drawing again https://github.com/wesnoth/wesnoth/commit/b0ec0b97062b51a7091fb5753440d1eeb4714f76
20170619 18:30:42< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 44de75dc6516 / / (5 files in 3 dirs): Moved renderer RAII helpers to their own file and added ones for clip rect and v https://github.com/wesnoth/wesnoth/commit/44de75dc65169cd43b4f097395897209cb767b7e
20170619 18:30:45< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering b7c788a3ec1e / src/video.cpp: CVideo: convert blit_surface to handle input as textures instead of blitting to  https://github.com/wesnoth/wesnoth/commit/b7c788a3ec1e0898126e3276b1b8202c3fcc54e8
20170619 18:30:48< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 83f8b009b099 / src/ (gui/widgets/widget.cpp sdl/rect.cpp): Use get_renderer() over *get_window() https://github.com/wesnoth/wesnoth/commit/83f8b009b099e8d35eb5447bd08578be18d5d381
20170619 18:30:51< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering f4cbc397dc3e / src/ (image.cpp image.hpp): Added a texture cache https://github.com/wesnoth/wesnoth/commit/f4cbc397dc3eb030dd633809ce9cda4b049e6ed0
20170619 18:30:54< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 6b86e79c0dcc / src/sdl/ (texture.cpp texture.hpp): Texture: converted internal handling to use a shared_ptr https://github.com/wesnoth/wesnoth/commit/6b86e79c0dcc52e0d515b4bff6e6db0c59fdcf07
20170619 18:30:57< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 81ea7eb272ee / src/gui/core/canvas.cpp: GUI2/Canvas: disable surface manip w/ accelerated rendering for stretch resize m https://github.com/wesnoth/wesnoth/commit/81ea7eb272eebfa39aff29b37609037aeaea084c
20170619 18:31:00< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 60da3eaa7e52 / src/sdl/render_utils.hpp: Fixed raii renderer rect setters https://github.com/wesnoth/wesnoth/commit/60da3eaa7e52b29cda53f000d250cd689c14bba3
20170619 18:31:03< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 4c24ed69740b / src/ (display.cpp sdl/utils.cpp sdl/utils.hpp): Enabled drawing of minimap in-game https://github.com/wesnoth/wesnoth/commit/4c24ed69740baa603797a91555681d78ebda1b22
20170619 18:31:06< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering cf9eaec03a34 / src/ (7 files in 4 dirs): Moved GUI2's per-frame drawing trigger to global scope https://github.com/wesnoth/wesnoth/commit/cf9eaec03a3488ba3694fac317c6bd2fc806ef0c
20170619 18:31:06< zookeeper> so will this take so long that i should still look up the mute command?
20170619 18:31:09< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 06d316045503 / data/gui/widget/image_default.cfg: GUI2/Image: added tiled-to-size image definition https://github.com/wesnoth/wesnoth/commit/06d31604550357c090a321fc0a59c4666aced57e
20170619 18:31:12< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 29dcc50ff68f / src/gui/widgets/minimap.cpp: GUI2/Miniap: draw any provided canvas data before the minimap https://github.com/wesnoth/wesnoth/commit/29dcc50ff68f2ac054023ce5becdc8e74f83ca32
20170619 18:31:15< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering bea9c279b91b / data/gui/widget/toggle_button_icon.cfg: GUI2: fixed some overlay drawing issues with icon toggle buttons https://github.com/wesnoth/wesnoth/commit/bea9c279b91baa5ea718b28725a531a85b0e3ebe
20170619 18:31:18< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering c38d3d86ba99 / data/gui/widget/image_default.cfg: GUI2/Image: added scaled-to-size definition https://github.com/wesnoth/wesnoth/commit/c38d3d86ba99b4ed892102624073b07e0db1ebc9
20170619 18:31:18< grzywacz> ...
