- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev] CVS and Automake/Autoconf
by Joe Hill April 28, 2003

April 28, 2003

On Mon, 28 Apr 2003 23:48:00+0200 Andreas Schweitzer<Andreas.Schweitzer(a)hs.uni-hamburg.de> wrote: > Hi, > > Since it probably won't help a lot, I only answer off-list. > > > I am running autoconf2.5-2.53 and automake1.6-1.63-1 on Mandrake > > 9.0. > > I'm compiling dillo with older and newer version of autoconf and > automake. So it *should* be ok. Yet, I do suspect some issues > with your version. > Thanks, you were right on the money. One thing, I tried to reply to you directly, but it looks like you have even more restrictive mail filters than I do ;), so I'll send this to the list just to mainly say thanks for the tip. I reinstalled automake and autoconf, this time with ./configure--prefix=/usr (seems I have to do that with everything on Mandrake), and all seems well. > > Any help appreciated, I am really excited about this project. I have > > .70 running now, but...I would like some of the bugfixes and newer > > featues. I tried installing dillo-0.7.1.2 from source, but it > > crashed every time I hit the bookmark icon. I'll keep on pushin' > > ahead if someone can give > > That should NOT be the case. You probably still have an old > bm_srv running. Exit all dillos, then do a "ps ax | egrep bm_srv" > and kill all matching processes. Then make sure, you have the latest > bm_srv12 (of dillo-0.7.1.2) in your $PATH, and no OLD bm_srv12 in > ~/.dillo/ > but why would that not happen with the rpm version that I installed? and it doesn't happen with the CVS ver I just built either. bah, it doesn't matter, anyway, right? It's workin now! Once again, thanks, this is one fantastic browser. The bookmark process isn't my favourite (no pun intended), but I think I saw on the list archives that might be changing? ++++ ..| ..\ /\ / ../ ..\ /\ / ../ ..| ..\/ ..\/ ../ ..\/ \/ ../ _| __/\__/ __/\__/ "The philosophers have only interpreted the world in various ways. The point however is to change it." -Chuck Marx > HTH, > Andreas > > -- > **************************** NEW ADDRESS > ****************************** Hamburger Sternwarte > Universitaet Hamburg > Gojenbergsweg 112 Tel. ++49 40 42891 > 4016 D-21029 Hamburg, Germany Fax. ++49 40 > 42891 4198

1 0

CVS and Automake/Autoconf
by Joe Hill April 28, 2003

April 28, 2003

When I try to build from CVS, ./autogen.sh gives me the following errors: aclocal: configure.in: 50: macro `AM_PATH_GTK' not found in library autoheader: error: AC_CONFIG_HEADERS not found in configure.in configure.in:8: error: possibly undefined macro: AM_INIT_AUTOMAKE If this token and others are legitimate, please use m4_pattern_allow. See the Autoconf documentation. configure.in:9: error: possibly undefined macro: AM_CONFIG_HEADER configure.in:29: error: possibly undefined macro: AM_PROG_CC_STDC configure.in:50: error: possibly undefined macro: AM_PATH_GTK configure.in:8: no proper implementation of AM_INIT_AUTOMAKE was found, configure.in:8: probably because aclocal.m4 is missing... configure.in:8: You should run aclocal to create this file, then configure.in:8: run automake again. configure.in: installing `./install-sh' configure.in: installing `./mkinstalldirs' configure.in: installing `./missing' configure.in:6: installing `./config.guess' configure.in:6: installing `./config.sub' configure.in:9: required file `./config.h.in' not found dpi/Makefile.am: installing `./depcomp' /usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/lang-compile.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/lang-compile.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/lang-compile.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/lang-compile.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/lang-compile.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/depend2.am: AMDEP does not appear in AM_CONDITIONAL/usr/local/share/automake-1.6/am/lang-compile.am: AMDEP does not appear in AM_CONDITIONAL a configure script is built, but of course it doesn't work: ./configure: line 1351: syntax error near unexpected token `dillo,' ./configure: line 1351: `AM_INIT_AUTOMAKE(dillo, 0.7.1)' I tried running aclocal as advised in the error message, however, more errors: aclocal: configure.in: 50: macro `AM_PATH_GTK' not found in library Since I have successfully built from CVS with other software using an autogen script, I am wondering if maybe this is just a version issue? I am running autoconf2.5-2.53 and automake1.6-1.63-1 on Mandrake 9.0. Any help appreciated, I am really excited about this project. I have .70 running now, but...I would like some of the bugfixes and newer featues. I tried installing dillo-0.7.1.2 from source, but it crashed every time I hit the bookmark icon. I'll keep on pushin' ahead if someone can give me a bit of a nudge about this autogen.sh thing. Thanks all, -J.

