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

patch: fix numeric ipv6 address handling

by 4winter＠informatik.uni-hamburg.de

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hey all :) there were two problems preventing the use of numeric ipv6 literals: 1. splitting the authority in hostname and port the attached patch fixes the handling of numeric ipv6 addresses entered into the url bar. a_Url_hostname got the colons delimiting the 8 groups of hexadecimal digits of an ipv6 address confused with a port number. 2. joining hostname and port for the Host: header a_Http_make_query_str used to build the Host: header by concatenating the url->hostname with the port number in case it's not the default port. Unfortunately this breaks with numeric ipv6 addresses since an ipv6 literal has to be enclosed with square brackets when used in urls and Host: header fields (according to my experiments and my interpretation of rfc 2616 section 14.23 ("Host header") and 3.2.2 ("http URL")). I would therefore suggest to use the original authority string obtained from the url. Patch is attached. Justus -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFI6AEJoPmwNWhsaZYRAkg1AKCMHotNlq93MMe2A4d/mcki3w8ScwCgqUWY 94TggAK4U7NNwLrXC+RPsq8= =Ajfy -----END PGP SIGNATURE-----

Oct. 5, 2008

viewport losing focus?

by corvid＠lavabit.com

If I start dillo and open <TABLE width=2600> <TR><TD>text </TABLE> I can press right arrow, but then when I press left arrow, they stop working and I see the bottom half of "Label" written over the top of the back and forward buttons, suggesting that tabs may be involved somehow.

Oct. 4, 2008

patch: print numeric addresses instead of pointer

by jcid＠dillo.org

On Sat, Oct 04, 2008 at 06:24:05PM +0200, Justus Winter wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi everyone :) > > while debugging ipv6 support I noticed that the debug output of > Dns_server(..) was not very helpful since it printed the pointer to the > address list instead of the addresses. Not yet committed, mainly because the feature freeze for release. (although the small fix went in). > The attached patch fixes this and > also adds the function > > void a_DilloHost_to_String(DilloHost *host, char *dst, size_t size) If you finally get to a complete fix, please name exported functions in dns.c with the "a_Dns_" prefix. e.g. a_Dns_host_to_string Patches are more than welcomed! -- Cheers Jorge.-

Oct. 4, 2008

patch: fix dns to ipv6 resolution

by jcid＠dillo.org

Hi Justus, On Sat, Oct 04, 2008 at 06:19:35PM +0200, Justus Winter wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi everyone :) > > I noticed that ipv6 support in dillo2 is somewhat broken. With this tiny > patch hostnames are properly resolved to ipv6 addresses and browsing > over ipv6 works fine but entering a numeric address does not. Committed. -- Cheers Jorge.-

Oct. 4, 2008

patch: close windows from menubar

by jcid＠dillo.org

On Sat, Oct 04, 2008 at 08:24:00PM +0200, Johannes Hofmann wrote: > On Fri, Oct 03, 2008 at 07:05:47PM -0400, Jorge Arellano Cid wrote: > > On Fri, Oct 03, 2008 at 07:02:47PM +0000, corvid wrote: > > > Jorge wrote: > > > > On Fri, Oct 03, 2008 at 01:07:36PM -0400, Jorge Arellano Cid wrote: > > > > > Hi, > > > > > > > > > > Committed. > > > > > > > > > > On Thu, Oct 02, 2008 at 06:58:25PM +0000, corvid wrote: > > > > > > BTW, I notice that ctrl-Q and alt-Q, listed in the menu, don't do anything. > > > > > > > > > > Yeah. > > > > > Even MenuBar::handle() doesn't see them. > > > > > > > > Seems like an FLTK bug/feature. > > > > > > > > If you try ./menu in FLTK's test/ directory, the CTRL shortcuts > > > > for the "Edit" menu don't work. > > > > > > > > Time to ask or to STR? > > > > > > I see that if I uncomment the > > > // if (widget->active() && widget->test_shortcut(false)) { > > > // setitem(p, menu, item); > > > // lastkey = 0; > > > // goto EXECUTE; > > > // } > > > in MWindow::handle(int event) in Menu_popup.cxx, ctrl-Q and alt-Q work. > > > > OK, it's clear it was deactivated by some reason... > > > > From: > > > > 1.- the mini_bug.xpm FLTK2 fix. > > 2.- the focused textarea callback issue. > > 3.- Alt-X crash in the menu. > > 4.- hardcoded 128 tabs. > > 5.- leaving-the-window nav_expecting problem. > > 6.- redraw storms (TabGroup pager) > > 7.- CTRL+q ALT+q don't work from File menu. > > > > We have: > > > > 1, 2) have patches (for FLTK), and we could wait for a new release > > or code workarounds for them in dillo. > > 3) Fixed > > 4) Easy to code a 128 limit in dillo and avoid the segfault. > > Patch attached. Committed. BTW, I introduced the "WORKAROUND:" tag (akin to "BUG:" and "TODO:"). -- Cheers Jorge.-