20170619 18:31:21< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 7b57b20e19cd / data/gui/widget/toggle_button_icon.cfg: GUI2/Toggle Button: added medium-sized icon button definition https://github.com/wesnoth/wesnoth/commit/7b57b20e19cda6cf4c6727043355dff94193807d
20170619 18:31:21< vultraz_iOS> eh, just leave it
20170619 18:31:24< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 7b7b65ef5a9e / / (7 files in 6 dirs): Added static prototype for GUI2 in-game theme https://github.com/wesnoth/wesnoth/commit/7b7b65ef5a9e0a8525e32a53358005e8551514d6
20170619 18:31:27< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering aede8582122f / / (19 files in 6 dirs): Decoupled drawing buffer code from display class https://github.com/wesnoth/wesnoth/commit/aede8582122f8b23d53030d7a534e55c02792dcd
20170619 18:31:30< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 56eb6b751644 / src/units/unit.cpp: Unit: fixed hp and xp colors having 0 alpha https://github.com/wesnoth/wesnoth/commit/56eb6b751644079c2f56809c174151fe2c5f682e
20170619 18:31:33< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering d2ec0c9fcf6a / src/ (13 files in 6 dirs): Initial refactor for main-screen accelerated rendering https://github.com/wesnoth/wesnoth/commit/d2ec0c9fcf6aa3a3dd101d5e58c8760ea5eea51b
20170619 18:31:36< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 85cc78f8639b / src/gui/core/event/handler.cpp: GUI2: moved event handling into main context https://github.com/wesnoth/wesnoth/commit/85cc78f8639b912963c3e9e11b83fa3bcfe6bbd8
20170619 18:31:39< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 0e89b7ed2ae1 / src/gui/widgets/text_box_base.cpp: Removed set_is_dirty call (do squash later) https://github.com/wesnoth/wesnoth/commit/0e89b7ed2ae199548f5926d1be39cdace91b48a4
20170619 18:31:42< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 8aa028bda1fa / src/units/drawer.cpp: Draw units under bars/orbs/etc https://github.com/wesnoth/wesnoth/commit/8aa028bda1fa85c3f537c68e2e0b6bea5a802b1f
20170619 18:31:45< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 4cfd11a8cbe1 / src/ (video.cpp video.hpp): Video: expand render_copy to take flip arguments https://github.com/wesnoth/wesnoth/commit/4cfd11a8cbe142d80a938b88ba342b5ac006d93a
20170619 18:31:48< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 2cb7ab18b6e9 / src/gui/core/canvas.cpp: GUI2/Canvas: made use of CVideo::render_copy https://github.com/wesnoth/wesnoth/commit/2cb7ab18b6e958164251cc7b04b4e0b49e90a6d9
20170619 18:31:51< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 99614d332cfd / src/play_controller.cpp: Fixed play_controller not joining current event context https://github.com/wesnoth/wesnoth/commit/99614d332cfdd5d69ceef5b1872b82f84d131b7e
20170619 18:31:54< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering b3ff442e5809 / src/units/drawer.cpp: Uint: fixed bad bar alpha on unit mouseover https://github.com/wesnoth/wesnoth/commit/b3ff442e5809230711c7d97e5d7d13ec9f7f7a3f
20170619 18:31:57< vultraz_iOS> ah i forgot to squash that one
20170619 18:31:57< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 3d157dd6dfc7 / src/ (13 files in 4 dirs): Another huge chunk of progress https://github.com/wesnoth/wesnoth/commit/3d157dd6dfc7062c889264b4d2fe800c0fbf533c
20170619 18:32:00< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering eba8770b1bfa / src/ (13 files in 4 dirs): Removed draw() calls into the display class https://github.com/wesnoth/wesnoth/commit/eba8770b1bfa142a9f76d41cc985ad81b05d72c4
20170619 18:32:03< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 3ea1ae8cbd78 / src/sdl/ (window.cpp window.hpp): SDL/Window: added getter for output size (pixels) https://github.com/wesnoth/wesnoth/commit/3ea1ae8cbd78b3e36deed2243da141f7e7f514aa
20170619 18:32:06< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering e81843335e73 / src/video.cpp: Video: removed reliance on framebuffer for window dimensions https://github.com/wesnoth/wesnoth/commit/e81843335e7315e89174b7f58ea89cd6e295d0bf
20170619 18:32:09< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering aeb33e5d70ef / src/ (events.cpp video.cpp video.hpp): Video: (almost) completely removed framebuffer surface https://github.com/wesnoth/wesnoth/commit/aeb33e5d70ef284d13e0353f9d8c096a59c09525
20170619 18:32:12< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering f8e9a123394f / src/ (tests/utils/fake_display.cpp video.cpp video.hpp): Video: removed unused arguments from make_test_fake https://github.com/wesnoth/wesnoth/commit/f8e9a123394fab61496af9ef1ac4f13d3fdec314
