Re: [Dillo-dev]checking cache data with HEAD From: Miklos Magyari - 2001-11-29 15:15 hi, > I would love to see this fixed, but worry slightly if this will slow > down Dillo overall because we'd need to contact the server every time, > which seems to take the most time. this delay is the reason why I suggested having a config directive to turn this on/off. Of course, it would be theoretically possible to maintain a list of sites that should always check for updates and the rest shouldn't, but this seems to be over-complicated for me and doesn't fit in dillo's design goals. Miki Re: [Dillo-dev]checking cache data with HEAD From: Lars Clausen - 2001-11-29 15:03 On Thu, 29 Nov 2001, Miklos Magyari wrote: > hi guys, >=20 > I took a quick look in the docs and the source, and it _seems_ that > dillo doesn't check for updates for a cached object automatically. Is > there any reason for it, or just simply not implemented yet? > I suggest at least having a new directive in dillorc for cache > behaviour. If the user wants to always check if the cache is up-to-date, > we should use http's HEAD method to compare the cached and actual > timestamp of an object. I would be happy to make some effort > implementing this feature if you like the idea. I agree a lot! Any kind of news page (slashdot, The Register, Bulletin Boards) suffer from lack of updates, I have to manually reload them all the time. I would love to see this fixed, but worry slightly if this will slow down Dillo overall because we'd need to contact the server every time, which seems to take the most time. -Lars --=20 Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| H=E5rdgrim of Numenor "I do not agree with a word that you say, but I |------------------------= ---- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbas= ket? Re: [Dillo-dev]checking cache data with HEAD From: Miklos Magyari - 2001-11-29 14:35 > we should use http's HEAD method to compare the cached and actual > timestamp of an object. or, alternatively (what I forgot to mention) we could use conditional GET with If-Modified-Since directive in the request header. cheers, Miki [Dillo-dev]checking cache data with HEAD From: Miklos Magyari - 2001-11-29 09:16 hi guys, I took a quick look in the docs and the source, and it _seems_ that dillo doesn't check for updates for a cached object automatically. Is there any reason for it, or just simply not implemented yet? I suggest at least having a new directive in dillorc for cache behaviour. If the user wants to always check if the cache is up-to-date, we should use http's HEAD method to compare the cached and actual timestamp of an object. I would be happy to make some effort implementing this feature if you like the idea. cheers, Miki [Dillo-dev]Link to Japanese support patch seems to be down. From: - 2001-11-28 09:09 Hi, all: While I attempted to consult the [Japanese support patch] in the section of [i18n & l10n] in Dillo Links page (http://dillo.so....net/Links.html) , I found that the link http://www.ee.ed.ac.uk/~rjt/dillo/japanese.html could not be reachable. I am a user form Taiwan and about to make efforts to Chinese support for Dillo. So I hope to refer the Japanese support patch in advance. Does anyone knows the link or give me that patch? Thanks. Regards, Jim Huang -- ¡° Origin: ¦¨¥\¤j¾Ç¸ê°T¤uµ{¾Ç¨t[¾K¸ê¤ßBBS]  ¡» From: ccns.ncku.edu.tw [Dillo-dev][patch] Let Dillo display Traditional Chinese pages. From: - 2001-11-28 08:32 Hi, all: I am a new user to Dillo Web Browser from Taiwan. I found it really nice for searching or browser. However, I also noticed that Dillo couldn't display Chinese Web page well, resulting from the fact that Dillo is not built in the manner of i18n programming paradigm. So, I did a patch to Dillo to let Dillo support Traditional Chinese (the language used widely in Taiwan) display and its encoding, Big-5. As follows: in src/dw_style.c: static void Dw_style_font_realize (DwStyleFont *font, gboolean try_all) { char fontname[256], *ItalicChar; ItalicChar = prefs.use_oblique ? "o" : "i"; sprintf (fontname, "-*-%s-%s-%s-*-*-%d-*-75-75-*-*-iso8859-1", font->name, font->bold ? "bold" : "medium", font->italic ? ItalicChar : "r", font->size); ++ strcat (fontname, ++ ",-arphic technology co.-ar pl mingti2l big5" ++ "-medium-r-normal-*-*-160-*-*-c-*-big5-0"); -- font->font = gdk_font_load(fontname); ++ font->font = gdk_fontset_load(fontname); Here, I suppose that Taiwan Linux users all have "ar pl mingti2l big5" font provided by Arphic Technology. It seems to be stupid but it works like this: http://ccns.ncku.edu.tw/~jimchyun/misc/dillo-big5.jpg Regards, Jim Huang -- ¡° Origin: ¦¨¥\¤j¾Ç¸ê°T¤uµ{¾Ç¨t[¾K¸ê¤ßBBS]  ¡» From: ccns.ncku.edu.tw [Dillo-dev]Here's a URL parsing hack... From: Paul Chamberlain - 2001-11-28 03:44 Attachments: url.diff I grew tired of the javascript links that didn't work even though it was obvious to me that there was a URL embedded in that javascript. I hacked the url parser so that it looks for an embedded " if all else fails. At the very least this fixes a bunch of the links when I view news.yahoo.com. It's okay to post hacks here, right? -- Paul Chamberlain, tif@ti... Re: [Dillo-dev]Javascript ? From: Clinton Ebadi - 2001-11-28 01:18 > Currently I'm concerned with rfc2396 (details at bug-track), and > there're still some patches waiting to be merged before releasing > dillo-0.6.3. > Please say one is the cookies patch? (Dillo with cookies is pretty much