Dropdown (above) does not show latest version (2.9.5) on program load, just keeps eating up memory until aborted. (32-bit version, x86 on same machine runs fine) Reproducible: Always Steps to Reproduce: 1. Installs fine 2. Run Krita (or double-click on executable 3. Shows no status, just keeps loading forever Actual Results: Had to Ctrl-Alt-Del and choose Task Manager to abort Expected Results: Should open program with welcome screen
Which version of windows? One guy had this issue with Krita on windows 10. (Also, I assume you meant the 64bit leaks and that 32bit doesn't leak?)
Thanks to wolthera for addressing this issue. As I said in the bug report, the 32-bit version runs fine (on the same machine where the 64bit version fails). I'm running Windows 7 (Professional) on an Asus Core i7 with 8gb memory. Video card is nVidia GT 530 w/ 2gb memory. Happy to provide any other details, ..don johnson On 6/27/2015 6:32 AM, wolthera wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > wolthera <griffinvalley@gmail.com> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > CC| |griffinvalley@gmail.com > > --- Comment #1 from wolthera <griffinvalley@gmail.com> --- > Which version of windows? One guy had this issue with Krita on windows 10. > > (Also, I assume you meant the 64bit leaks and that 32bit doesn't leak?) >
Created attachment 93528 [details] attachment-22472-0.html Hey, folks - FYI, memory leak still exists in latest release of the 64-bit version (krita_x64_2.9.6.0.msi). I keep hoping you'll get this fixed. Thanks, Don Johnson Oakland, CA -------- Original Message -------- Subject: Re: [krita] [Bug 349536] Memory leak on latest Windows 64-bit version Date: Sat, 27 Jun 2015 08:46:09 -0700 From: Donald Johnson <sfdeejay@mindspring.com> To: bug-control@bugs.kde.org Thanks to wolthera for addressing this issue. As I said in the bug report, the 32-bit version runs fine (on the same machine where the 64bit version fails). I'm running Windows 7 (Professional) on an Asus Core i7 with 8gb memory. Video card is nVidia GT 530 w/ 2gb memory. Happy to provide any other details, ..don johnson On 6/27/2015 6:32 AM, wolthera wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > wolthera <griffinvalley@gmail.com> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > CC| |griffinvalley@gmail.com > > --- Comment #1 from wolthera <griffinvalley@gmail.com> --- > Which version of windows? One guy had this issue with Krita on windows 10. > > (Also, I assume you meant the 64bit leaks and that 32bit doesn't leak?) >
Created attachment 93529 [details] attachment-25255-0.html After 64-bit version (2.9.6.0) failed, thought I'd try the 32-bit version. Now, BOTH 32 and 64 bit versions fail w/ memory leaks (never show a program screen, just keeps up eating all available memory). Should I give up on Krita for a while? ..dj On 7/8/2015 11:57 AM, Donald Johnson wrote: > Hey, folks - FYI, memory leak still exists in latest release of the > 64-bit version (krita_x64_2.9.6.0.msi). > I keep hoping you'll get this fixed. > Thanks, > Don Johnson > Oakland, CA > > > -------- Original Message -------- > Subject: Re: [krita] [Bug 349536] Memory leak on latest Windows > 64-bit version > Date: Sat, 27 Jun 2015 08:46:09 -0700 > From: Donald Johnson <sfdeejay@mindspring.com> > To: bug-control@bugs.kde.org > > > > Thanks to wolthera for addressing this issue. > As I said in the bug report, the 32-bit version runs fine (on the same > machine where the 64bit version fails). > I'm running Windows 7 (Professional) on an Asus Core i7 with 8gb memory. > Video card is nVidia GT 530 w/ 2gb memory. > Happy to provide any other details, > ..don johnson > > On 6/27/2015 6:32 AM, wolthera wrote: > >https://bugs.kde.org/show_bug.cgi?id=349536 > > > > wolthera<griffinvalley@gmail.com> changed: > > > > What |Removed |Added > > ---------------------------------------------------------------------------- > > CC| |griffinvalley@gmail.com > > > > --- Comment #1 from wolthera<griffinvalley@gmail.com> --- > > Which version of windows? One guy had this issue with Krita on windows 10. > > > > (Also, I assume you meant the 64bit leaks and that 32bit doesn't leak?) > > > > >
delete you kritarc in %appdata%/roaming, seeing as both version of Krita started to memory leak all of the sudden, this may be worth a try.
No luck. Deleted the entire directory (C:\Users\Don\AppData\Roaming\Krita) The kritarc file you mention does not appear after a 64-bit install, but was left behind from an earlier 32-bit install (which WAS working). Now, neither version of the latest release (2.9.6.1) fires up (even with that kritarc file deleted). Is it possible that I have conflicting versions of runtime files (like perl, or Python), or wrong version(s) of Microsoft's .net files? Thanks for continuing to search for a solution. ..dj On 7/13/2015 9:08 AM, wolthera wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > --- Comment #5 from wolthera <griffinvalley@gmail.com> --- > delete you kritarc in %appdata%/roaming, seeing as both version of Krita > started to memory leak all of the sudden, this may be worth a try. >
We had someone today who had more luck with the zip files than the msi's. You can download the zips from the release announcements, maybe it is something with the way we make installers?
