- Dillo-dev - lists.mailman3.com

Test mail

by rogerx.oss＠gmail.com

Just a reply email to check dillo-dev recievery... > On Wed, May 23, 2012 at 10:01:27AM -0400, Jorge Arellano Cid wrote: > Just a test email to check dillo-dev delivery... > >-- > Cheers > Jorge.- > >_______________________________________________ >Dillo-dev mailing list >Dillo-dev at dillo.org >http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev -- Roger http://rogerx.freeshell.org/

May 25, 2012

just a test msg

by corvid＠lavabit.com

test

May 24, 2012

Test mail

by jcid＠dillo.org

Just a test email to check dillo-dev delivery... -- Cheers Jorge.-

May 23, 2012

current plans

by itsme_410＠yahoo.com

--- On Thu, 4/19/12, corvid at lavabit.com <corvid at lavabit.com> wrote: > From: corvid at lavabit.com <corvid at lavabit.com> > Subject: Re: [Dillo-dev] current plans > To: "Globe Trotter" <itsme_410 at yahoo.com> > Date: Thursday, April 19, 2012, 12:11 AM > > the current plans page has a > number of items which are said to be at least > > in prototype: is there a way to access these > prototypes? assuming that > > they are still part of the plans to move them to > dillo? > > > > any time for a new major release yet? > > My understanding for prototype items are: > > Throttle rendering: Jorge must've done something at some > point, but I > don't know what. > Graphical plugins framework: Maybe this refers to something > from dpi > ancient history. > Image backgrounds: I think I remember Jorge mentioning > trying something > quickly a year or two ago, but I never saw the code or > anything. > Javascript: In addition to the link, I can say that I > think...Jeremy? was > interested in getting something working, but may have been > too busy to do > so. > > As for a release, I for one have been on vacation of > late... > > Thanks very much! I would like to request that you consider releasing more often so that more users may continue to be interested in dillo.....

April 19, 2012

current plans

by itsme_410＠yahoo.com

the current plans page has a number of items which are said to be at least in prototype: is there a way to access these prototypes? assuming that they are still part of the plans to move them to dillo? any time for a new major release yet? many thanks!

April 15, 2012

Display bug introduced in rev. 2368

by Johannes.Hofmann＠gmx.de

Hi Alexander, On Sat, Mar 31, 2012 at 11:04:26AM +0200, Alexander Voigt wrote: > Dear Dillo developers, > > I just noticed a minor page rendering issue, which was introduced in > revision 2368: The "I'm Feeling Lucky" button at google.com is no more > visible. > > How to reproduce: > ================ > > (1) hg update -r 2367 > (2) visit google.com and note the "I'm Feeling Lucky" button > (3) hg update -r 2368 > (4) visit google.com and note the absence of the > "I'm Feeling Lucky" button > > I noticed this issue with the german google page (google.de) but I > think it also happens on the US variant (google.com) Thanks for the detailed bug report. I backed out the inline-block support until this is resolved. Cheers, Johannes

April 7, 2012

Suggestion regarding XHTML handling

by singpolyma＠singpolyma.net

Somebody claiming to be Johannes Hofmann wrote: >On Fri, Mar 23, 2012 at 10:54:35PM -0500, Stephen Paul Weber wrote: >> Might I suggest actually doing basic namespace handling when a page is >> delivered using an XML MIME type? >> ... instead of seeing the Atom as something it >> cannot process or otherwise handling it, the rendering engine just turns all >> the tags into inline flow content and shows the text contents in what is >> essentially line noise. >> >I don't see what could be done differently here. Firefox renders it >pretty much the same way as Dillo does. Am I missing something? Yes, but that doesn't mean dillo can't be better than firerox :) Stuff from a namespace we understand (right now just XHTML, but in the future maybe SVG, MathML, whatever. or not) should get processed according to the rules of that namespace. Namespaces we don't understand, and for which we have no processing rules, should be ignored completely. And, I feel that if a server is serving with an XML MIME-type, we can trust that they expect XML-style processing rules like this, since tagsoup sites are not served with such a MIME. -- Stephen Paul Weber, @singpolyma See <http://singpolyma.net> for how I prefer to be contacted edition right joseph

March 31, 2012

Display bug introduced in rev. 2368

by Hole.destructor＠gmx.de

Dear Dillo developers, I just noticed a minor page rendering issue, which was introduced in revision 2368: The "I'm Feeling Lucky" button at google.com is no more visible. How to reproduce: ================ (1) hg update -r 2367 (2) visit google.com and note the "I'm Feeling Lucky" button (3) hg update -r 2368 (4) visit google.com and note the absence of the "I'm Feeling Lucky" button I noticed this issue with the german google page (google.de) but I think it also happens on the US variant (google.com) Kind regards, Alexander

March 31, 2012

Suggestion regarding XHTML handling

by Johannes.Hofmann＠gmx.de

On Fri, Mar 23, 2012 at 10:54:35PM -0500, Stephen Paul Weber wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA256 > > Might I suggest actually doing basic namespace handling when a page is > delivered using an XML MIME type? For example, this page: > <http://aggregator.singpolyma.net/feeds/denver/> contains both XHTML and > embedded (namespaced) Atom, but instead of seeing the Atom as something it > cannot process or otherwise handling it, the rendering engine just turns all > the tags into inline flow content and shows the text contents in what is > essentially line noise. > > I know that in a "normal" HTML page you want the heuristics to do this sort > of thing in case some bozo has dumped something that maybe looks like a > namespace into the tag soup (and namespaces aren't officially supported in > HTML anyway), but if the MIME-type is an XML one, the UA should be able to > process it as XML and ignore or otherwise handle namespaces it doesn't > understand. I don't see what could be done differently here. Firefox renders it pretty much the same way as Dillo does. Am I missing something? Cheers, Johannes