feature complete for me, just add the ability to upload files and I will be the happiest man alive). > The plan is to restart working the simple-plugins after 0.6.3 > release. > > Note that this scheme is far from what's commonly known as a > plugin. It's more close to a fast-cgi interface than a dlopen > handled module. Why not use libguile? It is the official extension language of the GNU Project, and would allow fast plugin development. Best part is that you can load C code from it with (load-extension "libfoo" init_function) (the .so and version stuff is added automatically). This is only in unstable guile from CVS though (which is _much_ improved over 1.4). So, people could write their plugins in C if they wanted and simply load it into scheme and attach it to dillo, or they could write the plugin entirely in scheme (or some combination of the two). There is the problem of added memory used by guile though (last time I checked guile used something like 8MB of vm, with about 2MB resident at any time). ------------------------------- #umbclinux on irc.openprojects.net lamer.hackedtobits.com unknown_lamer@ha... AIM: unknownlamer ICQ: 41087914 Re: [Dillo-dev]Javascript ? From: mallum - 2001-11-27 16:24 on Tue, Nov 27, 2001 at 08:52:18AM -0300, Jorge Arellano Cid wrote: > > Note that this scheme is far from what's commonly known as a > plugin. It's more close to a fast-cgi interface than a dlopen > handled module. > cool. Im guessing this would make it possible to write / prototype plugins in languages like Python or Perl ? -- mallum Re: [Dillo-dev]Javascript ? From: Jorge Arellano Cid - 2001-11-27 11:55 Hi, > > Song Yun Zhao schrieb am 27.11.01: > > Just wondering if Dillo has Javascript support? > > If not, how feasible would it be to add it? > I does not have JS support and as far as I am concerned > I'd be more than glad to keep it out of dillo - or at > least to keep it in a way that I can block out the code > with a '--disable-javascript' while configuring. My feelings exactly! Javascript is out of the plans. > Eventually there is an effort to include such features > with the planned plugin architecture. Not Javascript, just other features. Cheers Jorge.- Re: [Dillo-dev]Javascript ? From: Jorge Arellano Cid - 2001-11-27 11:55 Hi, > on Tue, Nov 27, 2001 at 09:35:29AM +0100, Bjoern Weber wrote: > > > Eventually there is an effort to include such features > > with the planned plugin architecture. > > > > May I enquire on what the current state of the plugin architecture is > ? Hows it fare on the list of priority's ? Sure. Currently I'm concerned with rfc2396 (details at bug-track), and there're still some patches waiting to be merged before releasing dillo-0.6.3. Also there's the whole subject of source forge and whether or not moving away... The plan is to restart working the simple-plugins after 0.6.3 release. Note that this scheme is far from what's commonly known as a plugin. It's more close to a fast-cgi interface than a dlopen handled module. Greetings Jorge.- Re: [Dillo-dev]Javascript ? From: mallum - 2001-11-27 11:16 on Tue, Nov 27, 2001 at 09:35:29AM +0100, Bjoern Weber wrote: > Eventually there is an effort to include such features > with the planned plugin architecture. > May I enquire on what the current state of the plugin architecture is ? Hows it fare on the list of priority's ? -- mallum [Dillo-dev]dillo on QNX From: Bjoern Weber - 2001-11-27 08:37 Good morning, as I wrote before, I'm testing out dillo on the QNX RTP 6.1 and so far as I'm concerned it runs smooth, yet not as stable as on other platforms. I'm just wondering if anyone else out there is interested in running dillo on that platform or has been testing it so we can corporate our efforts to mark out the last flaws. Greetings, Bjoern --=20 Sometimes, even at infinity, the largest distance can be much too close... =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= =5F=5F=5F=5F Geben Sie Ihren Lottotipp gerne auf den letzten Dr=FCcker ab=3FBeim WEB.DE Lottoservice gibt's keine Warteschlangen. http://tippen2.web.de/=3Fx=3D9 Re: [Dillo-dev]Javascript ? From: Bjoern Weber - 2001-11-27 08:35 > Song Yun Zhao schrieb am 27.11.01: > Just wondering if Dillo has Javascript support=3F > If not, how feasible would it be to add it=3F I does not have JS support and as far as I am concerned I'd be more than glad to keep it out of dillo - or at=20 least to keep it in a way that I can block out the code with a '--disable-javascript' while configuring. Eventually there is an effort to include such features with the planned plugin architecture. Greetings, Bjoern --=20 Sometimes, even at infinity, the largest distance can be much too close... =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F Berufsunf=E4higskeitversicherung von Mamax bei WEB.DE.