[Bug 22811] New: graphical rendering regressions on cad programs (most likely since 1.1.44)

http://bugs.winehq.org/show_bug.cgi?id=22811

Summary: graphical rendering regressions on cad programs (most
                    likely since 1.1.44)
           Product: Wine
           Version: 1.2-rc1
          Platform: x86
        OS/Version: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: -unknown
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

Created an attachment (id=28168)
 --> (http://bugs.winehq.org/attachment.cgi?id=28168)
screenshot illustrating the rendering problem.

Please do not ship 1.2 before this regression is fixed.

Attached screenshot illustrates the issue. In cad program, text labels appear
in the wrong place (note the "vss") label that overlaps other stuff, when it
should be going downwards, symmetrically to the "vdd" one.

Historically, this bug has come out every now and then, apparently being
re-inserted and re-fixed multiple times.

Unfortunately, it makes impossible to use CAD applications for anything
serious, since you cannot deliver mangled drawings.

A way to test is to use the LTspice scad3.exe program (that can be downloaded
for free). A link for the download is 
http://www.linear.com/designtools/software/switchercad.jsp.

Please test by opening the attached schematics.

In case a non-wine ddl can temporarily fix the issue, please be so kind to
report it.

[Bug 22811] graphical rendering regressions on cad programs (most likely since 1.1.44)

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #1 from Sergio Callegari   2010-05-22 03:17:34 ---
Created an attachment (id=28170)
 --> (http://bugs.winehq.org/attachment.cgi?id=28170)
File to open with LTSpice to see the problem

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Dmitry Timoshkov  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |regression
            Summary|graphical rendering         |Graphical rendering problem
                   |regressions on cad programs |in cad program
                   |(most likely since 1.1.44)  |

--- Comment #2 from Dmitry Timoshkov   2010-05-22 03:39:38 ---
Please post the result of your regression testing:
http://wiki.winehq.org/RegressionTesting

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |download
                URL|                            |http://www.linear.com/desig
                   |                            |ntools/software/switchercad
                   |                            |.jsp

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Vitaliy Margolen  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |trivial

--- Comment #3 from Vitaliy Margolen   2010-05-22 19:58:37 ---
http://bugs.winehq.org/page.cgi?id=fields.html#bug_severity

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #4 from Sergio Callegari   2010-05-23 16:57:24 ---
For regression testing... unfortunately older package archives for my distro
only arrive at 1.1.38, which is fine.  I would tend to say that the bug got
introduced in 1.1.44 or 1.1.43.

For severity, sorry for the mistake... but I believed that for the particular
case of a CAD program, not to be able to draw properly was to be intended as
loss of functionality (normal severity). I asked whether the drawing was done
by some dll, just to see if some other dll could have been a temporary
workaround (and downgrade the bug to minor).

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #5 from Dmitry Timoshkov   2010-05-24 00:08:10 ---
For the regression testing you need the Wine source, not the precompiled
package.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Roderick Colenbrander  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #6 from Roderick Colenbrander   2010-05-27 12:38:30 ---
I have used LTSpice on Windows frequently as well and I remember seeing such
issues there as well so now and then. I quickly installed LTSpice here and the
app is rendered properly even using the schematic you showed. I must say that
the font which the app selected here is different from the one in your
screenshot. I really think it is just a font issue.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #7 from Roderick Colenbrander   2010-05-27 13:38:55 ---
Created an attachment (id=28326)
 --> (http://bugs.winehq.org/attachment.cgi?id=28326)
screenshot

Just a screenshot to show that it runs fine. I don't think it is a Wine bug at
all.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Roderick Colenbrander  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|regression                  |

--- Comment #8 from Roderick Colenbrander   2010-05-27 13:41:00 ---
There is no proof this is an actual regression, so removing the regression
keyword.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #9 from Sergio Callegari   2010-05-28 03:56:26 ---
Hi,

Roderick, can you please check you attached the right screenshot?
Your http://bugs2.winehq.org/attachment.cgi?id=28326 shows exactly the same
problem as mine.

The "Vss" text should go *below* the blue wire at the bottom of the operational
amplifiers (the triangular shapes in the diagram), not *above*.

In other terms, the Vss text should be rendered *to the left* of the alignment
point (i.e. the bottom extreme point of the blue wire under the triangular
shape), *not to the right* of the alignment point (please, read "to the left"
as "to the bottom" and "to the right" as "to the top" if you prefer, since the
text is vertical).

In your attachment, the problem is only mitigated by the fact that by using a
smaller font the bit of text that gets the wrong alignment does not happen to
overlap with other text, but it is still misplaced taking its reference point
on the wrong side.

Hence, if your attachment is the correct one, please have it interpreted as a
confirmation of the bug.

Otherwise, if by mistake you attached a wrong screeenshot, can you be so kind
to upload the right one and maybe specify your fonts setting, so I can try to
reproduce here?

