CS4 bug: brush cursor display is incomplete

[_Tony_Tx_]

Just installed CS4 under WinXP, 4G RAM.

I've set my brush cursors to show full size. The cursor displays correctly for brush sizes up to about 150 px. Beyond 150 px, the circle representing the brush size gets truncated: at 170 px, it's down to a half-circle, at 200 px it's down to one-third of a circle, and at 300 px all that remains is a small curved line segment.

[_Derrick_Moore_]

Yes, adding the AllowOldGPUs_ON.reg registry key, I get complete cursors on
both monitors.

[_Derrick_Moore_]

Adam, FYI
On my home machine I have Vista64, OpenGL on, no AllowOldGPUs_ON reg edit. I
get complete cursors, image display and smooth zooming on both monitors. It
works. Also, on the XP machine the image title flickers during zooming, it
does not in Vista. Both machines use the same card, same driver version.

[Guest]

Statusupdate: Using "hacked" drivers from http://www.laptopvideo2go.com/ for my 8400M GS now and everything works like it should.

I still strongly believe that Adobe should do something for those unlucky users having a Mobile NVidia card not being able to get updated drivers from their manufacturers. I don't believe that using these "hacked" drivers is the way to go for everyone (especially for non tech-savvy).

[_William_R__Miller_]

I also had the same cursor truncation issue. I finally just got around to looking up this posting. The registery setting change that was provided worked for me.

I do agree that this issue NOT being fixed pretty much renders photoshop useless. 😞 I'm realy glad the registery setting was posted and worked.

Good luck to all of those stil experiencing issues!

[Chris Cox]

Cursor truncation is an NVidia driver bug. If it is not fixed in the latest drivers for your card, please contact NVidia.

[Mark Clarkson]

NVIDIA did finally release a driver for the FX 3450/3500 which fixes the cursor clipping problem.

Yay!

[Guest]

Chris: I don't like that you guys always refer to NVidia. You are perfectly right - The bug is on NVidia's side, but Notebook manufacturers will not release new drivers for the mobile GPUs and NVidia is not allowed to directly give drivers for these special mobile versions. I hope you don't expect every non-tech savvy user to use hacked drivers...

Contacting NVidia is NOT the problem's solution as NVidia does not give support to mobile GPUs. Contacting Dell, HP, Sony (whatever) will also most probably not yield to a result. The latest drivers for my 8400M GS official from dell are from late 2007. I hope you don't expect all users to use hacked drivers. What about publishing an official FAQ page in the adobe knowledge base with some tips for users how to circumvent this problem (e.g. the registry hack).

[akribie]

It is VERY disappointing to see Adobe passing the buck on this cursor problem.

Here we have a graphics card manufacturer that has a serious market share offering certain API calls, and Adobe relying on some arcane document (not specified here) which lays down something different.

If I took this approach when developing web sites, very few of the end users would get acceptable page rendering due to variations in adherence to the supposed 'standards'.

Surely Adobe should take the more realistic view that whatever the standards might say, a big company like NVidia cannot be ignored if its drivers don't fully comply. Especially when the problem was flagged up during beta.

And while I'm having a little rant, can I also say that the new cursor in CS4 is very much less user friendly (ie it disappears against certain backgrounds) than that in CS3 when doing grayscale retouching (of which I do a great deal). Surely it can't be that difficult to include an option in Preferences to use the old cursor code. That would be a pragmatic and welcome workaround to both issues.

Meanwhile, even after all the months since the problem was first exposed, there is no sign of an update to NVidia drivers. Doesn't say much for working level cooperation between two supposed industry leaders, does it.

[Guest]

<rev. lovejoy>oh, dear lord!</rev. lovejoy>

[Guest]

surely if the problem lies with nvidia, nvidia needs to fix the problem. you see, in computers, specifications exist so that companies creating products can be assured of interoperability. when one company breaks spec, all it does is inconvenience the user because they couldn't be bothered doing it right.

if you designed a website "to spec" (whatever, html, xhtml) but then "spiced it up" a little with something no one else uses, something "proprietary", would you expect everyone else to change for your website? (unless you're microsoft the answer is no... but even then, i notice i now have no problems browsing MSDN with firefox.)

[Guest]

in other words, write letters, make phone calls, scream and jump up and down until nvidia fixes their bugs. or HP, or dell, or lenovo. if you don't get satisfaction, what does that say about your choice in hardware vendors? if they can't be bothered to fix their bugs why would you think it's acceptable to rant at adobe for that?

sorry for MY rant, but that's how i see it.

