- Dillo-dev - lists.mailman3.com

ComplexButton crashing

by place＠gobigwest.com

Johannes wrote: > Ahh, now I understand. layout is a new layout specific to the > button. It is created in ComplexButton::init (). > So the purpose of ComplexButton is to create buttons that can have arbitrary > HTML rendered on them. > > I wonder if we really need to duplicate the button code from fltk > for this. > > For now I think your fix is good. Perhaps we can make the whole > stuff simpler in the future. > > Did you already manage to add an dw::Image to a ComplexButton? Yes. The two problems that remained were: Small images were only visible when the control panel was not visible. The recurring problem of dw's idea of offsets being different than that of fltk. And big images were usually (but not always -- some timing thing) offset inside their buttons by an amount that seemed to depend on the size of the alt text.

March 26, 2008

ComplexButton crashing

by Johannes.Hofmann＠gmx.de

On Wed, Mar 26, 2008 at 04:42:16PM +0000, place wrote: > Johannes wrote: > > place wrote: > > > In FltkComplexButtonResource::createNewWidget, the button's child is a view. > > > When the button is destroyed, since the button is a Group, its children > > > get destroyed. > > > Later on, the layout is being destroyed, and it deletes TopLevel. > > > The toplevel widget destructor calls layout->removeWidget(), which tries > > > to iterate through the views. > > > > Ok I'm just trying to understand the ComplexButton button stuff > > - rather good class name btw :-) > > Why is there a view as child of the button? > > What is this view supposed to display? > > I have no idea whether it's necessary. Maybe the layout in the button > doesn't get shown otherwise? > > > Anyway, first I'd suggest: > > > > diff -r f1f343e2c024 dw/fltkui.cc > > --- a/dw/fltkui.cc Tue Mar 18 18:05:51 2008 +0100 > > +++ b/dw/fltkui.cc Wed Mar 26 16:57:13 2008 +0100 > > @@ -423,13 +423,15 @@ > > if (!relief) > > button->box(::fltk::FLAT_BOX); > > > > - button->begin (); > > lastFlatView = > > new FltkFlatView (allocation->x + reliefXThickness (), > > allocation->y + reliefYThickness (), > > allocation->width - 2 * reliefXThickness (), > > allocation->ascent + allocation->descent > > - 2 * reliefYThickness ()); > > + > > + button->add (lastFlatView); > > + > > if (layout) > > layout->attachView (lastFlatView); > > return button; > > > > > > > > because this begin () / end () stuff is rather ugly, especially if > > the matching end () is missing. > > Yes! At first, without the end(), I assumed the begin() was extraneous. > > > Sounds reasonable, but I'd really like to understand that > > ComplexButton stuff first. Is there something in doc/ I did not see? > > How did you know it is supposed to be used for image inputs? > > In dw/ui.hh, it says > > * <tr><td>image <td>dw::core::ui::ComplexButtonResource > * <td>dw::core::ui::ResourceFactory::createComplexButtonResource, > * width a dw::Image inside and relief = false. > Ahh, now I understand. layout is a new layout specific to the button. It is created in ComplexButton::init (). So the purpose of ComplexButton is to create buttons that can have arbitrary HTML rendered on them. I wonder if we really need to duplicate the button code from fltk for this. For now I think your fix is good. Perhaps we can make the whole stuff simpler in the future. Did you already manage to add an dw::Image to a ComplexButton? Cheers, Johannes

March 26, 2008

patch: URL data as a Dstr

by place＠gobigwest.com

This is part one in a series for file inputs. Parts two and three are written. Parts four and five are contemplated. Maybe a part six if I feel like it. Part three is the one that makes it at least appear to work, though I'll need help in testing at that point. HTTPS note: If https does remain a dpi, somebody will have to decide what they want to do with Capi_dpi_build_cmd() if http_query can contain NULLs.

March 26, 2008

ComplexButton crashing

by place＠gobigwest.com

Johannes wrote: > place wrote: > > In FltkComplexButtonResource::createNewWidget, the button's child is a view. > > When the button is destroyed, since the button is a Group, its children > > get destroyed. > > Later on, the layout is being destroyed, and it deletes TopLevel. > > The toplevel widget destructor calls layout->removeWidget(), which tries > > to iterate through the views. > > Ok I'm just trying to understand the ComplexButton button stuff > - rather good class name btw :-) > Why is there a view as child of the button? > What is this view supposed to display? I have no idea whether it's necessary. Maybe the layout in the button doesn't get shown otherwise? > Anyway, first I'd suggest: > > diff -r f1f343e2c024 dw/fltkui.cc > --- a/dw/fltkui.cc Tue Mar 18 18:05:51 2008 +0100 > +++ b/dw/fltkui.cc Wed Mar 26 16:57:13 2008 +0100 > @@ -423,13 +423,15 @@ > if (!relief) > button->box(::fltk::FLAT_BOX); > > - button->begin (); > lastFlatView = > new FltkFlatView (allocation->x + reliefXThickness (), > allocation->y + reliefYThickness (), > allocation->width - 2 * reliefXThickness (), > allocation->ascent + allocation->descent > - 2 * reliefYThickness ()); > + > + button->add (lastFlatView); > + > if (layout) > layout->attachView (lastFlatView); > return button; > > > > because this begin () / end () stuff is rather ugly, especially if > the matching end () is missing. Yes! At first, without the end(), I assumed the begin() was extraneous. > Sounds reasonable, but I'd really like to understand that > ComplexButton stuff first. Is there something in doc/ I did not see? > How did you know it is supposed to be used for image inputs? In dw/ui.hh, it says * <tr><td>image <td>dw::core::ui::ComplexButtonResource * <td>dw::core::ui::ResourceFactory::createComplexButtonResource, * width a dw::Image inside and relief = false. PS I subscribed to the list under a new address, but my first msg is still awaiting moderator approval, so the old address lives on a bit longer...