--------

For what concerns the regression status, I can assure that it is a regression,
since former wine versions were definitely ok (with the exception of a few of
them, since this rendering issues has been several time re-inserted/fixed as
others have confirmed).  If I remember properly 1.41 was OK.
I am sorry that I cannot compile wine on the machine I am on, to fully bisect.
If someone can provide a 1.1.41 or 1.1.42 64bit deb, I would really be glad to
help test.

If fixing this is a relatively easy matter of reverting/fixing some recent
patch, it would really be great if it could be done as long as wine is RC,
rather than putting in 1.2 a bug that was not there until a month ago in the
development version. But of course I cannot judge the difficulty of addressing
the issue, nor the current priority, hence in any case thanks for the great
work. I can always keep a 1.1.38 around.

BTW, is it possible to have 2 wine versions installed at the same time (so I
can keep the one doing CAD rendering right and still test the 1.3.x branch when
it is started?)

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #10 from [e-mail]  2010-05-30 01:01:50 ---
Created an attachment (id=28408)
 --> (http://bugs.winehq.org/attachment.cgi?id=28408)
Confirmation of Sergio's bug

In the center, the lower part of the triangle, "VSS" isn't displayed as a
continuation of the line downwards, as it should be.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

[e-mail] changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #11 from [e-mail]  2010-05-30 01:08:05 ---
Hello

I can confirm this behavior since 1.1.44 until now - rc2: please see the
screenshot "Confirmation of Sergio's bug". You can see that I use fixed font
and it renders with the same wrong allignment. Version 1.1.43 is working.

As a side note, there are a few minor glitches, as well, like font antialiasing
comes by default (since rc1) and can't be turned off with winetricks (my
preference, nothing else), or editing a text/spice directive (right-click on
the text portion) shows the dialog which, if resized, doesn't get the text area
larger, only the outer geometry, and a few others, but they are minor and "if
they work, they don't need fixing". At least, until someone decides they are
worthy.

Regards,
Vlad.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #12 from [e-mail]  2010-05-30 01:16:37 ---
Created an attachment (id=28409)
 --> (http://bugs.winehq.org/attachment.cgi?id=28409)
Confirmation of Sergio's bug 2

This one was made with 1.1.43. It shows proper rendering of "VSS" node, as well
as antialiasing turned off -- rc1&2 make turning it off impossible(?)

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #13 from [e-mail]  2010-05-30 01:26:51 ---
My apologies for the mistakes(?), I really don't know "bug world". Here are the
complete names of the files used:

wine1.2_1.1.43-0ubuntu1~lucidppa4_amd64.deb   <--- last working
wine1.2_1.2~rc2-0ubuntu1~lucidppa1_amd64.deb  <--- last one in repositories

Regards,
Vlad.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

joaopa  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #14 from joaopa   2010-05-30 01:31:34 ---
Telling "bug is between wine1.1.43 and wine 1.2.rc2" is quite useless. You must
give the commit which causes the regression.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Wylda  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #15 from Wylda   2010-05-30 02:51:38 ---
(In reply to comment #14)
> Telling "bug is between wine1.1.43 and wine 1.2.rc2" is quite useless.

It's not that bad ;) It has all the needed for me... Download, example file and
we know good and bad versions. It's enough to process the bug.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #16 from Wylda   2010-05-30 03:17:12 ---

Easy... The hardest part is usually to understand whats the problem. So those
who are new here, look how to fill correctly a bug report. It's easy, fun and
takes just a few minutes even with filling all the details in the report.

1. Confirming, please consider UNCONFIRMED->NEW and KEYWORDS: +REGRESSION

2. I did a regression test between 1.1.43 and 1.1.44:

commit d92ed5bd079fc83ba30e17bb9776ad8db8c316a7
Author: Huw Davies 
Date:   Thu May 6 14:05:39 2010 +0100

gdi32: Add support for ETO_PDY and improve world transform support.

:040000 040000 c309831774078d979c1e4a80f351eab104cdd549
cb4d53abec93b81920b9b35e64eb28941d134059 M      dlls

3. No other bug report suffers from this commit.

4. Revert of this patch on top of wine-1.2-rc2 makes that problem go away.

5. Adding author of this patch to CC.

--private keyword: bisected

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Wylda  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Jeff Zaroyko  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |regression
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1

--- Comment #17 from Jeff Zaroyko   2010-05-30 03:27:06 ---
Confirming as per comment 16.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #18 from [e-mail]  2010-05-31 23:39:09 ---
(In reply to comment #14)
> Telling "bug is between wine1.1.43 and wine 1.2.rc2" is quite useless. You must
> give the commit which causes the regression.

I did say this:
> > > I can confirm this behavior since 1.1.44 until now - rc2 (...)
> > > Version 1.1.43 is working.

