May 2008 - Dillo-dev - lists.mailman3.com

[patch] dillo and huge pages
by Johannes.Hofmann＠gmx.de May 19, 2008

May 19, 2008

Hi, to test how dillo behaves on huge pages, I downloaded the mysql manual as a single html page: http://downloads.mysql.com/docs/refman-5.1-en.html.tar.gz With this page I noticed that the drawRegion in FltkViewBase can grow pretty large which slows down loading the page. Attached patch limits the size of the drawRegion to 16 rectangles. With this patch (and HUGE_FILESIZE in cache.c increased to 16M) dillo works pretty well on this huge page. Better than dillo1 and firefox! One problem is that clicking on anchor links (to anchors within the page) reloads the page which take some time. Also memory usage gets pretty high (250M with images disabled). I'm experimenting with reducing the size of struct Word in textblock.hh by moving the highlight stuff out of struct Word and into Textblock. This brings memory usage down to 199M. However, I don't know the highlighting stuff enough to be sure I don't break anything. This probabely falls into the "premature optimization" category anyway. Cheers, Johannes

2 2

patch: dw configure args: rtfl and doublebuffer
by corvid＠lavabit.com May 19, 2008

May 19, 2008

I saw in the bug tracker that --disable-rtfl still turns on rtfl msgs. Did some reading. The third argument to AC_ARG_ENABLE is always run if the option was specified, regardless of the value. If the option is not specified, the fourth argument is run. --disable-FEATURE just becomes --enable-FEATURE=no. (I'll send a separate patch to Jorge to allow compilation with rtfl.)

2 1

Bug that aborts Dillo
by jcid＠dillo.org May 18, 2008

May 18, 2008

Hi there, I reviewed my personal notes on dillo segfaults and all of them but one are fixed now. This is: BUG: http://episteme.arstechnica.com/groupee/forums/a/tpc/f/96509133 /m/151006647731/r/264004847731 Aborts dillo. If you wget that URL to a local file, it's easier to debug because the problem is with a requested image. A negative number is propagated down from image.cc:199 #13 0x0807d083 in dw::Image::sizeAllocateImpl (this=0x81eca98,allocation=0xbfc69fb8) at image.cc:201 (gdb) p *allocation $4 = {x = 366, y = 941, width = 3, ascent = 2, descent = 1} (gdb) p dx $5 = 8 (gdb) p dy $6 = 26607 (gdb) allocation->width - dx = -5. from there width=-5 wreaks havoc. Please someone more seasoned in the (dx,dy) offsets give it a look. -- Cheers Jorge.-

2 2

Dillo freeze
by jcid＠dillo.org May 17, 2008

May 17, 2008

Hi there, Dillo freezes on this page: http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.24.6.tar.bz2.sign Dillo bug or FLTK2? -- Cheers Jorge.-

3 6

patch: some more window deletion
by corvid＠lavabit.com May 16, 2008

May 16, 2008

I thought I was going to have to remove the viewport from the UI because both UI and layout would want to delete it, but it appears that TopGroup gets informed of the deletion and everything turns out fine.

2 1

tabs - what are they?
by Johannes.Hofmann＠gmx.de May 15, 2008

May 15, 2008

Hi, while thinking about tabs I realized that I'm not really sure what they are. * On the one hand they could be implemented just as a GUI alternative to multiple bowser windows. * On the other hand they could be something like temporary bookmarks that are easily accessible e.g. via buttons below the panel. That way one would only need one layout and view. The only issue here would be that one would expect a separate navigation stack for each "tab". What do you think? Cheers, Johannes

2 2

Future of dillo2
by 4winter＠informatik.uni-hamburg.de May 13, 2008

May 13, 2008

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Diego wrote: >> With my recent patch (dpid-cleanup-and-DPIPATH.dillo2) one can use the >> attached script to run dillo in-tree which is nice if you just want to >> try it out but should also be useful for development purposes (easily >> manage different branches without the need to install them somewhere). > > The patch seems good to me, but i have not look it very deep or test it. Some more testing turned out that it's not enough to add another path to search for dpis. With no dpidrc fill_services_list fails: [dpid]: dpid.c:467: fill_services_list: Fatal error [dpid]: - Can't find the directory for dpis. This seems very non-plugin-like to me. What's the point in being able to have dpis in ~/.dillo/... but to actually use them you have to add them to your ~/.dillo/dpidrc. I mean it's okay up to this point, but consider you have one custom dpi in ~/.dillo/... : * in order to use it you have to create a custom dpidrc too * judging by a quick look at dpid.c/register_service I'd say that you have to copy the content of the system wide dpidrc file too since your custom one takes preference over the system wide one * now there's a package installed containing one more dpi. It might be doable to update the system wide dpidrc via a post-install script within the package (though thats error prone, you have to remove your entries on package removal etc.) * but what about the custom dpircs? You cannot have the post-install script modify those as well (well, you could, but naaa) > The idea is that each dpi can have its own files in its directory > (preferences, data, cache files ... ). I could go with preferences and data files, but caching feels like a no-no to me since on most installations the user is unable to write into those directories. Justus - -- gpg key fingerprint: C82D 382A AB38 1A54 5290 19D6 A0F9 B035 686C 6996 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFIKgZ9oPmwNWhsaZYRAjL+AJ9SjhWBUNh5WcO51GkJLaqtDyjFEwCeMQkz jkTcvl8xIbP3EedZP8T72fc= =xAX5 -----END PGP SIGNATURE-----