March 26, 2008

ComplexButton crashing

by Johannes.Hofmann＠gmx.de

On Thu, Mar 20, 2008 at 03:12:15PM +0000, place wrote: > For you dw people... > I am working on getting ComplexButton to work (image inputs), Great! > and it would segfault when I'd close the bw. > > In FltkComplexButtonResource::createNewWidget, the button's child is a view. > When the button is destroyed, since the button is a Group, its children > get destroyed. > Later on, the layout is being destroyed, and it deletes TopLevel. > The toplevel widget destructor calls layout->removeWidget(), which tries > to iterate through the views. Ok I'm just trying to understand the ComplexButton button stuff - rather good class name btw :-) Why is there a view as child of the button? What is this view supposed to display? Anyway, first I'd suggest: diff -r f1f343e2c024 dw/fltkui.cc --- a/dw/fltkui.cc Tue Mar 18 18:05:51 2008 +0100 +++ b/dw/fltkui.cc Wed Mar 26 16:57:13 2008 +0100 @@ -423,13 +423,15 @@ if (!relief) button->box(::fltk::FLAT_BOX); - button->begin (); lastFlatView = new FltkFlatView (allocation->x + reliefXThickness (), allocation->y + reliefYThickness (), allocation->width - 2 * reliefXThickness (), allocation->ascent + allocation->descent - 2 * reliefYThickness ()); + + button->add (lastFlatView); + if (layout) layout->attachView (lastFlatView); return button; because this begin () / end () stuff is rather ugly, especially if the matching end () is missing. > > To make the problem go away, I added: > fltkcomplexbutton.hh: ~ComplexButton() {remove_all();} > but I don't know any of this code well enough to know what _should_ > be going on. Sounds reasonable, but I'd really like to understand that ComplexButton stuff first. Is there something in doc/ I did not see? How did you know it is supposed to be used for image inputs? Cheers, Johannes

March 26, 2008

removeIdle: *** [unknown] This should not happen! *** Aborted

by jcid＠dillo.org

On Tue, Mar 25, 2008 at 07:10:57PM +0100, Johannes Hofmann wrote: > On Thu, Mar 20, 2008 at 07:44:52PM +0000, place wrote: > > I think I mentioned this problem some time ago, but playing with > > ComplexButton makes it much worse (when meta charset causes a repush). > > > > The backtrace shows that it's unhappy because IdleFunc doesn't have an equals(). > > > > To make the problem go away, I changed the remove in FltkPlatform::removeIdle() > > to a removeRef, but, again, I don't know this code well enough to know what > > should be going on. > > > > #0 0x40344c81 in kill () from /lib/libc.so.6 > > #1 0x401b04a1 in pthread_kill () from /lib/libpthread.so.0 > > #2 0x401b087b in raise () from /lib/libpthread.so.0 > > #3 0x403448f8 in raise () from /lib/libc.so.6 > > #4 0x40345f00 in abort () from /lib/libc.so.6 > > #5 0x0809e054 in misc::assertNotReached () at misc.hh:67 > > #6 0x080ad6f1 in object::Object::equals (this=0x81ab118, other=0x81ab118) > > at object.cc:49 > > #7 0x080aa182 in container::untyped::List::remove0 (this=0x81c1b90, > > element=0x81ab118, compare=true, doNotDeleteAtAll=false) > > at container.cc:254 > > #8 0x080891c7 in container::untyped::List::remove (this=0x81c1b90, > > element=0x81ab118) at ../lout/container.hh:169 > > #9 0x080891eb in container::typed::List<dw::fltk::FltkPlatform::IdleFunc>::remo > > ve (this=0x81c1960, element=0x81ab118) at ../lout/container.hh:376 > > #10 0x08086968 in dw::fltk::FltkPlatform::removeIdle (this=0x81c1b68, idleId=1) > > at fltkplatform.cc:319 > > #11 0x0809feaa in ~Layout (this=0x81c1bf8) at layout.cc:117 > > > > > > As far as I can see it should be removeRef () here, as you > suggest. While looking at it, I found a buglet in layout.cc. Committed both. -- Cheers Jorge.-

