Steve Dwire wrote:
>What do people think? Is a source code patch an acceptable workaround
>for the 1.0 release to end users? Or should the patch become part of
>the 1.0 product, in spite of the ugly #ifdef and the
>duct-tape-and-baling-wire solution it contains?
>
>Personally, I'd prefer to see the official 1.0 release do "the best it
>can" with this issue, rather than sticking the poor end users with an
>"Oh, you can patch your source and rebuild the product" workaround. Is
>that likely without an official issue filed?
>
>
I'm seeing the problem hit by someone (on a team of a half dozen
developers) maybe every couple of days. The biggest problem is that if
it happens in a bit of the tree where you've actually done some work,
the simplest "trash the whole directory and re-checkout" workaround can
lose you changes and you actually need to be more careful. I'm not sure
if this is the "correct" workaround either.
If the source patch reduces the occurrences that would be great. To my
mind, I'd not be able to advocate using Subversion on Windows unless
this issue were fixed. Even if there's a fairly simple workaround, the
"checksums mismatch" that is displayed after the update failure is
really scary looking and would seriously reduce peoples confidence in
the system.
Best regards,
Mike.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Oct 20 22:04:35 2003