2018. november 17., szombat

My very last talk with Frode Hegland

A Reply to Lorand


Loránd Kedves17 days ago
After closing with "over and out" the last time, it's my turn to be short(er).

- Yes, I write text. Sorry, I could not make a working Memex implementation alone in "spare" time. Yet, Dust Construct was quite close, even though I did not know about Memex when I built it.
- No hyperbole, pen and pencil Mr Hegland. A working Java application that managed its own knowledge graph, built data panels on the fly and generated part of its own kernel code from text templates. You shrugged it off every time saying you focus on text. Should I try it again? Challenge rejected. I trashed Construct for a good reason. You have seen that too.
- My statements are only one and essential step back from running source code. They should be specific for a person who can build the recommended Xanadu prototype. That is not me anymore.

---

Our conflict is much deeper than "text or graph". From my perspective, you are in the gang of shallow strong guys kicking me out and breaking my back again. I have no personal problem with you but I have a conflict with a system that you are part of, regardless of dreaming that you are against it. Just like in the Matrix:


You think you keep the heritage of Douglas Engelbart by repeating his words and mimicking parts of NLS "in high res". I think you break his dreamed revolution by "reaching for the low hanging fruits". I only wanted to leave a message at this memorial place with my first comment at your Tools for Thought post. I hope I will still be in a proper physical and mental condition to talk when someone understands it.

I close the circle repeating the Call for Mission Control... https://medium.com/predict/...

(links appear in darker grey on dark grey, hard to notice...)


Frode Hegland Mod Loránd Kedves17 days ago
Lorand, I am neither shallow, strong or against you. I simply ask you to be clear about what you propose. I have no problems with statements being the key unit of communicated thought. I simply accept that the interface between humans will always require interpretation since we don't even have fool proof objective access to our own memories.


Loránd Kedves Frode Hegland17 days ago
Can you accept that I can't explain my understanding and implementation of Memex (which is by definition NOT a "fool proof..." but the opposite), NLS or bootstrapping to you? It required me almost 30 years of programming, designing systems, trashing and reimplementing modules, and I can't transfer that network into your brain. If you really want to help, please just pass the link to people like Alan Kay or Ted Nelson, because they don't talk to people from the streets (and they are right doing so). That's all.


Frode Hegland Mod Loránd Kedves17 days ago
I do do have a working knowledge of the systems you mention but I’m glad you accept that you can’t just transfer that network into my brain. That is exactly the design issue, how to deal with that fact.

I’ll be happy to pass along a link. Which link is it you would like me to pass along?


Loránd Kedves Frode Hegland17 days ago
The Call for Mission Control on Medium. It seems that finding a link in your comment section is hard not only for me in the current color scheme... here it comes again :-)

https://medium.com/predict/...

... please add my apology for the harsh tone (and accept if for yourself too). Shouting for decades while everyone is going to the other way and complaining about it at the same time has its price.

Thanks!


Frode Hegland Mod Loránd Kedves17 days ago
I read your piece again and all I find is a tirade against the ways things are and a hope for something better. I can agree with this, it looks like an Engelbart C-Level task. I fully support this. If you are to get any support however, you really need to find a way to specify some specific things you feel can be done to actually move the needle forward in a specific and positive way. Do you have that written up somewhere?


Loránd Kedves Frode Hegland17 days ago
Following the links under "references" in my tirade may help...

The technical parts are: Dust Short, Software as a Data Structure, by the way a more current roadmap is in my introduction at Doug'50, look for DSL because all is about atomic core languages.
The civilization ideas: Taste of Luck, Open Source Civilization and Dawnworld (although this is 120 pages, it takes a strong motivation to read - but guess how much it took to write, then translate to English, etc.) I also moved some of my 800-some articles from my blog and LinkedIn to Medium, and had some interesting talks there in the past few days.

I know, less is more, but I really have spent decades on this...


Frode Hegland Mod Loránd Kedves17 days ago
I appreciate you have sent decades on it but what specific steps do you want to take?


Loránd Kedves Frode Hegland16 days ago
And finally understood the essence of this question.

First I thought "it should be obvious!!!" Join a team of brilliant people, make Memex together, build a transparent global network of knowledge, save the world. But looking at the actual environment, rationally thinking about specific steps, to my greatest surprise, I am not that sure. I mean, at all. My 30 years long "quest for the Holy Grail" has ended and I am free.

No further step is needed.

