Hi Rick, per the instructions in https://docs.google.com/document/d/10S8ESUvwhEOOBEKr-hn97y8eRTYczavizsUNv5Gvcg8/edit?tab=t.0, would you be willing to review this?
The guidance in that doc says "Ask //OWNERS for review.", but the only actual owners listed there are file://ATL_OWNERS, where there is a comment saying "Note: this list is not for rubber-stamping mechanical changes that span the code base. Please reach out to owners of top-level directories instead.":
https://source.chromium.org/chromium/chromium/src/+/main:OWNERS;l=4-5;drc=677b9a0b2118af9dda2a25e1786e272ef9a029d5
If that comment is indeed correct when it comes to things that would be considered LSCs, should we update the LSC instructions to say that for LSCs where only 1 - 3 CLs are needed, the CLs should be split by top-level owner and then reviewed accordingly?
Note that in gerrit this change seems relatively small, but the presubmit checks encouraged using the LSC process for this:
```
This change contains 169 files.
Consider using the LSC (large scale change) process.
See https://chromium.googlesource.com/chromium/src/+/HEAD/docs/process/lsc/lsc_workflow.md.
```
For this change, I can abandon this CL and split it up instead if you'd prefer.
Thank you!
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
| Owners-Override | +1 |
LGTM
Entirely mechanical and I can't see any likely downside. So OO+1
Hi Rick, per the instructions in https://docs.google.com/document/d/10S8ESUvwhEOOBEKr-hn97y8eRTYczavizsUNv5Gvcg8/edit?tab=t.0, would you be willing to review this?
The guidance in that doc says "Ask //OWNERS for review.", but the only actual owners listed there are file://ATL_OWNERS, where there is a comment saying "Note: this list is not for rubber-stamping mechanical changes that span the code base. Please reach out to owners of top-level directories instead.":
https://source.chromium.org/chromium/chromium/src/+/main:OWNERS;l=4-5;drc=677b9a0b2118af9dda2a25e1786e272ef9a029d5If that comment is indeed correct when it comes to things that would be considered LSCs, should we update the LSC instructions to say that for LSCs where only 1 - 3 CLs are needed, the CLs should be split by top-level owner and then reviewed accordingly?
Note that in gerrit this change seems relatively small, but the presubmit checks encouraged using the LSC process for this:
```
This change contains 169 files.
Consider using the LSC (large scale change) process.
See https://chromium.googlesource.com/chromium/src/+/HEAD/docs/process/lsc/lsc_workflow.md.
```For this change, I can abandon this CL and split it up instead if you'd prefer.
Thank you!
Sorry for the contradiction. In this case I don't think the CL is worth splitting up - only took me a couple minutes to review in its entirety. I think the presubmit check is a rough heuristic of complexity that was just very wrong in this case (since most of the changes are renames). I think we will be improving how we handle LSCs, so I'll add this to a list of problems to fix with an overhaul.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Rick ByersHi Rick, per the instructions in https://docs.google.com/document/d/10S8ESUvwhEOOBEKr-hn97y8eRTYczavizsUNv5Gvcg8/edit?tab=t.0, would you be willing to review this?
The guidance in that doc says "Ask //OWNERS for review.", but the only actual owners listed there are file://ATL_OWNERS, where there is a comment saying "Note: this list is not for rubber-stamping mechanical changes that span the code base. Please reach out to owners of top-level directories instead.":
https://source.chromium.org/chromium/chromium/src/+/main:OWNERS;l=4-5;drc=677b9a0b2118af9dda2a25e1786e272ef9a029d5If that comment is indeed correct when it comes to things that would be considered LSCs, should we update the LSC instructions to say that for LSCs where only 1 - 3 CLs are needed, the CLs should be split by top-level owner and then reviewed accordingly?
Note that in gerrit this change seems relatively small, but the presubmit checks encouraged using the LSC process for this:
```
This change contains 169 files.
Consider using the LSC (large scale change) process.
See https://chromium.googlesource.com/chromium/src/+/HEAD/docs/process/lsc/lsc_workflow.md.
```For this change, I can abandon this CL and split it up instead if you'd prefer.
Thank you!
Sorry for the contradiction. In this case I don't think the CL is worth splitting up - only took me a couple minutes to review in its entirety. I think the presubmit check is a rough heuristic of complexity that was just very wrong in this case (since most of the changes are renames). I think we will be improving how we handle LSCs, so I'll add this to a list of problems to fix with an overhaul.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
3 is the latest approved patch-set.
No files were changed between the latest approved patch-set and the submitted one.
Rename _unittests.cc and _browsertests.cc files
Rename tests currently using _unittests.cc and _browsertests.cc names to
remove the 's', per:
https://chromium.googlesource.com/chromium/src/+/main/styleguide/c++/c++.md#tests-and-test_only-code
This makes it easy to filter out tests when using Codesearch
(-path:test.cc) and grep, and also at least one existing presubmit test
already makes assumptions about this convention:
https://source.chromium.org/chromium/chromium/src/+/main:PRESUBMIT.py;l=7980;drc=4926c951de95096e23a4c5867c4da5df1efcca7c
This CL renames existing test files, and another CL will add a presubmit
check that warns when new files mistakenly use the _unittests.cc or
_browsertests.cc suffix:
https://chromium-review.googlesource.com/c/chromium/src/+/7509278
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |