February 2004 - Dillo-dev - lists.mailman3.com

Re: [Dillo-dev]solaris 2.6 works

by Peter Sinn

Hello, Am 1. February 2004 17:52 Uhr schrieb Riccardo: > with the latest configure patches to correctly link, the solaris version > of dillo finishes build and seems to work pretty as much as 0.7.3 did. yes, I can confirm that it works. I had posted prematurely, just after posting I found that by substituting WAIT_ANY with -1 and by adding -lsocket -lnsl to the compile directives everythig runs just smoothly. Thanks for the quick fix. Peter

Feb. 2, 2004

[Dillo-dev]AIX problems

by Riccardo

Hello, since upgrading to 1.2.10 did not help, I pathed manually every file including gtk.h by doing #undef func_data before. - I notice I get 'stcasecmp' implicit declaration warning. - the vsdnprintf was already reported and shouldn't be a matter. at the final linking inside "src" I get errors about duplicate symbols (.strcmp, strcmp, .pov) This time I got the WAIT_ANY error on aix too (I wonder why I didn't get it last time!, maybe because I am using gnu-make ?) and subsituted it with -1. Compile ends. As the last time dillo crashes with an illegal instruction and GDB doesn't return me a useful backtrace. I am quite clueless. Can I compile dillo with debug? -Ric

Feb. 1, 2004

Re: [Dillo-dev]user agent and OS

by Riccardo

Kelson Vibber wrote: > > On Sunday 01 February 2004 9:03 am, Riccardo wrote: > > I'd like to see dillo reporting the OS in the User-agent, this can be > > interesting for statistics (and a reason ov proudness for lesser-known > > platforms, heh). > > I suggested this sometime last year, and provided a small patch, but Jorge > eventually rejected the idea for privacy reasons. Almost every browser I know of can do this. Some have a switch for it, a rpeference option. Maybe a compile-time option. I'd like a way to do it. Of course it is a minor thing regards to display problems or portability. -R

Feb. 1, 2004

Re: [Dillo-dev]user agent and OS

by Kelson Vibber

On Sunday 01 February 2004 9:03 am, Riccardo wrote: > I'd like to see dillo reporting the OS in the User-agent, this can be > interesting for statistics (and a reason ov proudness for lesser-known > platforms, heh). I suggested this sometime last year, and provided a small patch, but Jorge eventually rejected the idea for privacy reasons. -- Kelson Vibber www.hyperborea.org

Feb. 1, 2004

Re: [Dillo-dev]AIX : illegal instruction

by Riccardo

Andreas Schweitzer wrote: > > At least check where the illegal instrucution comes from ... > the backtrace stopped there! that was the problem. > maybe that's why I kept the "patch" to myself :-) ... I used > an older pre-packaged version of GTK and was too lazy to upgrade. > > And since Ricardo told us in another mail that he isn't using > 1.2.10 either .... I removed 1.2.9 and recompiled 1.2.10 and run also "make check" before installing. I tried building from a clean 30-Jan 0.8.0 tarball and it didn't work, I still get errors in the headers. I try applying your patch manually again and see if it helps. -Ric

Feb. 1, 2004

Re: [Dillo-dev]AIX : illegal instruction

by Andreas Schweitzer

Hi, On Sun, Feb 01, 2004 at 05:04:17PM +0100, Riccardo wrote: > Andreas Schweitzer wrote: > > About the WAIT_ANY : > AIX didn't complain... the fix was for solaris. :-) ... I only grep'ed through /usr/include and did not find it ... > > > But startup fails, GDB gives me the following: > > > (gdb) bt > > > #0 0x0 in ?? () from /usr/local/build/dillo-0.8.0-pre/src/./dillo > > > > and now I'm lost ... is this all the output ? > Yes, that's all. > > Maybe I should make clean and log the build output? Maybe some compiler > warnings could be instructive. At least check where the illegal instrucution comes from ... On Sun, Feb 01, 2004 at 12:27:11PM -0300, Jorge Arellano Cid wrote: > On Sun, 1 Feb 2004, Andreas Schweitzer wrote: > > <problems about func_data and GTK in AIX ...> > AIX defines func_data in timer.h, and GTK also used it. > > Is an AIX/GTK problem and was solved in GTK+-1.2.10. > (http://www.aoindustries.com/docs/glib-1.2.10/ChangeLog) maybe that's why I kept the "patch" to myself :-) ... I used an older pre-packaged version of GTK and was too lazy to upgrade. And since Ricardo told us in another mail that he isn't using 1.2.10 either .... Cheers, Andreas -- **************************** NEW ADDRESS ****************************** Hamburger Sternwarte Universitaet Hamburg Gojenbergsweg 112 Tel. ++49 40 42891 4016 D-21029 Hamburg, Germany Fax. ++49 40 42891 4198

