so I just tried a few experiments.  For some reason, none of the changes in
third_party\WebKit do an automatic try when uploading.
When I try to do gcl try with --root src\third_party, the change doesn't
apply on the server since it tries to patch a diff with
src\third_party\WebKit... in the filenames in the src\third_party directory.

I tried without --root, and it doesn't work because then the filenames are
under WebCore directory (see
http://codereview.chromium.org/download/issue223008_1.diff) don't have
src\third_party in them.

Something seems to be fishy in how the filenames are generates when in the
WebKit directory.  Marc-Antoine knows this best, so I'm handing it off.

On Wed, Sep 23, 2009 at 10:13 AM, Dimitri Glazkov <[email protected]>wrote:

> BTW, I trying to be snarky in my request. More like begging :)
>
> :DG<
>
> On Wed, Sep 23, 2009 at 10:08 AM, John Abd-El-Malek <[email protected]>
> wrote:
> > Marc-Antonie knows the try scripts best, so to avoid volunteering him
> > directly, I'll say that I'm sure he can answer any questions from whoever
> > does this.  I don't think it'll be much work.
> >
> > On Wed, Sep 23, 2009 at 10:04 AM, Dimitri Glazkov <[email protected]>
> > wrote:
> >>
> >> I think this is a great idea! Do we have a Python/gcl/rietveld expert
> >> who can tackle this?
> >>
> >> :DG<
> >>
> >> On Wed, Sep 23, 2009 at 9:49 AM, John Abd-El-Malek <[email protected]>
> >> wrote:
> >> > I didn't know this was possible.  It seems it will get a lot more
> usage
> >> > if
> >> > it "just works", i.e. the try script grabs these settings
> automatically
> >> > from
> >> > a codereview.settings file.  If we start by putting this file in
> >> > third_party\WebKit, then people who start with their patch there (also
> >> > to
> >> > upload) can use this transparently.  What do you think?
> >> > On Wed, Sep 23, 2009 at 7:42 AM, Marc-Antoine Ruel <
> [email protected]>
> >> > wrote:
> >> >>
> >> >> On Tue, Sep 22, 2009 at 9:35 PM, Adam Barth <[email protected]>
> >> >> wrote:
> >> >> > On Tue, Sep 22, 2009 at 6:25 PM, John Abd-El-Malek <
> [email protected]>
> >> >> > wrote:
> >> >> >> Is this even possible?  i.e. I had uploaded a WebKit patch on
> >> >> >> codereview but
> >> >> >> none of the patchsets got run on the try server
> >> >> >> http://codereview.chromium.org/178030/show
> >> >> >
> >> >> > It is possible:
> >> >> >
> >> >> > aba...@zenque:~/svn/kr/src/third_party/WebKit$ gcl try
> scriptcontext
> >> >> > --use_svn --svn_repo=svn://svn.chromium.org/chrome-try/try --bot
> >> >> > layout_win,layout_mac,layout_linux --root src/third_party
> >> >>
> >> >> I know the format is not very user friendly but it is definitely
> >> >> possible. You probably can skip --use_svn and --svn_repo or at least
> >> >> use the --http stuff instead, see go/chrometryserver.
> >> >>
> >> >> You could want to use both the layout tests and the normal tests so
> >> >> you can send the patch to all the slaves at the same time, e.g.
> >> >> --bot win,linux,mac,layout_win,layout_mac,layout_linux
> >> >>
> >> >> If the layout tests try slaves get overused, I'll add more slaves.
> >> >>
> >> >> M-A
> >> >
> >> >
> >
> >
>

--~--~---------~--~----~------------~-------~--~----~
Chromium Developers mailing list: [email protected] 
View archives, change email options, or unsubscribe: 
    http://groups.google.com/group/chromium-dev
-~----------~----~----~----~------~----~------~--~---

Reply via email to