[akribie]

But WHOSE standard is it when one of the few major graphics card manufacturers clearly doesn't adhere to it? Unless the standard is accepted by those who make the hardware, it does end users no favours to rely on the simplistic assumption that you can ignore known non-compliance.

When beta testers tell you there is a new problem with new code, regardless of whose fault it is, surely it is not unreasonable to expect a workaround, not a trenchant statement that it is someone else's fault and then to do nothing effective about it (I measure efficiency as output/input, not by the effort in between).

That's enough from me. If you follow the Adobe line that the rest of the world is out of step and can be ignored as non-compliant, so be it.

[Wolf Eilers]

>But WHOSE standard is it ...

http://www.opengl.org/about/overview/

[Guest]

>But WHOSE standard is it

wolf posted it. but more than that, the manufacturers AGREED to it, and then made mistakes (or "updates", or "customizations" whatever you want to call it) when implementing it. that's like breaking a contract. "oh yea, we support the entire opengl standard. see? got a pretty logo certification stamp right on our box! of course, we do do SOME things a 'little' differently..."

without following standards nothing would work together. that's kinda the agreement you make when you get into the game. sure make whatever you want, but if no one knows what the h%ll you're doing, there's gonna be problems. you can believe what you want about my feelings for adobe, but you're wrong.

>surely it is not unreasonable to expect a workaround, not a trenchant statement that it is someone else's fault and then to do nothing effective about it

i've seen adobe do several workarounds in the past. ps7 had a problem with terabyte sized drives due to a microsoft api bug. they worked around it in CS1. i saw them work around a kpt3 bug because no one was updating that popular plugin anymore. but sometimes it's just not possible. there are some things that underly everything else and they NEED to work as advertised. and there's nothing you can do to work around them, except to disable the features that need the problem pieces.

an argument can be made if it was wise for adobe to go down the opengl route at all. i'm not getting into that, but i might come down on the side that it was probably not wise, for the gains and changes they made. that's not the issue. the issue is now that they HAVE made that change, the video manufacturers need to step up to the plate and fulfill their side of the bargain.

sorry david j, i'm not following any "adobe line". i'm speaking as a pro developer myself (not having anything to do with adobe) who knows the need to follow standards and specs when you're interacting with products that also agree to follow that standard.

[Guest]

Dave, your post is well-reasoned, as usual.

The fact remains that it's entirely possible that Adobe made some mistakes. I'm sure they had a set date for the release of Creative Suite 4, so maybe some of the components (PS CS4) weren't quite fully cooked. Just speculation on my part, but certainly within the realm of possibility, given the power of the bean counters.

[akribie]

Thanks for the thoughtful responses, which is what I had hoped to provoke.

Nothing is ever straightforward and in millions of lines of code some bugs are inevitable. Overall, I am very impressed with PS CS4 and several little things that don't make the headlines have been a great help to things I do very often. So I'm not complaining about CS4, just challenging the response to known bugs.

I still consider that display of the cursor in PhotoShop is so fundamental to the working of the program that serious effort is justified to overcome cursor issues exposed during testing. And if the culprit (if that is a fair description of the graphics card manufacturer concerned) won't or can't fix it by release date, there should be a fall back position within the Adobe empire to avoid the negative customer experiences currently the subject of this thread.

Something not too dissimilar crippled the Premiere titler on secondary monitors for several versions of the program. That, too, was probably blamable on the graphics driver, but not getting it resolved for several years did nothing for Adobe's reputation in the eyes of those affected.

It is fixing the problem, if necessary by an Adobe compromise, that is lacking, whatever the rights and wrongs of the OpenGL standards compliance issues in this case. And, given Adobe's apparent line on this matter, the lack of any mention of the problem in the issues list for the latest NVidia drivers does nothing for my expectation of anything changing for a while yet.

[Guest]

>The fact remains that it's entirely possible that Adobe made some mistakes.

I agree 100% nick. and adobe has admitted to finding some problems. but there are others like the 1/2 cursor problem that are fixed in some versions of drivers. the fact that the laptop makers don't want to "waste" time fixing their drivers isn't adobe's fault. that's all i was getting at.

>I'm sure they had a set date for the release of Creative Suite 4, so maybe some of the components (PS CS4) weren't quite fully cooked.

I also 100% agree with that. everyone's had more than enough of my rants about adobe corporate bean counters and suits, i'm sure, but i want to make clear that i have nothing but the highest respect for the intelligence and talent of the development team and engineers.

