Re: Eurekas! The sea of nodes, half clones more

2012-03-04 Thread Edward K. Ream
On Sat, Mar 3, 2012 at 1:00 PM, Seth Johnson seth.p.john...@gmail.com wrote: I will be very interested in how this shapes up and how you start looking at collaboration, distribution and versioning.  (BTW, I would designate those organizations by unique keys for their use type, link type and

Full text search, Whoosh

2012-03-04 Thread Ville M. Vainio
Next up in my bigdash.py plugin / global search work is integrating the Whoosh full text search engine. I hope to implement it today or tomorrow. http://packages.python.org/Whoosh/ This will make search strings more google-like, maybe we could even customize it so that parents with matching

Re: Full text search, Whoosh

2012-03-04 Thread Kent Tenney
So Cool. Not only for Leo, but I look forward to learning more about Whoosh, after looking at several full-text search solutions, Whoosh seems like the sweet spot: pure Python, powerful but not too heavy, capable but not too complex ... Thanks, Kent On Sun, Mar 4, 2012 at 8:04 AM, Ville M.

Re: Eurekas! The sea of nodes, half clones more

2012-03-04 Thread Edward K. Ream
On Mar 3, 12:00 pm, Kent Tenney kten...@gmail.com wrote: OK, I guess I'll have to _see_ it to get it. Here's another picture. Suppose we implement chapters (or even hoists) in the new scheme. Different chapters = different links. So you could, in a chapter, insert or delete nodes, without

Re: Eurekas! The sea of nodes, half clones more

2012-03-04 Thread Ville M. Vainio
v.viewDict is an implementation detail. An alternative is to introduce 'edges' as separate first class construct, and then bundle edge type in that construct (color, half-clone, backlink, normal tree edge...). This scheme (of edges overall) was discussed earlier, when talking about database

Re: Eurekas! The sea of nodes, half clones more

2012-03-04 Thread Seth Johnson
The table in which Ville stores edges can be extended with columns representing the use type, node type and particular use (also columns for state) within which nodes are being organized -- that is, it can store edges with a specification of the context in which they are being used (Ed's distinct

Ville, please fix print statements in bigdash.py

2012-03-04 Thread Edward K. Ream
The 2.x only print statements cause a unit test to fail. Please fix them in future. I've fixed them at least twice, but my fixes keep getting reverted. Edward -- You received this message because you are subscribed to the Google Groups leo-editor group. To post to this group, send email to

Re: open-with starting to work: it's time to *really* fix it

2012-03-04 Thread Edward K. Ream
On Mar 3, 9:36 am, Edward K. Ream edream...@gmail.com wrote: Thanks for the confirmation.  I am going to fix the present problems, probably by having the contextmenu plugin skip the dialog when the update is to a temp file created by the vim or xemacs plugins. Done in the trunk at rev 5066.

Schedule for Leo 4.10 b1

2012-03-04 Thread Edward K. Ream
I expect to release 4.10 b1 this Friday, March 9. Only relatively minor bugs remain: - 6 or 7 numbered bugs having to do with how Leo handle's url's. I'll attempt yet another unification of the URL code. - 1 or 2 numbered bugs dealing with focus issues. There is also a considerable

Re: Ville, please fix print statements in bigdash.py

2012-03-04 Thread Ville M. Vainio
Sure thing. I don't know why your changes get reverted, none of my bzr pulls have modified the files. On Sun, Mar 4, 2012 at 10:38 PM, Edward K. Ream edream...@gmail.com wrote: The 2.x only print statements cause a unit test to fail.  Please fix them in future. I've fixed them at least twice,

Bug when using a '@url - node' which contains a file-URL?

2012-03-04 Thread Viktor Ransmayr
Hello Edward, I'm trying to keep the source and destination content of a work-log entry accessible in the same outline using @rst and @url nodes. I see a behaviour when 'clicking' on the @url node, which I think is a bug. - Here's the log of the 'rst3 - command' execution as well as the