20170619 18:32:15< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 737a50cdc551 / src/ (display.cpp display.hpp game_display.cpp units/drawer.cpp): Got a bunch more components drawing https://github.com/wesnoth/wesnoth/commit/737a50cdc551b510ce49bea33b92ff276bd8b6c6
20170619 18:32:18< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 0f56a42b6aa8 / src/display.cpp: Enabled more stuff drawing https://github.com/wesnoth/wesnoth/commit/0f56a42b6aa8e0169c1131a3427aed965ad27606
20170619 18:32:21< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 7048bfbee5d7 / src/ (5 files in 2 dirs): Removed most call to raise_draw_event https://github.com/wesnoth/wesnoth/commit/7048bfbee5d7553f2e90dcd488a9a376edc3ed54
20170619 18:32:24< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering fdce26922b10 / src/video.cpp: Video: enabled VSync for renderer (experimental) https://github.com/wesnoth/wesnoth/commit/fdce26922b10d0a65543bd9a2eac1d22fc8016fc
20170619 18:32:27< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 03c8ab410ed5 / src/units/drawer.cpp: Units: fixed bad HP/XP bars fill rect offset https://github.com/wesnoth/wesnoth/commit/03c8ab410ed52ee16789a26e5b3adabb33ff27b1
20170619 18:32:30< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 150f8cc38387 / src/gui/core/canvas.cpp: GUI2/Canvas: fixed mirrored images being flipped on the wrong axis. https://github.com/wesnoth/wesnoth/commit/150f8cc38387039761f11f0691ef27bb7620a2cc
20170619 18:32:32< vultraz_iOS> man i really have done a lot of work
20170619 18:32:33< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 17b19f71352c / src/display.cpp: Cleaned up the implementation of display::scroll() https://github.com/wesnoth/wesnoth/commit/17b19f71352cf6eaeeb14f3ad9bced55b66a608a
20170619 18:32:36< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 7ec043f326e2 / src/actions/move.cpp: Removed an unused variable https://github.com/wesnoth/wesnoth/commit/7ec043f326e2260040272e8a07091f3d939424d9
20170619 18:32:39< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering c74deaf9f64e / src/ (image.cpp image.hpp): Extended implementation of the texture cache https://github.com/wesnoth/wesnoth/commit/c74deaf9f64e6729143c5fc2b8da0a0f794492b2
20170619 18:32:42< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 483e5b3bd749 / src/display.cpp: Trim ToD images to hex https://github.com/wesnoth/wesnoth/commit/483e5b3bd749595bfd5f4d469a36a72dbbcf0e80
20170619 18:32:45< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 961f625a6046 / src/display.cpp: Temporarily disable map label position update on map scroll https://github.com/wesnoth/wesnoth/commit/961f625a6046c700ad736147887dae51628d7e1f
20170619 18:32:48< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 00fded0a16f1 / src/ (gui/core/canvas.cpp gui/widgets/widget.cpp video.cpp): Removed software rendering fallback https://github.com/wesnoth/wesnoth/commit/00fded0a16f122132d51450b3cab6d7bab7f5bcf
20170619 18:32:51< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering d8b583c87499 / src/gui/core/ (canvas.cpp canvas_private.hpp): GUI2/Canvas: refactor surface use out of image rendering https://github.com/wesnoth/wesnoth/commit/d8b583c8749913333d25b5148218715dea870492
20170619 18:32:54< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 515b42b602e4 / / (13 files in 6 dirs): Some initial code (disabled for now) for setting up an OGL context https://github.com/wesnoth/wesnoth/commit/515b42b602e49ce68eefdc35b8a4bd48e43371a2
20170619 18:33:08< vultraz_iOS> zookeeper: ok it's done
20170619 18:34:04< zookeeper> next time, >30 seconds :p
20170619 18:40:47< vultraz_iOS> gfgtdf: your resolution changes mean the small titlescreen panels are chosen for me even with a maximized window :/
20170619 18:41:55-!- gfgtdf [~chatzilla@x4e368ba3.dyn.telefonica.de] has joined #wesnoth-dev
20170619 18:42:06< gfgtdf> vultraz_iOS: hmm probably didnt update the cfg file of that
20170619 18:42:49< vultraz_iOS> panel_box_display.cfg
20170619 18:42:52-!- TadCarlucci [~lundberg@74.193.219.119] has joined #wesnoth-dev
20170619 18:43:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170619 18:43:46< TadCarlucci> I don't see what I'm doing wrong. I do what Travis does (export CC and CXX) setting for clang and when I make VERBOSE=1 it's still using gcc.
