On Mon, Aug 2, 2010 at 7:55 AM, Edward K. Ream <[email protected]> wrote:
> I am beginning to feel that something basic is missing from Leo.

I am having thoughts about Leo that I have never had before. That's
exciting in itself.

1. One set of thoughts involves re-imagining what a .leo file is.  For
example, instead of leoPyRef.leo and leoPy.leo, suppose we have
leoPy.leo be the "official",
changed-only-when-a-new-file-is-added-to-the-project "view" of the
project.

I want bzr to backup all my development work, so I want an
ekr_leoPy.leo (and ekr_clones, see point 2) as the vehicle to do that.
 People would be free to see my view of the data, but I would request
that people not change either leoPy.leo or ekr_leoPy.leo.  Instead,
people would create, say, ville_leoPy.leo or kent_leoPy.leo, and if
desired, add those .leo files the the bzr repository. In the unlikely
event that Ville wants to see my view of leoPy.leo, he can open
ekr_leoPy.leo.

2.  This is just one example of a new distinction: local/personal vs.
global data (or .leo files).

To continue, ekr_leoPy.leo would use something like ekr_clones.txt to
contain what is now in leoProjects.txt and leoNotes.txt, etc.  This
should eliminate conflicts in ekr_clones.txt--I do not expect to share
ekr_clones.txt data even though it would be backed up in the bzr
repository.

So now the bzr repository has *two* uses:  sharing (global) data and
backing up (personal) data.  It's a useful distinction, imo.

3. This suggests the following question:  what if clones are primarily
a local (personal, not often shared) kind of data?

Again, this is a new *thought*.  It might lead somewhere, or not.

There is one exception that immediately arises:  clones stand in for
methods in languages like html.  They are not simply part of a view of
data: they are an essential part of the html file.  I don't think this

Aside: gnx's are required in thin external files no matter what.  In
spite of saying that sentinels are not the problem, I keep thinking
that gnx's might be removed in some cases.  But that's simply wrong!

Perhaps personal my_clones.txt files are all that is needed to reduce
the present clone-based conflicts.  I think these files would be
useful in the repository.

4. The question still remains: what to do about LeoDocs.leo?  I agree
with Ville that we must use thin external files to hold most of the
data that presently is contained in the .leo file.  It may be that
something new is required, but more likely the present mechanisms
suffice.  I'll look into this asap.

In short, there are two largely separate issues: the leoPyRef.leo
question and the LeoDocs.leo question.

Edward

P.S. I plan to revisit caching asap.  It may be that there are gaps in
the present scheme that are causing confusion.  For sure I want to
verify that caching is reliable in all cases.

EKR

-- 
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.

Reply via email to