=20 Jetzt informieren! http://bu.web.de [Dillo-dev]Javascript ? From: Song Yun Zhao - 2001-11-26 23:49 Hi, Just wondering if Dillo has Javascript support? If not, how feasible would it be to add it? thanks Re: [Dillo-dev]list items (was: Dillo Freezing) From: Sebastian Geerken - 2001-11-26 15:44 On Wed, Nov 21, 2001 at 11:18:48PM +0100, Hugo Hallqvist wrote: > On Wed, 21 Nov 2001 19:41:00 +0100 > Sebastian Geerken wrote: > > > Hi, > > > > the freeze bugfix also solved the problems I had with list items, now > > list items are special DwPage's. There will be some enhances, probably > > within a new widget DwListItem. > > After a little more browsing. I noticed pages with numbered lists are not incremented correctly. > Look for example: http://www.blocket.se Fixed. Sebastian Re: Re: [Dillo-dev]Cookies for dessert.. : ) From: Paul Chamberlain - 2001-11-26 15:41 Lars Clausen wrote: > > What kind of mechanism will the end user have to accept/deny cookies ? > I've been thinking about how to do a good dialog to ask the user. I'm > leaning towards a standalone program that works on the cookie control > file. That way, Dillo isn't bloated with that interface, and those who > just want to use the file for control can do that. *delurking* I have also thought about using standalone programs to keep dillo from getting bloated while adding features. I thought that it could be done by implementing a "local CGI" feature, then writing CGI-like programs that output HTML forms and modify a user's local config files. It seems entirely reasonable to me to implement the entire "Preferences" UI in HTML this way. There are some drawbacks. You'd want to implement some security around this so mischevious remote web-pages couldn't try to launch local commands. Doing this means multiple executables, perhaps even multiple languages. -- Paul Chamberlain, tif@ti... Re: [Dillo-dev]tag parsing. From: Clinton Ebadi - 2001-11-26 00:19 > On Sun, 25 Nov 2001, Hugo Hallqvist wrote: > > > Is this valid html? I saw a link on linuxtoday.com today that had > > one of these in it. and it showed up unparsed. < a > > href=">mirrors.kernel.org > > While the HTML specs don't give a formal parsing rule for tags, I > doubt very much that it's correct. Unfortunately, the specs are in > the SGML specs, which seems to cost money to look at. I don't think that SGML cares about whitespace in tags...that probably should be valid html. I think that the XML standard says something about this (white space not mattering, making that valid), and since every valid xml document is a valid sgml document, that would mean that dillo isn't parsing correctly. I'll go and look it up in the standard now...(I wish I still had my annonated standard)...I can't find anything that explicitly says that is wrong, but I attempted to validated a document that uses that and I got this error: < a href=">mirrors.kernel.org. Error: end tag for element "A" which is not open; try removing the end tag or check for improper nesting of elements So, that is improper html. Linuxtoday is at fault. ------------------------------- #umbclinux on irc.openprojects.net lamer.hackedtobits.com unknown_lamer@ha... AIM: unknownlamer ICQ: 41087914 Re: [Dillo-dev]tag parsing. From: Lars Clausen - 2001-11-25 22:58 On Sun, 25 Nov 2001, Hugo Hallqvist wrote: > Is this valid html? I saw a link on linuxtoday.com today that had one of > these in it. and it showed up unparsed. < a > href=3D">mirrors.kernel.org While the HTML specs don't give a formal parsing rule for tags, I doubt very much that it's correct. Unfortunately, the specs are in the SGML specs, which seems to cost money to look at. -Lars --=20 Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| H=E5rdgrim of Numenor "I do not agree with a word that you say, but I |------------------------= ---- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbas= ket? [Dillo-dev]tag parsing. From: Hugo Hallqvist - 2001-11-25 21:29 Is this valid html? I saw a link on linuxtoday.com today that had one of these in it. and it showed up unparsed. < a href=">mirrors.kernel.org -- //Hugo Hallqvist - hugha495 at student dot liu dot se Re: [Dillo-dev]About (leaving) SourceForge :-( From: Clinton Ebadi - 2001-11-23 17:48 > What do we need to move (or distribute) dillo? > > * Download bandwidth and accurate counters > * mailing list (searchable and browseable) > * CVS > * Web site with a CGI enable-policy that lasts! > * ssh shell account for Web site administration. > > > Savannah's future looks promising, but it doesn't seem to > provide now what we need. Anyway, if they have (at least): > > * Web site with a CGI enable-policy that lasts! > * ssh shell account for Web site administration. > > I'd go there, and also set: > > * mailing list (if searchable and browseable) > > Probably downloads can be located elsewhere cause they only > provide http downloads from a directory, but no stats AFAIK. > > CVS can be set within the same machine as downloads or in > Savannah. I'm not sure which is better. Savannah has viewCVS and > that's a plus, but having downloads and CVS at the same domain > seems like a good distribution policy... > > Comments, suggestions, offerings etc, are very welcomed. > > Ah, if we finally decide to move, it'd be great to release > 0.6.3 from the new site, probably with Livio's new design. > > I'd go as stated, with CVS and downloads maintained by the > Debian-br crew! In a few months, I am planning to start a small online company (music store that only carries non-copy protected cds). If all goes well, I should have a server colocated somewhere before February or March...it will have plenty of spare bandwidth and disk space, so when I get it set up (filing incorporation papers, getting the other people in in to purchase their shares, then setting up the server and installing my code on it, and finally getting the first set of CDs), dillo can go there if you want. I plan for the server to be a mid- range 1U p3 box with about 256MB of ram and a nice 40GB hard drive, all running Debian. ------------------------------- #umbclinux on irc.openprojects.net