- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev] ipv6 and dns threads

by Damien Couderc

Hi all, this time i take care to cc the list :) This mail include the complete reply i sent to Jorge plus some comments added since i looked recent changes. On Sun, 16 Mar 2003 02:21:52 +0100 Damien Couderc <coudercd(a)nerim.net> wrote: > Hi Jorge :) > > First i'm sorry to be a bit late to reply, i'm really busy with a > new project actually. > > For the sparc64 we use gcc 2.95.3 by default, but we also have a > port of 3.2.2 (but i don't remember if i tested with 3.2.2 so i'll > check it). Still not done. > > > You will found at the following URL patches from the openbsd > > > port: > > > http://www.openbsd.org/cgi-bin/cvsweb/ports/www/dillo/patches/ > > > > > > Some of them fix ipv6 and threaded name resolution bugs. > > There're modified versions for those patches in our CVS now! Yep dns.c seems okay for me :) > > In the specific case of the IPv6 one, I don't understand why: > > > > - dns_server[i].th1 = (pthread_t) -1; > > + dns_server[i].th1 = GPOINTER_TO_INT((pthread_t) -1); > > because 'dns_server[i].th1' is of type 'pthread_t', not int! > > This was fixing a warning of the compiler, so not really important. > It is also no longer used as we comment G_DNS_THREADED in the latest > patches. > > > In the case of the http.c patch, I modified it a bit because > > 'sa_len' is not supposed to be used from sockaddr type (at least > > that's what libc's texinfo says). > > Oh yes linux seems to not have it. > You could give a look to SA_LEN() macro which is in libc but of > course it is linux specific. > Without starting flamewar you should not take linux as (only) > reference if you want to make something portable. Okay so, i've took some time to read your changes and something took my attention : IO/http.c : socket_len = sizeof(struct sockaddr_in6); The fact is that the structure sockaddr_in6 can be really big and cause problems and i'm not sure that the use of sizeof is safe. I think the best way should be to use a define that use sa_len on bsd compliant systems and the SA_LEN() macro on linux. Damien

March 16, 2003

Re: [Dillo-dev] another bug in bm_serv ?

by slymer＠slymer.org

On Sun, 16 Mar 2003 13:00:51 +0100 Andreas Schweitzer <Andreas.Schweitzer(a)hs.uni-hamburg.de> wrote: > or having the single quote (') Hmm I have tested further a bit, it seems that it's because of the ('). -- Registered Linux user #298569 It's time to boot, do your boot ROMs know where your disk controllers are?

March 16, 2003

Re: [Dillo-dev] another bug in bm_serv ?

by Andreas Schweitzer

On Sun, Mar 16, 2003 at 01:21:11PM +0200, slymer(a)slymer.org wrote: > > I think bm_serv has a problem adding a bookmark when the title > of the page is more than 8 chars in length. or having the single quote (') ? > For example, I have visited the LAME "encoder" home page and I > wanted too bookmark it. The title of the resulting bookmark is > "LAME Ain" instead of "LAME Ain't an MP3 Encoder". Also if after I can reproduce this. > adding it I try to set the name of the bookmark right bm_serv > will act a bit funky by using 100% of my CPU and not sending the > results of the modify command to dillo. I have to close dillo I cannot reproduce this, though. Latest version of bm_srv12.c on a Linux 2.4.20, gcc3.2.2 system. Cheers Andreas -- **************************** NEW ADDRESS ****************************** Hamburger Sternwarte Universitaet Hamburg Gojenbergsweg 112 Tel. ++49 40 42891 4016 D-21029 Hamburg, Germany Fax. ++49 40 42891 4198

March 16, 2003

another bug in bm_serv ?

by slymer＠slymer.org

I think bm_serv has a problem adding a bookmark when the title of the page is more than 8 chars in length. For example, I have visited the LAME "encoder" home page and I wanted too bookmark it. The title of the resulting bookmark is "LAME Ain" instead of "LAME Ain't an MP3 Encoder". Also if after adding it I try to set the name of the bookmark right bm_serv will act a bit funky by using 100% of my CPU and not sending the results of the modify command to dillo. I have to close dillo for bm_serv to behave normaly again and also if I check the bookmarks I see that modifying the bookmark's name worked afterall. -- Registered Linux user #298569 Let me do my TRIBUTE to FISHNET STOCKINGS ...

