- Dillo-dev - lists.mailman3.com

are we officially at feature freeze yet?

by jcid＠dillo.org

On Mon, Aug 01, 2011 at 04:08:32PM +0000, corvid wrote: > Jorge wrote: > > [ok] UI_TEMPORARILY_SHOW_PANELS > > Escape isn't hiding the findbar now. I try here with: Viewport Panels Focus Works -------------------------------------------------------- Empty Hidden Findbar OK Empty Shown Location OK Empty Shown Findbar OK HTML page Shown Findbar OK HTML page Shown Main OK HTML page Shown Location OK HTML page Hidden Findbar OK HTML page Hidden Main OK -------------------------------------------------------- and it works for every listed case. -- Cheers Jorge.-

Aug. 1, 2011

are we officially at feature freeze yet?

by corvid＠lavabit.com

Jorge wrote: > [ok] UI_TEMPORARILY_SHOW_PANELS Escape isn't hiding the findbar now.

Aug. 1, 2011

are we officially at feature freeze yet?

by jcid＠dillo.org

Hi, Now we have: [ok] moveToWidget(). Jorge has code for this. [ok] downloads dpi string escaping. [ ] the panel problems when resizing from very small. [ok] alt text wrapping (which looks like a nice feature). [ok] remove large panel. [ok] https dialog storm (BUG#868). [ok] The findbar input doesn't know about Ctrl+{a,e,d,k}. [ ] something other than "ctab%d" for initial tab label. [ ] single close tab button with tooltip. [ ] valgrind logs (needs case by case detail in this list) [ok] segfault with Ctrl-s (web search) [ok] UI_TEMPORARILY_SHOW_PANELS [ok] leave link into a native FLTK widget [ok] remove pointeronlink(), panelmode_cb_i() Just committed patches for the last three items! -- Cheers Jorge.-

Aug. 1, 2011

Re: What's pending in dillo-3.0 for a release

by dennisn＠dennisn.dyndns.org

On Fri, 29 Jul 2011 16:12:12 -0400, Jorge Arellano Cid wrote: > On Fri, Jul 08, 2011 at 10:25:04AM -0400, Dennis Nezic wrote: > > On Wed, 29 Jun 2011 11:37:21 -0400, Jorge Arellano Cid wrote: > > > Hi there, > > > > > > Fltk-1.3 is out there, and we've worked our pending issues list. > > > What's still pending? > > > > /me hopes the Choice5-popup dialog, regarding unverified > > certificates, has been fixed. And/or perhaps an option similar to > > wget's "--no-check-certificate" is implemented :). > > Please check the just committed patch for this (BUG#868). > > Quoting the log: > > "Workaround: request to only check the root URL's certificate (https). Would it save non-verified ones -- or would it ask during every runtime? :)

Aug. 1, 2011

valgrind & undeleted windows

by corvid＠lavabit.com

http://starurchin.org/dillo/valgrind/5125438759603857c01a7368aa40cd64873f5e… appears to have to do with the // TODO: free memory //delete window(); in CustTabs::remove_tab()

Aug. 1, 2011

valgrind & Cancel the expected URL after offering a download (BUG#982)

by corvid＠lavabit.com

This one appears to stem from that fix http://starurchin.org/dillo/valgrind/6120526f753eba6e5646cbfe382a68d971c39d…

Aug. 1, 2011

style bug, perhaps

by corvid＠lavabit.com

I reload the following, and the link border around the first image disappears. If the problem's in the style attr copying/comparison/etc., I'm failing to see it. I can see that the style has border type of none and border width of 0. I started to try to watch ":link img, :visited img {border: 1px solid}" get applied, but... <html> <body> <a href="whatever.html"> <img src="whatever.jpg"> </a> <button type='submit' name='button'"> <img src="something.jpg" alt="Search"> </button> </body> </html> BTW, Jeremy's valgrind logs have an interesting variety of 1.3 bugs for us now.

