regressions running Photoshop?

[Dan Kegel]

Hmm.  I just tried running Photoshop CS2 trial and Photoshop 5.5 trial,
and both failed on current wine.

CS2 complained "not enough DOS memory",
and 5.5 complained
lcms: Error #12288; Too many tags (2025813777)

PS6 works, though.

[Dmitry Timoshkov]

"Dan Kegel"  wrote:
> Hmm.  I just tried running Photoshop CS2 trial and Photoshop 5.5 trial,
> and both failed on current wine.
>
> CS2 complained "not enough DOS memory",
> and 5.5 complained
> lcms: Error #12288; Too many tags (2025813777)
>
> PS6 works, though.

This looks like an installer problem. Photoshop CS2 installed with rc1
works, the one installed with today's git doesn't.

[Dmitry Timoshkov]

"Dmitry Timoshkov"  wrote:
> "Dan Kegel"  wrote:
> > Hmm.  I just tried running Photoshop CS2 trial and Photoshop 5.5 trial,
> > and both failed on current wine.
> >
> > CS2 complained "not enough DOS memory",
> > and 5.5 complained
> > lcms: Error #12288; Too many tags (2025813777)
> >
> > PS6 works, though.
>
> This looks like an installer problem. Photoshop CS2 installed with rc1
> works, the one installed with today's git doesn't.

The culprit is:

4046075462c00f4479f185d1c0514584ff851223 is first bad commit
commit 4046075462c00f4479f185d1c0514584ff851223
Author: Andrew Talbot 
Date:   Tue May 13 22:41:58 2008 +0100

cabinet: Remove order-of-evaluation dependencies.

In particular the following change:

-        n -= (e = (e = ZIPWSIZE - ((d &= ZIPWSIZE-1) > w ? d : w)) > n ?n:e);
+        d = max(d & (ZIPWSIZE - 1), w);
+        e = min(ZIPWSIZE - d, n);
+        n -= e;

I'll send a patch.

[Dan Kegel]

On Thu, May 15, 2008 at 3:33 AM, Dmitry Timoshkov
 wrote:
> The culprit is:
>
> 4046075462c00f4479f185d1c0514584ff851223 is first bad commit
> commit 4046075462c00f4479f185d1c0514584ff851223
> Author: Andrew Talbot 
> Date:   Tue May 13 22:41:58 2008 +0100
>
> cabinet: Remove order-of-evaluation dependencies.
>
> In particular the following change:
> -
>    n -= (e = (e = ZIPWSIZE - ((d &= ZIPWSIZE-1) > w ? d : w)) > n
>
> ?n:e);
> -
>    d = max(d & (ZIPWSIZE - 1), w);
> -
>    e = min(ZIPWSIZE - d, n);
> -
>    n -= e;
>
> I'll send a patch.

Thanks!  I feel bad - Susan had identified that as the culprit for
a Dragon Naturally Speaking regression, and I emailed the author
rather than the list.  Might have saved you an hour if I had sent it
to the list.
- Dan