- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev] Sigh... CVS out of sync 5 minutes after I upload my patches

by Frank de Lange

Indan Zupancic wrote: >>It is as if the devil plays with it... 5 minutes after I upload my >>latest patches, CVS gets updated and the patch does not apply anymore. >> >> > >Oh so funny. :) > >I saw you updated your patch today, and the first thing I thought was: Why >didn't Frank say that he used the CVS version of yesterday, because CVS >didn't change for two days, and _if_ CVS would be updated today, there >will be problems. > > That is what the website is for. The name on the patch is the date it is released, the CVS date is the last date CVS is changed. The patch is always based on CVS for the release date (right before I release a patch I update CVS and reconcile any differences, if the differences are big I delay the patch to test the changes), but CVS is not always updated on that day (as you noticed). FYI, the CVS date/time I give is the date/time stamp for the src/ directory. Using CVS features is nice, but have you noticed how many (or few...) people even know how to get source from CVS in the first place? I want the patch to get exposure to as many people as possible, and including the source on which it is based is one way of doing that. Another way would be to release a patched source tree, but I prefer the patch + clean approach. >Could you reply to the CVS offer from Trent Jarvi and say what you think >about a CVS branch? See: > >http://lists.auriga.wearlab.de/pipermail/dillo-dev/2003-November/001470.html > > I noticed that posting, have been busy lately hence did not post reply. As there are only two active patch developers at this moment I think we do not need a CVS branch. It should be possible to synchronise releases for two patches. Of course, having a branch on the main Dillo CVS or elsewhere would not hurt, and I do not oppose it. I just think we can do without it for the moment. Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

Nov. 23, 2003

Re: [Dillo-dev] [PATCH] Version 15 of tab/frame/kbnav/etc patch, now with keywords and more...

by Frank de Lange

Paul Pelzl wrote: >On Fri, Nov 21, 2003 at 04:46:31PM +0100, Frank de Lange wrote: > > >>The next version (15) of my tab/frame/keyboard navigation patch is >>ready (see bottom of this message for download instructions). >> >> > >I've got a couple of early bugs to report: > >1) when using kbnav to move from a link to an entry box in a form, the > most recently selected link is not properly deselected. Example: > google.com. > > If this is what I think it is, it actually seems to be a widget bug. I've discussed this with Sebastian before. Relevant part of discussion plus link to bug description: >Another corner case (to be tried with the new patch): >> >> - highlight a link which is followed by an embedded GTK+ widget, >> so that the embedded widget is just outside the visible part of the >> document. >> - move focus to the embedded widget. The widget scrolls into view and >> is focused, >> as it should. >> - but... the link still appeats highlighted. This even though the >> unhighlight >> function is called. >> - cover the 'highlighted' link with another window. When it is >> uncovered, the >> highlight is gone - as it should. >> >> It appears the page is not redrawn correctly when focus is handed over >> to an embedded GTK+ widget. > > I'm not sure, but I assume that this is the another symptom of a known bug (#202). More on this bug at http://www.geocrawler.com/archives/3/702/2002/3/50/8045756/ It might be something else though (I've noticed this many times, and just thought of it as the described widget bug) so I'll have a look at it. >2) the "keywords" user pref is apparently not read in a robust manner. > If I separate keywords by a newline instead of a space, I can > generate a segfault upon typing any URL in the location bar. > Hmmm, I'll have a look at that... Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

Nov. 23, 2003

Re: [Dillo-dev] Confused

by Indan Zupancic

Hi, > Can someone tell me where to download a source with the https, tab patches > without having to track down what and or where to patch? Or send me a > copy:) thanks Currently I have a sort of installer that downloads Dillo, the tabs and https patches, and applies them, Can get it from: http://www.xs4all.nl/~dorinek/dillo/install-dillo.sh But it isn't finished yet, nor up to date (Frank has a newer tabs+frames patch). I have a better installer, but I need to get my https patch working first. It stopped working after I changed it a bit :(. There is a weird bug that I can't seem to find, will hopefully fix it today and then update my https patch and the installer. Greetings, Indan

Nov. 23, 2003

Confused

by Donald McKnight

Can someone tell me where to download a source with the https, tab patches without having to track down what and or where to patch? Or send me a copy:) thanks Thanks Donald ----- Original Message ----- From: "Indan Zupancic" <indan(a)nul.nu> To: <dillo-dev(a)dillo.org> Sent: Saturday, November 22, 2003 10:18 AM Subject: [Dillo-dev] [patch] Cleanup of IO.c > Hi, > > Just cleaned IO.c up, so nothing fundamental changed. Besides some > cleanups in multiple functions this are the two major changes: > > Removal of a_IO_write_chunks, mainly because it isn't used anywhere and it > doesn't add any real functionality, it does the same only in a different > way. Thanks to CVS it is easy to add it back when you really needed, if > you want. > > Removal of the ValidIOs list. This is the logical next step after > g_source_remove was discovered, because the extra indirection isn't needed > anymore. > > All in all the result is a much smaller IO.c (400 lines instead of 700), > smaller Dillo (about 3K with -O2), and O(1) behaviour instead of O(n) in > the IO engine. > > The changes have really no negative points as far as I can tell. > > Link to the patch: > > http://www.xs4all.nl/~dorinek/dillo/IO-cleanup.patch.gz > > Greetings, > > Indan > > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev(a)lists.auriga.wearlab.de > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