March 16, 2003

Re: [Dillo-dev] ipv6 and dns threads

by Jorge Arellano Cid

Damien, > You will found at the following URL patches from the openbsd port : > http://www.openbsd.org/cgi-bin/cvsweb/ports/www/dillo/patches/ > > Some of them fix ipv6 and threaded name resolution bugs. There're modified versions for those patches in our CVS now! In the specific case of the IPv6 one, I don't understand why: - dns_server[i].th1 = (pthread_t) -1; + dns_server[i].th1 = GPOINTER_TO_INT((pthread_t) -1); because 'dns_server[i].th1' is of type 'pthread_t', not int! In the case of the http.c patch, I modified it a bit because 'sa_len' is not supposed to be used from sockaddr type (at least that's what libc's texinfo says). Well, the important thing is to test them, as I don't have a *BSD machine here. Thanks Jorge.-

March 15, 2003

[Patch]: SpamSafe and <base href=..> bug fix

by Melvin Hadasht

Hi, The SpamSafe (-l command line) option is meant to refrain Dillo from retrieving remote resources (images). This works unless the original HTML contains a <base href=...> entity. In which case, the remote resources are unconditionally retrieved. In a older message (Message-Id: <20030221093243.435de32f.melvin.hadasht(a)free.fr>), I posted a corresponding test case. The attached patch fixes this misbehaviour by transmitting the URL_SpamSafe flag to the new BaseURL. The patch is against current CVS. Cheers. -- Melvin Hadasht

March 15, 2003

dillo officially supported in Sylpheed-Claws

by Melvin Hadasht

Hi, Current cvs version of Sylpheed-Claws mail client (0.8.11claws14) officially supports Dillo to"safely" view html mail within the mail client window using the embedding facility of Dillo and a plugin technique in Sylpheed-Claws. The forthcoming 0.8.12claws will be the first public release to provide the Dillo plugin. (no timeline known, but it's a matter of weeks). I wrote "safely" between quotes, because IMHO there is a bug in Dillo when using the "-l" (or SpamSafe) command line option, as I pointed out in a previous mail (Message-Id: <20030221093243.435de32f.melvin.hadasht(a)free.fr>). I'll shortly post a patch for it in this ML. Thank you for your work. Cheers -- Melvin Hadasht

March 15, 2003

Re: [Dillo-dev] dillo on FreeBSD with --enable-ipv6

by Harold Gutch

On Sat, Mar 15, 2003 at 02:14:45PM -0400, Jorge Arellano Cid wrote: > Considering that, I decided to commit a modified version of the > patch, but as I don't have a BSD machine to test, you'll have to > help me. Just get the code from the CVS and send some feedback! I just tested your version - it works fine, thanks! bye, Harold

March 15, 2003

Re: [Dillo-dev] dillo on FreeBSD with --enable-ipv6

by Jorge Arellano Cid

Hi there! On Sat, 15 Mar 2003, Harold Gutch wrote: > On Sat, Mar 15, 2003 at 02:08:29PM +0100, Andreas Schweitzer wrote: > > Hi, > > > > > when enabling IPv6 on FreeBSD, dillo completely stops to work. > > > With such a binary, not even IPv4 works anymore. I'm not really > > > > I think I expereineced this, too. But since I don't > > have an IPv6 connection, I didn't know, that IPv4 is > > still supposed to work. > > > > > sure if this is a bug in dillo, or if it's in FreeBSD, but the > > > third parameter to connect() apparently needs to be sizeof(struct > > > sockaddr), which it originally is not (in src/IO/http.c:307). > > > The attached patch makes dillo work again if I compile it with > > > IPv6-support. All of this might apply to the other BSDs as well, > > > and also to other OSes. Also, I don't quite think that this > > > patch should break anything, so it should be safe to apply it. > > That's wrong, I just noticed that my patch unbreaks IPv4, but it > still doesn't fix IPv6. I guess I only tried it with IPv4 > webservers just before :(. > > > > Compare to the patches for OpenBSD : > > http://www.openbsd.org/cgi-bin/cvsweb/ports/www/dillo/patches/ > > they have already been pointed to on this mailing list. And > > I think, your patch is essentially part of theirs ;-) > > Thanks for pointing me out to that! Will the according changes > be incorporated into dillo before the next release? I finally had some time to look at it, and I hesitate to include the patch as it is, because my texinfo documents for libc say: "The one piece of information that you can get from the `struct sockaddr' data type is the "address format designator". This tells you which data type to use to understand the address fully." so "sa_len" is out. Considering that, I decided to commit a modified version of the patch, but as I don't have a BSD machine to test, you'll have to help me. Just get the code from the CVS and send some feedback! Cheers Jorge.-