Feb. 1, 2004

[Dillo-dev]user agent and OS

by Riccardo

Hello, I'd like to see dillo reporting the OS in the User-agent, this can be interesting for statistics (and a reason ov proudness for lesser-known platforms, heh). Like netscape on my sun box reports: User-Agent: Mozilla/4.79 [en] (X11; U; SunOS 5.6 sun4u) while dillo only: User-Agent: Dillo/0.8.0-pre Since I write a small webserver with its log-analyzer I noticed this. Btw, some advertising... it is a small webserver, quite portable across unix and it weigths around 30k depending on the options and on the platform. It makes a nice match with dillo! With both you can have fun of the www on a 25Mhz computer again :) http://sourceforge.net/projects/pserv/ The matched log-analywer "p-reporter" reports dillo as a rocgnized web-browser :) -Ric

Feb. 1, 2004

[Dillo-dev]solaris 2.6 works

by Riccardo

Hello, with the latest configure patches to correctly link, the solaris version of dillo finishes build and seems to work pretty as much as 0.7.3 did. The problems I currently encounter are: - color problems. I have 256 colors. If the colors are already reserved say from Netscape or another applicaiton, images (but only those) are displayed with mangled colors, they don't get remapped to the neares color available. If the full palette is available, images are dithered as expected. - authenticaiton problem with www.blender.org to the forums. I insert my data, login.. I get a page with a meta-refresh which should me login, Dillo gives me the link, I follow it and I result not logged in. The specs for "wait any" on solaris are the following: If pid is equal to (pid_t)-1, status is requested for any child process. I think so that a cast of -1 to the correct type wouldn't hurt. Riccardo PS: I try building and isntalling glib/gtk 1.2.10 instead of the current 1.2.9 on AIX. One never knows, it could help.

Feb. 1, 2004

Re: [Dillo-dev]AIX : illegal instruction

by Riccardo

Andreas Schweitzer wrote: > > Hi, > > I followed the progress on the mailing list, but was always > too late ;-) ... > > About the WAIT_ANY : AIX 4.3.3 (and Ricardo's 4.2.X) indeed > does not define WAIT_ANY. I checked man-pages on several OS's > and neither says anything about this being standard, but all > agree that "-1" is standard, so "-1" is probably the way to go. AIX didn't complain... the fix was for solaris. Unfortunately it seems thjat I don't have the full developer manpages on AIX, so Ic ouldnt' check further. > *If* that ever wants to go into dillo, a cleaner way would be to > enclose it in "#ifdef _AIX" - but I don't know about AIX5 It should go into dillo... maybew check for AIX4 ? Is it possible? It should > > But startup fails, GDB gives me the following: > > (gdb) bt > > #0 0x0 in ?? () from /usr/local/build/dillo-0.8.0-pre/src/./dillo > > and now I'm lost ... is this all the output ? Yes, that's all. Maybe I should make clean and log the build output? Maybe some compiler warnings could be instructive. -Riccardo

Feb. 1, 2004

Re: [Dillo-dev] building on AIX, gcc, dillo 0.7.3

by Jorge Arellano Cid

On Sun, 1 Feb 2004, Riccardo wrote: > Jorge Arellano Cid wrote: > > > > > Some Solaris seem not to have it. :( > > I decided to replace WAIT_ANY with -1 as in dillo.c; commited. > > > > Please test it on aix and sparc. Use the Jan30 tarball and > > just replace WAIT_ANY with -1 in: > > > > dpi/downloads.c and > > dpid/dpid.c > Yes, I checked solaris doc and it says to use (pid_t)-1. > > Build stops howewer with: > gcc -I/usr/local/include/glib-1.2 -I/usr/local/lib/glib/include -g -O2 > -D_REENTRANT -D_THREAD_SAFE -Wall -W -Wno-unused-parameter > -Waggregate-return -R/usr/local/lib -L/usr/local/lib -R/usr/lib > -L/usr/lib -L/usr/local/lib -lposix4 -o dpid dpi.o dpi_service.o > dpi_socket_dir.o dpid.o dpid_common.o main.o misc_new.o -L/usr/local/lib > -lglib -lpthread > Undefined first referenced > symbol in file > socket dpid.o > accept main.o > bind dpid.o > listen dpid.o > connect dpid.o > ld: fatal: Symbol referencing errors. No output written to dpid > > > this alomst certainly means that -lsocket is not included (maybe -lnsl > too). I thought configure should take care of this? It usually does. I included a check for those. Please download: http://www.dillo.org/test/configure use it in Jan30 tarball and tell us how it goes. Cheers Jorge.- PS: Sorry Riccardo for the duplicated email, my mistake.

Feb. 1, 2004