halide
Adaptive Dark colorscheme for Stmt HTML. Ability to programmatically export conceptual stmt files.
Colorscheme based on gruvbox-material.
Also improved assembly syntax highlighting (while reducing one HTTP fetch, as the highlight rules are now embedded in the generated file.).
I'm not going to reject this, but let the record show that IMHO a light color scheme is preferable to a dark one; is there a way to have this say "prefer the system setting"?
It follows your system preferences. It's one of those CSS media queries. You'll see in the diff.
It follows your system preferences. It's one of those CSS media queries. You'll see in the diff.
Can it be made into a checkbox that either enables or disables dark mode? It can use the system preference by default, but I don't want to have to change my whole system preference just to get the page to switch palates.
Can it be made into a checkbox that either enables or disables dark mode?
Fair request; working on it.
Updated: added a toolbar with various settings. Browser persists those settings automatically (no cookie stuff). Color theme can be set to auto, or manual modes. Auto mode also live switches with the system preferences. @alexreinking @steven-johnson See video.
https://github.com/halide/Halide/assets/845012/be572828-5ddc-4f15-842d-9e0577b5029f
@steven-johnson Added those python bindings AFAICT. Not tested that tho. Just a copy-paste-and-modify.
@steven-johnson Did this OSX buildbot fail twice on exactly the same test? This thing doesn't seem to fail on other things? It's surprising that this PR would be related but I'm not ruling anything out here. Do you have more knowledge on this failure? Please see #8332 regarding this as well.
@steven-johnson Did this OSX buildbot fail twice on exactly the same test? This thing doesn't seem to fail on other things? It's surprising that this PR would be related but I'm not ruling anything out here. Do you have more knowledge on this failure? Please see #8332 regarding this as well.
Not sure, let's retry again
Failure does seem unrelated...
2024-06-28 17:09:27.935 generator_aot_gpu_multi_context_threaded[91557:5585717] Metal API Validation Enabled
Assertion failed: (buf2_result.all_equal(6)), function run_kernels_on_thread, file gpu_multi_context_threaded_aottest.cpp, line 237. 473/474 Test #631: generator_aotcpp_gpu_multi_context_threaded ................Subprocess aborted***Exception: 32.16 sec 2024-06-28 17:09:27.944 generator_aotcpp_gpu_multi_context_threaded[91558:5585726] Metal API Validation EnabledAssertion failed: (buf1_result.all_equal(val + 2)), function run_kernels_on_thread, file gpu_multi_context_threaded_aottest.cpp, line 236.
Can this be merged, or should something change?
@steven-johnson Can you have a look if this is okay? Side question: I'm guessing many folks are taking vacations and activity here is lower because of that?