March 29, 2012

Compiling on wheezy

by abe＠deuxchevaux.org

Hi, On Tue, Mar 27, 2012 at 04:35:23PM -0300, Jorge Arellano Cid wrote: > FYI, the last Debian wheezy update requires -fPIC for dillo > to compile (at least in my 64bit machine). Yeah, and it seems to happen on amd64 (x86_64) only so far: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=664949 > CFLAGS="-g -O0 -fPIC" CXXFLAGS="-g -O0 -fPIC" ./configure --enable-ssl Adding CPPFLAGS=-fPIC sufficed for me. We're still investigating why it's actually problem and wasn't one about a month ago... Kind regards, Axel -- /~\ Plain Text Ribbon Campaign | Axel Beckert \ / Say No to HTML in E-Mail and News | abe at deuxchevaux.org (Mail) X See http://www.asciiribbon.org/ | abe at noone.org (Mail+Jabber) / \ I love long mails: http://email.is-not-s.ms/ | http://noone.org/abe/ (Web)

March 27, 2012

Compiling on wheezy

by jcid＠dillo.org

Hi there, FYI, the last Debian wheezy update requires -fPIC for dillo to compile (at least in my 64bit machine). e.g. CFLAGS="-g -O0 -fPIC" CXXFLAGS="-g -O0 -fPIC" ./configure --enable-ssl HTH. -- Cheers Jorge.-

March 27, 2012

dillo on maemo

by Johannes.Hofmann＠gmx.de

here is a nice video showing Dillo on maemo: http://youtu.be/v2jH_PXeA2A Cheers, Johannes

March 26, 2012

Suggestion regarding XHTML handling

by singpolyma＠singpolyma.net

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Might I suggest actually doing basic namespace handling when a page is delivered using an XML MIME type? For example, this page: <http://aggregator.singpolyma.net/feeds/denver/> contains both XHTML and embedded (namespaced) Atom, but instead of seeing the Atom as something it cannot process or otherwise handling it, the rendering engine just turns all the tags into inline flow content and shows the text contents in what is essentially line noise. I know that in a "normal" HTML page you want the heuristics to do this sort of thing in case some bozo has dumped something that maybe looks like a namespace into the tag soup (and namespaces aren't officially supported in HTML anyway), but if the MIME-type is an XML one, the UA should be able to process it as XML and ignore or otherwise handle namespaces it doesn't understand. - -- Stephen Paul Weber, @singpolyma See <http://singpolyma.net> for how I prefer to be contacted edition right joseph -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) iQIcBAEBCAAGBQJPbUV7AAoJENEcKRHOUZze+aMQAKDSNe3QZzzsIFYZkH2fvWE+ 03gRZy7XtwzqQMBQENAuZgK1orUGR8KO8t69re7f+GFTed36lE5QrqtY2tKPmtrl BD0W7d0bbZpAzIk0akS44l9+ICy2X+iKMxhdzeELcEABxlRLqfgYwS3zZ+XjvOyK VGOj8ejDs1n4G2SO6IC28DXJxa8cijkhfN7QzAO1rUpHSlPV8xO6vCIOultr05Nc VoybKhBTetDV4FsRX4hpwd5/ToAl6JYWvkwaSEny7Pc30H8EXthIzASnf+K2rPwh v31aNWJgrf6DwXhnX1SZLYYFkePpuiLzOAWGbljacJlnkI9r6guXDUIq0U0mCdLD G1fuvnaLCSv78yz7MwDS/hDwxx7rElXk3w7BbwzA71tbuJocBTq3clX4Ho+dL84L +pYCXWjWbt5BPiIgHDiPzj+48AS4kvW9y4a/yCPoAJFExZDdoNdSlFGUvGmc2xwa 9Rcvz7lX6DQ5iZXXBLpVi5IrwMFEaXKd+0dUbsaXD30a6n6r+8o6sDMSh+CIDu20 dVhB58ff7j7cwVsPSQtHZmkjq4x3KxMiIm/gtYijkxh48lsUwMiyIBpw76gUZDPp N22Q4c5bGiHWRUhYHTwuu1C04Dx0wHeWoaKDGaPV9YyEyPq1EiZZTz2Vzd7AMbUT K1x3M4Fep0SUuLsVusle =ldnT -----END PGP SIGNATURE-----

March 24, 2012

thanks for display:block

by Hole.destructor＠gmx.de

Hi Jorge. > Since what dillo version do you see differences? > Could you please post a few example URLs? The improvements were introduced in revision 2357. So, if I update to revision 2356 (`hp update -r 2356') the yellow LaTeX code bits (within a <pre> environment) on the page http://iktp.tu-dresden.de/~avoigt/feynmf.xhtml does not look so nice (there is no border around the code field and the yellow background field color is not rectangular). If I update to revision 2357 (`hp update -r 2357') these code bits are surrounded by a border (as intended) and the yellow background has the form of a rectangle, as it should be. Kind regards, Alexander