1 0

Re: Dillo-dev digest, Vol 1 #100 - 6 msgs
by linux seaq April 28, 2003

April 28, 2003

> Onto another topic: > > A last minute leak fix (iterators) left dillo a bit unstable. > We knew the iterators had a leak and it's great to have found it. > Now it will take some extra time to try to fix the problem and, > after that, stress-test the Copy&Paste code again. > > We haven't got an accurate procedure to reproduce the crash > (yet). This explanation may help: > > <q> > I think I have found a relatively new segfault bug : > You need some sort of thumbtags like the screenshots on > http://www.incagold.com/products/Speed/speedthief.php > > Now you have to move down the mouse and at the same time click > on one image (i.e. click while still aiming for the image). > If the button-push occures outside the image and the button-release > inside the image, chances are high to get a segfault. This > is a bit hard to reproduce, but I couldn't get an older version > with selection to behave like that. It also happened a few > times to me last night (after the last 'cvs up') but never before. > </q> > HI i'm having that problem when the page is not entirely downloaded and i'm scrolling down the page and selecting some text, it segfaults.. it happens sometimes not always, so the conditions may vary.. i`m using the cvs from 06/04 when i used an earlier cvs update (just after select cvs commit) it ocurred*a lot more times* than now. --- Andres Mauricio Mujica SEAQ SERVICIOS CIA LTDA www.seaq.com.co

1 0

Copy&Paste bug BINGO!!!
by Jorge Arellano Cid April 26, 2003

April 26, 2003

Hi there, After some hard code revisions, I decided to "eat" Valgrind's manual, install it and BINGO!: <q> ==6797== ==6797== Invalid memory access of size 4 ==6797== at 0x80662B3: a_Selection_button_press (selection.c:142) ==6797== by 0x805EB8F: Dw_page_send_selection_event (dw_page.c:1486) ==6797== by 0x805EE47: Dw_page_button_press (dw_page.c:1499) ==6797== by 0x806474D: Dw_widget_mouse_event (dw_widget.c:740) ==6797== Address 0x40D714F0 is 0 bytes inside a block of size 52 free'd ==6797== at 0x40148050: free (in /usr/local/lib/valgrind/valgrind.so) ==6797== by 0x403C1FE8: g_free (in /usr/lib/libglib-1.2.so.0.0.10) ==6797== by 0x8057CA1: a_Dw_ext_iterator_new (dw_ext_iterator.c:171) ==6797== by 0x8066286: a_Selection_button_press (selection.c:136) </q> What does this mean? As a_Dw_ext_iterator_new(it) may _free_ 'it', 'it' should no longer be used after that call. That was the case with the page (right-click test case). I commited a short _code_change_ that avoids the problem. It is not a "patch" but just illustrative code. I believe the functions that may have side effects on their arguments should use an API that somehow reflects that. For instance: DwExtIterator* a_Dw_ext_iterator_new (DwIterator *it); and be used like: a_Dw_ext_iterator_new(&it); so 'it' can be NULL at return time. The same applies for a_Dw_iterator_free() and friends. BTW, the first approach I wrote was to force a_Dw_iterator_free to make NULL its argument). Well, that's it. As I don't know the iterators code as well as Sebastian, I'll let him make the decisions. Hope this helps Happy :) Jorge.-

2 2

Bookmarks
by Lars Clausen April 25, 2003

April 25, 2003

Are there any plans to reintroduce a bookmarks menu? Any patches for it? While the page looks nifty for the editing, I really don't like getting an extra bookmark page into my history stack every time I want to go to a bookmark. In other news: Yay! Cut-and-paste! One of my main reasons to use other browsers is now gone! Thanks! Thanks, -Lars -- Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor "I do not agree with a word that you say, but I |---------------------------- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbasket?

4 3

Delayed release
by Jorge Arellano Cid April 25, 2003

April 25, 2003

Hi there, I just commited the code that cleanups the bookmarks server. It works like this: <q> /* * If this is our first run, kill the (old) bookmarks server, * and create the '~/.dillo/.072' file to flag what we did. */ static void Dillo_check_bm_srv12(char *dir) [...] </q> Yes, it uses the "[b]m_srv12" trick and some other precautions. Please test it: the whole point is that when installing the new version of dillo (072), the old bookmarks server may keep running and survive the updated. The new code detects the first dillo-072 run and issues a kill command to "bm_srv12". Onto another topic: A last minute leak fix (iterators) left dillo a bit unstable. We knew the iterators had a leak and it's great to have found it. Now it will take some extra time to try to fix the problem and, after that, stress-test the Copy&Paste code again. We haven't got an accurate procedure to reproduce the crash (yet). This explanation may help: <q> I think I have found a relatively new segfault bug : You need some sort of thumbtags like the screenshots on http://www.incagold.com/products/Speed/speedthief.php Now you have to move down the mouse and at the same time click on one image (i.e. click while still aiming for the image). If the button-push occures outside the image and the button-release inside the image, chances are high to get a segfault. This is a bit hard to reproduce, but I couldn't get an older version with selection to behave like that. It also happened a few times to me last night (after the last 'cvs up') but never before. </q> We'll be working on it. Cheers Jorge.-

4 3

Cookie locks?
by Lars Clausen April 25, 2003

April 25, 2003

I understand unrestricted access to cookie jars can be bad for your health, but it's annoying that you can't have two Dillos using cookies at the same time. I usually have a Dillo running permanently, and when I get a URL in my mail, I can click on it to open another Dillo (quickly!) -- but that Dillo can't use cookies, which is frequently (e.g. in bugzilla) necessary. Could we do a write-lock on it? -Lars -- Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor "I do not agree with a word that you say, but I |---------------------------- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbasket?

3 3

[PATCH] Pth library support for threads
by Patrice Mandin April 24, 2003

April 24, 2003

Hello dillo coders, I made a simple patch to add pth support for threads for Dillo. So platforms that do not have the pthread library can also use it. I am not very clever regarding thread programming, it may needs a bit more of work. I tested it on very simple pages and it worked for me. -- Patrice Mandin WWW: http://membres.lycos.fr/pmandin/ Programmeur Linux, Atari Spécialité: Développement, jeux

1 0

Vera
by Andreas Zwinkau April 23, 2003

April 23, 2003

ave There is this new truetype font Bitstream Vera, which looks very nice IMHO. Example: http://www.bitstream.com/categories/products/fonts/vera/ Download: http://ftp.gnome.org/pub/GNOME/sources/ttf-bitstream-vera/1.10/ Is it possible to use it in Dillo? I tried to add this in the dillorc vw_fontname="Bitstream Vera Sans" fw_fontname="Bitstream Vera Sans Mono" But then everything had the same font size. <H1> the same size and weight as <H§3> etc. -- Andreas Zwinkau | web: andi.dasstellenwirinsinternet.de | mail: andi(a)buxach.de | jabber: beza1e1(a)amessage.de

4 6

Unix domain sockets
by Jorge Arellano Cid April 23, 2003

April 23, 2003

Hi there! This question is for Unix domain sockets savvy guys. When a server that listens on a UDS has started and its socket's filename is removed from the filesystem. Is it notified somehow, maybe with a SIGPIPE signal or something? or not? I ask because when dillo-0.7.2 is out, the old bm_srv12 MUST be removed (but it may remain sleeping in memory when the binary is updated). It'd be good to be able to remove it indirectly by making it react to an 'unlink' call on its socket (well, from this release and on). For the 0.7.1.2 bm_srv12, I'm considering `fuser -k bm_srv12` and a hackish: <dpi cmd='bye' (without the closing ">" an exit(1) is forced!) Anyone? :-) Cheers Jorge.-

11 15