20170619 18:43:54-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 18:44:12< vultraz_iOS> ?
20170619 18:44:15< vultraz_iOS> what are you doing?
20170619 18:44:39< irker410> wesnoth: gfgtdf wesnoth:master fbc7652a0cb4 / data/gui/widget/panel_box_display.cfg: fix panel_box_display.cfg window_width,window_height https://github.com/wesnoth/wesnoth/commit/fbc7652a0cb4cd10e04dfad76203c778efd7a072
20170619 18:44:51< TadCarlucci> trying to test that cmake/clang bug on Travis. And I can't get my local cmake to use clang.
20170619 18:44:59< gfgtdf> vultraz_iOS: try thst ^
20170619 18:45:17< gfgtdf> that*
20170619 18:45:53< vultraz_iOS> yeah that works thanks
20170619 18:47:58< vultraz_iOS> DeFender1031: ping
20170619 18:48:21< irker410> wesnoth: Charles Dang wesnoth:accelerated_rendering 15752f19e15d / src/ (image.cpp image.hpp): Allowed IPFs to work again when using the texture cache https://github.com/wesnoth/wesnoth/commit/15752f19e15d550e1b674cd4a7535a71f44920dd
20170619 18:49:27< vultraz_iOS> DeFender1031: ^ just restores IPFs to working order
20170619 18:51:55< vultraz_iOS> (you'll need to do some git stuff since i force-pushed... not sure what the best method is..)
20170619 18:57:04-!- TadCarlucci [~lundberg@74.193.219.119] has quit [Quit: Leaving]
20170619 18:57:06< gfgtdf> zookeeper: i was wondering: unlike standard unit filter, standard terrain filter so not support thigns liek $this_unit right? so there is no need to delay variable substitution until he filter is actualyl called ona a specific tile, 
20170619 19:00:09< DeFender1031> vultraz_iOS, i dunno what that means... will a normal fetch not do it?
20170619 19:00:36< vultraz_iOS> gfgtdf: what's the best method for DeFender1031 to update his local branch now that i force-pushed to remote?
20170619 19:01:09< gfgtdf> vultraz_iOS: does he has local changes on it or does he just want to overwrite it ?
20170619 19:01:19< vultraz_iOS> overwrite i think
20170619 19:01:51-!- JyrkiVesterinen [~jyrki@89-166-101-10.bb.dnainternet.fi] has quit [Quit: Going to bed]
20170619 19:02:08< gfgtdf> vultraz_iOS: you coudl do something liek update-ref, another possiblty is to first do a 'git reset HEAD~100 --hard' and then a normal git pull (-r)
20170619 19:02:33< vultraz_iOS> DeFender1031: ^
20170619 19:02:34< gfgtdf> vultraz_iOS: assuming that you forcpozsn didn't change more than 100 commits.
20170619 19:03:05< vultraz_iOS> no, only 64 commits :P
20170619 19:03:27< vultraz_iOS> though i did rebase w/ master
20170619 19:04:12< gfgtdf> vultraz_iOS: well if the othe thing doesnt work i'd use update-ref. It's possibel that there are better ways though.
20170619 19:04:48< zookeeper> gfgtdf, yeah, i think so. unless the filter uses a lua_function= with side-effects, maybe?
20170619 19:05:48< DeFender1031> i did a normal fetch, then checked out 515b42b602e4, which gave me a warning, before seeing here that it might be more complicated. After that i did reset HEAD~100 --hard and then another fetch, which did nothing, and checking out 515b again didn't give a warning either.