March 21, 2012

DNS code changes

by Johannes.Hofmann＠gmx.de

On Tue, Mar 20, 2012 at 06:07:08PM +0100, Georg Potthast wrote: > Sorry for posting three times I had the impression my emails got > lost somehow. It seems there was some issue with the mailing list. > > DOS is basically a single-tasking system. So you cannot fork a new > process and threads do not work well. A thread, as it is currently > implemented for DOS/DJGPP, should yield when it is done and a > different thread can continue. So I try to avoid threads and use > function calls instead if I can. We still use threads only for asynchronous DNS resolution and this can be disabled as before with --disable-threaded-dns. Let us know if there are new issues with the currently committed version. Cheers, Johannes

March 20, 2012

make fail on OS X 10.6.8

by corvid＠lavabit.com

Jorge wrote: > On Sat, Feb 25, 2012 at 05:22:25AM +0000, Phil Dobbin wrote: > > -----BEGIN PGP SIGNED MESSAGE----- > > Hash: SHA1 > > > > Hi, all. > > > > I'm trying to install Dillo v.3.0.2 from source on OS X 10.6.8 & after > > successfuly installing fltk-1.3.0 into /usr/local & running configure & > > make I get this error: > > > > `ld: duplicate symbol Fl_Printer::class_id in > > /usr/local/lib/libfltk.a(Fl_Printer.o) and > > ../dw/libDw-fltk.a(libDw_fltk_a-fltkplatform.o) > > collect2: ld returned 1 exit status > > make[3]: *** [dillo] Error 1 > > make[2]: *** [all-recursive] Error 1 > > make[1]: *** [all-recursive] Error 1 > > make: *** [all] Error 2` > > > > I've tried Googling it (this is where I got the email address for this > > list. Apologies if I should've sent it to dillo-users. If that's the > > case, please let me know & I'll make amends) but no luck. > > IIRC, sometime ago corvid added a few lines in fltkplatform.cc to > avoid linking of unused printer code. You may expect his answer, but > in the meanwhile, you may try commenting that line: > > - const char *Fl_Printer::class_id = "Fl_Printer"; > + // const char *Fl_Printer::class_id = "Fl_Printer"; > > or if that fails, > commenting the whole stub (lines 35 to 42 in fltkplatform.cc). For the moment, I can at least point us back to the original discussion at http://lists.auriga.wearlab.de/pipermail/dillo-dev/2011-June/008449.html The search engine suggests that ld complaining about duplicate symbols is fairly consistently an osx thing...

March 20, 2012

DNS code changes

by mailbox＠georgpotthast.de

Sorry for posting three times I had the impression my emails got lost somehow. DOS is basically a single-tasking system. So you cannot fork a new process and threads do not work well. A thread, as it is currently implemented for DOS/DJGPP, should yield when it is done and a different thread can continue. So I try to avoid threads and use function calls instead if I can. Georg

March 20, 2012

thanks for display:block

by jcid＠dillo.org

Hi Alexander, On Sun, Mar 18, 2012 at 03:49:46PM +0100, Alexander Voigt wrote: > Dear Dillo developers, > > I just want to thank you and Johannes for recently implementing the > "display:block" CSS property for the elements address, center, div, > h1, h2, h3, h4, h5, h6, ol, p, ul, pre. For me many pages look much > better now! Since what dillo version do you see differences? Could you please post a few example URLs? Thanks for the appreciative feedback! -- Cheers Jorge.-

March 20, 2012

dillo icon

by jcid＠dillo.org

On Sun, Mar 11, 2012 at 05:54:28PM -0700, Globe Trotter wrote: > Does dillo have a formal icon? If not, I would like to suggest > the hummingbird on the webpage which is graceful, elegant, > lightweight and spunky. Before some other undeserving browser > claims it! What do you all think? Personally, I like the hummingbird for the web site, but prefer John Grantham's one [1] for launch bars and menus. Anyway, it'd be great to know the canonical way to pack an icon in our tarball. Does anybody know how it's done? (so the distro packager can find it easily). [1] http://www.dillo.org/Icons/ -- Cheers Jorge.-

March 20, 2012

make fail on OS X 10.6.8

by jcid＠dillo.org

Hi, On Sat, Feb 25, 2012 at 05:22:25AM +0000, Phil Dobbin wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi, all. > > I'm trying to install Dillo v.3.0.2 from source on OS X 10.6.8 & after > successfuly installing fltk-1.3.0 into /usr/local & running configure & > make I get this error: > > `ld: duplicate symbol Fl_Printer::class_id in > /usr/local/lib/libfltk.a(Fl_Printer.o) and > ../dw/libDw-fltk.a(libDw_fltk_a-fltkplatform.o) > collect2: ld returned 1 exit status > make[3]: *** [dillo] Error 1 > make[2]: *** [all-recursive] Error 1 > make[1]: *** [all-recursive] Error 1 > make: *** [all] Error 2` > > I've tried Googling it (this is where I got the email address for this > list. Apologies if I should've sent it to dillo-users. If that's the > case, please let me know & I'll make amends) but no luck. IIRC, sometime ago corvid added a few lines in fltkplatform.cc to avoid linking of unused printer code. You may expect his answer, but in the meanwhile, you may try commenting that line: - const char *Fl_Printer::class_id = "Fl_Printer"; + // const char *Fl_Printer::class_id = "Fl_Printer"; or if that fails, commenting the whole stub (lines 35 to 42 in fltkplatform.cc). HTH. -- Cheers Jorge.-

