- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev] Re: BUG#494
by Trent Jarvi Dec. 15, 2003

Dec. 15, 2003

> > > Hi, > > Some *BSD don't provide a reentrant gethostbyname: > > <q source='manpage of a BSD'> > BUGS > These functions use static data storage; if the data is needed for future > use, it should be copied before any subsequent calls overwrite it. Only > the Internet address format is currently understood. > </q> > > The only one I know that doesn't work is openBSD >= 3.3. > (produces random lookups) > > Does anyone know the situation better? > It is for disabling threaded dns lookups at configure time. > > Would gethostbyname_r() solve this? -- Trent Jarvi taj(a)www.linux.org.uk

2 1

*BSD and gethostbyname
by Jorge Arellano Cid Dec. 13, 2003

Dec. 13, 2003

Hi, Some *BSD don't provide a reentrant gethostbyname: <q source='manpage of a BSD'> BUGS These functions use static data storage; if the data is needed for future use, it should be copied before any subsequent calls overwrite it. Only the Internet address format is currently understood. </q> The only one I know that doesn't work is openBSD >= 3.3. (produces random lookups) Does anyone know the situation better? It is for disabling threaded dns lookups at configure time. Cheers Jorge.- PS: That's it. A thread follows with other minutiae... ---------- Forwarded message ---------- Date: Sat, 13 Dec 2003 14:26:58 +0100 (MET) From: Anders Gavare <md1gavan(a)mdstud.chalmers.se> To: Jorge Arellano Cid <jcid(a)dillo.org> Subject: Re: [Dillo-dev] Re: BUG#494 On Sat, 13 Dec 2003, Jorge Arellano Cid wrote: > On Sat, 13 Dec 2003, Anders Gavare wrote: > > How typical: now I found this already mentioned in dillo's README file. > > Sorry to have bothered you. :-) Still, this is something that should > > be detected and taken care of by the configure script... > > Thanks for the info. > > Can you send me a patch for configure.in? Unfortunately I don't think I'm familiar enough with neither Dillo nor the configure system to add such a functionality. It took several days for me to track down the widget bug (#494), and several hours of testing to figure out this gethostbyname() bug. > Or at least the contents of the detected $target variable on > your machine (an extra line with "echo $target" after > AC_CANONICAL_SYSTEM will tell). sparc64-unknown-openbsd3.3 The same problem should exist in other BSDs as well, but I've not actually seen the symptoms. The manpages of FreeBSD 4.5 and NetBSD-current (after 1.6.1) have the following in their gethostbyname(3) manpages: BUGS These functions use static data storage; if the data is needed for future use, it should be copied before any subsequent calls overwrite it. Only the Internet address format is currently understood. Imho, if Dillo aims for platform independance, which is one of the things I like best about Dillo, this feature (multithreaded DNS lookups) would be better off disabled _unless_ the underlying platform explicitly allows it. Not the other way around. That's my opinion, anyway. Anders

1 0

BUG#494
by Jorge Arellano Cid Dec. 13, 2003

Dec. 13, 2003

Hi, Some time ago I read this bug and now the CVS has a bunch of fixed for signed/unsigned conflicts that may cause trouble on 64bit machines. Please test our CVS and tell me if the bug persists. You can send me the screenshot (not to dillo-dev please) Cheers Jorge.-

2 4

statically linked dillo
by Jorge Arellano Cid Dec. 12, 2003

Dec. 12, 2003

Hi, ---------- Forwarded message ---------- Date: Thu, 11 Dec 2003 23:30:05 +0000 From: Victor M. <voremo(a)safe-mail.net> To: jcid(a)softhome.net Subject: dillo Can dillo be compiled into a static, self-contained executable to run under X? (.i.e. that it can be pressed on and run from a publication CD, with no need for GTK libraries on the target computer...?) Victor --------------------------------------- Does anyone here know/done it? Cheers Jorge.-

