Basically, only

1. layout/style/nsCSSKeywordList.h
2. layout/style/nsCSSProps.cpp
3. layout/style/nsStyleConsts.h
4. servo/components/style/values/computed/length.rs

are needed for reviewing. The rest are just replacing -moz-max-content with max-content and -moz-min-content with min-content (in test files and comments).

Yeah sorry I was missing the context from bugzilla. I'd prefer to avoid prefs if possible (in particular to hide the unprefixed value) because they need to affect not only parsing but serialization, but just so i understand what is exactly the proposed behavior?

• Pref that toggles between prefixed and unprefixed.
• Pref that hides the prefixed version.
• Pref that hides the unprefixed one.

(1) and (3) are tricky because you need to either have extra enum variants and code handling it or switch in both parsing and serialization to prefix or not, and that includes the computed style serialization which uses the keyword tables. Not impossible, but not super excited about duplicating those tables just for this.

Re. (2) I think that is probably not worth guarding behind a pref...

In D7536#173875, @emilio wrote:

• Pref that toggles between prefixed and unprefixed.
• Pref that hides the prefixed version.
• Pref that hides the unprefixed one.

I think xidorn suggests (2) and (3) together, and mats suggests (3). The serialization could be a problem and I didn't notice it. So now, I prefer to adding the alias without any pref. I will update this patch by adding the alias, instead of the pref.

In D7536#179442, @emilio wrote:

This looks ok, code-wise, but I'd rather fix the computed value... Please ensure that there's a bug on file. Do you plan to fix it?

Also, landing a test for this given it looks like there's none would be great.

Filed Bug 1496558. I could take a look at it. If it is pretty easy, I could fix it ASAP before landing this bug. Besides, I will add a test for getComputedStyle() to make sure we get the correct results if using keywords. Thanks for this review.

In D7536#173929, @boris wrote:

In D7536#173875, @emilio wrote:

• Pref that toggles between prefixed and unprefixed.
• Pref that hides the prefixed version.
• Pref that hides the unprefixed one.

I think xidorn suggests (2) and (3) together, and mats suggests (3).

Actually, I said in comment 14 that I'm fine with just shipping the unprefixed values,
without a pref or -moz- variants. Chrome supports the unprefixed values, so I think
the web-compat risk is low. Shipping -moz- aliases is fine too though and slightly
lower risk. Adding a pref for this seems overkill to me.

In D7536#193084, @mats wrote:

Actually, I said in comment 14 that I'm fine with just shipping the unprefixed values,
without a pref or -moz- variants. Chrome supports the unprefixed values, so I think
the web-compat risk is low. Shipping -moz- aliases is fine too though and slightly
lower risk. Adding a pref for this seems overkill to me.

So now what I can do is "try to make sure we are ready to ship". i.e. Add more tests in block size dimension and finish Bug 1496558. Therefore, we could ship this without any concern.

Only did rebase and updated wpt because some tests are unexpected pass after the rebase.

This version still use alias and doesn't use pref.

Code analysis found 4 defects in this patch:

• 4 defects found by clang-format

You can run this analysis locally with:

• ./mach clang-format -p path/to/file.cpp (C/C++)

For your convenience, here is a patch that fixes all the clang-format defects: https://queue.taskcluster.net/v1/task/TofS2t1URN6D6tsbDQShLQ/runs/0/artifacts/public/results/clang-format-PHID-DIFF-suqtfkikhoell4rkload.diff (use it in your repository with hg import or git apply)

If you see a problem in this automated review, please report it here: https://bit.ly/2IyNRy2

Code analysis found 4 defects in this patch:

• 4 defects found by clang-format

You can run this analysis locally with:

• ./mach clang-format -p path/to/file.cpp (C/C++)

For your convenience, here is a patch that fixes all the clang-format defects: https://queue.taskcluster.net/v1/task/YwengBkTTFKdQdJgOjGqKg/runs/0/artifacts/public/results/clang-format-PHID-DIFF-pnzusgbeutc5xz6eyzgj.diff (use it in your repository with hg import or git apply)

If you see a problem in this automated review, please report it here: https://bit.ly/2IyNRy2