July 2010 - Dillo-dev - lists.mailman3.com

[patch]: fix .hgignore

by onepoint＠starurchin.org

corvid wrote: > Jeremy wrote: > > The problem was that .hgignore contains a line "dillo", which > > ignores every file with the string "dillo" anywhere in its path. > > Many other lines in .hgignore are similarly inappropriate. It > > looks as though these lines were written to be interpreted as > > glob-style patterns, but in fact mercurial interprets every line > > as a regexp. > > Looking in http://www.selenic.com/mercurial/hgignore.5.html it seems > that specifying "syntax: glob" is also a possibility. Yes, glob syntax might well be better, but either way the current .hgignore is not working as it should and could do with fixing. Regards, Jeremy Henty

July 31, 2010

[patch]: fix .hgignore

by corvid＠lavabit.com

Jeremy wrote: > I was freaking out a bit for a while there because I had created some > files with names like dillo_hg_* and "hg status" was not reporting > them as untracked, even though I had not "hg add"-ed them. > > The problem was that .hgignore contains a line "dillo", which ignores > every file with the string "dillo" anywhere in its path. Many other > lines in .hgignore are similarly inappropriate. It looks as though > these lines were written to be interpreted as glob-style patterns, but > in fact mercurial interprets every line as a regexp. Looking in http://www.selenic.com/mercurial/hgignore.5.html it seems that specifying "syntax: glob" is also a possibility.

July 31, 2010

[patch]: fix .hgignore

by onepoint＠starurchin.org

I was freaking out a bit for a while there because I had created some files with names like dillo_hg_* and "hg status" was not reporting them as untracked, even though I had not "hg add"-ed them. The problem was that .hgignore contains a line "dillo", which ignores every file with the string "dillo" anywhere in its path. Many other lines in .hgignore are similarly inappropriate. It looks as though these lines were written to be interpreted as glob-style patterns, but in fact mercurial interprets every line as a regexp. So here's patch to make .hgignore a little more sensible. Regards, Jeremy Henty

July 30, 2010

[patch]: Capi_filters_allow(): improve the logging

by corvid＠lavabit.com

Jeremy wrote: > Another skirmish in my eternal war against uninformative log messages! > :-) Patch attached. Rearranged slightly and committed. > BTW, dillo's loging is quite a mischmasch of different formats. Would > there be any interest in patches to make things more consistent: > eg. by always having the function name at the beginning of MSG calls? > That would help debugging, I think. I, for one, would at least be open to the idea. I see that compiling png.c gives me warnings now. I'll have to see what that's all about...

July 30, 2010

[patch]: Capi_filters_allow(): improve the logging

by onepoint＠starurchin.org

Another skirmish in my eternal war against uninformative log messages! :-) Patch attached. BTW, dillo's loging is quite a mischmasch of different formats. Would there be any interest in patches to make things more consistent: eg. by always having the function name at the beginning of MSG calls? That would help debugging, I think. Regards, Jeremy Henty

July 30, 2010

On startup, go to websites.

by rogerx＠sdf.org

On Wed, Jul 21, 2010 at 01:19:25PM +0200, Johannes Hofmann wrote: >you can pass multiple URLs on the command line, e.g. > >dillo dpi:/bm/ http://www.google.com http://forecast.weather.gov/MapClick.php?zoneid=AKZ223 > >and if you don't want to type that all the time you can do something >like: > >alias dillo="dillo dpi:/bm/ http://www.google.com http://forecast.weather.gov/MapClick.php?zoneid=AKZ222" This is weird. I setup the alias within $HOME/.bashrc, but dmenu completely ignores it and just runs /usr/bin/dillo instead. Executing dillo from terminal using Bash does execute the alias. Guess DMenu doesn't realize it was executed through a Bash shell (anymore). I'll create my own $HOME/bin/dillo-whatever script as usual instead.

July 30, 2010

[patch]: fix VPATH builds

by onepoint＠starurchin.org

Jorge Arellano Cid wrote: > On Sat, Jul 24, 2010 at 04:26:35AM +0100, Jeremy Henty wrote: > > > > What's the decision on this? I've not heard anythng. > > Committed. > > Sorry for the delay. I've had little spare time, and when the > spot arrived, the hg server was down. No problem. Personally I am amazed and impressed that anyone with a family can find any spare time to contribute at all! Jeremy Henty

July 26, 2010

[patch]: fix VPATH builds

by jcid＠dillo.org

On Sat, Jul 24, 2010 at 04:26:35AM +0100, Jeremy Henty wrote: > > What's the decision on this? I've not heard anythng. Committed. Sorry for the delay. I've had little spare time, and when the spot arrived, the hg server was down. -- Cheers Jorge.-