2 1

[PATCH] I18N
by Kiyo Dec. 12, 2003

Dec. 12, 2003

Hi. I made an I18N patch for Dillo. This is simple I18N. Japanese, Chinese, Korean, Greek, etc. Only LOCALE character are displayed. Not support "complicated languages". (such as Arabic) Internal code is mainly processing by UTF-8. UTF-8 wrapper of a part of APIs of GTK+1 is included. This wrapper will be unnecessary if Dillo ports to FLTK2 or GTK+2. But, I think that others will be probably needed. A part of this patch referred to Mr. Robert Thomson's Japanese patch. It is better to also gettextize. But, it has not applied to this patch. http://teki.jpn.ph/pc/software/dillo-cvs2003121203-i18n.diff.bz2 md5sum: 5a1e0428d0e049a081dd4721d48b1de2 and... For Core team, may not be desirable... This is large patch for Dillo-0.7.3 . I18N(UTF-8 support and gettext) Xft antialiasing(multibyte support) Setting Tool(gtk+) frame&tab(ver.10 with I18N fix) gzip compression transmission SearchBox UserAgent changer Referer meta-refresh SSL BASIC authorization Fix page scroll Character entity references(Probably almost is supported. but it's dependent on LOCALE.) and some various changes. Various persons' patches are unified and an original improvement is added. http://teki.jpn.ph/pc/software/dillo-0.7.3-i18n-misc-20031208.diff.bz2 md5sum: 0d3bdf88a3e61a4659e708ac4a971c1f #I'm hoping that both user and developer can become happy :-) --- Kiyo<kiyo(a)teki.jpn.ph> http://teki.jpn.ph/

1 0

latest network patch
by Jorge Arellano Cid Dec. 11, 2003

Dec. 11, 2003

Hi there! A network patch that covers several aspects was just commited: + Connected the dpi stream to the cache using the CCC! + Fixed the cache API by introducing the new call a_Capi_get_buf(). + Fixed a race condition and a multiple request problem. + Enabled the https dpi (it doesn't work yet, just a development version). + Removed a cpu-hog when the network is unreachable. Besides what's stated above, it has improved a bit the network experience (for instance on sites with lots of small images. For ex. newspapers). Please test it and tell me how it works. Anders: Please try it and tell us if it also produces the locks you've seen. Does 0.7.3 also lock? Ferdi: It has the proper bindings between the cache and dpi, so HTML served over FTP works, among others! Some https sites are also viewable... Madis: This patch has the framework to implement https in a similar way than the ssl gateway. The connection between dillo and the https dpi is done by the generic dpi handler. Note that although it may be easy to add the POST method using wget to the current https dpi (wget >= 1.9.x using --post-data and the DilloUrl), making a native https plugin (avoiding wget) as in the ssl gatway is preferred. This allows for a simple way to implement connection caching. The current https dpi is a filter, not a server as it should to implement it this way. This is also easy to do. Regards Jorge.-

1 0

BUG#494
by Andrew A. Gill Dec. 10, 2003

Dec. 10, 2003

Good Ol' Jorge wrote: > > Some time ago I read this bug and now the CVS has a bunch of > fixed for signed/unsigned conflicts that may cause trouble on > 64bit machines. Please test our CVS and tell me if the bug > persists. Alpha EV56 running gcc 3.3.2 (glibc 2.3.2) reports no errors yet, sir! In fact, Slashdot now seems to be readable, thanks to something in Dillo CVS (with old Dillo it's still unreadable). -- |Andrew A. Gill |I posted to Silent-Tristero and| |<superluser(a)frontiernet.net> |all I got was this stupid sig! | |alt.tv.simpsons CBG-FAQ author | | | (Report all obscene mail to Le Maitre Pots)| |Nothing here yet: <http://www.needsfoodbadly.com> Temporary sig: -- Full speed ahead!

1 0

navigation
by montegue＠myrealbox.com Dec. 9, 2003

Dec. 9, 2003

i would like to see in addition to the forward and backward navigation buttons, one that goes to the parent directory, like konqueror does, and galeon if i am not mistaken :) thanks for a terrific application -- GR I have seen the future and it is just like the present, only longer. -- Kehlog Albran

1 0

Re: dpid
by Jorge Arellano Cid Dec. 9, 2003

Dec. 9, 2003

On Sat, 6 Dec 2003, Ferdi Franceschini wrote: > Hi Jorge, > > On Fri, Dec 05, 2003 at 09:25:20AM -0300, Jorge Arellano Cid wrote: > > > > Hi Ferdi, > > > > I've catched an API problem inside the capi/cache that may be > > the cause for the fork-hog. This note is just to keep you from > > wasting time reviewing it. I'll write to you soon. > Great. It _was_ the cause! There was a race condition hidden behind the lack of dpi-cache connection (the part we blamed) and a slightly broken API (the cloaked problem). As a matter of fact this bug had remained unnoticed maybe for years. In my patched tree (which is under test), the bug is fixed (both parts) and dillo works even better than before. > I think I have a new solution to the inf loop caused when a connection is not > accepted on a dpi socket for any reason. This new solution does not require > any code changes to the plugins. I will post more info to dillo-dev but I need > to experiment a bit first. That's good news. Cheers Jorge.-

1 0

dead link behavior
by Peter Milne Dec. 5, 2003

Dec. 5, 2003

Greetings all, Was wondering if it is me or if it was normal. If I click a link and it turns out to be a dead link.....everything locks up hard and I have to 'xkill' Dillo. Thanks, -- ~~~Pete~~~ http://milneweb.com http://nomorevirus.com

1 1