Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(374)

Issue 9844015: Allow a commit message to be specified to merge-to-branch.sh (Closed)

Created:
8 years, 9 months ago by danno
Modified:
8 years, 8 months ago
CC:
v8-dev
Visibility:
Public.

Description

Allow a commit message to be specified to merge-to-branch.sh R=jkummerow@chromium.org Committed: https://code.google.com/p/v8/source/detail?r=11181

Patch Set 1 #

Patch Set 2 : Handle persisting empty variables correctly #

Total comments: 1
Unified diffs Side-by-side diffs Delta from patch set Stats (+21 lines, -8 lines) Patch
M tools/common-includes.sh View 1 1 chunk +9 lines, -2 lines 1 comment Download
M tools/merge-to-branch.sh View 5 chunks +12 lines, -6 lines 0 comments Download

Messages

Total messages: 6 (0 generated)
danno
8 years, 9 months ago (2012-03-28 07:02:41 UTC) #1
Jakob Kummerow
lgtm
8 years, 9 months ago (2012-03-28 07:44:04 UTC) #2
danno
please take another quick look
8 years, 9 months ago (2012-03-28 09:48:22 UTC) #3
Jakob Kummerow
Still LGTM. We'll have to extend it when someone wants to store "__EMPTY__" in a ...
8 years, 9 months ago (2012-03-28 09:54:05 UTC) #4
Sven Panne
https://chromiumcodereview.appspot.com/9844015/diff/3002/tools/common-includes.sh File tools/common-includes.sh (right): https://chromiumcodereview.appspot.com/9844015/diff/3002/tools/common-includes.sh#newcode90 tools/common-includes.sh:90: local VALUE="$(cat $FILE)" I'm just curious: Why is this ...
8 years, 9 months ago (2012-03-28 11:40:39 UTC) #5
Jakob Kummerow
8 years, 9 months ago (2012-03-28 11:43:49 UTC) #6
On 2012/03/28 11:40:39, Sven Panne wrote:
> I'm just curious: Why is this detour via __EMPTY__ necessary? The echo/cat
combo
> seems to do the right thing even for empty strings, but probably I'm missing
> something...

Persisting and restoring an empty variable would indeed work just fine; but the
check whether restoring was successful would fail. We could have removed the
check, but it has been useful to find errors in the past, so we decided to use a
sentinel value for "successfully restored but still empty" instead.

Powered by Google App Engine
This is Rietveld 408576698