July 25, 2010

Can't copy/paste, only urls

by thomas-forum＠orgis.org

Am Fri, 23 Jul 2010 17:33:37 -0800 schrieb Roger <rogerx(a)sdf.org>: > I now changed it to: > #middle_click_drags_page=NO > > Reran Dillo and I can now copy/paste into Dillo's web page forms without > the output from the copy buffer always going into the Dillo url entry box! Uh. That's collateral damage from a change I prompted ... I have to say sorry for actually seeing this misbehaviour but somehow managing not to raise the issue. As the user who wanted the pasting into the general window area to cause loading of an URL, as it happens to work in other browsers, I want to emphasize that I also think that this URL loading should happen only if there was no other sensible area for pasting underneath the cursor. This is how it usually works: Pasting into some form field causes insertion of the text there, pasting without anything special there means pasting to the browser as such, which triggers loading of the text as URL. Is this tricky to achieve with FLTK? I don't have any time to devote to this myself (apart from testing a bit), nor do I have experience with the dillo code. I have hope that it's not seriously complicated, but if the focus management happens to explicitly not anticipate such situations, it might non-trivial. But again, this is how other browsers work -- it would suit dillo to fulfill such basic expectations. Alrighty then, Thomas. PS: Sorry again, for not settling this one earlier...

July 24, 2010

[patch]: fix VPATH builds

by onepoint＠starurchin.org

What's the decision on this? I've not heard anythng. Regards, Jeremy Henty On Sun, Jul 18, 2010 at 10:39:21AM +0100, Jeremy Henty wrote: > > OK, the problem with VPATH builds is fixed but I think it can be > improved. > > Firstly, because -I is a preprocessor directive, it should be added to > AM_CPP_FLAGS. > > Secondly, the Makefile uses .. in various places to navigate from the > current build directory to where various headers and libraries are. I > think it is better to locate other directories from the top of the > build and source trees. So I would prefer to use -I$(top_srcdir) > instead of -I$(srcdir)/.. , and $(top_builddir)/path/to/lib instead of > ../path/to/lib . > > Patches attached, tested as before with in-source and external builds. > > Regards, > > Jeremy Henty > # HG changeset patch > # Parent fec840b80fd3e19e99b9b0e3a56e7be1290a6fa5 > > diff -r fec840b80fd3 src/Makefile.am > --- a/src/Makefile.am Sat Jul 17 20:18:15 2010 +0100 > +++ b/src/Makefile.am Sat Jul 17 20:23:48 2010 +0100 > @@ -1,9 +1,10 @@ > AM_CPPFLAGS= \ > + -I$(top_srcdir) \ > -DDILLO_SYSCONF='"$(sysconfdir)/"' \ > -DDILLO_DOCDIR='"$(docdir)/"' \ > @LIBJPEG_CPPFLAGS@ > AM_CFLAGS = @LIBPNG_CFLAGS@ > -AM_CXXFLAGS = -I$(srcdir)/.. @LIBPNG_CFLAGS@ @LIBFLTK_CXXFLAGS@ > +AM_CXXFLAGS = @LIBPNG_CFLAGS@ @LIBFLTK_CXXFLAGS@ > > SUBDIRS = IO > > # HG changeset patch > # Parent d813265f53f1d3b2e0ae23f7dd88f4fed2ed2361 > diff -r d813265f53f1 -r 000ab53976d1 src/Makefile.am > --- a/src/Makefile.am Sat Jul 17 20:23:48 2010 +0100 > +++ b/src/Makefile.am Sat Jul 17 20:40:06 2010 +0100 > @@ -11,13 +11,13 @@ > bin_PROGRAMS = dillo > > dillo_LDADD = \ > - ../dlib/libDlib.a \ > - ../dpip/libDpip.a \ > + $(top_builddir)/dlib/libDlib.a \ > + $(top_builddir)/dpip/libDpip.a \ > IO/libDiof.a \ > - ../dw/libDw-widgets.a \ > - ../dw/libDw-fltk.a \ > - ../dw/libDw-core.a \ > - ../lout/liblout.a \ > + $(top_builddir)/dw/libDw-widgets.a \ > + $(top_builddir)/dw/libDw-fltk.a \ > + $(top_builddir)/dw/libDw-core.a \ > + $(top_builddir)/lout/liblout.a \ > @LIBJPEG_LIBS@ @LIBPNG_LIBS@ @LIBFLTK_LIBS@ @LIBZ_LIBS@ @LIBICONV_LIBS@ > > dillo_SOURCES = \ > _______________________________________________ > Dillo-dev mailing list > Dillo-dev(a)dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

July 24, 2010