Oct. 4, 2008

patch: close windows from menubar

by Johannes.Hofmann＠gmx.de

On Fri, Oct 03, 2008 at 07:05:47PM -0400, Jorge Arellano Cid wrote: > On Fri, Oct 03, 2008 at 07:02:47PM +0000, corvid wrote: > > Jorge wrote: > > > On Fri, Oct 03, 2008 at 01:07:36PM -0400, Jorge Arellano Cid wrote: > > > > Hi, > > > > > > > > Committed. > > > > > > > > On Thu, Oct 02, 2008 at 06:58:25PM +0000, corvid wrote: > > > > > BTW, I notice that ctrl-Q and alt-Q, listed in the menu, don't do anything. > > > > > > > > Yeah. > > > > Even MenuBar::handle() doesn't see them. > > > > > > Seems like an FLTK bug/feature. > > > > > > If you try ./menu in FLTK's test/ directory, the CTRL shortcuts > > > for the "Edit" menu don't work. > > > > > > Time to ask or to STR? > > > > I see that if I uncomment the > > // if (widget->active() && widget->test_shortcut(false)) { > > // setitem(p, menu, item); > > // lastkey = 0; > > // goto EXECUTE; > > // } > > in MWindow::handle(int event) in Menu_popup.cxx, ctrl-Q and alt-Q work. > > OK, it's clear it was deactivated by some reason... > > From: > > 1.- the mini_bug.xpm FLTK2 fix. > 2.- the focused textarea callback issue. > 3.- Alt-X crash in the menu. > 4.- hardcoded 128 tabs. > 5.- leaving-the-window nav_expecting problem. > 6.- redraw storms (TabGroup pager) > 7.- CTRL+q ALT+q don't work from File menu. > > We have: > > 1, 2) have patches (for FLTK), and we could wait for a new release > or code workarounds for them in dillo. > 3) Fixed > 4) Easy to code a 128 limit in dillo and avoid the segfault. Patch attached. Cheers, Johannes

Oct. 4, 2008

Javascript engine

by 4winter＠informatik.uni-hamburg.de

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Jorge Arellano Cid wrote: > Disclaimer: I'm a newbie on JavaScript. JavaScript is a very cool language. Don't believe anyone telling you otherwise ;) > Does anybody know about a lightweight javascript engine we > can use with dillo? > > From what I've seen, SEE (Simple ECMAScript Engine), looks > like a good candidate. This depends entirely on your criteria :) SEE spidermonkey google v8 sloc ~31000 ~91000 ~77000 language c c c++/c license bsd {m,g,lg}pl bsd 100000 objects / s 5.6 0.4 0.2 1000000 objects / s ? 3.3 1.3 Slocs counted via sloccount, tests run on my laptop (core2duo 1.6GHz) run multiple times and averaged using a quick glance. Test script for 1000000 objects attached. see-shell didn't complete the 1000000 object run within 5 minutes of burning my cpu so I killed it. And there are some more things to consider like feature completeness. The mozilla guys are doing some sexy stuff like E4X and JavaScript 1.7 comes with syntactic sugar like array comprehension. The SEE site specifies that they are JavaScript 1.5 compliant, and although it seems actively maintained (latest 3.0 snapshot is dated 2008-02-10) I am not sure if they can keep up with the latest developments. Googles V8 does some interesting tricks like avoiding costly property lookups and generating native machine code, but this reduces portability (not sure if code generation is necessary or an add-on). The site states that it runs on "Windows XP and Vista, Mac OS X 10.5 (Leopard), and Linux systems that use IA-32 or ARM". I built an 32bit executable on linux/x86_64 that worked fine. I would suggest to pick one engine based on the ease of embedding and to develop a clean interface for scripting languages while integrating the chosen engine into dillo. If this is doable (not sure it is, but one can hope) it should be relatively easy to use the very same interface for other implementations and or languages. How cool would it be to hand configure parameters like --with-language javascript:spidermonkey - --with-language lua --with-language python :) And who knows what the next big thing of the web 3.0 will be? Grusszeugs, Justus -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFI57OToPmwNWhsaZYRAhtDAJ41dDSliA6jcKSUhw+k2/OrhY1QTgCdFtAD BnM+CV8oujgZLHz7NjY9fqw= =D1fS -----END PGP SIGNATURE-----

