[AFS3-std] Using Gerrit for draft review

[David Boyes]

(Late response -- it's been a busy month)

We tried this over in the tn3270e working group, and it didn't work so
well. The comment chains became very unwieldy very quickly -- especially
if the points that needed discussion were complex. We eventually dropped
back to mail to resolve the complex issues, and pretty much determined
that gerrit was a fine tool for code and structured stuff, but didn't
really lend itself well to human-oriented texts.

So, just a field experience with it, FWIW. Interesting experiment, but I
don't think it'll really help. We'd need a serious collaborative writing
tool with versioning to start doing joint composition or editing tasks.
Framemaker has some of what we'd need, but it's bloody expensive and a
PITA to learn how to use. Not probably worth it.

Let's stick to plain text. Easier for everyone.

--d b


On 12/7/10 2:14 PM, "Jeffrey Altman" <jaltman@secure-endpoints.com> wrote:

>Given the recent experience of OpenAFS using gerrit for code review I
>can't help but think that gerrit would be an excellent tool for the use
>of I-D review.
>
>Verified would be passing idnits without error or indicating an
>implementation generated from the proposed standard.  Code Review would
>be used to indicate up or down votes in favor of the current version.
>
>Most importantly, gerrit would permit discussion over specific portions
>of the text to be held within the body of the text itself.
>
>Is this something the group would like to try?
>
>Jeffrey Altman
>