Hi,
Some time ago I decided to rearrange my images. Before I started I took backup of all images and the db. When I started I used digiKam 1.5.0 so the first bullet happend with this version. I also use version-matching kipi-plugins. I have built digiKam/kipi-plugins myself on openSuse 11.3. At the moment I use KDE4.5.3. #1: In the middle of the move of one album with around 300 images, digiKam crashed. When I restarted digiKam the album was present in both locations but only with, the same, 79 images in them. On disk all the images were moved to the new location. Is there a way to recover from this? Or do I need to use the backup? The rest seems to be true both for 1.5.0 and 1.6.0. #2: When an album has been moved to a new collection, it's still available in the old collection. All thumbnails are there and the images can be filtered with their tags. It's not until another album is moved, deleted or created the moved album is deleted from it's original space. Is this a known fault/feature? On disk the images are moved to the new space as expected. #3: (I'm still using sqlite) When moving albums like above the db gets bigger and bigger. When I started the db was just under 16MB and when I was finished it was just over 24MB. Only thing done was moving albums. I haven't looked in the tables so I don't know if there are any duplicate entries or such. When 1.6.0 was delivered I saw that there was a bug fixed that prevented the db to be cleaned. I upgrade to 1.6.0 in hope of that was the problem. But it seemed not to help in this situation, the db still increases. Any comments on that? Can I clean the db? If that's the problem. /Anders _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
> #1: > In the middle of the move of one album with around 300 images, digiKam > crashed. When I restarted digiKam the album was present in both > locations but only with, the same, 79 images in them. On disk all the > images were moved to the new location. > Is there a way to recover from this? Or do I need to use the backup? > #2: > When an album has been moved to a new collection, it's still available > in the old collection. All thumbnails are there and the images can be > filtered with their tags. It's not until another album is moved, > deleted or created the moved album is deleted from it's original > space. Is this a known fault/feature? I have seen a bug report or two about such a problem. I dont know yet what's wrong; it's a very strange behavior, as normally, digikam scans all folders at startup and will cleanup after any crash or move (and of course, it also scans when a file move is detected while running). You dont have Scan at Startup disabled? > #3: > (I'm still using sqlite) > When moving albums like above the db gets bigger and bigger. When I > started the db was just under 16MB and when I was finished it was just > over 24MB. Only thing done was moving albums. I haven't looked in the > tables so I don't know if there are any duplicate entries or such. > > When 1.6.0 was delivered I saw that there was a bug fixed that > prevented the db to be cleaned. I upgrade to 1.6.0 in hope of that > was the problem. But it seemed not to help in this situation, the db > still increases. > > Any comments on that? Can I clean the db? If that's the problem. I also presumed that the problem is now fixed. Please note that the old entries are not deleted immediately, only after a certain time or a certain number of application runs. Marcel _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
2010/12/8 Marcel Wiesweg <[hidden email]>:
> >> #1: >> In the middle of the move of one album with around 300 images, digiKam >> crashed. When I restarted digiKam the album was present in both >> locations but only with, the same, 79 images in them. On disk all the >> images were moved to the new location. >> Is there a way to recover from this? Or do I need to use the backup? >> #2: >> When an album has been moved to a new collection, it's still available >> in the old collection. All thumbnails are there and the images can be >> filtered with their tags. It's not until another album is moved, >> deleted or created the moved album is deleted from it's original >> space. Is this a known fault/feature? > > I have seen a bug report or two about such a problem. I dont know yet what's > wrong; it's a very strange behavior, as normally, digikam scans all folders at > startup and will cleanup after any crash or move (and of course, it also scans > when a file move is detected while running). > > You dont have Scan at Startup disabled? I now see that I have that option