Remove mozCurrentTransform/mozCurrentTransformInverse usage
#15281
Conversation
|
Please note: Based on local testing I'm expecting smaller movement in some test-cases, since it appears that /botio test |
From: Bot.io (Windows)ReceivedCommand cmd_test from @Snuffleupagus received. Current queue size: 0 Live output at: http://54.193.163.58:8877/d525ec7d81f1c59/output.txt |
From: Bot.io (Linux m4)ReceivedCommand cmd_test from @Snuffleupagus received. Current queue size: 0 Live output at: http://54.241.84.105:8877/3ff6d6eee85cb14/output.txt |
From: Bot.io (Linux m4)FailedFull output at http://54.241.84.105:8877/3ff6d6eee85cb14/output.txt Total script time: 25.82 mins
Image differences available at: http://54.241.84.105:8877/3ff6d6eee85cb14/reftest-analyzer.html#web=eq.log |
From: Bot.io (Windows)FailedFull output at http://54.193.163.58:8877/d525ec7d81f1c59/output.txt Total script time: 29.89 mins
Image differences available at: http://54.193.163.58:8877/d525ec7d81f1c59/reftest-analyzer.html#web=eq.log |
|
I checked few failing ref tests and I didn't see anything really wrong: most of time some pixels are slightly different and it could be caused by some rounding errors. |
|
Oh I missed your second message. |
|
Interestingly enough, when testing locally on Windows I got a lot fewer "failures" (~100) and in addition the movement in each case was generally smaller compared to the bots. Possibly the particular graphics back-end, as used by the browser, affects things here. |
|
Does this also, at least partially, fix #7799? |
Partially yes, but the issue also talks about tracking the transform internally (to avoid querying the canvas) and that's not attempted here at all. |
These canvas-context properties are Mozilla-specific, and has obviously never been implemented anywhere else. Currently they are in the process of being removed, see [bug 1782651](https://bugzilla.mozilla.org/show_bug.cgi?id=1782651) and [bug 1294360](https://bugzilla.mozilla.org/show_bug.cgi?id=1294360), which in practice means that in e.g. Firefox Nightly the `addContextCurrentTransform`-function is now being used in the *built-in* PDF Viewer (which was obviously never intended). We should thus be able to replace these Mozilla-specific properties with `CanvasRenderingContext2D.getTransform()`, which is available in all browsers that we currently support: https://developer.mozilla.org/en-US/docs/Web/API/CanvasRenderingContext2D/getTransform#browser_compatibility
Snuffleupagus
force-pushed
the
getTransform
branch
from
bfddf0f
to
358a060
Compare
|
Looks good, and gets rid of quite a lot of code. Let's do this! /botio makeref |
From: Bot.io (Linux m4)ReceivedCommand cmd_makeref from @timvandermeij received. Current queue size: 0 Live output at: http://54.241.84.105:8877/41369f649c146f3/output.txt |
From: Bot.io (Windows)ReceivedCommand cmd_makeref from @timvandermeij received. Current queue size: 1 Live output at: http://54.193.163.58:8877/f2e4c5b9f3ff1d9/output.txt |
From: Bot.io (Linux m4)SuccessFull output at http://54.241.84.105:8877/41369f649c146f3/output.txt Total script time: 22.21 mins
|
From: Bot.io (Windows)SuccessFull output at http://54.193.163.58:8877/f2e4c5b9f3ff1d9/output.txt Total script time: 23.00 mins
|
These canvas-context properties are Mozilla-specific, and has obviously never been implemented anywhere else. Currently they are in the process of being removed, see bug 1782651 and bug 1294360, which in practice means that in e.g. Firefox Nightly the
addContextCurrentTransform-function is now being used in the built-in PDF Viewer (which was obviously never intended).We should thus be able to replace these Mozilla-specific properties with
CanvasRenderingContext2D.getTransform(), which is available in all browsers that we currently support: https://developer.mozilla.org/en-US/docs/Web/API/CanvasRenderingContext2D/getTransform#browser_compatibility