Hi, El 08/07/11 08:43, Edward K. Ream escribió:
[...] > > Rather, the problem is one of data synchronization. Of source, that > would be the point of, say, a bzr commit hook: it would ensure that > "helper" files get committed along with the "main" file. > > At first blush, this seems like a straightforward enterprise, and in > some sense it is. However, we have now added yet another layer of > complexity. When using Leo one would be forced to install all the > leo-centric hooks. I just don't like this: it makes using bzr (or > git) a requirement for using Leo. This doesn't seem like moving in > the right direction. > > Perhaps I am being just a bit timid or pessimistic. One could > *imagine* a one-click installer that installs bzr/git and all Leonine > hooks, but my experience with installers is that it's a big job and > I'm not very good at it. I hesitate to contemplate any approach that > makes the installer be an essential part of Leo. > > Edward > My approach won't be to build a one click installer with a VCS/sccs included. Just one click installer for Leo that can bootstrap the rest of the proper environment on a personal basis. In fact that's what I'm planning to do after I have a more proper understanding about how to make scripts run in Leo and deal with the external environment as a part of my "living inside Leo" project. This is my plan: 1. Understand how can I make Leo to install fossil scm, it could be any other VCS, but the minimal self contained nature of Fossil making it a "github in a box", with included webserver, web interface, distributed wiki and blog, distributed bugtracking, and distributed source control, in 2 just Mb of space, make it a good candidate for this initial test. I will just install it from Leo and try to commit something, using maybe the work made by Ville or Kent in VCS and Leo. 2. Try to put some eye-candy on this integration making the web interface available _inside_ Leo, maybe using something about @view-rendered plugin and the included fossil web server or somekind of html rendering/navigation provided by qt. 3. Make this more available to non-technical users creating a one click installer of Leo as a portable app for windows (solving the installation of Leo on Gnu/Linux is trivial, and by the way is not only available on Ubuntu/Debian based distros, but also in Arch ones, as a side note to put in the Installation web page), using the ideas that make some pyqt apps portable like exposed on [1] and [2] [1] http://www.daniweb.com/forums/thread154586.html [2] http://projectproto.blogspot.com/2009_10_01_archive.html Of course I could start with 3, but I think that what Leo does better is to bootstrap understanding about heterogeneous computer systems, and because Leo deals with external data it needs a proper discourse about how to deal with it in the best way, that is, using VCS. This has been some kind of ad-hoc solution until now, where VCS discourse is in the head of the developer and not inside Leo, but this should not be the only place, so my idea is to understand this in the particular case of Fossil and then try to bring that understanding to Windows users (I'm not one myself, but most of my students and my girlfriend are) in an easy way by constructing this Leo portable app with this discourse inside, powered by Leo ;-) I think that these are the kind of incremental projects and proposals for the "moratorium" time that Edward and Ville are talking about. Cheers, Offray -- You received this message because you are subscribed to the Google Groups "leo-editor" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/leo-editor?hl=en.
