Re: Tear 0.2 - a simple maemo WebKit browser.
 

Whoa! I'm just wondering to contribute. In summer I made some programming with vala and webkit to make a browser for maemo based on webkit but since I had a x64 ubuntu installed I was unable to make the sdk work. Now I installed a 32 bit Ubuntu. I have some ideas i would like to share with you.

Re: Tear 0.2 - a simple maemo WebKit browser.
 

Okay, go ahead :) Do you want write access? Or just ideas? :)

Re: Tear 0.2 - a simple maemo WebKit browser.
 

juergbi was saying on #maemo that he was going to look at the build script in mud and updating them for the newer versions of libgee & vala.

Then, it should just be a case of ensuring that vala is in your Build-Depends. I've not yet seen how Vala is supposed to handle additional vapis which aren't installed in the system-wide folder.

Worst case: you should be able to bundle them with your source?

Re: Tear 0.2 - a simple maemo WebKit browser.
 

They're already in Garage SVN, i didn't specify them explicitly in the build script, but maybe i should.

Re: Tear 0.2 - a simple maemo WebKit browser.
 

I haven't write any code yet so svn write access doesn't matter so much. I've been watching the code and I see you are storing the bookmarks on a sqlite database. I think It would be better if we integrate with maemo's bookmarks. Some time ago, I send an email to the maemo-developers list if there was some way for accessing the bookmarks. As it seems, the piece of code for doing this is closed source and Nokia isn't providing any specification for accessing it. Fortunately, the file (/usr/share/bookmark-manager/bookmarks/MyBookmarks.xml) where they are stored, it's an xml and complies with the xbel standard, so it shouldn't be so difficult to write some library for accessing them. Related to this, I would like to see some support for plugins to add bookmarks support so we can later add del.icio.us and related sites support, or send a link through IM, email or even sms.

I've been trying to compile webkit svn version on maemo's sdk but it doesn't make for me. It should be great if webkit-gtk developers can fix it, and even more wonderful if they can add support for finger navigation if hildon extensions are activated at compile time.

Eventually, I think I would start writing a lib in vala for accessing xbel files.

Re: Tear 0.2 - a simple maemo WebKit browser.
 

I'm planning on including plugin support, since it shoud be quite easy in Vala. As for MicroB xml bookmarks - if they comply to a standard, isn't there already some C lib to support it and we can just create some bindings for it (if there are none already)?

I also didn't have any luck with latest webkit commpilation - of about a dozen revisions i've tried, only one compiled, but javascript engine crashes on any mention of js in the page. Still trying though, since the march one doesn't have zoom and can't retrieve favicons from it (i want to add them in the history lists). The favicon issue can be worked around, but the zoom is needed.

Re: Tear 0.2 - a simple maemo WebKit browser.
 

For what it seems, there isnt any lib in c for accessing xbel files. Also, Midori has an internal implementation we could use. Favicon icons should be necessary also for bookmarks.

Re: Tear 0.2 - a simple maemo WebKit browser.
 

Probably, it is libxml based anyway.

XBEL Documentation (not too big):
http://xbel.sourceforge.net/language...ment-structure

SQLite will probably be faster and more memory efficient than internal XBEL implementation, maybe some syncing with MicroB's bookmarks and also other sources will be best (like delicious for instance). Of course, you can prove me wrong :)

Re: Tear 0.2 - a simple maemo WebKit browser.
 

Thanks for Tear! Looks great and renders fast on my N810 with Diablo. Much better than MicroB, and it didn't crash yet.

Re: Tear 0.2 - a simple maemo WebKit browser.
 

You are right that probably using sqlite would be more memory efficient than xbel approach. Also, I'm thinking on refactoring the code so we could replace the bookmarks backend without much effort if we at some point in the future want an xbel backend. This would also make easy to make providers for del.icio.us and so. I'm going to send you my email so we can contact each other easily. Should we make a ToDo list so we know what to work on, without overlaping each other?