1 0

Future of dillo2
by jcid＠dillo.org May 13, 2008

May 13, 2008

Hi there, This question: > What do you think of reviving use of the bug tracker > for the various little things that are easy to forget about? brings about a bigger topic: the future of dillo2. The goals of the project are: 1.- The democratization of internet information access 2.- Personal security and privacy 3.- High software efficiency I believe the key to achieving them are: a.- A larger development team. b.- A user community finding it useful. The vitality and development pace of the project are highly related to a). Point b), has been strong in the past; this is: judging from the received email feedback: http://www.dillo.org/emails.txt and from some surveys like: > On Sat, Aug 25, 2007 at 03:28:55AM +0200, Diego . wrote: > > DesktopLinux.com has published the results of its anual (desktop)linux poll > > The interesting thing is in the 3rd question: Which web browsers do you > > frequently use on your Linux desktop(s)? > > > > > http://www.desktoplinux.com/cgi-bin/survey/survey.cgi?view=archive&id=08132… > > > > The interesting thing is that a frozen project have enough users to win > > over 3 fully developed browsers :) (optimistic without cure) Point a) is key because it directly relates to the pace with which we pursue the moving target of Web browsing. Point b) is very important because if we end up with a browser that only a few people are willing to use, there's no point in it besides personal motivations. So: * Objective 1, is the result of a) and b). * Objective 2, needs the work of a). * Objective 3, is mainly achieved! How do we do it? i. - Increase the development team by adding 6+ new developers. ii.- Getting more involved with other projects. We could: * Attract new developers with a dillo2 release, leveraging the excellent documentation that we have. * Planning for CSS, floating objects and TABS. (this are important features to a nice browsing experience) * Coordinate with FLTK2, DSL (Damn small linux) and other lightweight distros, Debian installer team, MINIX3, etc. I don't want to miss the opportunity to thank the team for the excellent work during these months. Without you dillo2 wouldn't be so close to a release. Now, we need to plan, grow the number of developers, and bring dillo2 closer to its goals. These are just my views on the subject. Please share your thoughts, ideas and plans in this thread, so we can agree on a plan for the future. -- Cheers Jorge.-

8 22

Future of dillo2
by jcid＠dillo.org May 13, 2008

May 13, 2008

On Mon, May 12, 2008 at 10:30:14PM +0000, corvid wrote: > > On Mon, May 12, 2008 at 09:35:46PM +0000, corvid wrote: > > > > http://www.dillo.org/Plans.html > > > > > > Did you commit/make/whatever it? > > > I thought "What changed?" and looked at a cached version, > > > and didn't see any differences. > > > > Not much here AFAIR the form elements and the rewrite of tables > > have now a green ball (finished). > > > > The bulk of it is adding what "should" be added. :-) > > I see the problem now. > > You didn't update the alt tags :) Yes! Now they're updated, and the extra items added, some sorting made and it looks more like what we have. I believe if we can complete the high priority items and stabilize the code we'd have a release candidate in our hands. If Sebastian completes the floating objects in textblock, that'd be great for the release, because dillo2 would show much better rendering in DIV-based and TABLE-based pages Comments? -- Cheers Jorge.-

1 0

Future of dillo2
by jcid＠dillo.org May 13, 2008

May 13, 2008

On Mon, May 12, 2008 at 11:40:41PM +0200, Diego . wrote: > 2008/5/12 Jorge Arellano Cid <jcid(a)dillo.org>: > > > On Mon, May 12, 2008 at 03:51:42PM -0400, Jorge Arellano Cid wrote: > > > [...] > > > I'll try to update some pages in the site as soon as I find > > > some time. > > > > Just updated: > > > > http://www.dillo.org/cvs.html > > > > with directions on how to get dillo running. Developers, > > packagers, and experienced user should be pointed here. > > > > Somebody please check the directions work. :-P > > > > I doubt about the "exit" command, but i have not test the directions. "exit" is used as "logout" to get out of root. > and: > > > > http://www.dillo.org/Plans.html > > > > There is a line not updated under the "Current Plans for dillo2" text: > "--updated February 2007" > > It can be confusing. Ooops. Fixed. -- Cheers Jorge.-

1 0