Leo's python development is nearing its end

[Edward K. Ream]

Leo 6.6 may be the last substantial release in Leo's history. At present, the 6.6 to-do list contains five items. There are no items at present on the 6.7 to-do list. Many open items remain, but I have little desire to do any of them.  Expect 6.6 final in a month or so.

The leojs project now seems like the future of Leo. Indeed, it melds Leo with vs code, an unbeatable combination imo.

I have no clear idea about what I'll do after Leo 6.6. I've been thinking about that question quite a bit lately :-) I have no desire to retire, but I have less desire for make-work.

Edward

[tbp1...@gmail.com]

In the past, many or most of the development on Leo has centered on making Leo more usable, such as having more features for editing, or viewing what you have edited, or operating with scripts on content in Leo.  Leo is so powerful that it should be be a very suitable platform for other kinds of applications, ones that aren't centered on Leo itself.

IMHO Leo's potential for host a range of useful applications is nearly untapped.  When someone with a good idea tries to develop it in Leo, there may turn out to be a need for some new capability of feature.  But without tasks like that it's hard to know what would be needed.  Maybe some means for debugging programs in other languages would be helpful, making Leo more like a conventional IDE.  But that would take a lot of work absent a clear need.

Some readers  might remember me writing about the bookmark manager I wrote years ago to run inside a browser.  It's really nice, but the internals are complicated javascript code, making it hard to modify, and saving and restoring data is really clumsy and annoying. It would be very hard to get it working for someone else.  Leo turns out to have the abilities of the usual bookmark managers that are built into browsers except it's way better.  I have a couple of scripts: one lets you insert a bookmark node using a URL in the clipboard.  The other takes the standard HTML bookmark file that most browsers can generate and produces an entire Leo tree with organizer and URL nodes.

I'm already using this, and I like it a lot.  I'm prototyping some of the other features of my old inbrowser manager.  When that's all working  I'll have something very useful to me, and one much easier to modify and extend than my old one.  All in Leo.

So that's one application, one feature that isn't focused on making Leo easier to use.  The recent photo slideshow script is one of these.  There must be an unlimited number of others, if only people can think of them.

[Paul]

Edward, I have followed your facsinating project for over twenty years. I can't count how many times you have announced finalization, bug-free code, or feature freeze.

I imagine you'll have second thoughts.

In any case, the world is full of wonders and I'm sure you'll find others to explore.

Paul

[Edward K. Ream]

On Friday, February 18, 2022 at 1:37:42 PM UTC-6 Paul wrote:

Edward, I have followed your fascinating project for over twenty years. I can't count how many times you have announced finalization, bug-free code, or feature freeze.

:-) What's different now:

- I have little or no interest in the remaining issues on the to-do list. That has never happened before.

- leojs provides a way for Leonistas to get the benefits of VS Code's almost endless features.

 

I imagine you'll have second thoughts.

What I need now are new thoughts.  I agree with Thomas that " Leo's potential for host a range of useful applications is nearly untapped."  Thomas and I are saying that Leo doesn't have to "improve" in order to explore Leo's untapped potential.

In any case, the world is full of wonders and I'm sure you'll find others to explore.

Yes, that's the plan. Thanks for your comments.

Edward

P.S. There is only one remaining issue on the 6.6 list, and that involves unit testing. Imo, we can say that Leo is code complete, not just for Leo 6.6, but for the future.

EKR

[Edward K. Ream]

On Friday, February 18, 2022 at 1:47:03 PM UTC-6 Edward K. Ream wrote:

 What's different now:

- I have little or no interest in the remaining issues on the to-do list. That has never happened before.

- leojs provides a way for Leonistas to get the benefits of VS Code's almost endless features.

Most of my work for the last 40+ years has involved vanilla editor features. The preface lists the features that aren't just standard features. I'll continue to support Leo for as long as I am able, but the real challenge is to create fundamentally new Leonine features. But I'm not looking for any new features, because Leo already matches my work flow perfectly.

In other words, I know of no big problems that need solving within Leo. Absent such problems, invention is unlikely. I'll say more about what I am likely to do next in another thread.

Edward

P.S. Most of the vanilla editor work could have been avoided had I used emacs as a base. In that case, the work involved would have been comparable to creating emacs org mode. But I don't regret that work! Features like @button, @shadow, and @clean could not have happened without Leo's other devs.

EKR

[Josef]

Does that mean I will have to learn TypeScript to write little Leo extensions in the future?

[Edward K. Ream]

On Thu, Feb 24, 2022 at 10:56 AM Josef <joe...@gmx.net> wrote:

Does that mean I will have to learn TypeScript to write little Leo extensions in the future?

No. The desktop version of Leo isn't going away.

Even in VS Code (leointeg), you should be able to write @button scripts in python.

Edward