Nov. 23, 2003

[patch] Cleanup of IO.c

by Indan Zupancic

Hi, Just cleaned IO.c up, so nothing fundamental changed. Besides some cleanups in multiple functions this are the two major changes: Removal of a_IO_write_chunks, mainly because it isn't used anywhere and it doesn't add any real functionality, it does the same only in a different way. Thanks to CVS it is easy to add it back when you really needed, if you want. Removal of the ValidIOs list. This is the logical next step after g_source_remove was discovered, because the extra indirection isn't needed anymore. All in all the result is a much smaller IO.c (400 lines instead of 700), smaller Dillo (about 3K with -O2), and O(1) behaviour instead of O(n) in the IO engine. The changes have really no negative points as far as I can tell. Link to the patch: http://www.xs4all.nl/~dorinek/dillo/IO-cleanup.patch.gz Greetings, Indan

Nov. 22, 2003

patch: toggle force_my_colors

by Paul Pelzl

Hi folks, I have attached a pair of Dillo patches. They add a toggle-able menu item to the right-button context menu, which enables/disables the force_my_colors preference on the fly. I find this useful for some occasional websites that Dillo renders rather poorly, as well as sites that simply choose a horrid color scheme. The first patch applies on top of Frank's monster patch v15; the second applies against Dillo CVS of 20031119. I hope others find it useful. Paul

Nov. 22, 2003

Re: [Dillo-dev] [PATCH] Version 15 of tab/frame/kbnav/etc patch, now with keywords and more...

by Paul Pelzl

On Fri, Nov 21, 2003 at 04:46:31PM +0100, Frank de Lange wrote: > The next version (15) of my tab/frame/keyboard navigation patch is > ready (see bottom of this message for download instructions). I've got a couple of early bugs to report: 1) when using kbnav to move from a link to an entry box in a form, the most recently selected link is not properly deselected. Example: google.com. 2) the "keywords" user pref is apparently not read in a robust manner. If I separate keywords by a newline instead of a space, I can generate a segfault upon typing any URL in the location bar. Other than that, it's looking pretty solid so far. I'll keep hammering... Paul

Nov. 22, 2003

Re: [Dillo-dev] Sigh... CVS out of sync 5 minutes after I upload my patches

by Indan Zupancic

> It is as if the devil plays with it... 5 minutes after I upload my > latest patches, CVS gets updated and the patch does not apply anymore. Oh so funny. :) I saw you updated your patch today, and the first thing I thought was: Why didn't Frank say that he used the CVS version of yesterday, because CVS didn't change for two days, and _if_ CVS would be updated today, there will be problems. So, the easy sollution is to rename your patch to "dillo-20031120-tabs.patch.gz" and everything will work just fine. Although not with the current CVS of course, but that will happen anyway if you wait long enough. Are the recent changes worth it to update your patch, that's the question. Apparently you already did that, but there is no reason to include the whole source. What Thorben said is exactly what my installer script does; the -D option is nice. Offtopic: Could you reply to the CVS offer from Trent Jarvi and say what you think about a CVS branch? See: http://lists.auriga.wearlab.de/pipermail/dillo-dev/2003-November/001470.html I think a seperate CVS branch would be useful, one hosted by the same server as the official Dillo CVS, or an external one, like Trent's. Where doesn't matter much I guess. At least the patches are centralized then.

Nov. 21, 2003

Re: [Dillo-dev] Sigh... CVS out of sync 5 minutes after I upload my patches

by Thorben Thuermer

On Fri, 21 Nov 2003 22:41:25 +0100 Frank de Lange <frank(a)unternet.org> wrote: > Anyone who wants to try tab/frame patch no. 15 can download the CVS > source on which it is based from the patch site: > > http://www.geocities.com/ikbenfrank/dillo-200311191409.tar.gz cvs -d :pserver:anonymous@dillo.org:/sfhome/cvs/dillo co -D 'Nov 19 14:09 2003' dillo why not use CVS's features? </offtopic>, before we get a new development model discussion > Cheers//Frank Thorben

Nov. 21, 2003

Sigh... CVS out of sync 5 minutes after I upload my patches

by Frank de Lange

It is as if the devil plays with it... 5 minutes after I upload my latest patches, CVS gets updated and the patch does not apply anymore. Well, this time I was prepared for that. Anyone who wants to try tab/frame patch no. 15 can download the CVS source on which it is based from the patch site: http://www.geocities.com/ikbenfrank/dillo-200311191409.tar.gz This is the UNPATCHED CVS code that was used to create the patch. To get a working tabbed/frames/etc browser, download this tarball and the patch, then apply the patch. From now on I will always include the CVS code on which the patch was based to get around these irritating synchronisation problems. Indan, if you want to update your HTTPS patch, may I suggest you use this source code tarball instead of current Dillo CVS? This way we are less dependent on the state of CVS... Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

Nov. 21, 2003