Quoting Andrey Repin <anrdaemon_at_freemail.ru>:
> Greetings, jhanley_at_dgtlrift.com!
>
>>> For different copies and purposes, may be. But still, I would stick to one
>>> way or another, reusing SVN auth cache as much as possible without
>>> unnecessary
>>> bloating.
>
>> It's not bloating, its implementing section 3.2.1 of RFC 3986
>
> You messing things.
> Bloating in my sentence was in reference to the way SVN storing auth cache.
> For svn://host and https://host it'll be two different entries in database.
Ahh, Yes. I see and agree - this example would clutter up the cred db,
but I believe the original author was just using the two URLs to the
same server for illustration purposes tto show that there is a
difference in operation. I don't picture individual users accessing
the same server by different protocol means.
You bring up a point & question that is beyond the scope of this
thread - would every protocol to the same resource generate additional
entries to the db? ie http, https, svn, svn+ssh, file, future
protocols? Additionally, what about hostname vs fqdn? ie would
http://oursvnserver/ vs http://oursvnserver.ourcompany.com/ generate
multiple entries?
-Jim
Received on 2009-04-08 15:28:38 CEST