Oct. 4, 2008

patch: print numeric addresses instead of pointer

by 4winter＠informatik.uni-hamburg.de

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi everyone :) while debugging ipv6 support I noticed that the debug output of Dns_server(..) was not very helpful since it printed the pointer to the address list instead of the addresses. The attached patch fixes this and also adds the function void a_DilloHost_to_String(DilloHost *host, char *dst, size_t size) to dns.{c,h}. Justus -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFI55iloPmwNWhsaZYRAmP9AJ4/6msoDF59IMdpzWh0dZQhIqFgEACcDbBC wiaiHuwdsbOhU8o0ViA7a6M= =m7XD -----END PGP SIGNATURE-----

Oct. 4, 2008

patch: fix dns to ipv6 resolution

by 4winter＠informatik.uni-hamburg.de

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi everyone :) I noticed that ipv6 support in dillo2 is somewhat broken. With this tiny patch hostnames are properly resolved to ipv6 addresses and browsing over ipv6 works fine but entering a numeric address does not. Justus -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFI55eXoPmwNWhsaZYRAoCbAJ41ZA66N5zDWetCia5r4OUZmLfwngCfc0WT i2J3z0qscSGR1tuitxAX254= =GJLN -----END PGP SIGNATURE-----

Oct. 4, 2008

patch: close windows from menubar

by jcid＠dillo.org

On Sat, Oct 04, 2008 at 12:37:34AM +0000, corvid wrote: > Jorge wrote: > > From: > > > > 1.- the mini_bug.xpm FLTK2 fix. > > 2.- the focused textarea callback issue. > > 3.- Alt-X crash in the menu. > > 4.- hardcoded 128 tabs. > > 5.- leaving-the-window nav_expecting problem. > > 6.- redraw storms (TabGroup pager) > > 7.- CTRL+q ALT+q don't work from File menu. > > > > We have: > > > > 1, 2) have patches (for FLTK), and we could wait for a new release > > or code workarounds for them in dillo. > > 3) Fixed > > 4) Easy to code a 128 limit in dillo and avoid the segfault. > > 5) Fixed. > > 6) Fixed. > > 7) FLTK problem, no patch yet. No CRASH with Dillo. > > Also I'm seeing that findbar problem. > Load a page. ^F to open the findbar. Click in the viewport. Try to click in > the findbar again. Can't type in it. (Works fine if you use ^F to return > to it, though.) Fixed in CVS. > > What do we do? > > > > a) Contact FLTK developers and wait for a patched release, then > > workaround the rest (if any) in dillo and release. > > b) Workaround in dillo and release. > > > > Comments? > > Trying to get them to fix bugs and apply those that Johannes has > given them is good, but at the same time I think I would act as if > nothing will come of it -- working around what can be worked around > easily and preparing for release. I'd hate to wait while nothing > happens...and nothing happens...and nothing continues to happen. > > And, if it becomes necessary, a patched FLTK2 snapshot can be placed > on dillo.org until they fix the official one. Could be. I'd like to provide a patched FLTK2 tarball from our site, just to keep workarounds out of dillo tree, but some people may want to try dillo2 with their distro-provided FLTK2. In that case we may provide a dillo2 tarball for our patched FLTK2, and also a dillo-workarounds patch for the official FLTK2. Anyway, it'd be good to ask the FLTK team... -- Cheers Jorge.-

Oct. 4, 2008