lamer.hackedtobits.com unknown_lamer@ha... AIM: unknownlamer ICQ: 41087914 Re: [Dillo-dev]About (leaving) SourceForge :-( From: Jorge Arellano Cid - 2001-11-23 16:44 Gentlemen, On Mon, 12 Nov 2001, Livio Baldini Soares wrote: > Unfortunately I feel compelled to bring out a question that might cause > polemic. Some of you may know that VA (Linux) is changing their "ways" and > becoming more and more comercial. (BTW, VA Linux is the owner of sourceforge). > And I've recently learnt that sourforge is now using non-free software :-( > Before reading on, maybe you should read this article (highly recommended): > > http://www.fsfeurope.org/news/article2001-10-20-01.en.html and also: http://www.advogato.org/article/376.html http://slashdot.org/developers/01/11/12/1456219.shtml After careful consideration of these references, previous posts on this thread and from my personal experiences, I've come to think that unfortunately there're underhanded politics coming from VA. This is hard to write, specially when I'm both aware and grateful for all that VA has given to the Open Source & Free SW community. OTOH, I can't overlook the following points: * After months of license discussion, Loic was asked to give up his rights on contributions to the SF codebase (that were given under the GPL, and expected to remain as such), but later, without any previous warning, it was made propietary to VA. * The SF codebase that's still free is about a year old and in an unmaintainable state. Considering that, in conjunction with the time it took for Loic to sign the final document (while discussing it), the hypothesis of it all being planned seems a very plausible one. * Once on their server, you don't own the project. It can't be deleted, and the only way around is to tag it "unmaintained". This is kind of insidious! It makes people think the project is stalled without developers, that the last release is in SF and that there's no need to look for it elsewhere... * Facilities for exporting projects were "temporarily" removed. * SF is using propietary SW. There's no problem if there's a need, but being artificial, it gives the impression that Free-SW is not good enough for the task. * You're still free to use the SF server, but you won't have the freedom to copy, modify, study and distribute the SW it runs. * Technical issue: Download counters, although being essential information (both for a project and SF), are not working properly since months (maybe more than five). It's bug 440446 in SF's bug track. It was posted after a previous fix of the counters, that were reported to even go backwards sometimes. About dillo ----------- I have no problems with moving dillo, even more, if that serves to solve the download counter and the mailing list problems, it's very welcome! Note: mailing lists at SF are unmaintained by now, as they're working on a new solution... With regard to dillo, dillo-dev was created by James McCollough, he's the maintainer, but all my attempts to reach him by email had failed :(. Thus, since mailman SW started asking for an admin password, I haven't access to the susbscribers list. SF doesn't provide a way to change this, and my support requests remain unanswered... What do we need to move (or distribute) dillo? * Download bandwidth and accurate counters * mailing list (searchable and browseable) * CVS * Web site with a CGI enable-policy that lasts! * ssh shell account for Web site administration. Savannah's future looks promising, but it doesn't seem to provide now what we need. Anyway, if they have (at least): * Web site with a CGI enable-policy that lasts! * ssh shell account for Web site administration. I'd go there, and also set: * mailing list (if searchable and browseable) Probably downloads can be located elsewhere cause they only provide http downloads from a directory, but no stats AFAIK. CVS can be set within the same machine as downloads or in Savannah. I'm not sure which is better. Savannah has viewCVS and that's a plus, but having downloads and CVS at the same domain seems like a good distribution policy... Comments, suggestions, offerings etc, are very welcomed. Ah, if we finally decide to move, it'd be great to release 0.6.3 from the new site, probably with Livio's new design. I'd go as stated, with CVS and downloads maintained by the Debian-br crew! Regards Jorge. Re: [Dillo-dev]" " problem From: Clinton Ebadi - 2001-11-23 15:51 > Hello, > > first of all I have to say thanks for your work. The Dillo is > absolutely super. Maybe I found small problem, which I believe is not > difficult to fix. The problem is with  . When I used " " in my > page I expected to see empty space but I saw string " " instead. I > tried it in Netscape and Mozilla and there is behavior proper (one > empty space). Bellow is simple code demonstrating this. Can somebody > fix it, or give me clue which part of source I can fix it myself. > Entities must start with & and end with ;, I believe the proper behavior is to take the entire page after that as an entity and not display anything. So, just add a ; after nbsp ( ) and it will work. ------------------------------- #umbclinux on irc.openprojects.net lamer.hackedtobits.com unknown_lamer@ha... AIM: unknownlamer ICQ: 41087914 Re: [Dillo-dev]" " problem From: Grigory Bakunov - 2001-11-23 08:41 Date |Fri, 23 Nov 2001 09:25:58 +0100 From |Vaclav Tvrdik Hello! VT> Hello, VT> first of all I have to say thanks for your work. The Dillo is absolutely VT> super. Maybe I found small problem, which I believe is not difficult to fix. VT> The problem is with  . When I used " " in my page I expected to see VT> empty space but I saw string " " instead. I tried it in Netscape and VT> Mozilla and there is behavior proper (one empty space). Bellow is simple VT> code demonstrating this. Can somebody fix it, or give me clue which part of VT> source I can fix it myself. VT> Thank you and best regards VT> Vaclav TVRDIK VT> PGP Key : 0xA983E45F VT> file : test.html VT> VT> test VT> > VT> VT> VT>