March 15, 2003

Re: [Dillo-dev] dillo on FreeBSD with --enable-ipv6

by Harold Gutch

On Sat, Mar 15, 2003 at 02:08:29PM +0100, Andreas Schweitzer wrote: > Hi, > > > when enabling IPv6 on FreeBSD, dillo completely stops to work. > > With such a binary, not even IPv4 works anymore. I'm not really > > I think I expereineced this, too. But since I don't > have an IPv6 connection, I didn't know, that IPv4 is > still supposed to work. > > > sure if this is a bug in dillo, or if it's in FreeBSD, but the > > third parameter to connect() apparently needs to be sizeof(struct > > sockaddr), which it originally is not (in src/IO/http.c:307). > > The attached patch makes dillo work again if I compile it with > > IPv6-support. All of this might apply to the other BSDs as well, > > and also to other OSes. Also, I don't quite think that this > > patch should break anything, so it should be safe to apply it. That's wrong, I just noticed that my patch unbreaks IPv4, but it still doesn't fix IPv6. I guess I only tried it with IPv4 webservers just before :(. > Compare to the patches for OpenBSD : > http://www.openbsd.org/cgi-bin/cvsweb/ports/www/dillo/patches/ > they have already been pointed to on this mailing list. And > I think, your patch is essentially part of theirs ;-) Thanks for pointing me out to that! Will the according changes be incorporated into dillo before the next release? bye, Harold

March 15, 2003

Re: [Dillo-dev] dillo on FreeBSD with --enable-ipv6

by Andreas Schweitzer

Hi, > when enabling IPv6 on FreeBSD, dillo completely stops to work. > With such a binary, not even IPv4 works anymore. I'm not really I think I expereineced this, too. But since I don't have an IPv6 connection, I didn't know, that IPv4 is still supposed to work. > sure if this is a bug in dillo, or if it's in FreeBSD, but the > third parameter to connect() apparently needs to be sizeof(struct > sockaddr), which it originally is not (in src/IO/http.c:307). > The attached patch makes dillo work again if I compile it with > IPv6-support. All of this might apply to the other BSDs as well, > and also to other OSes. Also, I don't quite think that this > patch should break anything, so it should be safe to apply it. Compare to the patches for OpenBSD : http://www.openbsd.org/cgi-bin/cvsweb/ports/www/dillo/patches/ they have already been pointed to on this mailing list. And I think, your patch is essentially part of theirs ;-) Cheers Andreas -- **************************** NEW ADDRESS ****************************** Hamburger Sternwarte Universitaet Hamburg Gojenbergsweg 112 Tel. ++49 40 42891 4016 D-21029 Hamburg, Germany Fax. ++49 40 42891 4198

March 15, 2003

dillo on FreeBSD with --enable-ipv6

by Harold Gutch

Hi, when enabling IPv6 on FreeBSD, dillo completely stops to work. With such a binary, not even IPv4 works anymore. I'm not really sure if this is a bug in dillo, or if it's in FreeBSD, but the third parameter to connect() apparently needs to be sizeof(struct sockaddr), which it originally is not (in src/IO/http.c:307). The attached patch makes dillo work again if I compile it with IPv6-support. All of this might apply to the other BSDs as well, and also to other OSes. Also, I don't quite think that this patch should break anything, so it should be safe to apply it. Oh, please CC me in any upcoming mails in this thread, since I'm not subscribed to the list. bye, Harold

March 15, 2003

Re: [Dillo-dev] Selection

by Madis Janson

On Fri, 14 Mar 2003, Sebastian Geerken wrote: > Hi! > > I've committed the text selection. There are still some glitches, look > at src/selection.c in the sources for the known ones. Please check it > out and test it. Thats great! Thank you!

March 15, 2003

Re: [Dillo-dev] Selection

by Andreas Schweitzer

Hi, > I've committed the text selection. There are still some glitches, look > at src/selection.c in the sources for the known ones. Please check it > out and test it. Wow - where did this come from ;-) Cool ! works quite well here. Preliminary tests also work on a FreeBSD box. Maybe it's a known glitch, but if you go to http://www.daemonnews.org and try to select anything in the top red bar or anything in the left column, it works ...hmmm ... "funky" ... Cheers Andreas -- **************************** NEW ADDRESS ****************************** Hamburger Sternwarte Universitaet Hamburg Gojenbergsweg 112 Tel. ++49 40 42891 4016 D-21029 Hamburg, Germany Fax. ++49 40 42891 4198

March 15, 2003

floating tables

by Joe Crayne

The floating tables patch is now available against current CVS with text selection. http://home.earthlink.net/~crayne2/floats.html Cheers Joe.-

March 15, 2003

(no subject)

by pekdon＠gmx.net

March 14, 2003

Selection

by Sebastian Geerken

Hi! I've committed the text selection. There are still some glitches, look at src/selection.c in the sources for the known ones. Please check it out and test it. Sebastian PS: Full screen mode is now done by clicking on the button in the lower right corner.

