Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
> --- Thomas DeWeese <[EMAIL PROTECTED]> wrote: > > > > At least one of the issues is with the > > PDFGraphics2D. > > in PDFGraphics2D.java:632 in draw(shape s). There > > is > > a check for a newTransform which inexplicably > > decides that > > if the new transform is the Identity transform there > > is > > no change. IIRC that was because the transform would have no effect. A transform in PDF appends to the current transform rather than setting the transform. > Thanks, Thomas, for taking a look at the code for us. > Andreas added your comments to the Bugzilla report so > they won't be lost. We'll get to these issues > (hopefully!) soon. > > Glen Keiron
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
--- Thomas DeWeese <[EMAIL PROTECTED]> wrote: > > At least one of the issues is with the > PDFGraphics2D. > in PDFGraphics2D.java:632 in draw(shape s). There > is > a check for a newTransform which inexplicably > decides that > if the new transform is the Identity transform there > is > no change. > Thanks, Thomas, for taking a look at the code for us. Andreas added your comments to the Bugzilla report so they won't be lost. We'll get to these issues (hopefully!) soon. Glen __ Do you Yahoo!? Exclusive Video Premiere - Britney Spears http://launch.yahoo.com/promos/britneyspears/
RE: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
> -Original Message- > From: Glen Mazza [mailto:[EMAIL PROTECTED] > Also, providing a link to the bug is indicated > here--asking Thomas to hunt through Bugzilla in order > to help us out--these problems are with our code, not > his--is somewhat rude. > Ok. I probably supposed that Thomas would be all too familiar with the bugzilla page... Just wanted to contribute a bit, get a few different eyes to look at the initial problem... If it came across somewhat rude: apologies for my brevity. > Furthermore, be careful about dragging users onto the > FOP-DEV list. Sometimes it can cause the priority of > what we're working on to be skewed to the detriment of > the project. > I'll take this into consideration, certainly. Anyway, Thanks Thomas for your input (which came just rolling in). Greetz, Andreas
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
Glen Mazza wrote: http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23883 Thomas, You needn't bother on this at this time--we have yet to find where Batik is wrong. So far, Squiggle has drawn the SVG correct *all* the time. But comments always welcome, and it's good for you to be aware of the issue. At least one of the issues is with the PDFGraphics2D. in PDFGraphics2D.java:632 in draw(shape s). There is a check for a newTransform which inexplicably decides that if the new transform is the Identity transform there is no change. In the second example posted you can in fact find the 'blue' rectangle as a very small open square in the lower right hand corner of the PDF. Also, there were changes to the problem yesterday that Andreas apparently didn't read--the problem is basically that the same PDF file (containing certain SVG elements such as polyline) generated by FOP renders fine in Acrobat 6.0 but not in 5.5. Don't have a clue on this one (it is possible that the 5.5 engine uses 16.16 math but the 6.0 uses FP?) Adding to the fun, the user yesterday found other problems where Squiggle runs fine but FOP fails, for *all* Acrobat versions. So this will be an ongoing issue for FOP--also, given the scope of changes needed, perhaps we may be able to make just to our 1.0 branch, the one used by the transcoder. I believe the above problem is responsible for the issue with the blue box. Why the text is 'shifted' is still a bit of a mystery to me.
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
Good--I was concerned that this was just me having the problem. Glen --- "J.Pietschmann" <[EMAIL PROTECTED]> wrote: > Glen Mazza wrote: > > I'll leave the 0.20.x branch alone until others > > complain about this problem, however. > > There is no problem in the maintenance branch, > because it was > fixed there (and unfixed, and refixed etc...) ages > ago. > > J.Pietschmann > > __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com
RE: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
--- "Andreas L. Delmelle" <[EMAIL PROTECTED]> wrote: > > Thomas, > > Can you perhaps have a look at bug 23883? In > embedded SVG, something's going > wrong with translate() when large numbers are used. > Apparently this works > fine for svg:text elements, but a polyline gets > drawn really ugly... > Andreas, If you're going to ask for someone's help, please *read* the bug again so you know the latest on the issue. There were changes yesterday that made your description of the problem out-of-date. Also, providing a link to the bug is indicated here--asking Thomas to hunt through Bugzilla in order to help us out--these problems are with our code, not his--is somewhat rude. Furthermore, be careful about dragging users onto the FOP-DEV list. Sometimes it can cause the priority of what we're working on to be skewed to the detriment of the project. Thanks, Glen http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23883 Thomas, You needn't bother on this at this time--we have yet to find where Batik is wrong. So far, Squiggle has drawn the SVG correct *all* the time. But comments always welcome, and it's good for you to be aware of the issue. Also, there were changes to the problem yesterday that Andreas apparently didn't read--the problem is basically that the same PDF file (containing certain SVG elements such as polyline) generated by FOP renders fine in Acrobat 6.0 but not in 5.5. Adding to the fun, the user yesterday found other problems where Squiggle runs fine but FOP fails, for *all* Acrobat versions. So this will be an ongoing issue for FOP--also, given the scope of changes needed, perhaps we may be able to make just to our 1.0 branch, the one used by the transcoder. Thanks, Glen __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com