Tried your suggestion, but no luck. FYI, I also added the \bin and \kde4 directories to the system's path variable. Didn't seem to make a difference. NO screens come up at all, just keeps devouring memory. I won't give up if you don't. Thanks again, ..dj On 7/13/2015 4:11 PM, wolthera wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > --- Comment #7 from wolthera <griffinvalley@gmail.com> --- > We had someone today who had more luck with the zip files than the msi's. You > can download the zips from the release announcements, maybe it is something > with the way we make installers? >
No... krita is pretty self-contained, so I don't really believe that that's the problem. On the other hand, on my win7/win8 test devices, which are all 64 bits, I cannot reproduce the problem. There must be something particular to your setup! Well, there is one class of applications that really can make a difference, and that's programs like sandboxie that interfere with applications' startup. But that's pretty rare.
Boud, Funny thing is, I am able to run Krita in another partition of the same computer, running Ubuntu 15. I appreciate your taking the time, but think I'm destined to stay with the old 32-bit version. Pity, as the machine, memory, graphics is a screamer and I've had almost 100% success with other programs. Thanks, Don Johnson On 8/11/2015 7:25 AM, Boudewijn Rempt wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > Boudewijn Rempt <boud@valdyas.org> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > CC| |boud@valdyas.org > > --- Comment #9 from Boudewijn Rempt <boud@valdyas.org> --- > No... krita is pretty self-contained, so I don't really believe that that's the > problem. On the other hand, on my win7/win8 test devices, which are all 64 > bits, I cannot reproduce the problem. There must be something particular to > your setup! Well, there is one class of applications that really can make a > difference, and that's programs like sandboxie that interfere with > applications' startup. But that's pretty rare. >
I really wish there were a way for me to figure out what is going on, short of using teamviewer to peek into your actual machine. It must be debuggable...
Could you perhaps try to create a second user account in Windows and try with that? There must be some local stuff installed or running that prevents krita from starting, because this isn't something that's happening to other people.
Good suggestion, Boudewijn. However, I already tried as "Administrator", without luck. I'll add another regular user and give it a try. Thanks, Don Johnson On 8/25/2015 5:46 AM, Boudewijn Rempt wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > Boudewijn Rempt <boud@valdyas.org> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > Resolution|--- |WAITINGFORINFO > Status|UNCONFIRMED |NEEDSINFO > > --- Comment #12 from Boudewijn Rempt <boud@valdyas.org> --- > Could you perhaps try to create a second user account in Windows and try with > that? There must be some local stuff installed or running that prevents krita > from starting, because this isn't something that's happening to other people. >
Boudewijn, I meant to ask you earlier about something weird I noticed: When using Task Manager to stop the program from chewing up memory, I noticed Krita is running as a "process" and not an "application". Could you pass this by some of your colleagues, as most programs do run as an application, not a process. (I did as you suggested, adding a brand-new normal user, and there was no change, just failure. Thanks again, Don Johnson On 8/25/2015 5:46 AM, Boudewijn Rempt wrote: > https://bugs.kde.org/show_bug.cgi?id=349536 > > Boudewijn Rempt <boud@valdyas.org> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > Resolution|--- |WAITINGFORINFO > Status|UNCONFIRMED |NEEDSINFO > > --- Comment #12 from Boudewijn Rempt <boud@valdyas.org> --- > Could you perhaps try to create a second user account in Windows and try with > that? There must be some local stuff installed or running that prevents krita > from starting, because this isn't something that's happening to other people. >
Hi Don, Just to clarify... This is an open source project -- we're not a company. Here's how Krita development currently is funded: https://krita.org/item/funding-krita/. Together with Stuart, who is the volunteer creating the Steam builds, I'm currently the only one who is making regular Windows builds available. So there's not really a way to pass this to my colleagues. I'm the project maintainer and one of the main developers, and so happen to to answer most bugs, I'm not the tech support guy :-) I've checked with 2.9.6.3 on Windows 8 and Windows 7 on my test laptops, and everything looks normal. When a process is starting up in Windows, it begins the Processes tab, then when it gets mapped to the task bar, it is moved into the applications tab of the taskmanager. We already knew that on your system, it never gets that far. That must have something to do with what else is around on your Windows install, but I have no idea what that could be :-(.
*** Bug 352778 has been marked as a duplicate of this bug. ***
I'm afraid I have to give up on this issue: maybe the new 3.0 builds work better? They are done with a completely different compiler in any case!