some people want so bad for opengl to go away.

[aeva@mastodon.gamedev.place]

some people want so bad for opengl to go away. buddy, XNA hasn't even gone away.

[hyratel@yiff.life]

@aeva despite numerous attempts by Microsoft to say "Directinput (generous but limited scope USB HID Joystick API Standard: 8 axis, 4 POV/Hat, 128 buttons) is going away! Use Xinput instead!", more gamepads than ever are being shipped with DI-compliant USB HID, rather than XI. And no Flight Stick over $100 is going to have Xinput at all. (DI is an easily-sidestepped API if you need SDL or raw HID, but there's few reasons not to use it in Windows) For historical context: DI reached Feature Complete around ..... 2005? and there's been no pressure/need to change it. the only thing it substantially lacks are Nice Names for buttons, they just have a number

[aeva@mastodon.gamedev.place]

@hyratel that's awesome

[lritter@mastodon.gamedev.place]

@aeva worst case: GL becomes a pure front-end layer for vulkan. it's the API that matters.

[aeva@mastodon.gamedev.place]

@lritter yup. ES2 is already there thanks to ANGLE, which makes it the most portable graphics API

[tomf@mastodon.gamedev.place]

@lritter @aeva Yes but which of the myriad versions? Which is, of course, the problem with OpenGL.

They should make an OpenGL 5.0 that is a 1:1 wrapper of Vulkan and there you go.

[breakin@mastodon.gamedev.place]

@TomF @lritter @aeva We need to find the perfect mix of Vulkan and glBegin.

[lritter@mastodon.gamedev.place]

@TomF @aeva ?? freeze it at 4.5 and we're good. it's done. it's complete.

[jupiter@mastodon.gamedev.place]

@lritter @aeva You have a great point here, IMHO these relatively high level middleware frontends, like WebGPU, DirectX11 or whatever OpenGL is becoming, are important for what I would call the majority of developers and projects. It was important to unlock something closer to the metal, yet it cannot undo the addiction we have to abstractions to get anything reasonable done, especially with more diverse platforms and features. I remember days where just about everything was using GLU and GLUT.

[tomf@mastodon.gamedev.place]

@lritter @aeva No, you need to throw away 90% of it first. That's the problem.

[tomf@mastodon.gamedev.place]

@breakin @lritter @aeva Exactly! See... someone gets it.

[lritter@mastodon.gamedev.place]

@TomF @aeva ??? GL (like XNA) is for people with a habit and existing GL software. there is nothing to remove, nothing to replace, nothing to add.

[lritter@mastodon.gamedev.place]

@TomF @breakin @aeva well i use SDL_gpu now and i'm reasonably happy.

[lritter@mastodon.gamedev.place]

@TomF @aeva maybe this is difficult to understand for programmers so let me try another way

imagine software that is *done*. all books you bought about it forever valid. a target that does not move. it's legacy, but in the original meaning.

there is no need to upset Graham Sellers any further. Let Him Have This.

[aeva@mastodon.gamedev.place]

@lritter @TomF graham sellers just wants us to be happy

[breakin@mastodon.gamedev.place]

@lritter @TomF @aeva Maybe Graham Sellers is excited for OpenGL 5.0, eager to cut the 90%? Have you asked?

edit: I mean eager to write the book for it!

[aeva@mastodon.gamedev.place]

@breakin @lritter @TomF tbh I've considered making a library called "GL5" where it's (mostly) just a bindings generator for an opinionated subset of OpenGL 4.5

[aeva@mastodon.gamedev.place]

@breakin @lritter @TomF however my plans for the immediate future are strictly ES2

[tomf@mastodon.gamedev.place]

@aeva @breakin @lritter "Ey, there's a line here..."

[lritter@mastodon.gamedev.place]

@breakin @TomF @aeva i'm serious this time. the man did so much tech support on twitter, helped and advised me a bunch of times, and only stopped because of harrassment. he deserves better.