Tom Lane wrote:
> Rob Butler <crodster2k@yahoo.com> writes:
> > That makes it sound as if you didn't do the same level
> > of testing on *this* release, like it didn't go
> > through all the tests or something.
>
> > How about "it does not have the extensive testing
> > history that other supported platforms in this release
> > have."
>
> Not bad, but it doesn't make the point that there's a lot of new
> platform-specific code for Windows in there. You want to point
> out not only that there's no history, but that there's new code to be
> suspicious of.
Agreed, but savepoints probably has more code. Do we warn about that
too?
I agree there is a disconnect between the initial clause and the testing
part, and mentioning the code actually does fix that, so new wording is:
Because Win32 is significantly different from the Unix platforms
supported in previous releases, there is much new Win32-specific
code that has not been tested extensively. Please test it
thoroughly before using it in production.
Also, we probably have more beta testers on Win32 than on all Unix
platforms combined, though they might not be testing as thoroughly as
Unix.
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073