disabled. Now it's to late to test that... :-) > >> #3: >> (I'm still using sqlite) >> When moving albums like above the db gets bigger and bigger. When I >> started the db was just under 16MB and when I was finished it was just >> over 24MB. Only thing done was moving albums. I haven't looked in the >> tables so I don't know if there are any duplicate entries or such. >> >> When 1.6.0 was delivered I saw that there was a bug fixed that >> prevented the db to be cleaned. I upgrade to 1.6.0 in hope of that >> was the problem. But it seemed not to help in this situation, the db >> still increases. >> >> Any comments on that? Can I clean the db? If that's the problem. > > I also presumed that the problem is now fixed. Please note that the old > entries are not deleted immediately, only after a certain time or a certain > number of application runs. Do you know the time frame? That could probably be an user decided value in the future. Or maybe have "Would you like to clean the db now?" button, somewhere. /Anders > > Marcel > > > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users > _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
> Do you know the time frame? > > That could probably be an user decided value in the future. Or maybe > have "Would you like to clean the db now?" button, somewhere. At least a week. In detail: 7 days and two complete scans, or 30 days and no complete scans Values were arbitrarily chosen. _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
2010/12/9 Marcel Wiesweg <[hidden email]>:
> > >> Do you know the time frame? >> >> That could probably be an user decided value in the future. Or maybe >> have "Would you like to clean the db now?" button, somewhere. > > At least a week. In detail: > 7 days and two complete scans, or > 30 days and no complete scans > Values were arbitrarily chosen. And by two complete scans you mean? Is it sufficient to scan at startup? 30 days has not passed anyway. /Anders > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users > _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
> > And by two complete scans you mean? Is it sufficient to scan at startup? Yes, the scan at startup is a complete scan (locks the database, full transaction etc.) _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Hi,
I filed those two bug reports on this problem ... https://bugs.kde.org/show_bug.cgi?id=261624 https://bugs.kde.org/show_bug.cgi?id=258775 It worked fine in 1.2, started to bug in 1.5, still bugs on 1.6 ... I've tested on fresh installs of Fedora 14 and Ubuntu 10.10 with the same problem ... Very easy to reproduce ! happens 100% of the time. This can easily lead to data loss !! Let me know if there's something I can do to help fix this. Vince. Vincent Tassy E : [hidden email] <mailto:[hidden email]> http://photo.tassy.net http://www.showmetheplanet.com Skype: tassyv MSN: [hidden email] Linkedin <http://fr.linkedin.com/in/vincenttassy>Picasa <http://picasaweb.google.com/timetre>Flickr <http://www.flickr.com/photos/timetre>Twitter <http://www.twitter.com/timetre> On 12/09/2010 03:56 PM, Marcel Wiesweg wrote: > >> >> And by two complete scans you mean? Is it sufficient to scan at startup? > > Yes, the scan at startup is a complete scan (locks the database, full > transaction etc.) _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
I compiled the 2.0.0b1 and the bug is still present ...
On 01/05/2011 08:50 PM, Vincent Tassy wrote: > Hi, > > I filed those two bug reports on this problem ... > > https://bugs.kde.org/show_bug.cgi?id=261624 > https://bugs.kde.org/show_bug.cgi?id=258775 > > It worked fine in 1.2, started to bug in 1.5, still bugs on 1.6 ... > > I've tested on fresh installs of Fedora 14 and Ubuntu 10.10 with the same problem ... > Very easy to reproduce ! happens 100% of the time. > > This can easily lead to data loss !! > > > Let me know if there's something I can do to help fix this. > > Vince. > > > Vincent Tassy > > E : [hidden email] <mailto:[hidden email]> > http://photo.tassy.net > http://www.showmetheplanet.com > > Skype: tassyv MSN: [hidden email] > Linkedin <http://fr.linkedin.com/in/vincenttassy>Picasa <http://picasaweb.google.com/timetre>Flickr <http://www.flickr.com/photos/timetre>Twitter > <http://www.twitter.com/timetre> > On 12/09/2010 03:56 PM, Marcel Wiesweg wrote: >> >>> >>> And by two complete scans you mean? Is it sufficient to scan at startup? >> >> Yes, the scan at startup is a complete scan (locks the database, full >> transaction etc.) _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Free forum by Nabble | Edit this page |