On Tue, Jan 5, 2010 at 12:47 PM, Christian Boos <[email protected]> wrote:
> On 1/5/2010 4:31 PM, Roger Oberholtzer wrote:
>> On Tue, 2010-01-05 at 10:52 +0100, Christian Boos wrote:
>>> On 1/5/2010 9:17 AM, Roger Oberholtzer wrote:
>>>>
[...]
>>>>
>>> Well, you need to tell Trac how the svn URLs should be mapped to URLs in
>>> a Trac source browser (or any other source browser for that matter), as
>>> there's no way it could figure that information by itself.
>>>
>>> What's not clear in
>>> http://trac.edgewall.org/wiki/TracIni#svn:externals-section ?
>>>
>> Well, it never really worked as expected. I have tried again with:
>>
>> [svn:externals]
>> 1 = http://sto-opq-src.scc.se/svn http://www.opq.se/rsoft/browser/$path
>>
>> An example external is now listed as:
>>
>> PIAFmisc in http://sto-opq-src.scc.se/svn
>>
>> Which seems to be what it is. The URL generated is
>> http://www.opq.se/rsoft/browser/PIAFmisc
>>
>
> All those are private or protected URLs, so I can't test...
>

The same for /me ... that's why I didn't follow

> But that configuration says that you must have a PIAFmisc folder reachable
> at:
> http://sto-opq-src.scc.se/svn/PIAFmisc
>
> When you go on http://www.opq.se/rsoft/browser, are you actually browsing
> that repository?
>
> In which case you should see an PIAFmisc entry, and there's no reason the
> above link wouldn't work.
> Maybe that Trac is set up to browse another repository...
>

That's something I found a bit bizarre ... I could not see the repos
alias, so if it's external ... but I supposed Robert was talking about
another repos (with an alias ;o)

>> However, I get this error when I follow that link:
>>
>>        No node PIAFmisc at revision 3
>>
>> I have a number of externals, and all give the same complaint about
>> revision 3. I did have ?rev=$rev in the svn:externals, but that made no
>> difference.
>>
>> In fact, all this looks odd to me. The svn externals tell the location
>> of the svn repository. So I do not need to tell trac that. Given that it
>> knows where the stuff is in the svn repo, what am I really telling it
>> here? Surely not where to fine something - it already knows that.
>
[...]
>
> I believe your misunderstanding comes from the fact that you assume that a
> Trac repository browser can work as soon as it is given the URL to a
> repository. But that's not the case, we don't access the svn repositories
> via URLs (or more technically, using the svn_ra layer), but directly via the
> filesystem (using svn_repos and svn_fs layers). See #493 for understanding
> why.
>

I suppose that if the repos is managed by Hg or another DVCS (once
more impls will be ready and available ;o) then the right way to
«connect» that repo should be to schedule an e.g. `hg clone` and use
the local copy, isn't it ?

> So you need to map a given svn URL to a Trac (or another source browser)
> that is configured to browse /that/ repository. In Trac, the configuration
> is the directory containing the svn repository, and from that, there's no
> way to infer the URL(s) by which this repository is reachable (or
> conversely, from an URL, we can't know the location of the repository in the
> filesystem). That's why we need this mapping.
>

;o) ... The WebDAV | HTTP access to the SVN is managed by the web
server, so Trac cannot know the details (unless someone mention how to
map  ;o)

-- 
Regards,

Olemis.

Blog ES: http://simelo-es.blogspot.com/
Blog EN: http://simelo-en.blogspot.com/

Featured article:
Better ignore for `dist` and `build` ... bah!
http://simelo.hg.sourceforge.net/hgweb/simelo/trac-gviz/rev/dd48a5c68159?revcount=100
-- 
You received this message because you are subscribed to the Google Groups "Trac 
Users" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.


Reply via email to