Alan Burlison wrote:
> Alan Coopersmith wrote:
> 
>>> However, this one doesn't work as good as a real hg web browser. And you 
>>> can not get all log change history like hg web browser does.
> 
> You can, OpenGrok cprovides history and inter-revision diffs.
> 
>> This is probably a topic for website-discuss - I agree it would be nice to
>> offer the hg web browser for viewing changesets/history in addition to the
>> OpenGrok search tool - I don't know where that lies in the website team's
>> very long to-do list though.
> 
> I assume that 'hg web browser' refers to 'hg serve'.  'hg serve' is not 
> suitable for production use, and is infeasible bearing in mind we have 
> 255 Mercurial repositories at present.  I'm afraid we therefore won't be 
> providing it - OpenGrok is the way we are providing this functionality.

Please explain why http it isn't suitable for production use ?

The functionality it provides is completely different to OpenGrok. 
OpenGrok is cscope for the web with some SCM knowledge.  The web based 
interface to a Mercurial repo is invaluable for researching exactly when 
changes happened and for assisting merging of project gates with the 
upstream repo.

They just don't provide the same functionality, it isn't in my opinion 
an either or but we need both.

I agree that the built in 'hg serv' is not what should be used but even 
the official Mercurial wiki says that:

http://www.selenic.com/mercurial/wiki/index.cgi/PublishingRepositories

Why can't we deploy using hgwebdir.cgi ?

It could be an optin mechanism controlled from the SCM console, since 
ideally I'd suspect that projects would only want their "master" gate 
published this way rather than all repos.

Pushes over http could be (probably should be) disabled.


-- 
Darren J Moffat
_______________________________________________
opensolaris-code mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/opensolaris-code

Reply via email to