March 20, 2012

DNS code changes

by johannes.hofmann＠gmx.de

On Sat, Feb 25, 2012 at 10:58:36AM +0100, Georg Potthast wrote: > Could you please leave the configure option " > --disable-threaded-dns" in the code since I use that for my DOS > version? I cannot use dpi. It's still in and will stay. Do you have problems with current tip? If so, what exactly is the problem? Cheers, Johannes

March 20, 2012

DNS code changes

by mailbox＠georgpotthast.de

Could you please leave the configure option " --disable-threaded-dns" in the code since I use that for my DOS version? I cannot use dpi. Georg

March 19, 2012

thanks for display:block

by Hole.destructor＠gmx.de

Dear Dillo developers, I just want to thank you and Johannes for recently implementing the "display:block" CSS property for the elements address, center, div, h1, h2, h3, h4, h5, h6, ol, p, ul, pre. For me many pages look much better now! Thank you, Alexander

March 18, 2012

dillo icon

by itsme_410＠yahoo.com

Does dillo have a formal icon? If not, I would like to suggest the hummingbird on the webpage which is graceful, elegant, lightweight and spunky. Before some other undeserving browser claims it! What do you all think? Best wishes!

March 12, 2012

DNS code changes

by mailbox＠georgpotthast.de

Could you please leave the configure option " --disable-threaded-dns" in the code since I use that for my DOS version? I cannot use dpi. Georg

March 2, 2012

another tooltip bug

by corvid＠lavabit.com

With the attached page, I put the cursor on the first image and let the tooltip appear (and don't have the cursor too far over to the left of the window or screen). Then I move the cursor down into the second image and let go of the mouse. For a second, the tooltip for the second image shows up in the proper position, but then it moves to the right so that not all of it is visible.

March 2, 2012

DNS code changes

by mailbox＠georgpotthast.de

Could you please leave the configure option " --disable-threaded-dns" in the code since I use that for my DOS version? I cannot use dpi. Georg

Feb. 25, 2012

make fail on OS X 10.6.8

by phildobbin＠gmail.com

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, all. I'm trying to install Dillo v.3.0.2 from source on OS X 10.6.8 & after successfuly installing fltk-1.3.0 into /usr/local & running configure & make I get this error: `ld: duplicate symbol Fl_Printer::class_id in /usr/local/lib/libfltk.a(Fl_Printer.o) and ../dw/libDw-fltk.a(libDw_fltk_a-fltkplatform.o) collect2: ld returned 1 exit status make[3]: *** [dillo] Error 1 make[2]: *** [all-recursive] Error 1 make[1]: *** [all-recursive] Error 1 make: *** [all] Error 2` I've tried Googling it (this is where I got the email address for this list. Apologies if I should've sent it to dillo-users. If that's the case, please let me know & I'll make amends) but no luck. Any help appreciated. Cheers, Phil... - -- But masters, remember that I am an ass. Though it be not written down, yet forget not that I am an ass. Wm. Shakespeare - Much Ado About Nothing -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.17 (Darwin) Comment: ??auto-key-locate cert pka ldap hkp://keys.gnupg.net Comment: GPGTools - http://gpgtools.org Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQEcBAEBAgAGBQJPSHARAAoJEKpMeDHWT5AD/hMIAInZu2mC96q3ZmYniyrnFJ+p hyNs8ELSsDOxzzeIps5XjiRzAto92l37ey+u0m3Ykt7rtv/vLaVj4zo+Bq0zaVDp 3znKMhIriaYzw4d7lRaTfLjxMETniZrh575/yXNpkQoCaBS7qSFWbYcRq/OCtuxE zbkHA+C7W9XPX3nyDZ6DWpCAXWDH68U/NECEksSqeNYmGUOHojhzun7/HOFkXXsT 1O2vFZY3PmNV4KdHQnedqnF4iEOBamNZS1sCU2DtSZUd3FUrhR9Pb3ImGovmtCFx vSGdZOL7/UYDtULCEHxxFB+8FoZgSm2UU42/+ac0PGKNcTOOjoBikjEBokf3Fvs= =YBHD -----END PGP SIGNATURE-----

