(share)deepwiki

[HaveF HaveF]

DeepWiki: Your AI-Powered Guide to GitHub Repositories: https://apidog.com/blog/deepwiki/

https://deepwiki.com/leo-editor/leo-editor

[Edward K. Ream]

On Saturday, April 26, 2025 at 10:54:42 PM UTC-5 iamap...@gmail.com wrote:

DeepWiki: Your AI-Powered Guide to GitHub Repositories: https://apidog.com/blog/deepwiki/

https://deepwiki.com/leo-editor/leo-editor

Holy Smoke!

I have some quibbles about the diagrams and high-level overviews, but so what? We can expect lots of improvements.

Edward

[Thomas Passin]

Interesting. I see that some parts are very obsolete. E.g., the layout system and the Python and PyQt requirements.

[HaveF HaveF]

On Sun, Apr 27, 2025 at 8:20 PM Thomas Passin <tbp1...@gmail.com> wrote:

I see that some parts are very obsolete.

I have no idea. This is generated doc based on 27 April 2025 (4c67af), which is a relatively new snapshot.

Of course, the generated doc has lots of defects, maybe this defect comes from the old but not deleted code. Just a guess.

[Thomas Passin]

I suspect it may have come from old, obsolete documentation rather than code.,

[Félix]

Thank you for sharing this!

[Thomas Passin]

I just tried it on my own GF4-Project. It seems to have done a fairly decent job overall -

https://deepwiki.com/tbpassin/gf4-project/

We may have to wait some time for a User's Guide, though ...

[Thomas Passin]

The DeepWiki pages have a  box at the bottom labeled "Ask Devin about Leo-editor".  Just for fun I asked it to comment on the quality and clarity of the architecture.  Here is the result:

https://deepwiki.com/search/comment-on-the-quality-and-cla_4c6638af-00b0-443d-b486-29937852d0fb

This query took a long time, perhaps 15 minutes or more. The summary at the end of the response is:

Overall Assessment

The architecture of Leo Editor demonstrates several qualities of good design:

1. Clear separation of concerns: The MVC pattern and modular organization make responsibilities clear.
2. Information hiding: Components encapsulate their implementation details.
3. Stability: The architecture has remained stable for over 20 years.
4. Extensibility: The plugin architecture allows for extension without modifying core code.
5. Transparency: The documentation is honest about fragile parts of the system.

The architecture's clarity is evident in how consistently it's described across different documentation sources. The principles are well-articulated and appear to be consistently applied throughout the codebase.

[HaveF HaveF]

1. Clear separation of concerns: The MVC pattern and modular organization make responsibilities clear.Help me write⌥ H

In my opinion, the large language model is just predicting the next token. So as long as the document mentions MVC, it will continue to list the benefits, no matter what the code actually is. We should still be aware of the potential for hallucinations from these models.

I agree that wikis can provide a different perspective, we have to check the code. That's all.

Btw, recently, I use lots of editors like http://cursor.com/ and https://windsurf.com/ , I encourage you to try them, after you create some design rules, and they can help build things fast.

HaveF

[Thomas Passin]

I just asked the question out of general curiosity, to see what it came up with.  Not a bad showing, all in all.

[Edward K. Ream]

On Saturday, April 26, 2025 at 10:54:42 PM UTC-5 iamap...@gmail.com wrote:

DeepWiki: Your AI-Powered Guide to GitHub Repositories: https://apidog.com/blog/deepwiki/

https://deepwiki.com/leo-editor/leo-editor

Thanks for all your comments. Despite the OMG nature of the summary, I remain deeply skeptical about AI.

Imo, it's a continuing challenge to remember the utterly alien nature of AI. They have no human bodies, emotions, or experiences. They are unlikely to serve us humanely. Instead, they serve their billionaire masters. See the reviews of Move Everything Forever in Nature and Science.

For a discussion of what the embodied mind entails, I recommend George Lakoff's  Philosophy in the Flesh.

Edward

[jkn]

For a discussion of what the embodied mind entails, I recommend George Lakoff's  Philosophy in the Flesh.

good recommendation! I must dig out my copy...

    J^n

[Edward K. Ream]

On Tuesday, April 29, 2025 at 4:40:22 AM UTC-5 Edward K. Ream wrote:

See the reviews of Move Everything Forever in Nature and Science.

Here is the correct link to the review in Nature.

Edward

[Thomas Passin]

Philosophy in the Flesh - wow - that takes me back a long way.  Amazon says I ordered it in 2001.  I thought it was even earlier than that.  My copy is in storage somewhere so I can't browse through it now.  Lakoff has many books that are good reading. If you don't realize how deeply important metaphors are in everyday thinking and speaking, go out and get one or another of them.

[jkn]

I think I laughed out loud when I saw the first book of his(*) in a shop and knew I had to read it just from the title

    "Women, Fire, and Dangerous Things"

(*)OK, co-written

[Thomas Passin]

Same here.

[jkn]

This is very Off-Topic, but I can't resist, and I am pretty sure this is mentioned in at least one of the Lakoff books:

the classification of Animals that Jorge Luis Borges claimed to have found in an old Chinese encyclopedia,

The Celestial Emporium of Benevolent Knowledge.

According to this, the Chinese classified animals into:

a) those belonging to the emperor
b) those that are embalmed
c) tame or trained ones
d) suckling pigs
e) mermaids and sirens
f) those that are fabulous
g) stray dogs
h) those included in the present classification
i)  frenzied ones
j) innumerable ones
k) those drawn with a very fine camelhair brush
l) other ones
m) those that have recently broken a water pitcher
n) those that from a long way off look like flies

[Thomas Passin]

What an excellent hierarchy!