- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev]Frame support in CVS

by Kelson Vibber

On Saturday 21 February 2004 5:19 am, Andreas Kemnade wrote: > I have tried to update Frank's patch. It is not well tested or cleaned > up yet but I simply need to take a deep breath. Thanks! I tried to do that a week or so ago, but didn't get far enough for it to be useful. The patch you posted works so far here (minus obvious things like the incomplete splash page). I've posted some test RPMs in the usual spot, if anyone's interested. > The patch has one big problem. It is too big. Not that it bloats dillo, > but it is hard to review such a big patch. Perhaps it can be split up > and then be committed piece by piece. Has anyone ideas how to split it? > Perhaps in the DilloDoc/BrowserWindow separation, the kbnav thing and > frame support? Agreed. My suggestion would be to pare it down to just what's needed for frames and tabs. Rip out the kbnav changes, the splash screen changes, and everything else. Once that's done, maybe separate the pieces further just to make the pieces easier to review. Maybe three chunks: the doc/window split, frames support, and tabs support. > So what's the way to get frames into cvs? I suspect that the smaller and easier-to-review the patch is, the better chance it'll have of getting accepted. (But I'm just an interested end-user and packager, not a core developer, so what do I know?) > I would not like to see Frank's effort lost. Neither would I! -- Kelson Vibber www.hyperborea.org

Feb. 23, 2004

Re: [Dillo-dev]Italian help translation

by Lupino

Jorge Arellano Cid wrote: > > Hi, list! > > I'm a new user of Dillo, and I found it GREAT! It's handy at > > home, my PII-400 likes it; at the university is EVEN BETTER (we > > have PI-133, so you can imagine how sluggish is netscape..). > > Yes I can! > > Thanks for the encouraging. > > Would you mind telling us more about the lab with PI-133 you're > deploying dillo at? It's a lab in which they install Windows in good computers, and Linux is installed only in OLD machines, namely all PI-133. So Linux is used only by a few people, including me (it's entertaining to round up a good and fast system with a PI-133, especially when you don't have root privileges ;) So I use wmaker/dillo, I have a 20MB home, but I use /tmp to compile my programs, and it works because I ./configure --prefix=/home/../me all the programs/libs. Thanks for your interest and your efforts. :) > > I'm writing to ask if I can translate Dillo help in Italian: is > > there someone else that is translating in italian? If there > > isn't, which files should I translate, and where should I send them? > > Get: > > http://www.dillo.org/dillo-help.html.in > > and send me the translated version. I'm going to download it! I'll send it translated as soon as I have time to do it (in this period I have exams...). -- [ Andrea Spadaccini a.k.a. Lupino / Lupin85 - from Catania ] [ MAIL: lupin85(at)email.it - ICQ#: 91528290 - Linux 2.6.3 ] [ Linux Registered User #313388 - Powered by Slackware 9.1 ]

Feb. 22, 2004

Re: [Dillo-dev]0.8.0 issue

by Jorge Arellano Cid

On Sat, 14 Feb 2004, Johannes Hofmann wrote: > Some more info on the BSD dpid problem: > Actually, there is no need for any changes in the dillo sources. The problem > comes from dpid beeing linked with the "-pthread" flag, which has strange > consequences on the inherited file descriptors. If "-pthread" is used, one > cannot close the inherited file descriptors anymore (one gets EBADF). > I don't know, if this is a bug or a feature of the BSD thread implementation. > > To avoid the problem, one can simple link dpid manually without the > "-pthread" flag and everything works fine, at least for me. Very interesting. AFAIU gcc has a way to set its behaviour with respect to the -pthread flag; that way it can be properly set for each OS. Now, in my experience BSD developers are very competent with regard to technical issues, so if it was done on purpose there may be good reasons behind. If that's the case (and not a misconfigured gcc), we can add a check in our configure.in. Cheers Jorge.-

Feb. 22, 2004

Re: [Dillo-dev]Italian help translation

by Jorge Arellano Cid

On Thu, 12 Feb 2004, Andrea Spadaccini wrote: > Hi, list! > I'm a new user of Dillo, and I found it GREAT! It's handy at > home, my PII-400 likes it; at the university is EVEN BETTER (we > have PI-133, so you can imagine how sluggish is netscape..). Yes I can! Thanks for the encouraging. Would you mind telling us more about the lab with PI-133 you're deploying dillo at? > > I'm writing to ask if I can translate Dillo help in Italian: is > there someone else that is translating in italian? If there > isn't, which files should I translate, and where should I send them? Get: http://www.dillo.org/dillo-help.html.in and send me the translated version. > Thanks for you great work! :-) Cheers Jorge.-

Feb. 22, 2004

Re: [Dillo-dev]progress on AIX?

by Jorge Arellano Cid

On Sat, 14 Feb 2004, Riccardo wrote: > Hello, > > I posted the buildlog I got on AIX... did someone had a look at it to > see if there are some warnings or errors that could explain the > brokeness of dillo on aix? Of course the right way to fix the socklen_t issue is to have: TYPE_SOCKLEN_T in configure.in, or to define it like this: AC_DEFUN([TYPE_SOCKLEN_T], [AC_CACHE_CHECK([for socklen_t], ac_cv_type_socklen_t, [ AC_TRY_COMPILE( [#include <sys/types.h> #include <sys/socket.h>], [socklen_t len = 42; return 0;], ac_cv_type_socklen_t=yes, ac_cv_type_socklen_t=no) ]) if test $ac_cv_type_socklen_t != yes; then AC_DEFINE(socklen_t, int) fi ]) I just want to know it it solves the problem before changing the configure script. Cheers Jorge.-