Thank you!


Frode Hegland Mod Loránd Kedves16 days ago
Your sarcasm is not useful


Loránd Kedves Frode Hegland16 days ago
It is already impossible for me to write a text that you can decode because I have a fundamentally different (yet quite effective) knowledge network in my brain and a very narrow communication channel. Do you really think I waste my time adding another layer of blur to that?

I mean exactly what I wrote. Our talk changed my core motivations. It feels much better now and I needed that.


Frode Hegland Mod Loránd Kedves16 days ago
What you write is the same as what we write do of course I agree. It is quite general and I apologise for trying to get something more specific and clear out of you. This discussion is not useful


Loránd Kedves Frode Hegland2 days ago
For the record...

We both know Xanadu's concept, demos and status, so I assumed we had a common ground.
I presented major architectural issues that I think caused its failure and recommended solutions.
I offered my help as a veteran architect and developer as part of my PhD research for free.
I mentioned that I did not want to work on it without talking about the changes with the "project owner", your friend, Ted Nelson. Because I could be wrong and did not want to waste time on something I did not understand correctly.

I don't know your professional background, but this is more than we generally have at a project kickoff in real life IT industry (at least according to my 20 years experience here). I did my best, it was not enough, happened many times before, no harm no foul.

Peace.


Frode Hegland Mod Loránd Kedves2 days ago
If you want me to forward an email to Ted I'll do that.


Loránd Kedves Frode Heglanda day ago
No thanks. I have finally learned my lesson here and don't have more time to spend on a discussion that the partner judges useless. Happened many times, no harm no foul but this should be the last.


Frode Hegland Mod Loránd Kedvesa day ago
I don't think your work is useless. I think it's crucial to explore different options.


Loránd Kedves Frode Hegland6 hours ago
Sorry, you (and those sweet likers) still don't get it... :-D

I have literally spent 20+ years of my life on this work. I mean solving "impossible" problems in real environments, not "exploring options" in presentations. Its value and the Dunning/Kruger effect around it scares the hell out of me. This is where those ridiculous "non-specific tirades" come from.
I was not looking for these useless "come on!" messages (they only make me laugh and pissed off at the same time to be honest) but real cooperation with experts. The Engelbart community with you at the reception was my last resort.
I have spent many years trying to connect to experts, with zero result. You finally convinced me that I would talk to the hand all the time in "useless discussions". I am grateful for that. I will make it or break it alone. Business as usual.

Take care.


Frode Hegland Mod Loránd Kedves3 hours ago
I think you should consider how your moaning and insulting of people, coupled with a complete lack of demonstration of what you are talking about, only arrogant references, might be the reason why you have zero result. You work and ideas might be important and brilliant, or not. You make it a struggle for 'us idiots who are so stupid we don't know it' to tell.


Loránd Kedves Frode Hegland37 minutes ago
For the record again:

I tried to show you a working garage-Memex. I used it as a thinking augmentation tool to design both itself and a real ETL web application. I have built it before I learned about Engelbart or Bush. You had no time and were interested in text. Fine, but to me far from "complete lack of demonstration".

I came back only because you started drawing graphs and wrote about Tools for Thought. That's what I do. And yes, it takes serious effort to understand, otherwise Google, Microsoft, IBM etc. would have done it. That can make others feel bad. Sorry, that is not my problem, making it is much worse.

I can show things but can't make people see them. After a long time that made me impatient and the whole thing became even more hopeless. That leads to our shared conclusion: I should stop talking. But I saved your excellent summary on me for later reference.

Thank you.

2018. november 16., péntek

Text vs graphs in knowledge management

Why work on Improving our Tools for Thought?


Loránd Kedvesa month ago
Dear Frode,

we had our short journey together but to me it seemed you focus on text and existing toolchain, while my obsession is what I call today (thanks to Gyuri!) idea projection: the process of direct interaction with the network of knowledge, and the requirements of an infrastructure that allows the continuous morphing of that network (that is merely impossible with what we have today).

Now the development is on hold due to personal financial and health issues, but I still write texts :-) sometimes. I guess you may like it... https://medium.com/@lkedves...



Frode Hegland Mod Loránd Kedvesa month ago
Hi Lorand. Yes you are right, I am focused on text, I feel that external symbols are a powerful way to manipulate our thoughts and communications. If you can find a way to directly interact with knowledge, that would indeed be amazing but I am a text guy and that is outside of my area.


