- Dillo-dev - lists.mailman3.com

[RFC] Quo Vadis Dillo

by Frank de Lange

Hi'all, Now that the last release of the patches seem to work (number 13 was less of a success...)... Where do we want to be tomorrow? What needs to be added/changed to improve Dillo? Should the tab/frame/etcetera + Indan's patch + ??? be forked into an 'experimental' version, somewhat like the Sylpheed MUA was forked into Sylpheed-claws? Talking about my patch, what does need improvement? Some ideas: - select contents of location bar/search field when they are focused (DONE, in next version) - make the 'search the web' popup disappear when popup_dialogs=NO, replace it with a non_popup version or use the location bar for search (like several other browsers do) - more key bindings (replacing current key handling where applicable) - remove compile options for tabs and frames, make tabs/frames standard? [ PDA/embedded developers: is this OK with you? ] - ??? Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

Nov. 11, 2003

Re: Keybd 3 patch

by Frank de Lange

Jorge Arellano Cid wrote: >Hi Frank, > > I just noticed the newes keyboard navigation patch in the >mailing list. Please tell me if this problems were fixed or if a >should wait the next version before aiming to a merge: > > These problems are gone. I have completely overhauled the internals so that focus always is in sync with the rest of Dillo. Which just happened to fix the problem Sebastian noticed... I'd say, wait a week or so to see if any issues crop up. If all seems OK commit the patch. [ LIST: please give keyboard navigation, on either version of the patch, a workout. This needs some testing if it is to be part of mainstream Dillo. Try it in frame sites, try navigatin between frames with Tab/Shift-Tab, try the keyboard bindings, etc. ] By the way, the patch also contains a fix to a problem with dpid/Makefile.am (which did not honour DESTDIR on installation, messing up make install for non-root users). Let me know when you apply this, because it will affect the tab/frame/kbnav patch s well. Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by Paul Pelzl

On Tue, Nov 11, 2003 at 10:37:41AM +0000, robert w hall wrote: > To reiterate > It would be _really nice_ if all these links to these very useful > patches were findable from the dillo main site :-) Perhaps some enterprising individual (not me) would be interested in hosting some sort of Dillo Wiki. Any patch authors could then post links on the wiki without having to bother website maintainers... Paul

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by Indan Zupancic

> Note: When Indian continues to maintain his https patch i'll probably do > not bother with mine, lazy as i am... Also the Indian's idea of cleaning > up ccc into something more easy to understand is nice (current ccc has at > least one bad property - it is _hard_ to understand what goes on in it). It is my intention to keep my https up to date, it will be in a couple of days. I'm more or less waiting till Jorge tells me that I may go ahead with writing a patch that does that IOData and GDK stuff, or that he wants to keep it as it is now. That way it's clear for me what I can do, and how the more final version of my https patch should look like. I started documenting how the CCC functions are used currently in Dillo. After having it clear what happens now it's easier to see how to improve it, and what's wrong with the existing implementation. I do not know how hard it is to manage an experimental CVS branch, it has advantages and disadvantages. One pro is that the patches always work with eachother, because they use the same code, but that also means that the patches won't work seperately, that is, there is just an extended Dillo version, and no real patches anymore. It is very good to work with a stable code base, because then it's clear that all the encountered bugs are caused by your code, and not by someone else's code. Also it's useful when the base code doesn't change that much, especially when the patch is young. It is easier to keep the code up to date when it has it's own cvs branch, but on the other hand it could also mean that if someone drops his code and stops working on it, that someone else need to keep it up to date. Altogether quite risky. Of course everything is much simpler when only a few people use the alternative branch, it shouldn't be a dump place for never updated patches (tiny ones are ok of course). I think it will be only succesful when the branch is used only for things that have a chance to come in the official Dillo, as a sort of testing and developing place. Otherwise it will be very hard to keep it in sync with the official Dillo version, and the changes will only become greater after a while. Greetings, Indan

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by robert w hall

In message <Pine.LNX.4.58.0311111201430.31566(a)kiku.itsise>, Madis Janson <madis(a)ats.cyber.ee> writes >you can use my patch for 0.7.3 until something better appears... > >http://www.zone.ee/myzz/dillo/0.7.3/dillo-0.7.3-mzz-wtab.diff.bz2 >http://www.zone.ee/myzz/dillo/0.7.3/dillo-0.7.3-mzz.html > To reiterate It would be _really nice_ if all these links to these very useful patches were findable from the dillo main site :-) -- robert w hall

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by Madis Janson

On Tue, 11 Nov 2003, Mike Eriksen wrote: > Sorry to interrupt... I'm a keen user of Dillo but no hacker. I'll not > comment on all the internals of Dillo since you guys are much more clever > on this than me. > > But, see - the reason why I like Dillo is that it is light and mean - and > can be tweaked into something I need. My need is a *light* browser which > can be fitted into a thin client but still allow you to read web mail. > This requires frame and https support. Other users of Dillo, like PDA > users, couldn't care less about frames and https. > you can use my patch for 0.7.3 until something better appears... http://www.zone.ee/myzz/dillo/0.7.3/dillo-0.7.3-mzz-wtab.diff.bz2 http://www.zone.ee/myzz/dillo/0.7.3/dillo-0.7.3-mzz.html It contains both tabs patch v6 and one https implementation (mine), but you should read sslgw/README to get it work, if you want to use it. Note: When Indian continues to maintain his https patch i'll probably do not bother with mine, lazy as i am... Also the Indian's idea of cleaning up ccc into something more easy to understand is nice (current ccc has at least one bad property - it is _hard_ to understand what goes on in it). > The key point here is that neither I nor the PDA guys really have > anywhere else to go than Dillo. >