20170619 19:06:00< DeFender1031> I assume this means that I haven't totally screwed everything up?
20170619 19:06:05< gfgtdf> zookeeper: amm i think it's safe to assume that people shouldn't ahev side effects in lua_function or relay on the order in whihc the locatios are checked in geneal
20170619 19:06:27< DeFender1031> (I already started sconsing in the meantime, because, well, hour-long compile, may as well get started.
20170619 19:06:29< DeFender1031> )
20170619 19:06:34< zookeeper> gfgtdf, sure, seems reasonable
20170619 19:06:45< DeFender1031> gfgtdf, ^^^
20170619 19:07:01< gfgtdf> DeFender1031: hmm hard to tell what it'll look after that without seeng it
20170619 19:07:53< gfgtdf> DeFender1031: i'D probably do 'checkout <your local branch name>'
20170619 19:08:25< gfgtdf> and then do a 'git update-ref HEAD 515b42b602e4'
20170619 19:08:49< gfgtdf> you'll probably need a 'git reset HEAD --hard' after that
20170619 19:09:31< DeFender1031> blarg.
20170619 19:10:35< zookeeper> gfgtdf, so i take it that you want to make SLF matching faster by making variable substitution in SLF only happen once instead of once per location checked, or something like that?
20170619 19:10:49< gfgtdf> zookeeper: yes that was my idea
20170619 19:13:03< zookeeper> gfgtdf, okay, sounds good to me. i can imagine one _might_ want to add something analogous to $this_unit at some point in the future, but it's hard to come up with any ideas for what that could be.
20170619 19:13:44-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 19:22:34< DeFender1031> gfgtdf, vultraz_iOS, https://xkcd.com/1597/
20170619 19:23:13< vultraz_iOS> indeed :P
20170619 19:23:36< vultraz_iOS> build failed?
20170619 19:38:45-!- EliDupree [~quassel@2604:a880:400:d0::9bb:2001] has quit [Remote host closed the connection]
20170619 19:38:55-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170619 19:39:13-!- EliDupree [~quassel@2604:a880:400:d0::9bb:2001] has joined #wesnoth-dev
20170619 19:39:29-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 19:50:09< DeFender1031> not at all
20170619 19:50:16< DeFender1031> just general frustration over git.
20170619 19:52:02< DeFender1031> though, after building, it seems as though I didn't actually get the updated code...
20170619 19:53:47< vultraz_iOS> :|
20170619 19:53:49-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Remote host closed the connection]
20170619 19:54:44< DeFender1031> okay... let's see what update-ref does...
20170619 19:55:55< DeFender1031> nothing. great.
20170619 19:56:08< gfgtdf> DeFender1031: did you do the 'git reset HEAD --hard' after update-ref ?
20170619 19:56:21< DeFender1031> okay, i checked out master, then checked out 515b... let's see what that does.
20170619 19:56:23< vultraz_iOS> DeFender1031: what does `git log -n1` give you?
20170619 19:56:48< DeFender1031> commit 515b42b602e49ce68eefdc35b8a4bd48e43371a2
20170619 19:57:02< DeFender1031> gfgtdf, no... should I have?
20170619 19:57:13< vultraz_iOS> you want 15752f1
20170619 19:57:26< DeFender1031> oh...
20170619 19:57:30< DeFender1031> that'd explain it.
20170619 19:58:10-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 19:59:13< vultraz_iOS> DeFender1031: 515b42b60 is the commit right before what you want
20170619 19:59:30< DeFender1031> i know
20170619 19:59:36< DeFender1031> i missed the last one in the channel
20170619 19:59:53< DeFender1031> i checked out the proper commit and now it looks like i'm in the state i need.
20170619 19:59:57< DeFender1031> sconsing now.
20170619 20:00:24< DeFender1031> will let you know in half an hourish if it worked or not.
20170619 20:07:13-!- mjs-de [~mjs-de@dorf.cc] has quit [Quit: On the road again]
20170619 20:13:47-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Remote host closed the connection]
20170619 20:14:14-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 20:17:49-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 20:21:47< DeFender1031> vultraz_iOS, for some reason, the fade is now jumping to black and then fading in from black each time rather than fading from one image to the next...