test of nbsp 
VT> VT> HUH. AFAIK for use entities you need to correct this   to   - how it sow in http://www.w3.org ------------------------------------------------------- Grigory Bakunov ASPLinux Support Team http://www.asplinux.ru Re: [Dillo-dev]" " problem From: Bo Lorentsen - 2001-11-23 08:37 Vaclav Tvrdik wrote: > first of all I have to say thanks for your work. The Dillo is absolutely > super. Maybe I found small problem, which I believe is not difficult to fix. > The problem is with  . When I used " " in my page I expected to see > empty space but I saw string " " instead. I tried it in Netscape and > Mozilla and there is behavior proper (one empty space). Bellow is simple > code demonstrating this. Can somebody fix it, or give me clue which part of > source I can fix it myself. I think it would be a good idea to terminate the " " with a simicolon, like this " ". Hope this help ! /BL [Dillo-dev]" " problem From: Vaclav Tvrdik - 2001-11-23 08:23 Hello, first of all I have to say thanks for your work. The Dillo is absolutely super. Maybe I found small problem, which I believe is not difficult to fix. The problem is with  . When I used " " in my page I expected to see empty space but I saw string " " instead. I tried it in Netscape and Mozilla and there is behavior proper (one empty space). Bellow is simple code demonstrating this. Can somebody fix it, or give me clue which part of source I can fix it myself. Thank you and best regards Vaclav TVRDIK PGP Key : 0xA983E45F file : test.html test
test of nbsp 
Re: [Dillo-dev]Cookies for dessert.. : ) From: William Kendrick - 2001-11-22 23:23 On Thu, Nov 22, 2001 at 04:26:57PM -0600, Lars Clausen wrote: > I've been thinking about how to do a good dialog to ask the user. I'm > leaning towards a standalone program that works on the cookie control > file. Excellent. :) -bill! Re: [Dillo-dev]Cookies for dessert.. : ) From: Lars Clausen - 2001-11-22 22:27 On Wed, 21 Nov 2001, hebertrich@ya... wrote: > Gentlemen ...=20 >=20 > What kind of mechanism will the end user have to accept/deny cookies ?=20 Currently, there is a text file that controls cookies. It will allow or deny or allow non-persistently cookies from a given domain, plus allow a default setting. > What kind of built i protection will Dillo have to help the surfer > do it's thing while at the same time reject what is obviously cookies > containing personally identifiable numbers or info ?=20 > In fact is such protection feasible ? In my opinion, not really. Very few cookies will contain the information themselves, all they need is an identifying number that will let them look into the database. There's no standard for the contents of cookies whatsoever. Many are also 'encrypted'. All we can say is where they are from, how long they will last, and who will be able to get them. > Will there be for example cookies off=20 > cookies on ( accept all) ask user > accept/deny for domains... ( lil window with > a list ? ) third party cookies on ( accept > all ) ask user display cookies ( scroll list > ,click and view the cookie ) > cookie warning ( ex this cookie is > illegal...shows user name / password .. I've been thinking about how to do a good dialog to ask the user. I'm leaning towards a standalone program that works on the cookie control file. That way, Dillo isn't bloated with that interface, and those who just want to use the file for control can do that. One of the few places where I need cookies is the Dagorhir bulletin board (www.dagorhir.com/dagorhir/infoboard). Its cookies contain the username and password in cleartext, but it's not an important password, so it's ok. Most places that have important passwords will either use https or just have an identifying cookie. Trying to figure out the 'legality' of the cookie (beyond standards conformance) is IMHO futile. > I guess what i want to say is try to do your best, like usual : )= =20 >=20 > Can automatic protection be made to look into the cookie for obvious=20 > UIN or other privacy issues ?=20 As explained above, not really. The big privacy issues aren't in the contents of the cookie, but in how it's used. -Lars --=20 Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| H=E5rdgrim of Numenor "I do not agree with a word that you say, but I |------------------------= ---- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbas= ket? [Dillo-dev]Font size problem with locales, and null char in text files From: Niklas Hoglund - 2001-11-21 21:27 I'm experiencing two bugs in Dillo 0.6.2. Firstly, if I have my default locale setting, LC_ALL=sv_SE, Dillo ignores the font_factor setting and displays tiny fonts. With LC_ALL=C it works OK. Secondly, and this is more an irritating thing than a bug, when displaying for example the changelog on Dillo's homepage every line begins with a null character (the one used when the font lacks some character). This only happens on text