Back to the talks Previous by track: Basic Calc functionality for engineering or electronics Next by track: GNU Emacs Greader (Gnamù Reader) mode is the best Emacs mode in existence Track: General

Blee-LCNT: An Emacs-centered content production and self-publication framework

Mohsen BANAN (he/him) - Pronunciation: MO-HH-SS-EN, http://mohsen.1.banan.byname.net

Format: 37-min talk ; Q&A: BigBlueButton conference room Etherpad: https://pad.emacsconf.org/2025-blee-lcnt
Etherpad: https://pad.emacsconf.org/2025-blee-lcnt
Status: TO_CAPTION_QA

Talk

00:05.760 Introduction
01:20.080 Scope: A complete multi-media content processing framework
02:10.320 Prior art and similar art
03:02.420 LaTeX-Beamer + Reveal.js with Blee and BISOS
03:57.160 Blee-LCNT novel concepts
05:12.520 Part of a bigger picture - part of a series
06:32.560 Nature of polyexistentials
12:52.640 Content processing - a ByStar/BISOS/Blee Capability Bundle (BCB)
14:23.120 ByStar containment hierarchy and ByStar capability bundles
14:31.280 Aggregated conviviality of ByStar capabilities
15:22.000 Parts list: integrated components
15:47.868 Resulting contents - output forms and formats
18:45.720 reveal.js
20:31.980 Generating the video
21:33.480 A unified single input -- a sequencef of frames
22:39.180 Abstractions to keep in mind
23:16.200 Frame control types
24:24.360 How outputs are generate from the inputs
26:25.200 Context for unified source walkthrough
27:46.480 One slide
29:24.080 Dynamic blocks
31:05.800 Internationalization - a non-Americanist perspective
33:42.280 Autonomous self-publication and federated re-publications
35:07.720 Ingredients of BISOS platforms and their progression
36:02.560 Moving forward
Duration: 36:41 minutes

Q&A

00:22.880 Q: Where do we find all the inputs and outputs you mentioned?
04:48.400 Making presentations easier to distribute
05:42.040 Reveal output
08:15.000 GitHub organizations
12:24.040 Challenge of DIY model and recipes
13:57.480 Dblocks everywhere
17:09.960 Q: What changes have you seen in the culture while developing all these things like libre-halal system and now blee-lcnt?
19:11.160 Intellectual property rights
23:43.560 Q: Given that large AI companies are openly stealing IP and copyright, thereby eroding the authority of such law (and eroding truth itself as well), can you see a future where IP & copyright flaw become untenable and what sort of onwards effect might that have?
Listen to just the audio:
Duration: 27:45 minutes

Description

In a sense this is yet another talk about how you can use Emacs to produce fancy presentations like this or write complex books and self-publish them. But our approach is fundamentally different.

Many talks at previous Emacs Conferences have described how Emacs and org-mode can be extended to facilitate content production by adding more to Emacs. Our approach is that of putting a smaller Emacs at the core of something bigger. That something bigger is an autonomy oriented digital ecosystem called "ByStar" which is uniformly built with a layer on top of Debian called BISOS (ByStar Internet Services OS).

At Emacs Conf-2024 the title of my talk was "About Blee" – https://emacsconf.org/2024/talks/blee. Blee (ByStar Libre-Halaal Emacs Environment) is that smaller Emacs packaging that positions Emacs at the core of BISOS and ByStar. BISOS and Blee are intertwined and ByStar is about autonomy oriented unified platforms for developing and delivering both internet services and software-service continuums.

This talk is about Content Production and Self-Publication capabilities of Blee and BISOS.

Blee-LCNT is LaTeX centric. The original text is always in COMEEGA-LaTeX – LaTeX augmented by Org-Mode. This is the inverse direction of exporting LaTeX from Org-Mode. For typesetting, the LaTeX syntax is far more powerful than org-mode. And with COMEEGA-LaTeX, you can also benefit from all that org-mode offers. The scope of Blee-LCNT is all types of content from presentations to videos to books to name-tags and business cards.