20170619 20:22:01< DeFender1031> don't have time now to dig deeper as to a possible cause.
20170619 20:22:13< vultraz_iOS> hmmm
20170619 20:22:19< DeFender1031> good news though is that the fade from black seems decently smooth.
20170619 20:22:21< vultraz_iOS> but is the fade smooth
20170619 20:24:18< vultraz_iOS> DeFender1031: can i have the addon again?
20170619 20:24:22< vultraz_iOS> I'll look into the alpha issues
20170619 20:25:22-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 20:26:02-!- horrowind [~Thunderbi@p2003008E6C2A64D6964452FFFE0220ED.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170619 20:30:29-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170619 20:31:02-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 20:32:07-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 20:32:24< vultraz_iOS> ok...
20170619 20:32:30< vultraz_iOS> hmmm
20170619 20:33:00< vultraz_iOS> DeFender1031: is this the fade granularity you intended?
20170619 20:33:15< vultraz_iOS> it doesn't seem absolutely smooth to me, but perhaps you intended it so
20170619 20:38:53< vultraz_iOS> hmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
20170619 20:39:03-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 20:44:47-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Remote host closed the connection]
20170619 20:47:22-!- sevu [~Shiki@141.57.60.71] has quit [Remote host closed the connection]
20170619 20:48:32-!- mjs-de [~mjs-de@x4e318eed.dyn.telefonica.de] has joined #wesnoth-dev
20170619 20:49:22-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 20:50:14< vultraz_iOS> for the record this method is now very inefficient 
20170619 20:53:20-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20170619 20:55:02< vultraz_iOS> very very inefficient..
20170619 20:55:03< vultraz_iOS> problem is
20170619 20:55:12< vultraz_iOS> I'm not sure how to expose the more efficient method! :(
20170619 20:56:29< vultraz_iOS> I yearn for anura's object system...
20170619 21:08:58-!- vn971 [~vasya@0896414046.static.corbina.ru] has joined #wesnoth-dev
20170619 21:09:21< vn971> Greetings. +1 to bugs, I think I've found another one.
20170619 21:09:30< vultraz_iOS> :(
20170619 21:10:17< vn971> A dwarf berserker (AI) attacks a Ghoul at his turn. He gets hits from the Ghoul but does NOT get poisoned.
20170619 21:11:16-!- atarocch [~atarocch@93.56.160.37] has quit [Remote host closed the connection]
20170619 21:11:51< zookeeper> vn971, got a save/replay or something reproducible?
20170619 21:16:22< Ravana_> last time I checked such situation it was from addon changing units type, but not changing undead-specific variables
20170619 21:18:15< gfgtdf> ye i also already observed what ravanna just said.
20170619 21:18:15< vn971> zookeeper: danm, can't reproduce myself. But I definitely saw it, checked twice, including a reload. Since this seems non-trivial, will try to extract a working case anyway. Stay tuned..
20170619 21:18:33< gfgtdf> was some randomleader no-mirrir addon i think.
20170619 21:19:22< gfgtdf> when i observed it mean
20170619 21:21:43-!- Netsplit *.net <-> *.split quits: DDR, MagneticUnicorn[, Ieuan, ChipmunkV[m]
20170619 21:21:50< vn971> Ravana_: I have an addon that "randomizes zombies", but it was disabled for this game. Active addons didn't change units. Maybe also the map did, this I do not know. It's Race for Dimaga Beach
20170619 21:22:06-!- Netsplit over, joins: DDR
20170619 21:22:08< vn971> *active addons do not change unit types.
20170619 21:23:18-!- yaiyan [~yaiyan@46.101.48.31] has joined #wesnoth-dev
20170619 21:23:20< vn971> uh sorry, disregard, it's not a bug. I messed up.:confused:
20170619 21:24:04< vn971> The unit had an additional weapon and just chose it. So it's all right.
20170619 21:24:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170619 21:24:16< zookeeper> whoops.
20170619 21:24:37-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 21:25:29< vn971> human error..)
