Symbie

[SamuraiJack1]

Symbie (http://symbie.org) - is the framework for one-page web-sites (actually applications). Its currently under active development, if someone wants to participate - PM me, or post on symbie.org forum.

[mask_hot]

Is there a downloadable version available?

it looks really promising!

[SamuraiJack1]

Not yet unfortunately, also no documentation - only the sources for now )

[SamuraiJack1]

The whole project is also migrating to Joose (http://code.google.com/p/joose-js/) - the most advanced implementation of class system for javascript - so the primary API is mutating..

[dolittle]

Can you explain why do you need Joose if you are using ext?
What is the license of symbie?

Thanks

[tof]

Hi,

I'm also interseted in Why and How you used Joose ?

Thanks :)

[SamuraiJack1]

Well, Joose is a class system implementation only, it doesnt offer any DOM manipulation capabilities. Meanwhile the class system implementation in Ext isn't perfect - see this thread for example:
http://extjs.com/forum/showthread.php?t=49083

And there are many cases (for example plugins on class level), when the elegant solution of problem is available with Joose. Joose also have established packaging system and joose community is currently developing a standard for distributing a javascript package.

Currently I'm integrating jScout in Joose, so it will have optimized loading features soon.

Then, I'll implement bridge from Joose to Ext - and all features of Joose will be available for Ext applications (Symbie sites are Ext applications also).

And then, Symbie will be refactored with Joose in mind, joosified ))

As about license - currently Symbie is GPL or commercial (free for active contributors).

[IT100]

As I'm looking for a solution on how to index javascript code I stumbled upon your
symbie solution. I found you are exploring server site generating code for robots?

I was wondering if you have any progress on this? I would be willing to join the effort if wanted.

I started this http://extjs.com/forum/showthread.php?t=53933 tread involving my one page demo site.

Mjlecomte directed me kindly to your solution...


Regards,

Chris.

[SamuraiJack1]

Yes, I have a certain progress with indexing. The solution will be based on "Mozilla Embedding Widget".
more info here:
http://www.mozilla.org/projects/embedding/
http://www.mozilla.org/unix/gtk-embedding.html

I'm using perl wrappers around this module (Mozilla::DOM and Gtk2::MozEmbed) and was able to successfully receive the html sources of URL from the perl program, so the "proof of concept" is done.

The next step to create a standalone "rendering service" for web-robots and browsers with disabled javascript. It will be a sub-project of Symbie, but I'm going to create it as universal configurable solution suitable for any javascript site.

You are welcome to join (PM me).

Regards

[SamuraiJack1]

New site, powered by Symbie:

http://zemfiras.com/

(photographer's portfolio, plus administration panel for albums creation)

[SamuraiJack1]

I'm glad to announce, that search engines indexation problem is solved and Symbie applications are now SEO friendly.

As planned, the created component is a small, standalone perl script, independed from the application logic. It operates with perl wrappers (Mozilla::DOM (http://search.cpan.org/%7Eslanning/Mozilla-DOM-0.21/) and Gtk2::MozEmbed (http://search.cpan.org/author/TSCH/Gtk2-MozEmbed-0.08/)) around the GtkMozEmbed (http://www.mozilla.org/unix/gtk-embedding.html).

Search engines spiders are determined by request's UserAgent header (currently with a simple regex matching /google|yandex|yahoo|bot/).

If the UserAgent match - then the request is handling with script, otherwise - as usual.
Generally speaking spiders will see the exact copy of the site, rendered on Gecko engine.

To test how it works, change your browser's UserAgent (http://johnbokma.com/mexit/2004/04/24/changinguseragent.html) for example to "google" and disable JavaScript. Then open any of "powered by Symbie" sites:
http://be-travel.ru
http://symbie.org
http://zemfiras.com

Mention, that you (and spiders) can also navigate through the sites via usual <a> links (best seen on be-travel.ru, click on photo's thumbnails)