Feb. 25, 2012

[exprimental patch] wakeup instead of polling for dns

by johannes.hofmann＠gmx.de

On Sun, Feb 19, 2012 at 01:53:13PM -0300, Jorge Arellano Cid wrote: > Hi Johannes, > > Please excuse the delayed answer. > > On Tue, Feb 14, 2012 at 09:44:24PM +0100, Johannes Hofmann wrote: > > On Sun, Feb 12, 2012 at 08:44:04PM +0100, Johannes Hofmann wrote: > > > Hi Jorge, > > > > > > On Fri, Feb 10, 2012 at 01:40:10PM -0300, Jorge Arellano Cid wrote: > > > > Hi Johannes, > > > > > > > > On Mon, Feb 06, 2012 at 09:45:44PM +0100, Johannes Hofmann wrote: > > > > > Hi, > > > > > > > > > > attached experimental patch replaces the polling for DNS answers > > > > > with the wakeup mechanism fltk provides. > > > > > What do you think? > > > > > > > > Well dns resolving is delicate, but not sacred ;) > > > > > > > > AFAIS the main advantage would be a near 0.1 second speedup on > > > > average when displaying non dns-cached pages, (0.2 best case). > > > > Which is significative. > > > > > > > > Please let me know if there's another advantage I miss. > > > > > > Well, I don't like the polling :-) > > > There might be an increased power consumption related to the > > > 5 times/s timer, but I'm not sure whether it really matters. > > > > > > > > > > > The cost is using an FLTK mechanism (awake) which we don't in > > > > the rest of the code. > > > > > > Yes, plus we need to switch on locking for all fltk just for the dns > > > thing. That's the major drawback I think. > > > > > > > > > > > It looks clean. It should be tested on different platforms/OS. > > > > > > > > BTW, dns.c has code for a blocking server (not threaded). Does > > > > this awake() call work OK when called from the same main thread? > > > > Is this by design or by chance? Maybe it would be safer to avoid > > > > calling awake() in this case. > > > > > > I didn't test it yet. I still want to try other approaches that > > > would avoid switching on the fltk threading stuff with the > > > Fl::lock() call. > > > > Ok, attached is another approach that avoids the Fl::lock() and > > instead uses a Fl::add_fd() which is already used in src/IO. > > Let me know what you think. > > I just re-read the patch, this time from HG, and it looks clean. > The pipe trick is very well tested within dpi so it should work > as well as it looks. Ok good. I just committed it as it no longer has the drawbacks of the initial patch I sent. Cheers, Johannes

Feb. 20, 2012

[exprimental patch] wakeup instead of polling for dns

by jcid＠dillo.org

Hi Johannes, Please excuse the delayed answer. On Tue, Feb 14, 2012 at 09:44:24PM +0100, Johannes Hofmann wrote: > On Sun, Feb 12, 2012 at 08:44:04PM +0100, Johannes Hofmann wrote: > > Hi Jorge, > > > > On Fri, Feb 10, 2012 at 01:40:10PM -0300, Jorge Arellano Cid wrote: > > > Hi Johannes, > > > > > > On Mon, Feb 06, 2012 at 09:45:44PM +0100, Johannes Hofmann wrote: > > > > Hi, > > > > > > > > attached experimental patch replaces the polling for DNS answers > > > > with the wakeup mechanism fltk provides. > > > > What do you think? > > > > > > Well dns resolving is delicate, but not sacred ;) > > > > > > AFAIS the main advantage would be a near 0.1 second speedup on > > > average when displaying non dns-cached pages, (0.2 best case). > > > Which is significative. > > > > > > Please let me know if there's another advantage I miss. > > > > Well, I don't like the polling :-) > > There might be an increased power consumption related to the > > 5 times/s timer, but I'm not sure whether it really matters. > > > > > > > > The cost is using an FLTK mechanism (awake) which we don't in > > > the rest of the code. > > > > Yes, plus we need to switch on locking for all fltk just for the dns > > thing. That's the major drawback I think. > > > > > > > > It looks clean. It should be tested on different platforms/OS. > > > > > > BTW, dns.c has code for a blocking server (not threaded). Does > > > this awake() call work OK when called from the same main thread? > > > Is this by design or by chance? Maybe it would be safer to avoid > > > calling awake() in this case. > > > > I didn't test it yet. I still want to try other approaches that > > would avoid switching on the fltk threading stuff with the > > Fl::lock() call. > > Ok, attached is another approach that avoids the Fl::lock() and > instead uses a Fl::add_fd() which is already used in src/IO. > Let me know what you think. I just re-read the patch, this time from HG, and it looks clean. The pipe trick is very well tested within dpi so it should work as well as it looks. -- Cheers Jorge.-

Feb. 19, 2012

[exprimental patch] wakeup instead of polling for dns

by Johannes.Hofmann＠gmx.de

On Sun, Feb 12, 2012 at 08:44:04PM +0100, Johannes Hofmann wrote: > Hi Jorge, > > On Fri, Feb 10, 2012 at 01:40:10PM -0300, Jorge Arellano Cid wrote: > > Hi Johannes, > > > > On Mon, Feb 06, 2012 at 09:45:44PM +0100, Johannes Hofmann wrote: > > > Hi, > > > > > > attached experimental patch replaces the polling for DNS answers > > > with the wakeup mechanism fltk provides. > > > What do you think? > > > > Well dns resolving is delicate, but not sacred ;) > > > > AFAIS the main advantage would be a near 0.1 second speedup on > > average when displaying non dns-cached pages, (0.2 best case). > > Which is significative. > > > > Please let me know if there's another advantage I miss. > > Well, I don't like the polling :-) > There might be an increased power consumption related to the > 5 times/s timer, but I'm not sure whether it really matters. > > > > > The cost is using an FLTK mechanism (awake) which we don't in > > the rest of the code. > > Yes, plus we need to switch on locking for all fltk just for the dns > thing. That's the major drawback I think. > > > > > It looks clean. It should be tested on different platforms/OS. > > > > BTW, dns.c has code for a blocking server (not threaded). Does > > this awake() call work OK when called from the same main thread? > > Is this by design or by chance? Maybe it would be safer to avoid > > calling awake() in this case. > > I didn't test it yet. I still want to try other approaches that > would avoid switching on the fltk threading stuff with the > Fl::lock() call. ... and here comes the patch.

