On Dec 13, 2004, at 9:34 AM, Scott Palmer wrote:
>
> On Dec 13, 2004, at 8:52 AM, Ben Collins-Sussman wrote:
>
>>
>> On Dec 13, 2004, at 6:34 AM, Henrik Vendelbo wrote:
>>
>>> WebDAV is actually considered the preferred access method if previous
>>> mailing list comments are to be belived.
>>
>> I sure hope that's not the impression we give. I think just as many
>> people use svn:// or svn+ssh:// as use http://. The book even goes
>> to great length to stress that no access method is "preferred" over
>> another.
>
> From reading this list I get the opposite impression. Save yourself
> some grief with the complex and error-prone setup of Apache (and BDB,
> and unix permission ugliness) and run a simple svnserve with a FSFS
> back end. (If the simpler security scheme is acceptable to you of
> course.) With that setup and using only the svn: access method, NOT
> svn+ssh: (set up a ssh tunnel if you need to), the chances of screwing
> up seem to go down a great deal.
Sure, if the criteria is "what's the simplest possible way to access
Subversion remotely?", then you'll almost always get the same answer:
'svnserve -d'. But there are many other possible criteria for
choosing an access method. That's what the table at the beginning of
chapter 6 is all about.
My point is that neither svnserve nor apache is a second-class citizen,
nor is one server is "more official" than another, or "more preferred".
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Mon Dec 13 19:22:29 2004