Re: [gedit-list] wrapping gedit-dev

[Sébastien Wilmet <swilmet gnome org>]

On Wed, Mar 26, 2014 at 10:51:49AM +0100, Matěj Cepl wrote:
> Well, even better would be if we could have finally 
> document-type-specific plugins (like Vim and Emacs have). Then 

gedit-code-assistance already do that. gnome-code-assistance has several 
backends for different programming languages.

> the LaTeXila wouldn’t have to have FAQ:
>
>     Why LaTeXila is an independent application instead of 
>     a gedit plugin?
>
>     An independent application is easier for the users. The 
>     LaTeX-specific features of LaTeXila are more well 
>     integrated. 
>
> I don’t think there is any good reason for that, just that we 
> cannot hide some plugins which are not applicable, so when using 
> more plugins, the result is a mess.

An independent application has other advantages, as explained in the 
wiki page.  The UI and the text editor part can be better suited for 
LaTeX needs.  For example there is no need to print the LaTeX sources, 
what is needed is to print the PDF with evince. Another example is the 
list of languages for syntax highlighting. This list is not needed in 
LaTeXila because only LaTeX is used (and if another file is opened, 
there is automatic detection that often works well).

A new argument in favor of an independent application for latexila: have 
a traditional menu and toolbars. The menu in latexila is quite big, with 
lots of LaTeX commands.

Also, when using only one general-purpose text editor for different 
tasks, the user probably wants to display different things in the side 
and bottom panels (or hide them). And depending on what is displayed in 
the panels, the panels size should be different. For that, gedit needs 
profiles: you could launch gedit with the profile LaTeX, launch another 
instance with the profile C, etc.

If creating a text editor like gedit (without the plugins system) is a 
matter of 1000 lines of code, more developers will create specialized 
text editors and IDEs.

Making the gedit source code more reusable is compatible with 
initiatives like gedit-code-assistance, or implementing a profile 
feature.

I'm more in favor of specialized text editors, the usability is 
potentially better IMHO. But it requires currently more work, because 
gedit hasn't been developed with this is mind. Fortunately the gedit 
code has a good architecture and some parts are already easily reusable.

Sébastien