RE: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
> -Original Message- > From: Glen Mazza [mailto:[EMAIL PROTECTED] > > Thanks, Tom, for your quick assistance. I didn't know > about the AWT threading issue you brought up. > Thomas, Can you perhaps have a look at bug 23883? In embedded SVG, something's going wrong with translate() when large numbers are used. Apparently this works fine for svg:text elements, but a polyline gets drawn really ugly... It seems like a very nasty one, and we may need your assistance in tracking this down. The poster mentioned the svg being properly shown in Batik Squiggle, but when the image is rendered to PDF the coordinates of the polyline get all screwed up. I've already had a look at the PDFTranscoder, but type double is being used for the translate itself, so this should pose no problems. I really hope you can shed some more light on this. TIA! Greetz, Andreas
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
Glen Mazza wrote: I'll leave the 0.20.x branch alone until others complain about this problem, however. There is no problem in the maintenance branch, because it was fixed there (and unfixed, and refixed etc...) ages ago. J.Pietschmann
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
Thanks, Tom, for your quick assistance. I didn't know about the AWT threading issue you brought up. I went ahead and added explicit System.exit() commands to Fop.java in 1.0. I was somewhat reluctant about this though because others haven't been complaining about this problem, and for some reason my computer at home was just listing these three threads (instead of the seven below) before exiting gracefully: Thread[main,5,main] Thread[Thread-0,5,main] Thread[Java2D Disposer,10,main] I'll leave the 0.20.x branch alone until others complain about this problem, however. Thanks again, Glen --- Thomas DeWeese <[EMAIL PROTECTED]> wrote: > > index = 0 Thread = Thread[main,5,main] > > index = 1 Thread = Thread[AWT-EventQueue-0,6,main] > > index = 2 Thread = > > Thread[SunToolkit.PostEventQueue-0,6,main] > > index = 3 Thread = Thread[AWT-Windows,6,main] > > index = 4 Thread = > > > Thread[EventQueueMonitor-ComponentEvtDispatch,5,main] > > index = 5 Thread = Thread[Thread-1,5,main] > > index = 6 Thread = Thread[Thread-2,5,main] > > __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
Thanks, Tom, for your quick assistance. I didn't know about the AWT threading issue you brought up. I went ahead and added explicit System.exit() commands to Fop.java in 1.0. I was somewhat reluctant about this though because others haven't been complaining about this problem, and for some reason my computer at home was just listing these three threads (instead of the seven below) before exiting gracefully: Thread[main,5,main] Thread[Thread-0,5,main] Thread[Java2D Disposer,10,main] I'll leave the 0.20.x branch alone until others complain about this problem, however. Thanks again, Glen --- Thomas DeWeese <[EMAIL PROTECTED]> wrote: > > index = 0 Thread = Thread[main,5,main] > > index = 1 Thread = Thread[AWT-EventQueue-0,6,main] > > index = 2 Thread = > > Thread[SunToolkit.PostEventQueue-0,6,main] > > index = 3 Thread = Thread[AWT-Windows,6,main] > > index = 4 Thread = > > > Thread[EventQueueMonitor-ComponentEvtDispatch,5,main] > > index = 5 Thread = Thread[Thread-1,5,main] > > index = 6 Thread = Thread[Thread-2,5,main] > > __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com
Re: Batik hanging on FOP 0.20.x nightly/1.0 dev release.
Hi Glen, The only thing I find surprising is that the old copies don't hang. It is a well known issue that when you touch AWT the Event Threads start and the only way to close the app is to call System.exit(0). Is it perhaps the case that older copies of FO called System.exit, but someone removed it? (probably because the application terminates cleanly when it doesn't touch SVG and hence AWT) Glen Mazza wrote: Team, Using the elements within an fo:instream-foreign-object is causing my work computer to having hanging threads (everything works fine at my home computer though). I'm concerned others may be getting this hanging thread problem on their machines. Results of the below FO (work computer): 0.20.5 release: works fine (1-yr. old Batik) 0.20.x nightly: hangs (Batik updated one month ago, due to API changes) 1.0 dev: hangs (Batik version of two weeks ago, also with nightly build) All three still generate a correct PDF document w/SVG, even though the app hangs (I just need to Ctrl-C to end FOP.) Here's the FO: http://www.w3.org/1999/XSL/Format"; xmlns:svg="http://www.w3.org/2000/svg";> Test FO w/SVG Running this at work without the (i.e., just an empty causes this to run fine. As soon as I add any SVG elements in though, the hanging occurs. (1) Will someone please run the above FO on a recent 1.0 build and let me know whether it exits cleanly on your machine? (2) Just before FOP exits (in FOP.java) I put in some debug statements to determine the thread counts: (run without SVG--no hanging): [INFO] 1.0dev index = 0 Thread = Thread[main,5,main] (run with SVG included--hanging): index = 0 Thread = Thread[main,5,main] index = 1 Thread = Thread[AWT-EventQueue-0,6,main] index = 2 Thread = Thread[SunToolkit.PostEventQueue-0,6,main] index = 3 Thread = Thread[AWT-Windows,6,main] index = 4 Thread = Thread[EventQueueMonitor-ComponentEvtDispatch,5,main] index = 5 Thread = Thread[Thread-1,5,main] index = 6 Thread = Thread[Thread-2,5,main] In the (unlikely?) event others are getting hanging threads w/the FO above, looking at the names of the threads above, is the hanging probably occuring with Batik threads or within FOP? I don't believe we're running multithreaded here--but am unsure where the problem is. Thanks, Glen __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com