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by Frank de Lange

Indan Zupancic wrote: > Having a central place for all Dillo patches is always useful, because > that way it's easier to find and use them for users. Centralised isn't the > same as semi-official, it's a united effort of the patch makers. And that > means they probably have something betters to do, say, improving their > patch and keeping it up to date with cvs. ;) Another thing we can do is go the 'sylpheed-claws' route: create an extended version of Dillo, which regularly syncs up with Dillo CVS (and, possibly, the other way around). That way it is easier to create and maintain those features which are missing from Dillo. The risk is of course that the extended version starts to divert so far from CVS that it is hard to use it as a base to patch CVS. I've thought about going that way with the tab/frame/keyboard_navigation patch, as it does not get that much exposure currently. I have not done it yet, but given the number of patches which languish because of lack of exposure it is probably a good start. An 'experimental' branch on CVS and the main Dillo site would work as well of course... Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

Nov. 11, 2003

Re: libpng detection again...

by William Trenker

Jorge Arellano Cid wrote: > <candidate line> > CPPFLAGS="$CPPFLAGS `$PNG_CONFIG --cppflags`" > </candidate line> Yes! Hooray! I've been struggling with this for months and at last I have Dillo back again. After applying this change to configure.in and re-running autoconf the ./configure worked and I was able to build Dillo without a hitch. Thank you very much. And thanks Dillo team for a great browser! Regards, Bill

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by Indan Zupancic

> When I tried to add Indans https patch, it > didn't work. Too old. Two weeks ago I tried the same. At that time the > problem was the opposite! Plain bad luck, Jorge just changed those files enough, and I was bussy with making a pc stable and upgrading it, all my Dillo files are backuped on another pc now, it's a slight mess. :) Also I want to merge the read/write IOData structures, and change to the GDK API, if Jorge and others agree with that. If that's going to change then updating my https patch now is a waste of time, because relatively much is going to change (the https patch will shrink incredibly :). > But in my humble oppinion either Dillo should release official minor > version (0.7.4, 0.7.5 etc) more often so privateers can offer patches > against these, or go along with this suggestion of Indans and support a > semi-official patch model. More often is not the point, Dillo is relatively close to 0.80 and far away from 0.73, so this seems to be just an unfortunate time for patches. Normally you should just use the 0.73 patches. There is no 0.73 https patch from me because I'm not happy with the current implementation, caused by all the hassle with keeping the two IOData structures synced and GTK's io channel annoyance. If there are fewer official Dillo versions it's easier to keep all the patches up to date. Actually, all what is needed is a cvs tag for each stable release, and maybe some extra tag for sort of beta versions that can be used for patches if it turns out that the official releases are not often enough for the patches. This way it's clear for the patches what Dillo version to use, so all can be relatively easily kept synchronized (at least for what Dillo version the patches use). Having a central place for all Dillo patches is always useful, because that way it's easier to find and use them for users. Centralised isn't the same as semi-official, it's a united effort of the patch makers. And that means they probably have something betters to do, say, improving their patch and keeping it up to date with cvs. ;) Greetings, Indan P.S. Seems like I just can't write short mails. :(

Nov. 11, 2003

Re: [Dillo-dev] HTTPS Patch

by Mike Eriksen

On Mon, 10 Nov 2003, Indan Zupancic wrote: [cut - a lot!] > > Making a web site (or page for it) requires time, if you're > > willing to do that, go ahead but beware that we do not support > >unofficial patches so you'd have to answer all the incoming > >questions too. > > Hmm, a web page, mind if I use the code of the splash screen as a basis? I > only want to make an installer so that it's easy for people who, for > instance, don't know what cvs is, to get a patched Dillo version. The goal > is to have some sort of centralized patch place, with patches that work > with eachother and are up to date. It requires time, true, but I'm willing > to spend time on Dillo. The installer has no priority though, would rather > spend my time to unite the IOData structures, switch to the GDK api, and > clean up the https patch for now. > > All the things I want to have different in Dillo I'm willing to change > myself. I'm not saying you should change this or that so or so, I ask if > you mind if I change or add some things a certain way, and if you agree > that it's better that way. > > Greetings, > > Indan Sorry to interrupt... I'm a keen user of Dillo but no hacker. I'll not comment on all the internals of Dillo since you guys are much more clever on this than me. But, see - the reason why I like Dillo is that it is light and mean - and can be tweaked into something I need. My need is a *light* browser which can be fitted into a thin client but still allow you to read web mail. This requires frame and https support. Other users of Dillo, like PDA users, couldn't care less about frames and https. The key point here is that neither I nor the PDA guys really have anywhere else to go than Dillo. I need to patch up Dillo with both Franks tab/frame patch and Indans https patch. Currently Franks patch is at ver. 14. The latest patch working on the most recent official Dillo release (0.7.3) is - as far as I remember - version 6. Today, when Franks new patch came out, I patched the Dillo cvs version with it. No problem. When I tried to add Indans https patch, it didn't work. Too old. Two weeks ago I tried the same. At that time the problem was the opposite! Honestly, this isn't a good situation. After following the mailing list for a couple of months I understand that the core developers have strong oppinions on the structure of the internals of Dillo. The is probably good since this created Dillo as it is in the first place. But in my humble oppinion either Dillo should release official minor version (0.7.4, 0.7.5 etc) more often so privateers can offer patches against these, or go along with this suggestion of Indans and support a semi-official patch model. The current model will not bring Dillo forward with the necessary speed, I fear. A lot of good ideas need to be tested and eventually absorbed if proven good. > (I wonder how many people wil read this 13K monster :) I did! And this post even looks short now ;-) Mike -- Thinstation FAQ maintainer http://thinstation.sourceforge.net - a light, full featured linux based thin client OS

Nov. 10, 2003