The above lines tell that version 1.1.43 is the last working one, as pointed
out in the comment of the second screenshot, too:
> > > This one was made with 1.1.43

I think I made it clear from what version the "glitches" (if "bug" is not
proper) appeared. Also, Wylda, you may know by heart what
> > > gdi32: Add support for ETO_PDY and improve world transform support.
> > > :040000 040000 c309831774078d979c1e4a80f351eab104cdd549
> > > cb4d53abec93b81920b9b35e64eb28941d134059 M      dlls

means, but to me it's a mystery, and I don't intend to change it, it makes life
prettier.
In short, all I did was to confirm Sergio's words and point out where the
change occurred. From this point on, entomologists should know what to do...

Regards,
Vlad.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #19 from Wylda   2010-06-01 01:51:05 ---
(In reply to comment #18)
> ...(if "bug" is not proper)

Yes, it's bug, it's even confirmed (Status: NEW) so this comment is redundant.
> Also, Wylda, you may know by heart what .... means, but to me...

Sleep well, i don't have a slightest idea as well. But if you want to help come
one in a while and give your apps a re-test under latest wine. Thus it won't be
forgotten in so many bugs and hopefully it will work as a reminder for those
who understand ;)

So no need to post here, if you won't test _new_ wine version (rc3 or latter ~ 
otherwise "redundant" comment as above).

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #20 from [e-mail]  2010-06-02 03:37:07 ---
 I suppose one could twist the meanings of the words and say that if I don't
debug, I don't help, or that I'm ignorant to the problems that arise from this,
etc. Wrong. There are many ways to help. Plus, I don't intend to be a debugger
(even if I respect those who do it because it is my impression that it's harder
to debug than to actually program) the same way I don't intend to be a priest.
I prefer light wear...
 I only decided to post aftwer I saw Roderick Colenbrander's comment, to help
removing the doubts concerning this problem, that's all. Incidentally, this is
the first time I participated in a bug discovery; there are others who are
knowing and willing to do so and do it far better than I could. The same way I
waited for other programs to be cleared, the same way I could have waited now,
if it wasn't for Roderick's post. Since I waited for a few versions for the bug
to be discovered, I could have helped debugging, if that was my purpose (the
fact that I use the unstable version doesn't make me a tester).
 On a more practical note, being an Ubuntu user, the repository that I use is
this one:
deb http://ppa.launchpad.net/ubuntu-wine/ppa/ubuntu lucid main
and, so far, it only has rc2 (I can't see rc3 in winehq downloads section,
either).
 These said, I wish good luck to everyone involved in these matters. It's a
demanding job, no less than any other involving creativity, such as this.

Regards,
Vlad.

PS: If characters on the screen could be heard, the tone of voice used would
be the most neutral one, no impurities, for all my posts.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #21 from [e-mail]  2010-06-12 00:38:22 ---
Today's updates included rc3, but the problem persists. Font antialiasing,
dialog resizing, too.

Regards,
Vlad.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #22 from Sergio Callegari   2010-06-12 03:56:56 ---
I confirm that RC3 still carries the bug.

It is unfortunate that this bug is not in the list scheduled for fixing before
1.2 goes gold, namely

http://bugs.winehq.org/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&am...

Of course it looks unfortunate because the specific commit that introduced the
regression is now known and reverting the commit would probably be easy.

Please, can the bug be added to the list?

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Wylda  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|-unknown                    |gdi32
            Version|1.2-rc1                     |1.1.44

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Juan Lang  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |1.2.0

--- Comment #23 from Juan Lang   2010-06-15 18:05:09 ---
Can't hurt to ask.  The commit doesn't have tests or reference another bug, so
while I expect it was done for a specific app, we don't know which it is.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Alexandre Julliard  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

--- Comment #24 from Alexandre Julliard   2010-06-17 11:19:17 ---
Fixed by 69d5a56103bf35f955fc4b76a5a9e0ccd8e07f38.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #25 from Wylda   2010-06-17 11:38:49 ---

Fixed indeed in wine-1.2-rc3-106-gae942d6. Thank you!

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Alexandre Julliard  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #26 from Alexandre Julliard   2010-06-18 12:47:38 ---
Closing bugs fixed in 1.2-rc4.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

--- Comment #27 from [e-mail]  2010-06-23 03:00:46 ---
Today's updates included rc4, with a nice surprise. Thank you.

[Bug 22811] Graphical rendering problem in cad program

http://bugs.winehq.org/show_bug.cgi?id=22811

Wylda  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
      Fixed by SHA1|                            |69d5a56103bf35f955fc4b76a5a
                   |                            |9e0ccd8e07f38
    Regression SHA1|                            |d92ed5bd079fc83ba30e17bb977
                   |                            |6ad8db8c316a7