Feb. 14, 2012

[exprimental patch] wakeup instead of polling for dns

by Johannes.Hofmann＠gmx.de

On Sun, Feb 12, 2012 at 08:44:04PM +0100, Johannes Hofmann wrote: > Hi Jorge, > > On Fri, Feb 10, 2012 at 01:40:10PM -0300, Jorge Arellano Cid wrote: > > Hi Johannes, > > > > On Mon, Feb 06, 2012 at 09:45:44PM +0100, Johannes Hofmann wrote: > > > Hi, > > > > > > attached experimental patch replaces the polling for DNS answers > > > with the wakeup mechanism fltk provides. > > > What do you think? > > > > Well dns resolving is delicate, but not sacred ;) > > > > AFAIS the main advantage would be a near 0.1 second speedup on > > average when displaying non dns-cached pages, (0.2 best case). > > Which is significative. > > > > Please let me know if there's another advantage I miss. > > Well, I don't like the polling :-) > There might be an increased power consumption related to the > 5 times/s timer, but I'm not sure whether it really matters. > > > > > The cost is using an FLTK mechanism (awake) which we don't in > > the rest of the code. > > Yes, plus we need to switch on locking for all fltk just for the dns > thing. That's the major drawback I think. > > > > > It looks clean. It should be tested on different platforms/OS. > > > > BTW, dns.c has code for a blocking server (not threaded). Does > > this awake() call work OK when called from the same main thread? > > Is this by design or by chance? Maybe it would be safer to avoid > > calling awake() in this case. > > I didn't test it yet. I still want to try other approaches that > would avoid switching on the fltk threading stuff with the > Fl::lock() call. Ok, attached is another approach that avoids the Fl::lock() and instead uses a Fl::add_fd() which is already used in src/IO. Let me know what you think. Cheers, Johannes

Feb. 14, 2012

[exprimental patch] wakeup instead of polling for dns

by Johannes.Hofmann＠gmx.de

Hi Jorge, On Fri, Feb 10, 2012 at 01:40:10PM -0300, Jorge Arellano Cid wrote: > Hi Johannes, > > On Mon, Feb 06, 2012 at 09:45:44PM +0100, Johannes Hofmann wrote: > > Hi, > > > > attached experimental patch replaces the polling for DNS answers > > with the wakeup mechanism fltk provides. > > What do you think? > > Well dns resolving is delicate, but not sacred ;) > > AFAIS the main advantage would be a near 0.1 second speedup on > average when displaying non dns-cached pages, (0.2 best case). > Which is significative. > > Please let me know if there's another advantage I miss. Well, I don't like the polling :-) There might be an increased power consumption related to the 5 times/s timer, but I'm not sure whether it really matters. > > The cost is using an FLTK mechanism (awake) which we don't in > the rest of the code. Yes, plus we need to switch on locking for all fltk just for the dns thing. That's the major drawback I think. > > It looks clean. It should be tested on different platforms/OS. > > BTW, dns.c has code for a blocking server (not threaded). Does > this awake() call work OK when called from the same main thread? > Is this by design or by chance? Maybe it would be safer to avoid > calling awake() in this case. I didn't test it yet. I still want to try other approaches that would avoid switching on the fltk threading stuff with the Fl::lock() call. Cheers, Johannes

Feb. 12, 2012

[exprimental patch] wakeup instead of polling for dns

by jcid＠dillo.org

Hi Johannes, On Mon, Feb 06, 2012 at 09:45:44PM +0100, Johannes Hofmann wrote: > Hi, > > attached experimental patch replaces the polling for DNS answers > with the wakeup mechanism fltk provides. > What do you think? Well dns resolving is delicate, but not sacred ;) AFAIS the main advantage would be a near 0.1 second speedup on average when displaying non dns-cached pages, (0.2 best case). Which is significative. Please let me know if there's another advantage I miss. The cost is using an FLTK mechanism (awake) which we don't in the rest of the code. It looks clean. It should be tested on different platforms/OS. BTW, dns.c has code for a blocking server (not threaded). Does this awake() call work OK when called from the same main thread? Is this by design or by chance? Maybe it would be safer to avoid calling awake() in this case. -- Cheers Jorge.-

Feb. 10, 2012

Other protocol handler dpi for mailto etc

by dennisn＠dennisn.dyndns.org

What ever happened to the (idea of a) general protocol handler for mailto: and irc: and other such links? It was mentioned once upon a time, at least, in: http://lists.auriga.wearlab.de/pipermail/dillo-dev/2004-November/002404.html

Feb. 7, 2012

[exprimental patch] wakeup instead of polling for dns

by Johannes.Hofmann＠gmx.de

Hi, attached experimental patch replaces the polling for DNS answers with the wakeup mechanism fltk provides. What do you think? Cheers, Johannes

Feb. 6, 2012

null URL

by corvid＠lavabit.com

Looking at http://starurchin.org/dillo/valgrind/14566b8d334f46a5fa74873a6ea3f68e0d418a… my guess is that something in nav stack cleaning made the const DilloUrl *url = a_History_get_url(NAV_TOP_UIDX(bw)); in a_UIcmd_page_popup() unhappy.