displayed with the fixed-width font. If it matters any, I use the Debian package of Dillo. Except for that, you have my kudos for making a nice little program. -- Niklas [Dillo-dev]Cookies for dessert.. : ) From: Richard - 2001-11-21 20:35 Gentlemen ... What kind of mechanism will the end user have to accept/deny cookies ? What kind of built i protection will Dillo have to help the surfer do it's thing while at the same time reject what is obviously cookies containing personally identifiable numbers or info ? In fact is such protection feasible ? Will there be for example cookies off cookies on ( accept all) ask user accept/deny for domains... ( lil window with a list ? ) third party cookies on ( accept all ) ask user display cookies ( scroll list ,click and view the cookie ) cookie warning ( ex this cookie is illegal...shows user name / password .. I guess what i want to say is try to do your best, like usual : ) Can automatic protection be made to look into the cookie for obvious UIN or other privacy issues ? Richard Hebert On 20 Nov 2001 19:33:35 -0600 Lars Clausen wrote: > On Tue, 20 Nov 2001, Clinton Ebadi wrote: > > > I noticed the patches for cookies and https (among other things) on > > http://haru.wom.ru/pub/dillo/patches/. Are there any plans to merge > > these into the main dillo source tree? Cookies would be useful to me > > (i.e. for viewing sites like kuro5hin where you have to be logged in > > to do important things like post). > [...] > > The cookie patch out there is only half of the story. Jorgen Viksell had > created a different cookie patch that to some degree complemented mine. We > have combined the two, and now have error checking, file-based cookie > control and persistence. We shall send out the full patch soon. > > -Lars > > -- > Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor > "I do not agree with a word that you say, but I |---------------------------- > will defend to the death your right to say it." | Where are we going, and > --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbasket? > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev@li... > https://lists.so....net/lists/listinfo/dillo-dev _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com Re: [Dillo-dev]list items (was: Dillo Freezing) From: Hugo Hallqvist - 2001-11-21 18:53 On Wed, 21 Nov 2001 19:41:00 +0100 Sebastian Geerken wrote: > the freeze bugfix also solved the problems I had with list items, now > list items are special DwPage's. There will be some enhances, probably > within a new widget DwListItem. I've tested the patch for a moment now, and it seems to handle every "problem-page", I've tested, fine. For example: http://www.tomshardware.com http://www.phatboydesigns.net/alloywm/index.html Very nice! Keep up the good work! -- //Hugo Hallqvist - hugha495 at student dot liu dot se [Dillo-dev]list items (was: Dillo Freezing) From: Sebastian Geerken - 2001-11-21 18:41 Hi, the freeze bugfix also solved the problems I had with list items, now list items are special DwPage's. There will be some enhances, probably within a new widget DwListItem. Sebastian Re: Dillo Freezing (was: Re: [Dillo-dev]DwPage Update) From: Sebastian Geerken - 2001-11-21 18:16 On Sat, Nov 17, 2001 at 11:54:04AM -0200, Livio Baldini Soares wrote: > Livio Baldini Soares writes: > > > > If there is any other info you'll like from me, just say so. I'll try to come > > up with a reproducible test, but my feeling is that this is like a race, or > > something very rare... > > > > Ok, it seems that this page causes the mentioned error about half the time: > http://so....net/softwaremap/trove_list.php?form_cat=87 > > If it doesn't well then try resizing the dillo window _while_ the page is > loading (or reloading). Here's what I got this time: > > Dw_table_sub_calc_col_widths (sub=0x82c63e8, width=1260419230, total_width=1260419229) > > Which is bogus, so Dw_table_size_request must be doing something wrong with > it's calculations. > > hope this helps, Yes, thanks, it helped a lot to track this bug down. It should finally be fixed, please post your experiences. Sebastian [Dillo-dev]it works! From: Clinton Ebadi - 2001-11-21 16:11 Cookies work. Yes indeed, they work well. The fact that I am sending this message from dillo proves it (i am away from my home box, and am using OpenWebMail on my mail host). This needs cookies to work, and it is working. So, any plans to merge this in to cvs? Woo, now I never have to leave dillo (I don't go anywhere that uses frames anyway). My one last reason for using anything other than dillo is gone :-) ! -- + hackedtobits.com -- the home of irresponsible journalism http://lamer.hackedtobits.com -- the home of the lamer Re: [Dillo-dev]DwPage Update From: Sebastian Geerken - 2001-11-21 15:40 On Fri, Nov 16, 2001 at 04:35:28PM -0200, Livio Baldini Soares wrote: > Yeeeaaaah! :-) This bug was kind of irritating me... Well, I don't know if > you're aware of it or not, but this also fixed bug #252! Yes, handling of breaks was changed a bit. Bug tracking engine is updated. Sebastian Re: [Dillo-dev]Cookies and https From: Lars Clausen - 2001-11-21 15:32 On Wed, 21 Nov 2001, Grigory Bakunov wrote: > Date |Tue, 20 Nov 2001 17:11:09 -0500 >>From |"Clinton Ebadi" >=20 > Hello! >=20 > I want to say - all who use old cookies patch > MUST update it from patch'o'matic http://haru.wom.ru/pub/dillo/patches/ > Now this patch is realy greate! it's not only save cookies > do ~/.dillo, it's can control ACCEPT/REJECT cookies. >=20 > Nice work guys! Thank you! It still has one problem: The order of the control file is important, so if you have DEFAULT: DENY at the top, no (new) cookies are allowed, regardless of what's below. -Lars --=20 Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| H=E5rdgrim of Numenor "I do not agree with a word that you say, but I |------------------------= ---- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbas= ket? Re: [Dillo-dev]Cookies and https From: Grigory Bakunov - 2001-11-21 07:33 Date |Tue, 20 Nov 2001 17:11:09 -0500 From |"Clinton Ebadi" Hello! I want to say - all who use old cookies patch MUST update it from patch'o'matic http://haru.wom.ru/pub/dillo/patches/ Now this patch is realy greate! it's not only save cookies do ~/.dillo, it's can control ACCEPT/REJECT cookies. Nice work guys! Now i only need to join http_auth and cookies patches to enjoy ;-) ------------------------------------------------------- Grigory Bakunov ASPLinux Support Team http://www.asplinux.ru Re: [Dillo-dev]Cookies and https From: Lars Clausen - 2001-11-21 01:33 On Tue, 20 Nov 2001, Clinton Ebadi wrote: > I noticed the patches for cookies and https (among other things) on=20 > http://haru.wom.ru/pub/dillo/patches/. Are there any plans to merge=20 > these into the main dillo source tree? Cookies would be useful to me=20 > (i.e. for viewing sites like kuro5hin where you have to be logged in=20 > to do important things like post). [...] The cookie patch out there is only half of the story. Jorgen Viksell had created a different cookie patch that to some degree complemented mine. We have combined the two, and now have error checking, file-based cookie control and persistence. We shall send out the full patch soon. -Lars --=20 Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| H=E5rdgrim of Numenor "I do not agree with a word that you say, but I |------------------------= ---- will defend to the death your right to say it." | Where are we going, and --Evelyn Beatrice Hall paraphrasing Voltaire | what's with the handbas= ket? Re: [Dillo-dev]Cookies and https From: Carl Wilhelm Soderstrom - 2001-11-20 22:50 > Note the "do other stuff" part...yes, I can use galeon or skipstone, > but then I can't do much else. Usually I am running xchat, xmms, > gaim, a bunch of wterms, and dillo. I can do that fine. If you > replace dillo with galeon, I have to lose gaim, xmms, and a few > wterms to make the box usable. point taken. I can't gainsay that dillo is a *lot* better than just about every other browser out there, in terms of memory use. Just saying that it's not hopeless without it. :) Carl Soderstrom. -- Network Engineer Real-Time Enterprises (952) 943-8700 Re: [Dillo-dev]Cookies and https From: Clinton Ebadi - 2001-11-20 22:38 > > I am in a bit of a tough situation > > with dillo--it is the only browser I can use on my box, since I have > > a mere 32MB of ram (and cannot install Netscape Navigator as it is > > non-free software). > > I used to run Galeon in 32MB of ram, with acceptable performance. it's > no speed demon, but it's not bad. was a lot better on a Cyrix 166/32MB > than Netscape was on a 486/100/32MB. on this same machine, Mozilla was > nearly unusable. > Note the "do other stuff" part...yes, I can use galeon or skipstone, but then I can't do much else. Usually I am running xchat, xmms, gaim, a bunch of wterms, and dillo. I can do that fine. If you replace dillo with galeon, I have to lose gaim, xmms, and a few wterms to make the box usable. > it makes a world of difference to use a lightweight window > manager/desktop environment. KDE and/or GNOME+Sawfish will crush > anything with less than 64MB; jettison those if you haven't already. I > use FVWM; but Blackbox or WindowMaker will work about as well. Galeon > takes up a bit more memory than Konqueror, but (IMHO) has a much > nicer user interface. give it a try. :) I run WindowMaker 0.70. I have no need for desktop evironments. I even have my WindowMaker settings tuned way down, so it uses a mere 700K of ram (including all of my dock apps). I have a solid color for a background, and make sure to minimize my background stuff to keep ram for myself (about the only daemons I run are boa and cron). ------------------------------- #umbclinux on irc.openprojects.net lamer.hackedtobits.com unknown_lamer@ha... AIM: unknownlamer ICQ: 41087914 Re: [Dillo-dev]Cookies and https From: Carl Wilhelm Soderstrom - 2001-11-20 22:19 > I am in a bit of a tough situation > with dillo--it is the only browser I can use on my box, since I have > a mere 32MB of ram (and cannot install Netscape Navigator as it is > non-free software). I used to run Galeon in 32MB of ram, with acceptable performance. it's no speed demon, but it's not bad. was a lot better on a Cyrix 166/32MB than Netscape was on a 486/100/32MB. on this same machine, Mozilla was nearly unusable. it makes a world of difference to use a lightweight window manager/desktop environment. KDE and/or GNOME+Sawfish will crush anything with less than 64MB; jettison those if you haven't already. I use FVWM; but Blackbox or WindowMaker will work about as well. Galeon takes up a bit more memory than Konqueror, but (IMHO) has a much nicer user interface. give it a try. :) Carl Soderstrom. -- Network Engineer Real-Time Enterprises (952) 943-8700 [Dillo-dev]Cookies and https From: Clinton Ebadi - 2001-11-20 22:03 I noticed the patches for cookies