Feb. 22, 2004

Re: [Dillo-dev]progress on AIX?

by Jorge Arellano Cid

On Sat, 14 Feb 2004, Riccardo wrote: > Hello, > > I posted the buildlog I got on AIX... did someone had a look at it to > see if there are some warnings or errors that could explain the > brokeness of dillo on aix? Yes I did. The only thing capable of producing an illegal instruction I can think of by reading the gcc compile log is: <q> main.c:298: warning: passing arg 3 of `naccept' from incompatible pointer type bookmarks.c:1824: warning: passing arg 3 of `naccept' from incompatible pointer type downloads.c:176: warning: passing arg 3 of `naccept' from incompatible pointer type </q> The third argument of accept() is of socklen_t * type. If the size of the pointed-to value is smaller than the actual type, memory corruption can happen. If AIX doesn't provide socklen_t, dillo blithely defines it as: <q> IO/http.c: #define socklen_t guint32 ../dpi/downloads.c: #define socklen_t guint32 ../dpid/dpi.h: #define socklen_t guint32 </q> Please try changing these to int. This is: #define socklen_t int The other warnings worth of considering at this point are: <q> ld: 0711-224 WARNING: Duplicate symbol: .strcmp ld: 0711-224 WARNING: Duplicate symbol: strcmp ld: 0711-224 WARNING: Duplicate symbol: .pow ld: 0711-345 Use the -bloadmap or -bnoquiet option to obtain more information. </q> Hope this helps Jorge.-

Feb. 22, 2004

[Dillo-dev]Resuming...

by Jorge Arellano Cid

Hi there, I'm back from my week off! It was very nice with plenty of good times in the southern part of Chile. A bit short but reassuring! In fact it was a few days ago that I started moving the large stack of emails and pending issues... Those still waiting for an answer, please be patient, just keep sending periodic remainders to the mailing list just in case I get lost in the email pile. ;) (email is kind of an entropic activity: the more you write, the more you receive). One of the already answered was a request from LinuxDevices. Now you can check their coverage of 0.8.0 release here: http://linuxdevices.com/news/NS6109819427.html Greetings to everyone Jorge.-

Feb. 22, 2004

Re: [Dillo-dev]GTK error

by Sebastian Geerken

On Sat, Feb 14, Riccardo wrote: > dillo 0.8.0 (well, the last pre series I think...) dies from time to > time with this error: > Gdk-ERROR **: BadMatch (invalid parameter attributes) > serial 6737 error_code 8 request_code 72 minor_code 0 > > this is on solaris, gtk 1.2.10 I'm not sure whether I will be able to help you at all, but for this kind of error, the following is helpful for others. If you are able to reproduce it (or it happens often enough so that the chance is there), try the following: 1. Start gdb with dillo, and run it with "--g-fatal-warnings". 2. When this error occurs, dillo will immediately abort, and you are back in gdb. 3. Type "bt" in gdb. If you are a bit familiar with the internals of dillo, try to find the relevant level on the stack (usually the top-most function of dillo itself), and print the values of the variables of which you think that they might help. Sebastian

Feb. 22, 2004

[Dillo-dev]CSS pre-release

by Sebastian Geerken

Hi! For those interested in the work on CSS, there is a pre-release at <http://www.dillo.org/download/dillo-0.8.0-css-1.tar.gz>. Don't expect too much, it is rather incomplete, even the splash screen cannot be displayed :-) (because of a syntax error in it). It is based on dillo 0.8.0, although much of the functionality in the latter is missing. See the file README for further informations. Sebastian

Feb. 22, 2004

Re: [Dillo-dev]Frame support in CVS

by Andreas Kemnade

On Thu, 05 Feb 2004 14:05:06 -0800 Kelson Vibber <kelson(a)pobox.com> wrote: > At 07:39 PM 2/3/2004, Jacky Lam wrote: > > I am new to dillo and I want to use dillo with frame support. I try > > dillo in CVS and find it only show some link pages for each frames > > respectively. Is it the current status? How about Frank patches? > > I don't know. I tried applying the latest version of Frank's patch > (version 16, from November 24), but ended up with about 30 reject files, so > it's clearly *way* out of sync. Unfortunately, Frank hasn't posted here > since early December. > > Does anyone else know what's going on with this? Even if frames aren't > being merged in yet, it would be nice to have a patch that's in sync with > the new release. > I have tried to update Frank's patch. It is not well tested or cleaned up yet but I simply need to take a deep breath. http://users.auriga.wearlab.de/~andi/dillo/dillo-frames.diff.gz Tested with yesterday's cvs. At least it seems to work at me. There are some issues with scrolling where I have disabled some pieces of code and marked them with a #warning. I think whoever maintains this patch in the future should be noticed about changes in the dillo source which might affect the patch. The patch has one big problem. It is too big. Not that it bloats dillo, but it is hard to review such a big patch. Perhaps it can be split up and then be committed piece by piece. Has anyone ideas how to split it? Perhaps in the DilloDoc/BrowserWindow separation, the kbnav thing and frame support? So what's the way to get frames into cvs? I would not like to see Frank's efford lost. Greetings Andreas Kemnade

Feb. 22, 2004