March 14, 2003

Re: https approaches?

by Ami Fischman

I read the dpi1 document. As far as I can see, it refers to DPI's having a "scope" which URLs can fall inside of, and this scope is always prefixed with "dpi:/". Are you saying that: 1) There is no restriction in the code to the dpi:/ prefix? or, 2) You have no objection to removing the current restriction to dpi:/? If 1, please explain (or point to any documentation/source code that explains). As near as I can see dillo/src/IO/Url.c:a_Url_get_ccc_funct() branches to a_Dpi_ccc only if dpi:/ is the prefix. Furthermore, dillo/src/capi.c:a_Capi_open_url() branches on whether url_str starts with "dpi:/". If 2, do you already have plans in place for maintaining a list of which plugins handle which prefixes? Cheers, -- Ami Fischman usenet(a)fischman.org

March 14, 2003

RE: [Dillo-dev] Re: https approaches?

by Eric Gaudet

Speaking of dpi, it seems that only one plugin (that must be called bm_something) can exist in the current implementation. When will we be able to register multiple plugins? Thanks Eric

March 14, 2003

Re: [Dillo-dev] Incorrect rendering of tables on fresco.org

by Sebastian Geerken

On Wed, Mar 12, TheStar wrote: > Also, is it possible to use dillo as a plug-in in non-gtk apps? (ie: porting) You may take a look at <http://www.freedesktop.org/standards/xembed.html>. Since dillo uses Gtk+ for embedding (although I don't know how far Gtk+ 1.2 complies to the specification), it's possible that using this protocol should do the job. Sebastian

March 14, 2003

Re: [Dillo-dev] Re: https approaches?

by Jorge Arellano Cid

On Fri, 14 Mar 2003, Ami Fischman wrote: > Thanks for the great summary. > > It looks to me like Jonathan Springer already implemented the BIO approach > for 0.6.5. The fact that it isn't integrated in 0.7 discourages me from > re-writing it for 0.7. > > Since implementing https as a pluging will require a change to the core > dillo code (AFAIK only dpi:/ URL's get handed to plugins), it would be nice > to get confirmation from the authors that this is OK (would be a candidate > for merging in) before spending the time. Jorge? Jorge wrote a full document explaining the dpi concept and also wrote a 30 page bookmarks server to illustrate it. Cheers Jorge.-

March 14, 2003

Re: https approaches?

by Ami Fischman

Thanks for the great summary. It looks to me like Jonathan Springer already implemented the BIO approach for 0.6.5. The fact that it isn't integrated in 0.7 discourages me from re-writing it for 0.7. Since implementing https as a pluging will require a change to the core dillo code (AFAIK only dpi:/ URL's get handed to plugins), it would be nice to get confirmation from the authors that this is OK (would be a candidate for merging in) before spending the time. Jorge? Cheers, -- Ami Fischman usenet(a)fischman.org

March 14, 2003

Re: [Dillo-dev] https approaches?

by Andreas Schweitzer

> Is anyone currently working on ssl support in dillo? To shed some light on this topic, I have searched the OLD mailing list archive since I did remember quite a bit. Here are some threads : http://sourceforge.net/mailarchive/message.php?msg_id=1975637 http://sourceforge.net/mailarchive/message.php?msg_id=1388644 note the last paragraph !, and follow the link to http://www.geocrawler.com/archives/3/702/2002/1/0/7681591/ Also, note Jonathan Springers work : http://sourceforge.net/mailarchive/message.php?msg_id=1601722 and its continuation at http://sourceforge.net/mailarchive/message.php?msg_id=1610063 BTW, the old archives should be searched from here : http://sourceforge.net/mail/?group_id=603 Could a webmaster change the link to "Old Archives" on http://dillo.auriga.wearlab.de/MList.html ? Finally, note Jorge's very recent mail in which he lists some priorities : http://lists.auriga.wearlab.de/pipermail/dillo-dev/2003-March/000229.html Cheers Andreas -- **************************** NEW ADDRESS ****************************** Hamburger Sternwarte Universitaet Hamburg Gojenbergsweg 112 Tel. ++49 40 42891 4016 D-21029 Hamburg, Germany Fax. ++49 40 42891 4198

March 14, 2003

Re: [Dillo-dev] https approaches?

by andi

Thu, 13 Mar 2003 15:51:09 -0800: Ami Fischman <usenet(a)fischman.org> > SSL support is one of the biggest lacks in dillo for me. To that end, > I'm thinking of adding it in. Without knowing the dillo codebase, I > see two obvious options: > 1) Add a src/IO/https.c file based on src/IO/http.c > 2) Move http.c to using openssl's BIO's which can support either > encrypted > or clear traffic, as well as filters. I think the dpi is, in favor of design, the best solution. I am no programmer, so don't count that much on me. But i have another idea. What about implementing every protocol as dpi? currently the dpi looks, for me as simple user: dpi:/bm/ but why not implementing it as bm://, so http:// would call the http-dpi, ftp:// would call the ftp-dpi, ... This would create a complete module-based client and dillo can be focused on displaying. http:// could be not much more than a wrapped wget, or i am wrong and this whole scheme is impossible? -- Andreas Zwinkau (andi.dasstellenwirinsinternet.de)

