I’m into astronomy. I’m currently doing an MSc in it. But mainly I like it as a hobby, usually just observing but also imaging. Most astrophotography software is written for windows (MaximDL, Registax, ImagesPlus), the pros will use IRAF but that’s overkill for most of what I do.
Earlier in the year I got a canon 350D (digital rebel XT to the yanks). And most useful processing tools for regular images are windows only also (canon’s software, picassa).
Now I know about UNIX tools. I’m probably more familiar with GIMP than any of the above mentioned tools. However since I’m stuck with needing to boot the laptop into windows occasionally, at the encouragement of some astronomy friends I decided to give Photoshop CS2 a whirl. I’ve used the free version that you get with scanners etc. Before, and to be honest I’d rather use gimp. But since adobe give you a 30 day trial I though I’d try it.
The download didn’t take too long, not as long as the install actually. But after a mug of tea it was time to fire up Photoshop. It started. It displayed its window. It displayed its splash screen and text about loading preferences, loading plugins etc. Then when it all seemed done I clicked on “File” to load an image.
The program appears to halt and the helpful “Program not responding” appears.
Perhaps the old version of photoshop that was installed was interfering. Remove all photoshops and install CS2 again. Same result.
Perhaps its just my user. Log in as another user. Same result.
Perhaps its services. Run msconfig start in diagnostic mode. Same result.
Delete any preferences files. Same result.
Fire up dtrace and see what the hell Photoshop is doing taking up 92% of my CPU and 55Mb of memory. Oh wait…
I dont know much about windows administration or debugging so off to the internet to find information on “windows xp diagnosis”. Google’s first match was this technet article.
This is a interesting, but useless document. Its entitled “Windows XP: The Rock of Reliability”. I’ll give XP one thing. Its only crashed on me once, a monumental improvement on earlier versions. However the author goes on to make some interesting comments. Some of which tell me things that I assumed windows had always done like having versioning in DLL’s. Some were new, they have a list of device drivers that they know will break your system (we just badpatch drivers like that, if the driver even made it to being a patch before being caught). Though this document told me lots about how XP attempts to stop you shooting yourself in the foot, and all the new ways they have to help you recover from the bulletwound, I don’t think the article deserves the title “Windows XP: The Rock of Reliability”. Neverthless a comment about diagnosis being a part of reliability spurred me on to a document that promised more help, Reliability Improvements in Windows XP Professional. I only have the home edition but it looked promising.
Here is what that document has to say about unresponsive programs:
“The application window has been improved in Windows XP to allow you to easily close applications that are unresponsive..”
Great. The rock of reliability has made it easier to close unresponsive programs. Thanks.
Lets try searching for “windows XP instrumentation” instead. The first match tells me how to back up the registry. The second how to use the system restore utility. Not quite what I need.
All I want is a simple way to truss what photoshop is doing on my CPU!
If Microsoft have made it easier to kill unresponsive programs surely it must be happening often enough that someone somewhere might actually want to examine why the programs are not responding. There must be some way of doing this in XP? Answers on a postcard…
Albert, thought youu knew html a bit better.. Those links need editing man. But anyways, the links,
First page, second point ‘First Line Of Defence’.. Why defence in the first place ? Theres something already to defend against ? Ok.. you’re (ms that is) is actually expecting an attack by just saying that statement.
Roll back drivers. Gee. Whiz.. They’ll probably try to patent a ‘rename file’ now.. Oh wait they have to reboot. No unix (yes I’m including solaris, aix, linux, hp-ux, *BSD*, plan9, macosX) needs that reboot.. If they do, sigh…
Jump back in time.. again with the save file versions.. more sigh, groan.
Too many to count and then list all 8. Yes they are not counting thumbs here folks. Eight fingers is all most people have. And two thumbs, but whos counting.
I’ve done enough here…
Later people.
>Those links need editing man.
The blog entry needs not to be cut and pasted in future… Fixed now. Thanks
Hmm… Mark Russinovich and Bryce Cogwell over at Sysinternals write some excellent free utilities…
You might want to check out Filemon, Regmon, CPUMon etc..
Try STrace for NT – see http://www.bindview.com/Services/RAZOR/Utilities/Windows/strace_readme.cfm
If you have MS Visual Studio, when the application hangs, you’re given an option to be thrown into Visual Studio’s debugger, which is to date the best I’ve ever worked with (only ASM-One’s debugger on Amiga comes close). Anyways, if you select to be thrown into the debugger, you’ll end up at the exact line of code (assembler) code that’s causing the hang. Now I know we’re used to system wide instrumentation on Solaris, but MS did kind of elegantly solve the problem. The catch is only that you have to have MS Visual Studio to do this.
Albert, I have exactly the same problem. CS2 takes about 15 minutes before it’s usable. It starts ok, but as soon as I go to click any menu or toolbar, it locks up with 100% CPU. It takes about 15 minutes to come good.
Did you ever find a solution to your problem?