LaTeX to HTML translation is done with HeVeA. For presentation/screen-casting, the original text is then augmented in layers by images, audio voice-overs, screen captures, videos and captions. The Beamer LaTeX file is then processed by both LaTeX and HeVeA. LaTeX produced slides are then absorbed in html by HeVeA as images. HeVeA output is destined to be dispensed by Reveal.js. The video is then just a screen capture of the autoplay of reveal file. Viewing presentations in their original Reveal form makes for an even richer experience.

All of this involves a whole lot of integration and scripting. But all of that has been done and you can get it all in one shot by just running one script.

To get started with BISOS, Blee, and ByStar, visit https://github.com/bxgenesis/start. From a vanilla Debian 13 installation ("Fresh-Debian"), you can bootstrap BISOS and Blee (with Emacs-30) in one step by running the raw-bisos.sh script. It produces "Raw-BISOS" which includes "Raw-Blee".

You can then add the LaTeX sources for your content as ByStar Portable Objects (BPO) to BISOS and process your content with Blee-LCNT.

All of this and more has been documented in a book that was produced by Blee-LCNT itself. The title of that book is:

Nature of Polyexistentials: Basis for Abolishment of the Western Intellectual Property Rights Regime And Introduction of the Libre-Halaal ByStar Digital Ecosystem

https://raw.github.com/bxplpc/120033/main/pdf/c-120033-1_05-book-8.5x11-col-emb-pub.pdf

Download: https://raw.github.com/bxplpc/120074/main/pdf/c-120074-1_05-book-a4-col-emb-pub.pdf DOI: https://doi.org/10.5281/zenodo.8003846

https://jangal.com/fa/product/252689/nature-of-polyexistentials

I welcome your thoughts and feedback, especially if you experiment with Blee, BISOS, ByStar, and the model and the concept of Libre-Halaal Polyexistentials.