Feb. 6, 2012

TitanTV does not show the default channel listing

by rogerx.oss＠gmail.com

> On Wed, Feb 01, 2012 at 04:03:48PM -0900, Roger wrote: >> On Wed, Feb 01, 2012 at 09:58:35PM +0100, Johannes Hofmann wrote: >> >>Ok, if it works without Javascript then there is hope to get it >>working with Dillo. >> >>Maybe it needs cookies. Do you have them enabled for this site? > >Seems I forgot to add titantv.com. :-/ > >$ cat .dillo/cookiesrc >DEFAULT DENY >.gmail.com ACCEPT_SESSION >.google.com ACCEPT_SESSION >.wikimedia.org ACCEPT_SESSION >.titantv.com ACCEPT_SESSION > > >However, titantv.com is still somehow pulling an old cached channel listing >which is non-existant on my account/cookie. > >"fgrep $HOME/.dillorc -e titantv" shows no other locations the domain is in. > > >Well, I'll play with what I got for now since I now have cookies enabled and >see if I can figure-out why Dillo (including ELinks) is still pulling an old >non-existent channel listing. OK. I'm now pulling my *real* default listing and not some ghost listing. I had to 1) enable cookies for the titantv.com domain and 2) login from Dillo once cookies were enabled. I also just performed a login on ELinks and am now also getting my *real* default channel listing and not some ghost channel listing. (shrugs) I was logging-in all the time the other day on ELinks and I don't know why it was showing a ghost channel listing -- weird hunch they're playing with the server-end stuff too. -- Roger http://rogerx.freeshell.org/

Feb. 2, 2012

TitanTV does not show the default channel listing

by rogerx.oss＠gmail.com

> On Wed, Feb 01, 2012 at 09:58:35PM +0100, Johannes Hofmann wrote: > >Ok, if it works without Javascript then there is hope to get it >working with Dillo. > >Maybe it needs cookies. Do you have them enabled for this site? Seems I forgot to add titantv.com. :-/ $ cat .dillo/cookiesrc DEFAULT DENY .gmail.com ACCEPT_SESSION .google.com ACCEPT_SESSION .wikimedia.org ACCEPT_SESSION .titantv.com ACCEPT_SESSION However, titantv.com is still somehow pulling an old cached channel listing which is non-existant on my account/cookie. "fgrep $HOME/.dillorc -e titantv" shows no other locations the domain is in. Well, I'll play with what I got for now since I now have cookies enabled and see if I can figure-out why Dillo (including ELinks) is still pulling an old non-existent channel listing. -- Roger http://rogerx.freeshell.org/

Feb. 2, 2012

TitanTV does not show the default channel listing

by Johannes.Hofmann＠gmx.de

On Tue, Jan 31, 2012 at 09:03:09PM -0900, Roger wrote: > > On Tue, Jan 31, 2012 at 08:32:09PM +0100, Johannes Hofmann wrote: > >On Mon, Jan 30, 2012 at 02:49:54PM -0900, Roger wrote: > >> Just a quick note, having a pain in the arse viewing titantv.com channel > >> listings. > >> > >> Seems Dillo keeps pulling an unwanted channel listing page instead of my > >> default channel listing. I even deleted all other channel listings and > >> restarted Dillo (and dpid, including rebooting) and I'm still getting a > >> unwanted listing. Changed user agent too. > >> > >> ELinks seems to be doing similar, but I once got it to display my default > >> channel listing without having to delete other listings. But now even ELinks > >> is now showing the same listing that Dillo displays. > >> > >> Only browser that seems to work is Seamonkey (Mozilla) here. > >> > >> From what I see, titantv.com is using MS .asp stuff and am guessing its the > >> server screwing around and trying to push more eye candy. > > > >Can you check whether it works with Seamonkey and Javascript disabled? > > I know titantv.com uses heavy javascript. > > Turning-off javascript does have a surprising result, it still shows my > preferred default channel listing, unlike ELinks or Dillo showing some other > old cached channel listing. Ok, if it works without Javascript then there is hope to get it working with Dillo. Maybe it needs cookies. Do you have them enabled for this site? Cheers, Johannes

Feb. 1, 2012

TitanTV does not show the default channel listing

by rogerx.oss＠gmail.com

> On Tue, Jan 31, 2012 at 08:32:09PM +0100, Johannes Hofmann wrote: >On Mon, Jan 30, 2012 at 02:49:54PM -0900, Roger wrote: >> Just a quick note, having a pain in the arse viewing titantv.com channel >> listings. >> >> Seems Dillo keeps pulling an unwanted channel listing page instead of my >> default channel listing. I even deleted all other channel listings and >> restarted Dillo (and dpid, including rebooting) and I'm still getting a >> unwanted listing. Changed user agent too. >> >> ELinks seems to be doing similar, but I once got it to display my default >> channel listing without having to delete other listings. But now even ELinks >> is now showing the same listing that Dillo displays. >> >> Only browser that seems to work is Seamonkey (Mozilla) here. >> >> From what I see, titantv.com is using MS .asp stuff and am guessing its the >> server screwing around and trying to push more eye candy. > >Can you check whether it works with Seamonkey and Javascript disabled? I know titantv.com uses heavy javascript. Turning-off javascript does have a surprising result, it still shows my preferred default channel listing, unlike ELinks or Dillo showing some other old cached channel listing. Yup. Just checking ELinks and Dillo just now, are still showing an old deleted channel line-up. -- Roger http://rogerx.freeshell.org/