Aug. 1, 2011

Support for title attribute in tooltips

by Johannes.Hofmann＠gmx.de

On Sat, Jul 30, 2011 at 11:38:17PM +0000, corvid wrote: > Rob wrote: > > You are all likely to cringe in horror when you see what I have been > > up to today. I basically wanted to add hover/tooltip functionality > > for elements which contain the title attribute. > > I remember discussing this with Johannes, and I think he had some reason > that he was reluctant to check for the title attr there...? I hope he > remembers because I can't seem to find record of it. I guess you are refering to this discussion: On Wed, Sep 30, 2009 at 05:09:10PM +0200, Johannes Hofmann wrote: > On Wed, Sep 30, 2009 at 04:31:42AM +0000, corvid wrote: > > Johannes wrote: > > > On Sun, Sep 27, 2009 at 06:47:31PM +0000, corvid wrote: > > > > Also, I tried moving the title attribute check > > > > to Html_parse_common_attrs because it's part > > > > of %coreattrs (as the html spec puts it), and > > > > styleengine's style0() gave me an assertion > > > > because now I'm doing too much style computation. > > > > What should I do in light of this? > > > > > > Currently you can call StyleEngine::setNonCssHints() only once per > > > element. So we would need to pass the CssPropertyList that holds the > > > NonCssHints until we have collected them all in html.cc and then set > > > them once. > > > > Another thing to think about if we change the attr stuff > > around is that gprof shows me > > (unoptimized, auto img loading off, remote css off, > > embedded css on) > > > > % cumulative self self total > > time seconds seconds calls ms/call ms/call name > > 7.63 2.26 2.26 259353 0.01 0.01 Html_get_attr2(DilloHtml*, char const*, int, char const*, int) > > 4.23 3.51 1.25 rgb_to_565d(unsigned char const*, unsigned char*, int, int) > > 2.40 4.22 0.71 329489 0.00 0.00 dw::Textblock::wordWrap(int) > > 2.27 4.89 0.67 194573 0.00 0.01 CssStyleSheet::apply(CssPropertyList*, Doctree*, DoctreeNode const*) > > 1.93 5.46 0.57 382 1.49 9.67 Html_write_raw(DilloHtml*, char*, int, int) > > 1.62 5.94 0.48 369132 0.00 0.00 lout::object::ConstString::hashValue(char const*) > > 1.32 6.33 0.39 149410 0.00 0.01 dw::core::Widget::queueResize(int, bool) > > 1.22 6.69 0.36 745628 0.00 0.00 Html_tag_compare(char const*, char const*) > > .. > > > > I remember Jorge saying something once about having experimented > > with a more sophisticated/complicated get_attr system long ago > > but it not having been worth the complexity at the time. > > > > Ah yes, I also thought about parsing attrs in one go and storing the > result in some data structure. > Whether this is really faster than the current approach needs to be > tested. I'm not opposed to check for a title attribute in every element, we just need to make sure it doesn't slow things down too much. Cheers, Johannes

July 31, 2011

moveToWidget

by jcid＠dillo.org

On Sat, Jul 30, 2011 at 02:04:42PM +0000, corvid wrote: > > If there's a link at the very top of the viewport and you move the > mouse cursor up to the panel, the status output isn't cleared as you > leave. Ah, yes. The attached patch fixes it. The interesting part is the amount of code that can be disabled in Layout::leaveNotify(). I've tested it for some time and have found no regressions so far. Please test. -- Cheers Jorge.-

July 31, 2011

Support for title attribute in tooltips

by corvid＠lavabit.com

Rob wrote: > > I remember discussing this with Johannes, and I think he had some reason > > that he was reluctant to check for the title attr there...? I hope he > > remembers because I can't seem to find record of it. > > > > BTW, strictly speaking, we shouldn't be calling Html_parse_common_attrs() > > for every element because I believe there were a few that don't have > > %coreattrs under ATTLIST in the spec. > > Of course you are right. I was too eager to send something in, and > was really swimming over my head. Going back to > Html_parse_common_attrs(), it is plain that this is for all elements > having common attributes, whereas there are a handful of elements that > do not allow for the title attribute, and are therefore excluded. > I'll go back to learning how this thing works, and hopefully Johannes > will have something to add about the conversation you mentioned in the > meantime. Well, it's equally true of the attrs already in there -- id, class, and style, so it's not necessarily a bad place. The declaration or whatever goes: <!ENTITY % coreattrs "id ID #IMPLIED -- document-wide unique id -- class CDATA #IMPLIED -- space-separated list of classes -- style %StyleSheet; #IMPLIED -- associated style info -- title %Text; #IMPLIED -- advisory title --" >

July 31, 2011