>I still consider that display of the cursor in PhotoShop is so fundamental to the working of the program that serious effort is justified to overcome cursor issues exposed during testing.

but that's exactly one of my points david. the cursor bug has been proven to be a driver issue. we're told it was fixed by many manufacturers before the cs4 release. but some makers (specifically integrated vid laptop makers) couldn't be bothered. that's not adobe's fault and you should be screaming bloody murder to the laptop makers to fix their bugs that are crippling your system.

>there should be a fall back position within the Adobe empire to avoid the negative customer experiences currently the subject of this thread.

can't deny that. but it's a huge duplication of effort to make the same thing work 2 different way. that's a design decision, to support a fall back, or not.

I can't speak to anything with premire, as i haven't updated my copy since version 6... :)

[dwterry]

>>I still consider that display of the cursor in PhotoShop is so
>>fundamental to the working of the program that serious effort is
>>justified to overcome cursor issues exposed during testing.

>you should be screaming bloody murder to the laptop makers to fix
>their bugs that are crippling your system.

While it seems so cut-and-dried to you ... the same is not true for the end user.

The end user who just paid good money for a software upgrade that doesn't work (whereas prior versions of the same software worked fine) sees only that Photoshop isn't working like it should.

The "screaming bloody murder" you recommended ideally needs to go in both directions.

Why?

Because some people don't have options for upgrading video drivers. And because if Adobe recognizes that fact, then they will look for ways around the issue. Software can do anything, given enough time and resources (i.e. priority).

Sure, you want it known that Adobe to be innocent. FINE. We've heard you. Blame nVidia. Now get the software working using a work around or there will be more screams of bloody murder.

-------------------

(to Adobe's credit, and for anyone who hasn't read the entire thread, one of Adobe's engineers did post a work around in this thread - which surprises me that people here are still taking the hard line stance of "scream bloody murder at nVidia" when they could take the softer approach of, "please contact nVidia for a real fix, and in the mean time, here is a nice work around that you can use while you wait", customers would be so much more relieved to hear this line than the other)

[Guest]

Dave,

The video card manufacturers did step up to the plate during beta testing. Proof lies in the abobe web page with all them tested cards. Correct? Adobe even stated that as beta testing progressed so did the updates to the drivers to make the software work perfect thus ready to launch. Then the launch date came and...........

So, when did things fall apart? Dave, don't get me wrong as I do agree with your point of view. What was the sweet spot to make these tested cards work during beta yet the same card not work on other peoples machines? I do understand that some hardware works different on other machines. However, after reading all these posts about people having high end hardware that is above and beyound what I am guessing the beta testers used, it makes leads me to believe there are other un-noticed issues causing the software not to run right or fails outright to install. DRM issues?

So yes I agree that the card manufacturers need to look at the driver issues. I also agree with the person who mention that maybe the software was not fully baked for release.

With that in mind we all come together as users in this tiny webspace trying to help each other out as much as we can until the software powers to be can iron things out.

[Guest]

>While it seems so cut-and-dried to you ... the same is not true for the end user.

 i am an end user.

[Guest]

> I also agree with the person who mention that maybe the software was not fully baked for release.

I agree, dave.

>With that in mind we all come together as users in this tiny webspace trying to help each other out as much as we can until the software powers to be can iron things out.

yup. i log in here every day to learn and to help out by passing that along. that's what this forum is all about.

[dwterry]

> i am an end user.

That changes nothing in what I said. All that says is that you are apparently one of the fortunate ones who is not affected by the problem. LUCKY YOU. Bow out. The issue is between those who are affected and Adobe - and you quite obviously don't appreciate what they are going through.

Take away your working machine. Replace it with a laptop and un-upgradeable video drivers. Pay good money for an upgrade to CS4. Start trying to use it. Then see how you feel. The tune will quickly change.

AGAIN: Adobe has posted a work around. It is mentioned right here in this thread. Adobe listened. Adobe responded. Why continue the "hard line". Download the work around. Begin using it now. Then go scream at those who maintain the video drivers.

[Guest]

wah. yawn.

[Hudechrome-sd9sPI]

I see that truncated brush as well, and it's not on any kind of late model board. It' not even a PCIe video but a PCI. So Open GL isn't a factor. It won't run at all.

The fact of the matter is that CS4 is not very kind to legacy stuff. Unfortunately, it doesn't seem kind to current stuff either. That or by the release date, everything out there is legacy.

[Guest]

> LUCKY YOU. Bow out.

Doesn't it occur to you mate that people without problems might

a. be able to help by describing what they did right?

b. have sympathy with the unfortunates that are having problems?

c. have sympathy with the Adobe engineers who don't set crazy Suite marketing deadlines?

If you can't be more tolerant, take your own advice!