Feb. 1, 2012

TitanTV does not show the default channel listing

by Johannes.Hofmann＠gmx.de

On Mon, Jan 30, 2012 at 02:49:54PM -0900, Roger wrote: > Just a quick note, having a pain in the arse viewing titantv.com channel > listings. > > Seems Dillo keeps pulling an unwanted channel listing page instead of my > default channel listing. I even deleted all other channel listings and > restarted Dillo (and dpid, including rebooting) and I'm still getting a > unwanted listing. Changed user agent too. > > ELinks seems to be doing similar, but I once got it to display my default > channel listing without having to delete other listings. But now even ELinks > is now showing the same listing that Dillo displays. > > Only browser that seems to work is Seamonkey (Mozilla) here. > > From what I see, titantv.com is using MS .asp stuff and am guessing its the > server screwing around and trying to push more eye candy. Can you check whether it works with Seamonkey and Javascript disabled? Cheers, Johannes

Jan. 31, 2012

TitanTV does not show the default channel listing

by rogerx.oss＠gmail.com

Just a quick note, having a pain in the arse viewing titantv.com channel listings. Seems Dillo keeps pulling an unwanted channel listing page instead of my default channel listing. I even deleted all other channel listings and restarted Dillo (and dpid, including rebooting) and I'm still getting a unwanted listing. Changed user agent too. ELinks seems to be doing similar, but I once got it to display my default channel listing without having to delete other listings. But now even ELinks is now showing the same listing that Dillo displays. Only browser that seems to work is Seamonkey (Mozilla) here.

Jan. 30, 2012

Ping

by rogerx.oss＠gmail.com

> On Mon, Jan 30, 2012 at 09:22:15AM -0300, Jorge Arellano Cid wrote: >Hi, > > This is just a test to check mailing list delivery works OK. Yup. It is. -- Roger http://rogerx.freeshell.org/

Jan. 30, 2012

No subject

by bogus＠does.not.exist.com

server screwing around and trying to push more eye candy. -- Roger http://rogerx.freeshell.org/

Jan. 30, 2012

Ping

by jcid＠dillo.org

Hi, This is just a test to check mailing list delivery works OK. -- Cheers Jorge.-

Jan. 30, 2012

Handling of HTML entities without terminating semicolon (bug 1043)

by qartis＠gmail.com

Currently any inline HTML entity will be converted if the whole entity name matches up to a non-alnum, even if there is no trailing semicolon. For example, foo&lt,bar would be converted, but foo&ltbar wouldn't. This means that an improperly encoded url like /foo.html?a=b&lang=en would not work, because &lang would be translated to a Left Angle Bracket. But we can't just require all HTML entites to have a terminating semicolon, because that would cause worse behavior on all the broken websites that rely on lazy entity termination. Webkit has a fairly good solution for this. Their entity list (http://trac.webkit.org/browser/trunk/Source/WebCore/html/parser/HTMLEntityN…) has defined duplicate elements without trailing semicolons for some tags only. By my count, 106/2125 tags have a duplicate zero-semicolon definition. Does anybody know how Webkit chose which tags are valid without a semicolon? Is that defined in an RFC somewhere? Another option would be to require entites to have a terminating semicolon when they're part of a tag attribute. That's assuming most pages that encode their tag attributes do it properly. For example, <a href="error.cgi?msg=Unauthorized&nbsp."> wouldn't behave as it used to.

Jan. 8, 2012

Happy new year!

by jcid＠dillo.org

Hi, It is with great pleasure that I come bearing good wishes for this new year (2012). 2011 ended being very good for he Dillo project: > Hi there, > > Please receive my best wishes and thanks to the team. > > Let's hope this year will help our project thrive again, in > both the distros and the collaboration agreement we're trying to > make. > > (-: 2011 :-) As you can see, those wishes became real during the year! Hopefully 2012 will see floats and a continuation of our collaboration agreement. Ah, and it may also materialize a very good event WRT the development team. :) Cheers! -- Cheers Jorge.-

Jan. 2, 2012

Wikipedia pages

by rogerx.oss＠gmail.com

> On Sun, Dec 25, 2011 at 09:00:52PM +0100, Georg Potthast wrote: >I hope there are still plans to get Dillo to display Wikipedia pages >properly. > >The NetSurf browser, as I found, has no problems with those: >http://www.netsurf-browser.org/about/screenshots/images/gtk-wikipedia.png > >I thought that the rendering engines of Dillo and NetSurf would be >comparable and not that much different. > >Georg Try just using http://en.mobile.wikipedia.org/ or http://en.m.wikipedia.org/ as a work around? -- Roger http://rogerx.freeshell.org/

Dec. 26, 2011

Wikipedia pages

by mailbox＠georgpotthast.de

I hope there are still plans to get Dillo to display Wikipedia pages properly. The NetSurf browser, as I found, has no problems with those: http://www.netsurf-browser.org/about/screenshots/images/gtk-wikipedia.png I thought that the rendering engines of Dillo and NetSurf would be comparable and not that much different. Georg

Dec. 25, 2011