and https (among other things) on http://haru.wom.ru/pub/dillo/patches/. Are there any plans to merge these into the main dillo source tree? Cookies would be useful to me (i.e. for viewing sites like kuro5hin where you have to be logged in to do important things like post). I am in a bit of a tough situation with dillo--it is the only browser I can use on my box, since I have a mere 32MB of ram (and cannot install Netscape Navigator as it is non-free software). Having to load konqueror to do something as simple as post a message to k5 is painful (especially since I can't do much else when running konqueror or mozilla). If no one else wants to do it, I will manually apply the patches to the stuff in cvs (since cookies was done for 0.62) and submit a cvs diff (just ask). Even without cookies, dillo still rocks. Uses less than 1MB of ram (woo, no dicache is great!) and renders most sites (one exception: any of the forums in forums.newdoom.com...) extremely well. My only complaints would be a lack of cookies, https, frames, and disk cache (you see, I am on a dial-up so...). I can live without all of those (except maybe cookies). I had an idea for implementing frames: since dillo has the html widget, why not just use a GtkLayout and embed multiple dillo widgets inside? Anyway, I can apply those patches and send a big diff if you'd like (just say what patches and I'll start working on it when I get the message). ------------------------------- #umbclinux on irc.openprojects.net lamer.hackedtobits.com unknown_lamer@ha... AIM: unknownlamer ICQ: 41087914 Re: [Dillo-dev]dillo strange behaviour (fwd) From: Jorge Arellano Cid - 2001-11-20 17:17 Hi everyone! On Tue, 16 Oct 2001, Grigory Bakunov wrote: > I found realy strange dillo behaviour on rpmfind.net site. > > look - at first search we have this html: > ----- HTML-phase-1 ------------------------------------ >
> > >
> ------------------------------------------------END------- > > so, realy good work form. after input into 'query' field text '12345' > and press submit we pass to rpmfind.net site next url query: > http://rpmfind.net/linux/rpm2html/search.php?query=1234&submit=Search+... > and now form in html look like this > > ------- HTML-phase-2--------------------------------- >
> > >
> ------------------------------------------------END----- > As all see - authors of rpmfind site place our 'query' string into form action tag. > so, after input into query field '4321' and press submit - we have VERY strange URL string: > http://rpmfind.net/linux/rpm2html/search.php?query=1234?query=4321&submit=Search+... > look! we have to 'query' variable in url. but RFC says it's not correct. > and at least - we have absolutely wrong html form text - > > ------- HTML-phase-3--------------------------------- >
enctype="application/x-www-form-urlencoded" method="GET"> > > >
> ------------------------------------------------END---- > > so, we need to remove dublicate variables when insert new variables by GET method. > Thanks for all who be so patient and read this message to the end. By the time this report was posted, rpmfind.net was down or unaccessible from here. But now it seems to be back again and the patch for BUG#255 I just commited, should also fix this bug! Note: to remove duplicate variables is NOT the rigth solution. It's a matter of resolving as RFC2396 suggests! BTW: the current URL resolver fails a lot when contrasted with RFC2396's examples, but it doesn't seem to cause much trouble because it happens in not-so-common cases. Anyway, now I'm playing with a parsing prototype that uses regular expresions. If there's a knowledgeable volunteer (versed in url.c, RFCs, RE, ...) willing to finish it please drop me note. If not, I'll keep working on it, because it is a very sensible part of the browser. Cheers Jorge.- Re: [Dillo-dev]Forgot (fwd) From: Grigory Bakunov - 2001-11-19 15:08 Date |Mon, 19 Nov 2001 11:22:00 -0300 (CLST) From |Jorge Arellano Cid Hello! JAC> (this is a fwd, because I got no answer to the original) huh. i always forget to answer. ;-) realy sorry. JAC> ---------- Forwarded message ---------- JAC> Date: Tue, 6 Nov 2001 21:03:42 -0300 (CLST) JAC> From: Jorge Arellano Cid JAC> To: Grigory Bakunov JAC> Subject: Re: [Dillo-dev]Forgot JAC> Grigory, >> Also i make small sites with all patches what i found at this moment. >> http://haru.wom.ru/pub/dillo/patches/ >> All patches cleaned for current cvs version. >> You can add your own patch by form on site. JAC> I gave the path-o-matic a try! Nice work. JAC> One of the reasons I had for not enabling sourceforge's one, is JAC> that currently I have more than enough things to worry about, and JAC> adding more is a NO-NO for me. JAC> Well, if you add an explanation that those patches are there JAC> for improvement, that they haven't been accepted yet, and that JAC> using them may break otherwise good behaviour, that they're JAC> posted there to let interested developers work them out without JAC> trashing the mailing list, and also ask the poster to explain JAC> briefly what they're known to break and warn that the submission JAC> MUST be done using a browser other than dillo (no file upload JAC> facility yet), I'll gladly add a link to it from somewhere JAC> convenient in the site. Realy all my problems in too short english practice - if somebody write this explanation for site - i place it on first page and make more frienly interface. also i want to add optional