March 14, 2003

VERY URGENT

by Dr. Mrs. Mariam Abacha

Dear Friend, I am Dr. Mrs. Mariam Abacha, wife to the late Nigerian Head of state, General Sani Abacha who died on the 8th of June 1998 while still on active service for our Country. l am contacting you with the hope that you will be of great assistance to me, I currently have within my reach the sum of 24 million U.S dollars cash which l intend to use for investment purposes outside Nigeria. This money came as a result of a payback contract deal between my husband and a Russian firm in our country's multi-billion dollar Ajaokuta steel plant. The Russian partners returned my husbands share being the above sum after his death. Presently, the new civilian Government has intensified their probe into my husbands financial resources which has led to the freezing of all our accounts, local and foreign, the revoking of all our business licences and the arrest of my First son. In view of this I acted very fast to withdraw this money from one of our finance houses before it was closed down. I have deposited the money in a security company with the help of very loyal officials of my late husband. No record is known about this fund by the government because there is no documentation showing that we received such funds. Due to the current situation in the country and government attitude to my family, I cannot make use of this money within, thus I seek your help in transferring this funds out of the sub-African region. Bearing in mind that you may assist me, l have decided to part with 20% of the total sum. Your URGENT response is needed. Please include your personal phone and fax number. Regards, Dr. Mrs. Mariam Abacha

March 14, 2003

Re: [Dillo-dev] https approaches?

by madis

On Thu, 13 Mar 2003, Ami Fischman wrote: > Hi there, > > SSL support is one of the biggest lacks in dillo for me. To that end, > I'm thinking of adding it in. Without knowing the dillo codebase, I see two > obvious options: > 1) Add a src/IO/https.c file based on src/IO/http.c > 2) Move http.c to using openssl's BIO's which can support either encrypted > or clear traffic, as well as filters. Imho it should be a plugin or something alike to avoid bloating dillo with ssl specific code. Current dpi does not support sending raw http to dillo, so maybe simplest solution would be to create pseudo-proxy for https, which sends https queries in unencrypted form to unix socket sitting in ~/.dillo/. Local proxy would then do the actual encryption (some authentication may be required to ensure that both dillo and proxy are actually runned by same user). -- mzz

March 14, 2003

https approaches?

by Ami Fischman

Hi there, SSL support is one of the biggest lacks in dillo for me. To that end, I'm thinking of adding it in. Without knowing the dillo codebase, I see two obvious options: 1) Add a src/IO/https.c file based on src/IO/http.c 2) Move http.c to using openssl's BIO's which can support either encrypted or clear traffic, as well as filters. 1 seems much more straightforward and will result in quicker https support, but will cause there to be a whole hell of a lot of code duplication between http.c & https.c. 2 seems like the more "moral" choice and will make dillo better, at the cost of making openssl necessary for dillo builds. Is anyone currently working on ssl support in dillo? Are you using either of the above two options? If the answer to the first question is negative, do people have feelings about code duplication vs. reliance on openssl? (of course, there's a third option of #ifdef'ing the hell out of http.c, adding BIO support that can be compiled out leaving the current plain socket-based code, but that would probably be more of a nightmare to maintain than 1) above). Cheers, -- Ami Fischman usenet(a)fischman.org

March 13, 2003

Re: [Dillo-dev] Incorrect rendering of tables on fresco.org

by Sebastian Geerken

On Thu, Mar 13, Thorben Thuermer wrote: > On Wed, 12 Mar 2003 21:25:53 +1100 > TheStar <thestar(a)tpg.com.au> wrote: > > > I have noticed that the tables at fresco.org are rendered > > incorrectly - in particular the "introduction" page. > > > > All the tables become _very_ thin - is this a non-standard page, or > > a to-be-worked bug in dillo? > > after some more looking at it, the problem is that the cell containing > the actual content is declared as: <td width="" valign="top"> and dillo > seems to give the empty width a lower priority than anything else, instead > of ignoring it like other browsers seem to. Dillo treats this as 0, as soon as the attribute "width" exists, which normally means that a column is rendered as narrow as possible (this is called "size hint" in the documentation). The latter is not actually the case, but dillo's table rendering is still neccessary to improve. (This is just how dillo behaves, not how it should behave.) Sebastian

March 13, 2003

Re: [Dillo-dev] Incorrect rendering of tables on fresco.org

by TheStar

On Thu, 13 Mar 2003 17:17:49 +0100 Thorben Thuermer <r00t(a)Informatik.Uni-Bremen.DE> wrote: > On Wed, 12 Mar 2003 21:25:53 +1100 > TheStar <thestar(a)tpg.com.au> wrote: > > > I have noticed that the tables at fresco.org are rendered > > incorrectly - in particular the "introduction" page. > > > > All the tables become _very_ thin - is this a non-standard page, or > > a to-be-worked bug in dillo? > > after some more looking at it, the problem is that the cell containing > the actual content is declared as: <td width="" valign="top"> and dillo > seems to give the empty width a lower priority than anything else, instead > of ignoring it like other browsers seem to. I'll alert the fresco team about that - thanks :)

