codereview-discuss Google Group

Re: Why not store comments directly in VCS?

gu...@python.org (Guido van Rossum) — Thu, 02 Oct 2008 01:40:46 UT

The Mondrian/Perforce way starts with p4 annotate, which is similar to
svn annotate, to find the revision; then you ask Mondrian for the
review of that revision. It's not so easy to duplicate this behavior
in Rietveld because there is not a direct tie between the revision and
the issue#. This is more a limitation of svn than of Rietveld.

Re: Why not store comments directly in VCS?

will...@gmail.com (willdye) — Thu, 02 Oct 2008 01:27:00 UT

I think it would be useful to be able to start from a given line of
code in the source, and find any code review comments which applied to
nearby lines. Note this does *not* necessarily involve inserting the
review comments directly into the source, but it does require
something akin to backwards hyperlinking. Obviously it would be

Re: side by side diff unusable for already-committed changes?

stefan.hueh...@openbravo.com (stefan.huehner@openbravo.com) — Tue, 30 Sep 2008 12:47:05 UT

On Sep 11, 3:13 pm, Andi Albrecht <albrecht.a...@googlemail.com>
wrote:

Hi, i just found this thread after logging issue 53. For the new user
of upload.py the semantics of the --rev on subversion are unclear. I
tired to use --rev=1234 and expected to get the diff for commit 1234.
For this behaviour is have to specify --rev=1233:1234. Perhaps we

Re: nickname in reviewers list

albrecht.a...@googlemail.com (Andi Albrecht) — Fri, 26 Sep 2008 19:59:23 UT

Yes, it's broken ATM. The nicknames are lowercased before the
datastore query. So in your case it searches for johna instead JohnA.
I've uploaded a patch on codereview (don't have the issue number
now...)

nickname in reviewers list

mart...@danga.com (Evan Martin) — Fri, 26 Sep 2008 18:23:56 UT

I notice there was some recent changes with nicknames in the reviewer list.
I think it might be a bit broken. Here's what I did, on codereview.appspot.com:
1) Upload a new review ([link]), click the
"edit issue" link. Typed john.abdelma...@gmail.com as the reviewer.
2) Clicked "publish+mail comments". It shows JohnA (his nickname on

Re: getting unusable auth token from Rietveld

gu...@python.org (Guido van Rossum) — Sat, 20 Sep 2008 16:06:26 UT

Yeah, you now have two separate accounts tied to the same email -- one
"hosted account" (the one created via your company) and one "google
account" (the one you later created to sign up for the public group).
That's the maximum you can have with one email though, you won't be
asked to create more accounts. :-) We've run into this several times

Re: getting unusable auth token from Rietveld

matt.er...@loomia.com (Matt Ernst) — Sat, 20 Sep 2008 15:56:51 UT

The upload.py I had was using "GOOGLE," but when I forced it to
"HOSTED_OR_GOOGLE" I was able to upload issues again. Thanks! I'm not
sure why it stopped working in the first place though, as I've been
using the same upload.py without trouble before.

I do have two Google accounts for matt.er...@loomia.com. The first was

Re: getting unusable auth token from Rietveld

gu...@python.org (Guido van Rossum) — Sat, 20 Sep 2008 02:38:05 UT

Perhaps it's the upload.py version you're using? Some versions have
GOOGLE, others have HOSTED, yet others have HOSTED_OR_GOOGLE for their
auth settings. This has varied over time. And it sometimes matters,
depending on how the auth for the app is set up and what kind of
account you are using. Also beware of having two accounts for

getting unusable auth token from Rietveld

matt.er...@loomia.com (Matt Ernst) — Sat, 20 Sep 2008 02:23:14 UT

I recently updated a private instance of Rietveld running at
loomiareview.appspot.com to r323 from r306. After the update, I found
that I was unable to upload new issues. I successfully get an auth
token from the application in _GetAuthToken, which is always saved in
~/.codereview_upload_cookies, but it doesn't work for upload. In Send

Re: side by side diff unusable for already-committed changes?

matt.er...@loomia.com (Matt Ernst) — Fri, 12 Sep 2008 02:07:16 UT

On Sep 11, 6:13 am, Andi Albrecht <albrecht.a...@googlemail.com>
wrote:

Thanks a lot! I just tried it and it's great. This has really made my
workflow easier.

Matt

Re: Atom feeds generation

gu...@python.org (Guido van Rossum) — Thu, 11 Sep 2008 13:42:18 UT

I Kaelog,
I'm really sorry, I have gotten far behind on my code review duties
(and others, for sure). I will look at it again. I think it is
actually pretty close to being acceptable...

Atom feeds generation

kae...@gmail.com (kaelog) — Thu, 11 Sep 2008 13:22:16 UT

Hi,
if developped a little patch for codereview for generating Atom feed from
issues.
[link]
Last patchset has no comments.
Is there a possibility that modification be included into rietveld?
Regards
Kaelog

Re: side by side diff unusable for already-committed changes?

albrecht.a...@googlemail.com (Andi Albrecht) — Thu, 11 Sep 2008 13:13:47 UT

Here's a patch that adds this feature for Subversion to upload.py:
[link]

Re: git-cl -- git integration with rietveld

mart...@danga.com (Evan Martin) — Wed, 10 Sep 2008 21:42:19 UT

[link]
[link]
Committed:
[link]

Re: git-cl -- git integration with rietveld

mart...@danga.com (Evan Martin) — Wed, 10 Sep 2008 17:39:06 UT

Yeah! I've wanted this as well.
Rietveld uploader. I can try to clarify.
They should be -- could you elaborate? "git cl config" asks for a default CC.
Other have asked me this as well. It's again from the Rietveld
uploader. I could have it internally set the flag saying "don't ask
me about this". I have two minds on this: