- Dillo-dev - lists.mailman3.com

Dillo on the iPAQ

by Ivan Gevirtz

I'm trying to install dillo on the iPAQ running familiar linux. I've tried it on several versions of familiar, and keep getting the same result: # ./dillo Setting locale to C Gtk-WARNING **: cannot open display: :0.0 # any hints on how to fix this? Thanks, Ivan

April 29, 2003

Re: [Dillo-dev] dillo-0.7.2 compile probs SuSE 8.1

by William Trenker

On Tue, 29 Apr 2003 21:38:36 -0500, Patrick Shanahan <paka(a)MyRealBox.com> wrote: > Having successfully compiled 0.7.0, 0.7.1 & 0.7.12, I am unable to > compile 0.7.2. I get the following errors which mean very little to > me. Any help or direction will be appreciated. > > checking png.h usability... no > checking png.h presence... yes > configure: WARNING: png.h: present but cannot be compiled > configure: WARNING: png.h: check for missing prerequisite headers? > configure: WARNING: png.h: proceeding with the preprocessor's result > configure: WARNING: ## ------------------------------------ ## > configure: WARNING: ## Report this to bug-autoconf(a)gnu.org. ## > configure: WARNING: ## ------------------------------------ ## <snip> Yes, I got that problem too when I tried to build Dillo 0.7.2 earlier today. Since it didn't work I downloaded the rpm (http://www.hyperborea.org/software/repackaged/dillo-0.7.2-1.rh7.i386.rpm) which installed but when I run Dillo it segfaults with a message saying the Dillo was built for a different version of libpng (which I've already reported on this list today). So there seems to be a common thread here. Bill -- Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/

April 29, 2003

Bad view page usiing overlib

by Diego Sáenz

I view this page bad (scroll down and see the table) http://www.riojaparty.com/riojaparty2003/modules.php?name=Content&pa=showpa… They are testing overlib, a javascript lib that is free software, so maybe the page changes fast. Before using that lib i saw the page well. The html of the table seems well to my brain, can be that dillo fail to ignore some javascript stuff or must i send a mail to the webmaster? Another question: What program do you use to test the html syntaxis of the pages? Thanks for this time. Diego.

April 29, 2003

0.7.2 installation problem

by William Trenker

If been using Dillo for several months now. I just downloaded http://www.hyperborea.org/software/repackaged/dillo-0.7.2-1.rh7.i386.rpm and installed it using this command: rpm -Uvh --nodeps dillo-0.7.2-1.rh7.i386.rpm (To be fair, I had to use the --nodeps option because without it rpm complains that I don't have libpng.so.2. But I've done other installs with this same arrangement and they work fine.) When I run dillo I get the Dillo 0.7.2 spash screen. I then click on the Change Log link and dillo loads the changelog page from the dillo web site. But when the graphic at the top of the page gets about 1/2 done dillo segfaults with this: # dillo Setting locale to C... dillo_dns_init: Here we go! Disabling cookies. Nav_open_url: Url=>about:splash< Nav_open_url: Url=>http://dillo.auriga.wearlab.de/ChangeLog.html< Dns_server [0]: dillo.auriga.wearlab.de is 0x80d33d8 Connecting to 134.102.206.165 Png_error_handling: Application was compiled with png.h from libpng-1.0.14 Segmentation fault I seem to have a problem with my libpng. Here is what I have installed: # l /usr/lib/libpng* lrwxrwxrwx 1 root root 10 Jan 8 13:59 /usr/lib/libpng.a - > libpng12.a lrwxrwxrwx 1 root root 19 Jan 8 14:42 /usr/lib/libpng.so - > libpng12.so.0.1.2.4* lrwxrwxrwx 1 root root 17 Jan 5 14:53 /usr/lib/libpng.so.2 -> libpng.so.3.1.2.1* lrwxrwxrwx 1 root root 17 Jan 8 16:10 /usr/lib/libpng.so.3 -> libpng.so.3.1.2.1* -rwxr-xr-x 1 root root 127212 Jun 10 2002 /usr/lib/libpng.so.3.1.2.1* lrwxrwxrwx 1 root root 19 Jan 8 14:42 /usr/lib/libpng.so.3.1.2.4 -> libpng12.so.0.1.2.4* -rw-r--r-- 1 root root 203154 Nov 3 00:00 /usr/lib/libpng12.a lrwxrwxrwx 1 root root 13 Jan 8 14:42 /usr/lib/libpng12.so -> libpng12.so.0* lrwxrwxrwx 1 root root 19 Jan 8 14:42 /usr/lib/libpng12.so.0 -> libpng12.so.0.1.2.4* -rwxr-xr-x 1 root root 176576 Aug 24 2002 /usr/lib/libpng12.so.0.1.2.4* Do you know where I can get a libpng that meets dillo's needs? By the way, Dillo 0.7.1 didn't have this problem, I use it all the time. Thanks for Dillo and any help you can provide, Bill -- Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/

April 29, 2003

Re: [Dillo-dev] CVS and Automake/Autoconf

by Joe Hill

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

April 28, 2003

CVS and Automake/Autoconf

by Joe Hill

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.

April 28, 2003

Re: Dillo-dev digest, Vol 1 #100 - 6 msgs

by linux seaq

> 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

April 28, 2003

Re: [Dillo-dev] Copy&Paste bug BINGO!!!

by Jorge Arellano Cid

Hi there, > > [...] > > 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); > > Should be > > DwExtIterator* a_Dw_ext_iterator_new (DwIterator **it); > ^^ Yeah, just a typo. > Another possibility would be to define wrapper macros: > > #define a_Dw_ext_iterator_new(it) _a_Dw_ext_iterator_new(&it) > > DwExtIterator* _a_Dw_ext_iterator_new (DwIterator **it); > > but that's a bit kludgy for my taste. Yep. > > The same applies for a_Dw_iterator_free() and friends. > > It should be quite obvious that functions ending on "_free", > "_destroy" etc. have a side-effect of their argument, so I'd keep this > and regard it as an exception. a_Dw_ext_iterator_new is another case, > and very confusing, so this one should be fixed. Hem, that's what I meant (rush writing). > > BTW, the > > first approach I wrote was to force a_Dw_iterator_free to make > > NULL its argument). > > Very useful for debugging! :-) That was the idea! > > Well, that's it. As I don't know the iterators code as well as > > Sebastian, I'll let him make the decisions. > > Well, first I'd use this change for the release, and fix it for > 0.7.3. A grep for "a_Dw_ext_iterator_new" shows that this is handled > correctly elsewhere. OK guys, please give Copy&Paste some hard testing. If it shows no problems, I'll probably make the final touches this Monday and try to make a release on Tuesday. > Perhaps it is cleaner to remove the side-effect completely, That sounds good and clean. > by letting > a_Dw_ext_iterator_new putting a clone on top of the stack, so that the > callers must free the simple iterator. This would be a bit slower, but > compared to that, a_Dw_ext_iterator_new creates anyway quite a couple > of DwIterator's. Has to be tested. Yes, probably 0.7.3 will require some profiling and tunning. > (Thorben: As far as I can say from a first gprof run, this is not the > critical code, so the problem is not made worse.) This problem certainly is worth an investigation. Let it be on 0.7.3. Cheers Jorge.-

April 26, 2003

Re: [Dillo-dev] Copy&Paste bug BINGO!!!

by Sebastian Geerken

On Fri, Apr 25, Jorge Arellano Cid wrote: > After some hard code revisions, I decided to "eat" Valgrind's > manual, install it and BINGO!: > > [...] > > 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); Should be 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. Another possibility would be to define wrapper macros: #define a_Dw_ext_iterator_new(it) _a_Dw_ext_iterator_new(&it) DwExtIterator* _a_Dw_ext_iterator_new (DwIterator **it); but that's a bit kludgy for my taste. > The same applies for a_Dw_iterator_free() and friends. It should be quite obvious that functions ending on "_free", "_destroy" etc. have a side-effect of their argument, so I'd keep this and regard it as an exception. a_Dw_ext_iterator_new is another case, and very confusing, so this one should be fixed. > BTW, the > first approach I wrote was to force a_Dw_iterator_free to make > NULL its argument). Very useful for debugging! :-) > Well, that's it. As I don't know the iterators code as well as > Sebastian, I'll let him make the decisions. Well, first I'd use this change for the release, and fix it for 0.7.3. A grep for "a_Dw_ext_iterator_new" shows that this is handled correctly elsewhere. Perhaps it is cleaner to remove the side-effect completely, by letting a_Dw_ext_iterator_new putting a clone on top of the stack, so that the callers must free the simple iterator. This would be a bit slower, but compared to that, a_Dw_ext_iterator_new creates anyway quite a couple of DwIterator's. Has to be tested. (Thorben: As far as I can say from a first gprof run, this is not the critical code, so the problem is not made worse.) Sebastian

April 26, 2003

Copy&Paste bug BINGO!!!

by Jorge Arellano Cid

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.-

April 25, 2003