Duplicating a vector layer does not produce an exact copy. Here is an original layer: http://files.progarm.org/2015-02-24-183033_1276x761_scrot.png And here is a duplicated layer: http://files.progarm.org/2015-02-24-183046_1276x761_scrot.png Switch between the images back and forth to see the difference. Also, here are both layers, original layer is colored red: http://files.progarm.org/2015-02-24-183132_1276x761_scrot.png Wolthera on #krita suggested that it might be a rounding error. Indeed, the error is much worse in the bottom right corner: http://files.progarm.org/2015-02-24-175430_532x523_scrot.png Reproducible: Always Steps to Reproduce: 1. Create a vector layer and draw some paths (polyline tool) 2. Duplicate it 3. Compare two layers 4. Huh?
Hi Aleks, Thanks for your report. I haven't been able to reproduce it myself yet, but if Wolthera can confirm it, it's confirmed. Would you be able to attach a .kra file with the broken duplicated vector layers so we can see what's inside?
We still need a kra file that shows the issue.
Unfortunately, I am unable to install krita currently. However, I am surprised that it is so hard to reproduce it. It was one of the easiest bugs to reproduce. Perhaps something interesting is happening here? For example, at the time I was using 32-bit system, could that play any role in it? But yeah, until problems with krita debian package are sorted out I can't help.
Ah, you're the person who couldn't run the appimage on Debian?
Created attachment 101668 [details] requested krita file (but does it have this bug?)
No. But good that you mentioned appimage! OK, so I believe I was able to reproduce it. Interestingly, the results are not as dramatic as in my original report. Perhaps indeed the difference may exist between 32-bit and 64-bit builds? I don't know. Could've been different image size or something else. But let's see. I created a vector image layer similar to the one in my original report. Then I duplicated it and changed the colors of these layers to be able to see the difference. Now, the difference is not that big, but it is still there. Here is a screenshot of the bottom right corner: https://files.progarm.org/2016-10-20-202555_481x395_scrot.png And here is a screenshot of top left corner: https://files.progarm.org/2016-10-20-202637_740x585_scrot.png So on bottom right it is clearly shifted, although not as much as on my original screenshots. Now, I saved the file (you can see it attached), but if I reopen it I see this: https://files.progarm.org/2016-10-20-203353_671x548_scrot.png Sooo… where did the bug go? It feels like it is still there on Y axis, but not on X… weird. Does it help?
Yes, strange... Could you share the .kra file with us?
It is already attached. But as I said, I'm not sure if it is going to be visible in the .kra file. Are there any problems reproducing it?
Ah, I missed that, sorry.
Ah, I missed that, sorry. Yes, with that file I can see the effect you mean.
*** Bug 358714 has been marked as a duplicate of this bug. ***
Note that we will be replacing the current odg implementation with svg, which means all of this code will be gone.
In the current version of my Vector branch, filling of the shapes in not supported. So i'll mark this bug as "later" so we could check when the patterns thing will be implemented properly.
Ouch, I commented in a wrong bug. This bug is actually fixed in my branch :)