Loránd Kedves Frode Hegland25 days ago
OK, but then please explain me why you put a graph to the top of the article as an illustration of Tools for Thought? That is not text, you can't use a text editor to author it (yet another markup or source language IDE is not a "text editor", but manages hidden properties assigned to text fragments).

My problem with your approach is that as long as you put an equal sign between thoughts, symbols and text, you can't even see the point: the also morphing language and syntax required to transfer (serialize! so no graph here) your knowledge to text. A real Tool for Thought (which I have already created and used to improve itself before learning about Doug, but was too deeply locked to one platform) should include changing ALL, even the rules it uses to fully describe itself. That is missing since Bush described and Doug partially implemented Memex.

Focusing on text (writing yet another editor) is a dead end street, as example see Ted Nelson's Xanadu progress in the past 50 years. By the way, I don't see the fundamental improvement in Liquid over Xanadu. What is it?


Frode Hegland Mod Loránd Kedves25 days ago
I don't think text needs to be in a column, I guess that is where we have different perspectives. I'm sorry you feel my work is a dead end street and no, I am not trying to go down my good friend Ted's route, it's a different direction.


Loránd Kedves Frode Hegland19 days ago
> I don't think text needs to be in a column, I guess that is where we have different perspectives. I'm sorry you feel my work is a dead end street…

On your level of extreme oversimplification

Not exactly.

On my level of TL;DR

I have been writing source code for living for 20+ years. I always hated to bury my understanding of a task into serial text of the chosen programming language and libraries. To me the real work is building the structure and interaction of components. That can be wrong for the first time, this is where design patterns, refactoring, Agile methodologies come in. They are as natural to me as breathing, in fact I used before learning about them. I created always better tools to support design and automatize coding until Dust Construct that I used to think about and build itself. At this point I first met Engelbart's ideas and realized that this is Augmenting Human Intellect.

I was already at the border of "machine text" and plans of "human text" and asked if thinking in general can have the same problem as source code. I have found Tim van Gelder, a philosopher who refers to Engelbart, derives argument maps from text to represent the real content. He says that a communication error is the difference between the knowledge graph that the reader decodes from the text and the one the sender wanted to transfer. We need tools making these graphs transparent to decrease errors caused by the weakness of text. I found Neil Postman, who explained the dangers of using modern information systems without understanding how much they change our thinking. And many, many others.

I say text is outdated because of the
- need of serialization,
- lack of exact definitions and structures,
- missing ability to follow morphing background.

I knew that you did not listen when I tried to explain this. Your not-even-wrong summary adds those months to the long list of bitter mistakes of mine. Thank you for making it clear.

> … and no, I am not trying to go down my good friend Ted's route, it's a different direction.

Sure. Xanadu is not only ahead of Liquid by 50 years, but almost solved the problem. Yet, Ted made some fundamental mistakes in the Why? (Vision) to What? (Architecture) translation. These are:
- The “atom” of text is NOT the character, but the statement. References exist among statements, not character ranges. This allows changing other statements or translating the text without breaking existing references.
- Store statements by using combined IDs: internal primary key + commit ID. You can change statements later and save them by the same internal but different commit ID. You keep the original references and can follow updates on Xanadu level.
- Trash EDL, understand that any given text is a hierarchy of statements, either local or remote. You need to store the links in their natural bidirectional form, and build the text from a root node to display. The statements are not locked into their owning text, but a flexible cloud open for reuse. By the way, this is the essence of ZigZag database.

This Xanadu prototype is a 6 months spare time project for a developer with my skills. I started it using Hungarian laws as data source for my research, but trashed due to the lack of support and feedback. I also close my PhD because I “lost” the time needed to get the required publication credits.

"Your good friend Ted" was quite close to see his 50-year-old dream working TODAY, beyond a quite weak 3D framework demo of squares with custom texture and colored rectangles. But, he does not talk with strangers on his YouTube channel and has you protecting him from the chance.

If anything, you’d better be sorry for this instead of my guessed feelings.

To quote the Oracle from Matrix:
"I'm sorry, kiddo. I really am. You have a good soul. And I hate giving good people bad news. Don't worry about it. As soon as you step outside that door, you'll start feeling better. You'll remember you don't believe in any of this [Lorand] crap."

Over and out.


Frode Hegland Mod Loránd Kedves18 days ago
Lorand, I have replied to you here: http://wordpress.liquid.inf...