Running the rendering tests again this week I encountered a regression in the rendering of filters-comptran-01-b (from W3C SVG test suite). This test contains four rectangles with the same gradient, one without any filter and three with three different component transfer filters.
Last week the four rectangles in the filters-comptran-01 test were still rendered with the wrong colours, but they were rendered. This week two of them are not rendered anymore... As far as I can tell the component transfer code wasn't touched during that time though, so I have absolutely no idea why this suddenly happens.
Now the question is, should a bug be reported? And if so, what is this the actual bug? As far as I can tell feComponentTransfer isn't supposed to work anyway, so filing a bug seems a bit pointless at this point. However, as it doesn't appear to be related to a change to the component transfer filter it might actually be an unrelated bug in disguise.
So, any ideas on what could be causing this?
Update on this, one(!) of the gradients suddenly magically reappeared :) This must have happened in the past few hours, but the SVN log only shows some change that made color profiles listed in a tooltip or something...
Any ideas? Any at all?
Jasper van de Gronde wrote:
Running the rendering tests again this week I encountered a regression in the rendering of filters-comptran-01-b (from W3C SVG test suite). This test contains four rectangles with the same gradient, one without any filter and three with three different component transfer filters.
Last week the four rectangles in the filters-comptran-01 test were still rendered with the wrong colours, but they were rendered. This week two of them are not rendered anymore... As far as I can tell the component transfer code wasn't touched during that time though, so I have absolutely no idea why this suddenly happens.
Now the question is, should a bug be reported? And if so, what is this the actual bug? As far as I can tell feComponentTransfer isn't supposed to work anyway, so filing a bug seems a bit pointless at this point. However, as it doesn't appear to be related to a change to the component transfer filter it might actually be an unrelated bug in disguise.
So, any ideas on what could be causing this?
Not saying that the cause is the same... but, the only thing I can tell you is that for me, I need to render things more than once (occasionally) for all objects to show up. If it was reproducible or if I could figure out why, I would file a bug report on it.
Cheers, Josh
Jasper van de Gronde wrote:
Update on this, one(!) of the gradients suddenly magically reappeared :) This must have happened in the past few hours, but the SVN log only shows some change that made color profiles listed in a tooltip or something...
Any ideas? Any at all?
Jasper van de Gronde wrote:
Running the rendering tests again this week I encountered a regression in the rendering of filters-comptran-01-b (from W3C SVG test suite). This test contains four rectangles with the same gradient, one without any filter and three with three different component transfer filters.
Last week the four rectangles in the filters-comptran-01 test were still rendered with the wrong colours, but they were rendered. This week two of them are not rendered anymore... As far as I can tell the component transfer code wasn't touched during that time though, so I have absolutely no idea why this suddenly happens.
Now the question is, should a bug be reported? And if so, what is this the actual bug? As far as I can tell feComponentTransfer isn't supposed to work anyway, so filing a bug seems a bit pointless at this point. However, as it doesn't appear to be related to a change to the component transfer filter it might actually be an unrelated bug in disguise.
So, any ideas on what could be causing this?
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
Another week, another combination of gradients... This is the fifth(!) combination so far. I double-checked and Josh is right that it's not reproducible, on the second run today all the gradients appeared again. However, it apparently happens often enough to happen regularly with this image. (The rendering of the gradients that are rendered seems to be unaffected btw., the either render of they don't.)
It feels like a kind of memory initialization issue, as it appears to be so random. But why it happens especially often with this image?
Josh Andler wrote:
Not saying that the cause is the same... but, the only thing I can tell you is that for me, I need to render things more than once (occasionally) for all objects to show up. If it was reproducible or if I could figure out why, I would file a bug report on it.
Cheers, Josh
Jasper van de Gronde wrote:
Update on this, one(!) of the gradients suddenly magically reappeared :) This must have happened in the past few hours, but the SVN log only shows some change that made color profiles listed in a tooltip or something...
Any ideas? Any at all?
Jasper van de Gronde wrote:
Running the rendering tests again this week I encountered a regression in the rendering of filters-comptran-01-b (from W3C SVG test suite). This test contains four rectangles with the same gradient, one without any filter and three with three different component transfer filters.
Last week the four rectangles in the filters-comptran-01 test were still rendered with the wrong colours, but they were rendered. This week two of them are not rendered anymore... As far as I can tell the component transfer code wasn't touched during that time though, so I have absolutely no idea why this suddenly happens.
Now the question is, should a bug be reported? And if so, what is this the actual bug? As far as I can tell feComponentTransfer isn't supposed to work anyway, so filing a bug seems a bit pointless at this point. However, as it doesn't appear to be related to a change to the component transfer filter it might actually be an unrelated bug in disguise.
So, any ideas on what could be causing this?
Fixed :) The code would only assign types for channels that were explicitly defined. Although the SVG standard doesn't seem to explicitly specify what has to be done with the other channels I've opted to simply set the rest to the identity transform.
Jasper van de Gronde wrote:
Another week, another combination of gradients... This is the fifth(!) combination so far. I double-checked and Josh is right that it's not reproducible, on the second run today all the gradients appeared again. However, it apparently happens often enough to happen regularly with this image. (The rendering of the gradients that are rendered seems to be unaffected btw., the either render of they don't.)
It feels like a kind of memory initialization issue, as it appears to be so random. But why it happens especially often with this image?
Josh Andler wrote:
Not saying that the cause is the same... but, the only thing I can tell you is that for me, I need to render things more than once (occasionally) for all objects to show up. If it was reproducible or if I could figure out why, I would file a bug report on it.
Cheers, Josh
Jasper van de Gronde wrote:
Update on this, one(!) of the gradients suddenly magically reappeared :) This must have happened in the past few hours, but the SVN log only shows some change that made color profiles listed in a tooltip or something...
Any ideas? Any at all?
Jasper van de Gronde wrote:
Running the rendering tests again this week I encountered a regression in the rendering of filters-comptran-01-b (from W3C SVG test suite). This test contains four rectangles with the same gradient, one without any filter and three with three different component transfer filters.
Last week the four rectangles in the filters-comptran-01 test were still rendered with the wrong colours, but they were rendered. This week two of them are not rendered anymore... As far as I can tell the component transfer code wasn't touched during that time though, so I have absolutely no idea why this suddenly happens.
Now the question is, should a bug be reported? And if so, what is this the actual bug? As far as I can tell feComponentTransfer isn't supposed to work anyway, so filing a bug seems a bit pointless at this point. However, as it doesn't appear to be related to a change to the component transfer filter it might actually be an unrelated bug in disguise.
So, any ideas on what could be causing this?
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
participants (2)
-
Jasper van de Gronde -
Josh Andler