March 26, 2008

Re: cache entry status

by jcid＠dillo.org

On Sun, Mar 23, 2008 at 05:29:00PM +0000, place wrote: > This is what I've ended up with. I suspect you'll regard it as overengineered... A bit too complex, yes. Although most of it stems from a not so clear design isolation between cache, dicache, capi and the rest. BTW, I suspect the dicache may have obsolete parts. > I couldn't bring myself to call it is_cached when it returned various flags, > so it's entry_status. > > I didn't know what to do about testing for Aborted. Aborted is not yet implemented. I committed a modified version (hopefully simpler). > English: I was looking at some docs the other day, and I started to notice > that you use "there're" sometimes... It's perfectly normal in speech, but it > looks odd when written. Committed. -- Cheers Jorge.-

March 26, 2008

removeIdle: *** [unknown] This should not happen! *** Aborted

by Johannes.Hofmann＠gmx.de

On Thu, Mar 20, 2008 at 07:44:52PM +0000, place wrote: > I think I mentioned this problem some time ago, but playing with > ComplexButton makes it much worse (when meta charset causes a repush). > > The backtrace shows that it's unhappy because IdleFunc doesn't have an equals(). > > To make the problem go away, I changed the remove in FltkPlatform::removeIdle() > to a removeRef, but, again, I don't know this code well enough to know what > should be going on. > > #0 0x40344c81 in kill () from /lib/libc.so.6 > #1 0x401b04a1 in pthread_kill () from /lib/libpthread.so.0 > #2 0x401b087b in raise () from /lib/libpthread.so.0 > #3 0x403448f8 in raise () from /lib/libc.so.6 > #4 0x40345f00 in abort () from /lib/libc.so.6 > #5 0x0809e054 in misc::assertNotReached () at misc.hh:67 > #6 0x080ad6f1 in object::Object::equals (this=0x81ab118, other=0x81ab118) > at object.cc:49 > #7 0x080aa182 in container::untyped::List::remove0 (this=0x81c1b90, > element=0x81ab118, compare=true, doNotDeleteAtAll=false) > at container.cc:254 > #8 0x080891c7 in container::untyped::List::remove (this=0x81c1b90, > element=0x81ab118) at ../lout/container.hh:169 > #9 0x080891eb in container::typed::List<dw::fltk::FltkPlatform::IdleFunc>::remo > ve (this=0x81c1960, element=0x81ab118) at ../lout/container.hh:376 > #10 0x08086968 in dw::fltk::FltkPlatform::removeIdle (this=0x81c1b68, idleId=1) > at fltkplatform.cc:319 > #11 0x0809feaa in ~Layout (this=0x81c1bf8) at layout.cc:117 > > As far as I can see it should be removeRef () here, as you suggest. While looking at it, I found a buglet in layout.cc. Cheers, Johannes diff -r f1f343e2c024 dw/fltkplatform.cc --- a/dw/fltkplatform.cc Tue Mar 18 18:05:51 2008 +0100 +++ b/dw/fltkplatform.cc Tue Mar 25 19:09:26 2008 +0100 @@ -316,7 +316,7 @@ void FltkPlatform::removeIdle (int idleI for(found = false, it = idleQueue->iterator(); !found && it.hasNext(); ) { idleFunc = it.getNext(); if(idleFunc->id == idleId) { - idleQueue->remove (idleFunc); + idleQueue->removeRef (idleFunc); found = true; } } diff -r f1f343e2c024 dw/layout.cc --- a/dw/layout.cc Tue Mar 18 18:05:51 2008 +0100 +++ b/dw/layout.cc Tue Mar 25 19:09:26 2008 +0100 @@ -473,7 +473,7 @@ void Layout::updateAnchor () if (anchor == NULL) { /** \todo Copy comment from old docs. */ - if (scrollIdleId && !scrollIdleNotInterrupted) { + if (scrollIdleId != -1 && !scrollIdleNotInterrupted) { platform->removeIdle (scrollIdleId); scrollIdleId = -1; }

March 25, 2008

Re: patch: concatenate Dstrs

by jcid＠dillo.org

On Tue, Mar 18, 2008 at 07:48:00PM +0000, place wrote: > Jorge wrote: > > Does anybody else have some experience/theory/article/knowledge on API > > design that can enlighten us? > > > > ;) > > I just meant "okay, never mind". After some investigation I chose to honor "minimal" in this case and committed the equivalent using Dstr_append_l(). With regard to entry_status()/is_cached(), I'm trying to find a good trade off between "minimal set" and "obvious meaning from first code read". ;-) BTW, I found good summary advice on API design here: http://blog.isnotworking.com/2007/05/api-design-guidelines.html -- Cheers Jorge.-

March 24, 2008

https plugin status

by onepoint＠starurchin.org

On Thu, Mar 20, 2008 at 09:37:13AM -0400, Jorge Arellano Cid wrote: > On Sat, Mar 15, 2008 at 08:21:23PM +0000, Jeremy Henty wrote: > > > > OK, what about my patch to redirect MSG debugging messages from > > stdout to stderr? Is that correct? > > Dpi programs are reading and writing from the stdin descriptor, > and all of them send debug messages to stdout. I think this is > correct, or at least standard by now. OK, I see. I wasn't expecting anything to *write* to the stdin descriptor so I assumed something was wrong. Regards, Jeremy Henty

March 24, 2008