20170619 21:25:51-!- MagneticUnicorn[ [magneticun@gateway/shell/matrix.org/x-bdbobfwjhxcuvwkr] has joined #wesnoth-dev
20170619 21:26:32-!- ChipmunkV[m] [chipmunkvm@gateway/shell/matrix.org/x-jjhqkratpzjxauii] has joined #wesnoth-dev
20170619 21:28:59-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 21:30:53< DeFender1031> vultraz_iOS, i played with different granularities until i found one that was a balance of semi decent and not lagging in 1.12. If these rendering enhancements really end up working well, then I may be able to increase the granularity.
20170619 21:31:40< DeFender1031> vultraz_iOS, despite the utter crap quality of my early WML code, the macro for generating the fade ought to be simple enough that you can adjust it to different granularities and see what kind of effect you get.
20170619 21:31:57< vultraz_iOS> DeFender1031: it's not anything on your end
20170619 21:32:02< vultraz_iOS> it's this:
20170619 21:32:17< vultraz_iOS> since the images use IPFs, surfaces must be generated
20170619 21:32:25< vultraz_iOS> so it generates one surface for each alpha level
20170619 21:32:25-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 21:32:38< DeFender1031> well, the pre-caching that i mentioned the other day could help with that.
20170619 21:32:42< vultraz_iOS> a texture is also created for each of those and added to the cache
20170619 21:32:47< DeFender1031> right.
20170619 21:33:28< DeFender1031> so does it suffer from the same sort of "lag on first view, smooth on subsequent views" issue that I noticed with the water?
20170619 21:34:09-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 21:34:16< vultraz_iOS> so here we have 10 surfaces, each with alpha modification that happens on the CPU before being uploaded to textures and then a different one being drawn in the same place every 50 ms
20170619 21:34:29< DeFender1031> right...
20170619 21:34:43< DeFender1031> well, I stand by my pre-caching suggestion.
20170619 21:35:15< vultraz_iOS> But in C++, I could instead take one texture for the base image, loaded directly (so no surfaces), and just call SDL_SetTextureAlphaMod every, say 50 ms.
20170619 21:35:53< vultraz_iOS> the texture is rendered every frame (or it should be), and any change to the alpha mod will just reflect next frame
20170619 21:36:31< vultraz_iOS> You can see how that would be more efficient :/
20170619 21:36:51< vultraz_iOS> But I don't know how best to expose such functionality to WML without it being so overly specific as to be bad design
20170619 21:37:14< vultraz_iOS> DeFender1031: see the problem :/
20170619 21:39:03-!- Rhonda [~rhonda@anguilla.debian.or.at] has quit [Remote host closed the connection]
20170619 21:42:34-!- Rhonda [~rhonda@anguilla.debian.or.at] has joined #wesnoth-dev
20170619 21:45:22< zookeeper> i don't know which fadeout you're having a problem with, but couldn't one workaround the IPF slowness by causing them to be pre-cached by incorporating those IPF chains as part of an image that is seen earlier?
20170619 21:46:21< zookeeper> terribly hacky of course, but still.
20170619 21:47:10-!- mjs-de [~mjs-de@x4e318eed.dyn.telefonica.de] has quit [Remote host closed the connection]
20170619 21:49:00-!- irker410 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20170619 21:54:15-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 21:55:20-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 22:02:56-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 22:06:39< DeFender1031> vultraz_iOS, yeah, WML doesn't lend itself to neatly abstracting graphical features, which again, is why I suggest the pre-caching approach.
20170619 22:08:03< DeFender1031> zookeeper, that's an interesting idea, but wouldn't it make sense for the game itself to just go through all the WML for a given game load and be aware of what keys refer to images and just pre-build the cache at load time?
20170619 22:09:03-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has quit [Remote host closed the connection]
20170619 22:15:40-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 22:15:58-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 22:16:39-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170619 22:16:43< zookeeper> DeFender1031, of course
20170619 22:17:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 22:20:00-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 22:21:44-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 22:26:20-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 22:28:02-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 22:29:41-!- zookeeper [zookeeper@wesnoth/developer/zookeeper] has quit [Ping timeout: 255 seconds]
20170619 22:34:50< gfgtdf> DeFender1031: i think for a lot of imagees it onyl decionded at runtime whether they are actualyl used
20170619 22:35:18< gfgtdf> DeFender1031: also some images/ifp are creates ar runtime by wml variabe concatination or lua
20170619 22:36:05-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 22:38:00-!- vn971 [~vasya@0896414046.static.corbina.ru] has left #wesnoth-dev ["ISON Rogach NickServ "]
20170619 22:41:57-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 22:44:17< DeFender1031> gfgtdf, for your first point, I think it's probably pretty rare for images not to be used, and probably better performance-wise to waste a little space in the cache for an image we never show than to have to load things on-the-fly in real time. For your second one, yes, obviously there will be edge cases which won't be possible to completely cover, but even a partial improvement here will help greatly.
