Mark Phippard wrote:
> cmpilato@localhost.localdomain wrote on 10/28/2004 11:08:37 PM:
>
>> Greg Hudson <ghudson@MIT.EDU> writes:
>>
>>> Here are the arguments I know of for putting the lock table in the FS
>>> back end:
>>
>> You forgot one:
>>
>> * If the community can't agree about whether to use Berkeley DB or
>> some flat-file system to implement the lock table.
>
> That was the one point I wanted to raise as well. We are nearing
> completion on our port of Subversion to OS/400. In all likelihood, we
> could not have done the port without the fsfs backend. So I am hoping and
> praying that the locking implementation will not require BDB, at least
> when using a fsfs backend.
I think that is guaranteed. Having provided the feature of a BDB-less server
in 1.1, compatibility rules forbid removing that feature again for any 1.x
release - and common sense suggests that it would be foolish to *ever*
remove that feature.
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Oct 29 15:12:23 2004