Ben Reser wrote:
> On Mon, Jun 21, 2004 at 11:34:21AM +0200, Branko ??ibej wrote:
>> O.K:, thanks. I think we can require >= 1.3.21 after the 1.0.x
>> branch is dead (which hopefully means soon after 1.1 is out)
>
> I'd rather not. 1.3.19 is fairly common. Will be fairly common for a
> long time to come. As we're actually releasing software rather than
> just developing it, we need to have a darn good reason for increasing
> the requirements. Cleaning out cruft out of the build system that
> supports multiple versions, does not seem to be the right reason to do
> this. We can support 1.3.19 and on with a minor amount of effort. I
> don't see any reason not to continue to do so.
We're talking about Windows only here, where installations of SWIG 1.3.19
can hardly be considered common. The change doesn't affect any other
platforms.
And the point of the change wasn't just to "clean out cruft," it was to
replace a fragile method of generating the runtime that relied on
assumptions about internal swig files with a more robust method that used
swig's command line interface. I could have kept backwards compatibility
with SWIG 1.3.19 on Windows in addition to doing this, I just didn't know it
was important.
- Russ
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Jun 21 20:54:04 2004