20170619 22:45:33< DeFender1031> I'm also thinking that perhaps we could offer some kind of [precache] tag or something where creators can specify a list of images or IFP strings which won't be covered automatically by the precache I'm suggesting.
20170619 22:45:58< gfgtdf> DeFender1031: unusd images are there all the time, just  think of a the unit and terrain stuff (which usuualy are most of the images i think). a usualy scnerio will maybe use <20% of all unit types in wesnoth, i don't know for terrian but i expect similar values.
20170619 22:46:08< DeFender1031> and obviously, anything not precached would still be generated on the fly, but at least this would improve most cases.
20170619 22:46:49< DeFender1031> gfgtdf, that's fair, but is it really that much overhead to include them in the cache? how much data are we really talking about?
20170619 22:47:48< DeFender1031> gfgtdf, also, the precacher could be written smart enough to know what campaign/scenario is being loaded and look only bother with units and terrains used therein.
20170619 22:48:06-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has joined #wesnoth-dev
20170619 22:48:19< DeFender1031> (obviously accounting for things like [unit], [replace_map], etc.)
20170619 22:49:25< gfgtdf> DeFender1031: not sure was just a rnadom thought. since data\core\images is less than 90mb here we are probably talking about < 90mb
20170619 22:49:37< DeFender1031> another approach which would probably be more complicated would be a JIT cacher which basically says "okay, i'm about to display this tag which involves these images, so we'll cache everything that's about to be loaded immediately even though we're 50 ms away from actually using the next frame"
20170619 22:49:50-!- Greg-Boggs [~greg_bogg@2601:1c2:f00:9780:75b7:100f:b3bc:77b5] has quit [Remote host closed the connection]
20170619 22:50:19< gfgtdf> DeFender1031: well i think, a basis scenario parser could make sense. I wonder though, are we taking here about master branch or acelerated_rendering branch ?
20170619 22:50:21< DeFender1031> gfgtdf, yeah, so I think 90 extra mb of cache in exchange for not having to delay and generate all sorts of crap on the fly is reasonable.
20170619 22:50:37-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20170619 22:50:45< DeFender1031> gfgtdf, or a separate branch entirely...
20170619 22:51:19< DeFender1031> I'm not really a good person to suggest how to structure the git repo for an idea though...
20170619 22:51:39< gfgtdf> DeFender1031: so you currently have problems with imageloading on current master ?
20170619 22:59:10< DeFender1031> gfgtdf, hard to tell. rendering on master is already molasses, so a bit of extra lag from image loading isn't really all that noticeable.
20170619 23:00:47-!- Greg-Boggs [~greg_bogg@c-76-115-139-154.hsd1.or.comcast.net] has joined #wesnoth-dev
20170619 23:06:16-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20170619 23:09:12-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170619 23:09:46-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170619 23:13:33-!- travis-ci [~travis-ci@ec2-54-224-118-55.compute-1.amazonaws.com] has joined #wesnoth-dev
20170619 23:13:34< travis-ci> wesnoth/wesnoth#14289 (accelerated_rendering - 15752f1 : Charles Dang): The build has errored.
20170619 23:13:34< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/244645204
20170619 23:13:34-!- travis-ci [~travis-ci@ec2-54-224-118-55.compute-1.amazonaws.com] has left #wesnoth-dev []
20170619 23:36:37-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20170619 23:49:50-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20170619 23:53:05-!- horrowind [~Thunderbi@p2003008E6C2A64D6964452FFFE0220ED.dip0.t-ipconnect.de] has quit [Quit: horrowind]
--- Log closed Tue Jun 20 00:00:06 2017