RESOLVED FIXED31904
[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904
Summary [bzt] Kill WebKitLandingScripts
Adam Barth
Reported 2009-11-25 23:51:23 PST
It is an evil static of doom that prevents us from testing download commands.
Attachments
step 1 (8.77 KB, patch)
2009-11-26 00:41 PST, Adam Barth
no flags
step 2 (4.59 KB, patch)
2009-11-26 00:41 PST, Adam Barth
no flags
step 3 (2.79 KB, patch)
2009-11-26 00:41 PST, Adam Barth
no flags
Step 4 (4.52 KB, patch)
2009-11-26 00:41 PST, Adam Barth
no flags
Step 5 (8.24 KB, patch)
2009-11-26 00:41 PST, Adam Barth
no flags
Step 6 (19.82 KB, patch)
2009-11-26 00:41 PST, Adam Barth
eric: review+
eric: commit-queue-
Adam Barth
Comment 1 2009-11-26 00:41:20 PST
Adam Barth
Comment 2 2009-11-26 00:41:27 PST
Adam Barth
Comment 3 2009-11-26 00:41:31 PST
Adam Barth
Comment 4 2009-11-26 00:41:34 PST
Adam Barth
Comment 5 2009-11-26 00:41:38 PST
Adam Barth
Comment 6 2009-11-26 00:41:42 PST
Eric Seidel (no email)
Comment 7 2009-11-26 20:16:52 PST
Comment on attachment 43901 [details] step 1 Why pass tool? 119 def setup_for_patch_apply(tool, options): Seems scm is cleaner there. This looks good though.
Eric Seidel (no email)
Comment 8 2009-11-26 20:17:47 PST
Comment on attachment 43902 [details] step 2 Calling error() is evil and wrong. We need to fix that at some point.
Eric Seidel (no email)
Comment 9 2009-11-26 20:18:19 PST
Comment on attachment 43903 [details] step 3 LGTM.
Eric Seidel (no email)
Comment 10 2009-11-26 20:25:38 PST
Comment on attachment 43904 [details] Step 4 Commands already have a self.tool now. No need to pass tool: self._create_changelogs_for_revert(tool.scm(), revision) 398 self._create_changelogs_for_revert(tool, revision) If you're going to move away from scm(), we should move to self.tool
Eric Seidel (no email)
Comment 11 2009-11-26 20:27:13 PST
Comment on attachment 43905 [details] Step 5 LGTM. Eventually we need to unify SCM.run_command with this too.
Eric Seidel (no email)
Comment 12 2009-11-26 20:34:16 PST
Comment on attachment 43906 [details] Step 6 I intentionally had this take an scm param: commit_message = commit_message_for_this_commit(scm) 135 commit_message = scm.commit_message_for_this_commit() It doesn't belong on SCM. Our commit message junk doesn't belong in scm.py. 38 # FIXME: The options should really live on each "Step" object. I LOVE the idea of a Step object, and even more that options would be automatically assembled from the set of step objects! rollout used to work just fine: comment_text = WebKitLandingScripts.build_and_commit(tool.scm(), options) 407 # FIXME: This function does not exist!! 408 # comment_text = WebKitLandingScripts.build_and_commit(tool.scm(), options) 409 raise ScriptError("OOPS! This option is not implemented (yet).") Not sure when it broke. 406410 self._reopen_bug_after_rollout(tool, bug_id, comment_text) is broken in the current bugzilla however, because there is no longer a "REOPENED" state in the popup. I guess we could add a big fixme next to commit_message_for_this_commit. It doesn't belong in scm.py. :( I'd like you to at least add a FIXME.
Adam Barth
Comment 13 2009-11-26 23:55:04 PST
Comment on attachment 43901 [details] step 1 Clearing flags on attachment: 43901 Committed r51431: <http://trac.webkit.org/changeset/51431>
Adam Barth
Comment 14 2009-11-26 23:55:38 PST
Comment on attachment 43902 [details] step 2 Clearing flags on attachment: 43902 Committed r51432: <http://trac.webkit.org/changeset/51432>
Adam Barth
Comment 15 2009-11-26 23:56:05 PST
Comment on attachment 43903 [details] step 3 Clearing flags on attachment: 43903 Committed r51433: <http://trac.webkit.org/changeset/51433>
Adam Barth
Comment 16 2009-11-26 23:56:31 PST
Comment on attachment 43904 [details] Step 4 Clearing flags on attachment: 43904 Committed r51434: <http://trac.webkit.org/changeset/51434>
Adam Barth
Comment 17 2009-11-26 23:56:59 PST
Comment on attachment 43905 [details] Step 5 Clearing flags on attachment: 43905 Committed r51435: <http://trac.webkit.org/changeset/51435>
Adam Barth
Comment 18 2009-11-27 00:04:48 PST
Note You need to log in before you can comment on or make changes to this bug.