On Wed, Dec 29, 2010 at 4:08 AM, Maciej Piechotka <[email protected]>wrote:
> On Tue, 2010-12-28 at 15:19 -0800, Sriram Ramkrishna wrote: > > > > It will be important to get people writing extensions before the > > release happens i think. (sorry went into marketing mode...!) > > I take the liberty of saying that as person who tried to write an > extension (some time ago): > > - Lack of basic documentation. Reverse engineering C/JavaScript to > write an extension (for example - how to use St) is very hard. > The Gnome shell guys only have limited resources. Owen and Jon are more than happy to help if someone wants to volunteer to do them. But their focus is on getting shell feature complete for the first release. Having done some of the work, it would be great to document what you've done thus far and then consult Jon and Owen on the limitations like reverse engineering. I know for a fact that a lot of components still haven't gotten the inspection work on it and they are looking for volunteers for that as well. Some are pretty easy others take some work. > > - Lack of basic tutorials. There no Hello World tutorial, video or > anything (at least linked from page). [I distinguish between > step-by-step tutorial and detailed documentation] > Yes, we do need to have that before 3.0. This would in fact be a great article on Gnome Journal if someone wants to help writing that for fun and glory. > - High overhead of code - if I remember correctly adding button > requires adding to 2 files (CSS and JS) - where button should be put/how > it behaves and how does it look like. Other technologies (gtk+ + > gtkbuilder, html+css+js) allows to work on the scaffolding first and > then work on details. Possibly 'sane' defaults and primitives would > help. > We probably want to move this to gnome-shell mailing list. > > - Lack of sandboxing. I'm not quite sure if and how it can be done but > last time error crashed whole shell. Since JavaScript gives > <del>little</del><ins>no</ins> compile-time checks and errors do happen > it would be helpful. > > Yeah, we probably want to move this over as well. It doesn't particularly address the current topic in regards to fallback to Gnome classic. sri
_______________________________________________ desktop-devel-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/desktop-devel-list