March 13, 2003

Re: [Dillo-dev] dillo source

by Kelson Vibber

At 08:29 AM 3/13/2003, Jorge Arellano Cid wrote: >On Mon, 10 Mar 2003 slymer(a)slymer.org wrote: > > Why doesn't dillo also come in tar.bz2 archives ? > For compatibility reasons. Why not provide both? I've seen a lot of projects that provide both a .tar.gz and a .tar.bz2 (and some of them even still provide a .tar.Z). You'll need a bit more storage for hosting, but if enough people choose the .bz2 version, you'll save on bandwidth. Kelson Vibber www.hyperborea.org

March 13, 2003

Re: [Dillo-dev] dillo source

by Jorge Arellano Cid

On Wed, 12 Mar 2003, higuita wrote: > [...] Nice signature! Cheers Jorge.-

March 13, 2003

Re: [Dillo-dev] dillo source

by Jorge Arellano Cid

On Mon, 10 Mar 2003 slymer(a)slymer.org wrote: > > Why doesn't dillo also come in tar.bz2 archives ? > For us dialupers even the 72Kb diff between gzipped and bzipped > counts. For compatibility reasons. There're lots of platforms in which dillo runs. We can assume they have gzip (and _maybe_ bzip2). If they all were to have bzip2, then there would be no problem. Cheers Jorge.-

March 13, 2003

Re: [Dillo-dev] Incorrect rendering of tables on fresco.org

by Thorben Thuermer

On Wed, 12 Mar 2003 21:25:53 +1100 TheStar <thestar(a)tpg.com.au> wrote: > I have noticed that the tables at fresco.org are rendered > incorrectly - in particular the "introduction" page. > > All the tables become _very_ thin - is this a non-standard page, or > a to-be-worked bug in dillo? after some more looking at it, the problem is that the cell containing the actual content is declared as: <td width="" valign="top"> and dillo seems to give the empty width a lower priority than anything else, instead of ignoring it like other browsers seem to. Thorben Thuermer

March 13, 2003

iso-8859-2

by Wojciech Karaś

I can't see iso-8859-2 fonts in my dillo. Is it possible to use *-iso-8859-2 fonts, to display pages? -- W.Karas tel:0048(12)411-52-22(home),0048(12)617-25-88(work) mailto:karas@uci.agh.edu.pl http://gollum.uci.agh.edu.pl visit 'aid for polish children' site http://www.pajacyk.pl

March 13, 2003

Re: [Dillo-dev] .7.1.2 ??

by Jorge Arellano Cid

On Wed, 12 Mar 2003, Frank W McCormick wrote: > Just dl'ed 0.7.1.2...but the dir inside says 0.7.1.1 . Built it > anyway. Which version do I have ?? The dir inside should be "0.7.1", and you have the "0.7.1.2" release. (that is: 0.7.1 with just one line changed). Cheers Jorge.-

March 12, 2003

.7.1.2 ??

by Frank W McCormick

Just dl'ed 0.7.1.2...but the dir inside says 0.7.1.1 . Built it anyway. Which version do I have ?? Thanks

March 12, 2003

Dillo (fwd)

by Jorge Arellano Cid

---------- Forwarded message ---------- Date: Mon, 10 Mar 2003 20:42:47 +0100 From: Michael Baeuerle <michael.baeuerle(a)gmx.net> Reply-To: micha(a)hilfe-fuer-linux.de To: jcid(a)softhome.net Subject: Dillo Hi, since a few month I use dillo for "daily surfing". Formerly I have used Netscape 4.x with disabled JavaScript on my old 60MHz SPARC machine - neither very fast nor very stable, but useable. But now with dillo: The small size and lightning fast startup (1s even on my ancient machine) makes me really happy! Thanks for writing this nice little piece of code (The world don't need bloated things like Mozilla). Regards, Micha -- Linux - May the source be with you! Note: Because of spam overflow, mails with invalid destination address or HTML content are no longer accepted!

March 12, 2003

Re: [Dillo-dev] Incorrect rendering of tables on fresco.org

by Melvin Hadasht

Hi TheStar, on Wed, 12 Mar 2003 21:25:53 +1100 TheStar <thestar(a)tpg.com.au> wrote: > Also, is it possible to use dillo as a plug-in in non-gtk apps? (ie: porting) Perhaps. Dillo can be started with the X11-Window ID of the window where it will be embedded. But I don't know of any non-gtk application that does this. AFAIK, newer xterms can be embedded in a gtk application (I did some tests) using the same mechanism. So I would say the other way (gtk app in a Xlib app) should be possible. Do-da-da-go-hv-i (Cherokee: Goodbye) -- Melvin Hadasht

March 12, 2003

Incorrect rendering of tables on fresco.org

by TheStar

I have noticed that the tables at fresco.org are rendered incorrectly - in particular the "introduction" page. All the tables become _very_ thin - is this a non-standard page, or a to-be-worked bug in dillo? I understand that this site probably uses floating tables, but why are the tables so thin? Also, is it possible to use dillo as a plug-in in non-gtk apps? (ie: porting) I am not a fresco developer :) I'm just interested...

March 12, 2003

Re: [Dillo-dev] space in urls

by John Bradford

> Is there some reason why spaces are dropped from URLs in dillo? Most other > browsers will convert such spaces to %20 (and similarly with other illegal > URI characters) automagically. It's a convinience which some webmasters > take for granted (and which breaks browsing with dillo). I've asked several webmasters to fix the problem, and so far they all have :-). Seriously, the problem with making a non standards-conforming browser is that it encourages people not to bother to follow the standards. John.

March 12, 2003

Re: [Dillo-dev] dillo source

by higuita

On Tue, 11 Mar 2003 09:59:41 -0500, Patrick Shanahan <paka(a)MyRealBox.com> wrote: > With the compression algorithms incorporated in modem transfers, the is > probably very little difference in transfer time. The bz2 files are > more tightly compressed and would allow for less compression during > transfer versus the opposite condition for gzipped files. err ... well... 8) this isnt true, the modem compression is very limited and is no diferent from the normal compression (you can even use the gzip algorithm) its limited because it can only compresses using a very small buffer, usually around 4Kb IIRC i dont know the block size for gzip, but for bzip, even in the fastest mode (-1), it is about 100Kb, for -9 its about 900Kb as we already know, compressing a compressed file is useless doesnt matter if is well compressed or badly compressed, the final file is still a very packed file that have very low repeat rate and so compresses badly no matter the options or algorithm used so when compressing text, computer side compresion (.gz, .bz2, etc) is always alot better than the modem compression because of the very low block size of the last one compressing a compressed file is useless for both computer side software and for modem side, no matter if the original file is well compressed or badly compressed if we are transfering only compressed files via a modem, we can even get a better speed by disabling the modem compression but usually isnt worth it because we transfer also text in the form of html pages, http, ftp, etc commands and finally the tcp/ip headers can be also a little compressed so using gzip or bzip2 is the same for the modem, the smaller of both will finish first all the time higuita -- Naturally the common people don't want war... but after all it is the leaders of a country who determine the policy, and it is always a simple matter to drag the people along, whether it is a democracy, or a fascist dictatorship, or a parliament, or a communist dictatorship. Voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked, and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same in every country. -- Hermann Goering, Nazi and war criminal, 1883-1946

March 12, 2003

Re: [Dillo-dev] feature request

by Andreas Schweitzer

On Tue, Mar 11, 2003 at 04:47:02PM -0700, Kenneth D. Weinert wrote: > I just wondered if there was a method to force dillo to treat a file > given on the command line as HTML. Not really an answer to your question, but ... > > I'd like to use dillo to view html attachments from my mail program, > but since the temporary file doesn't have .htm[l] as an extension > (actually, it is extension free) so dillo just displays the html > instead of interpreting it. ... reading the "user-agent" from your header, I figured you are using mutt. Then add the following into .mailcap : text/html; dillo %s; nametemplate=%s.html That's what I have :-) It may work for other MUA's as well. Optinally, add a test. Cheers Andreas -- **************************** NEW ADDRESS ****************************** Hamburger Sternwarte Universitaet Hamburg Gojenbergsweg 112 Tel. ++49 40 42891 4016 D-21029 Hamburg, Germany Fax. ++49 40 42891 4198

March 12, 2003

space in urls

by Ami Fischman

Is there some reason why spaces are dropped from URLs in dillo? Most other browsers will convert such spaces to %20 (and similarly with other illegal URI characters) automagically. It's a convinience which some webmasters take for granted (and which breaks browsing with dillo). Cheers, -- Ami Fischman usenet(a)fischman.org

March 11, 2003

feature request

by Kenneth D. Weinert

I just wondered if there was a method to force dillo to treat a file given on the command line as HTML. I'd like to use dillo to view html attachments from my mail program, but since the temporary file doesn't have .htm[l] as an extension (actually, it is extension free) so dillo just displays the html instead of interpreting it. I don't mind working on a patch if this capability doesn't already exist, but before embarking down that path I thought I'd check with the experts as I didn't see a reference to this "problem". Thanks. -- /~\ The ASCII Ken Weinert mc(a)quarter-flash.com \ / Ribbon Campaign 303-452-6603 (V) 303-705-4258 (F) X Against HTML GnuPG: 9274F1CE GnuPG available at http://www.gnupg.org/ / \ Email! 1D87 3720 BB77 4489 A928 79D6 F8EC DD76 9274 F1CE Press <CTRL>-<ALT>-<DEL> to continue ...

March 11, 2003

Re: [Dillo-dev] dillo binary size

by Raphael Barabas

On Tue, Mar 11, 2003 at 05:13:32PM -0500, Telent wrote: > Personally, my unstripped build is 5.5M; stripped, it's only 272K. doh! -rwxr-xr-x 1 raphi raphi 296164 Mar 11 23:49 dillo ahh, much better. :) raphi

March 11, 2003

Re: [Dillo-dev] dillo binary size

by Telent

> -rwxr-xr-x 1 raphi raphi 6358748 Mar 11 18:43 src/dillo Personally, my unstripped build is 5.5M; stripped, it's only 272K. -Sunny Raspet

March 11, 2003

Re: [Dillo-dev] nowrap and prefs.limit_text_width

by Sebastian Geerken

On Mon, Mar 10, Nicolas Kaiser wrote: > I'm having a little problem with sites that excessively make use of > "nowrap" tags, like <http://futurezone.orf.at> > > I toyed with limit_text_width (quote from dillorc: > > # Set this to YES, if you want to limit the word wrap width to the vieport > # width (may be useful for iPAQ) > limit_text_width=YES > > ), but I noticed that this option doesn't work as I would have expected, > because it doesn't force wrapping here (nowrap). "limit_text_width" is intended for table columns with explicitly or implicitly (e.g. by large banner images) set table widths. On futurezone, you have (IIRC; there was already some discussion exactly about this site) something like <td width=500 nowrap>, which is actually some kind of wrong HTML, since any correct interpretation would indeed not wrap the text (if I have understood it correct, strangely, most browsers *do* wrap the text). This is really a reason to send a mail to the webmaster. > Now I think in the context of this dillorc remark, it might be a good idea > to override 'nowrap' when limit_text_width is wanted. > > Am I cruelly breaking standards, or would that make sense? > Please have a look at the above site with limit_text_width=YES, and then > with the below patch. "nowrap" should be honored in any case (at least, it should not be turned off by *this* option), since it can be used for texts where wrapping makes the text unreadable. Sebastian

March 11, 2003

Re: [Dillo-dev] dillo binary size

by Simon Davis

Hi Melvin, I just checked the two binaries (0.7 and 0.7.2.1) and it seems that 0.7.2.1 is linking a few libraries in statically. It didn't occur to me to check ldd on both binaries. Thanks for the hint. SD --- Melvin Hadasht <melvin.hadasht(a)free.fr> wrote: > Hi Simon, > > on Tue, 11 Mar 2003 09:02:58 -0800 (PST) > Simon Davis <sd1324(a)yahoo.com> wrote: > > > Hi, > > > > After building dillo 0.7.1.2 and stripping the binary the file size is almost > > 1.6 megs. When I built 0.7 the final binary came out to be 289k. > > > > I built both using the uclibc library and gcc 3.2. > > > > Is there some special configuration option I need to add to make the binary > > smaller? > > > Maybe it was not dynamically linked? Try "ldd `which dillo`", it will show you > to which library it is dynamically linked. > > Cheers > > > -- > Melvin Hadasht > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev(a)lists.auriga.wearlab.de > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev __________________________________________________ Do you Yahoo!? Yahoo! Web Hosting - establish your business online http://webhosting.yahoo.com

March 11, 2003

Re: [Dillo-dev] dillo binary size

by Melvin Hadasht

Hi Simon, on Tue, 11 Mar 2003 09:02:58 -0800 (PST) Simon Davis <sd1324(a)yahoo.com> wrote: > Hi, > > After building dillo 0.7.1.2 and stripping the binary the file size is almost > 1.6 megs. When I built 0.7 the final binary came out to be 289k. > > I built both using the uclibc library and gcc 3.2. > > Is there some special configuration option I need to add to make the binary > smaller? Maybe it was not dynamically linked? Try "ldd `which dillo`", it will show you to which library it is dynamically linked. Cheers -- Melvin Hadasht

March 11, 2003

Re: [Dillo-dev] dillo binary size

by Raphael Barabas

On Tue, Mar 11, 2003 at 09:02:58AM -0800, Simon Davis wrote: > After building dillo 0.7.1.2 and stripping the binary the file size > is almost 1.6 megs. When I built 0.7 the final binary came out to > be 289k. interesting. i never looked at the size of my binary before: -rwxr-xr-x 1 raphi raphi 6358748 Mar 11 18:43 src/dillo new record :) it still starts amazingly fast compared to any other browser. i compile dillo cvs on solaris8/sparc64 using gcc 3.2.2, gtk 1.2.10. raphi

March 11, 2003