Transcript

[00:00:05.760] Introduction
Greetings. Salaam. This is Mohsen Banan. I am a software and internet engineer. The title of this presentation is "Blee-LCNT: An Emacs Centered Content Production and Self-Publication Framework". Blee stands for ByStar Libre-Halaal Emacs Environment. In last year's EmacsConf, I introduced Blee, BISOS and ByStar as concepts and as foundations. This year I want to focus on one concrete capability. Content Production and Self-Publication is a foundational Blee and BISOS Capability Bundle. Both this presentation and the Nature of Polyexistentials book were developed with Blee-LCNT. In this presentation I want to look at Emacs as a central ingredient for a usage environment that we can use to orchestrate production of quite fancy multi-media presentations.
[00:01:20.080] Scope: A complete multi-media content processing framework
Let's consider two different scopes. First, the scope of Blee-LCNT Capabilities Bundle, which is that of a complete multi-media content authorship, generation, publication and distribution framework. That complete scope is presented in this slide and it spans both black ink and violet ink. Second, the scope of this presentation, which is more limited. In this presentation I confine myself to the bullets is violet ink. Here, I focus on presentation and video as content types and their authorship and generation and their federated re-publication.
[00:02:10.320] Prior art and similar art
This is a common topic. It makes good sense for us to start with a review of prior art and similar art. I went through the past EmacsConf talks and found a good number of them that also deal with the topic of content generation. A few of these are included in black ink in this slide. Many of these have chosen the Babel, in other words Org-Mode+LaTeX as primary input. I prefer the inverse of that. I also looked for past talks which have used Reveal.js and LaTeX-Beamer. For example, Sacha's use of Reveal.js is shown in violet inK. And Ihor's use of Beamer is in teal ink.
[00:03:02.420] LaTeX-Beamer + Reveal.js with Blee and BISOS
This presentation is about a combination of Reveal.js and LaTeX-Beamer. For those who may not be familiar with Beamer and Reveal, here is a quick intro. Among academics, LaTeX-Beamer is the go-to tool for producing presentations. Reveal.js is recognized as the best of breed for dispensing HTML slide decks. For many, Reveal and Beamer live in different universes. Beamer is pdf oriented and Reveal is html oriented. Combining two powerful tools makes for an even more powerful tool. This Blee-LCNT Presentations combines the best of LaTeX-Beamer with Reveal.js.
[00:03:57.160] Blee-LCNT novel concepts
Beamer primarily functions as producer and Reveal functions as dispenser and multi-media enhancer. Here is how the combination works. LaTeX Beamer pdf result is dissected into named frame images which can then be inserted in Reveal.js. LaTeX Beamer frames can also be translated into html with HeVeA which can also be inserted in Reveal.js. Voice-overs for Beamer frames can be correlated to frame names and applied to image or html frames. Screen captures and image narrations as videos can be directly dispensed through Reveal. There are various additional novel concepts with regard to the way that we have integrated all of this together. Instead of Org-Mode+LaTeX, we do LaTeX+Org-Mode. Instead of Babel, we do COMEEGA, instead of the Literate model we introduce the Surrounded model. You shall see various examples of these shortly.
[00:05:12.520] Part of a bigger picture - part of a series
All of this is part of a bigger picture. A much bigger picture. My talks at EmacsConf 2021, 2022 and 2024 are related. This 2025 talk builds on those. Last year's talk "About Blee: enveloping our own autonomy directed digital ecosystem with Emacs" in particular, lays the foundations for this talk. If you have not seen that, it would make good sense to review it. In my previous talks I have been criticized of having a "prophetic" style. The scope of ByStar is lofty and immense. In many ways it is unbelievable. And EmacsConf talks are meant to be short. So, as a result, sometimes I end up being cryptic. Having accepted the "prophetic" criticism as legitimate, I now need to put a book on the table. With that book in place, moving forward, when needing to be cryptic, I shall cite Chapter and Verse.
[00:06:32.560] Nature of polyexistentials
I am delighted to announce the availability of my recent book, "Nature of Polyexistentials". The full title of my book is: Nature Of Polyexistentials--- Basis For Abolishment Of The Western Intellectual Property Rights Regime--- And Introduction Of The Libre-Halaal ByStar Digital Ecosystem. Knowledge, know-how, uses of know-how, ideas, formulas, software and information are inherently non-scarce. They are *polyexistentials*. Unlike monoexistentials which exist in singular, polyexistentials naturally exist in multiples. What is abundant in nature is being made artificially scarce through man-made ownership rules called copyright and patents. These mistaken ownership rules, the so called Western IPR regime, has immense ramifications on the shape and the direction of the American Digital Ecosystem. It would be an understatement to say that the American Digital Ecosystem has put humanity in danger. Two parts of the book, in particular are of immediate relevance. Part III, the ethics layer, focuses on contours of cures. Having dismissed the Western intellectual property rights (IPR) regime as an erroneous governance model for polyexistentials, I propose the Libre-Halaal model of governance of polyexistentials towards facilitating conviviality of tools. Part IV, the engineering layer, introduces the Libre-Halaal ByStar Digital Ecosystem. as an ethical alternative to the prevailing proprietary American digital ecosystem. The book also provides additional details about the content generation and publication facilities that I am presenting here. And the book itself, as content, was generated and published using the facilities that I am presenting here. You can think of this book as being in two volumes. Our focus are Blee and BISOS in Volume II. Volume I deals with the general concept of polyexistence and invalidity of IPR and our terminoloy of Libre-Halaal--- instead of the common but ill directed vocabulary of Free Software and Open-Source and FOSS. In Chapter 11, I introduce the very sensitive and potent vocabulary of Halaal and Libre-Halaal. The contents of this book belong to all of humanity and verbatim copying of it is unrestricted. If you want to read it, this book is yours. The "Nature of Polyexistentials" book is available both online and in print. This book is available as two editions. The US Edition and the International edition. The US Edition is written with a slightly milder Western unfriendly tone, while the International Edition includes additional original content in Farsi. I consider the International Edition to be the authoritative version. However, many readers in the US and Western countries may prefer the US Edition. I maintain separate Git repositories for each edition on GitHub: US Edition is at bxplpc/120033 and International Edition: bxplpc/120074 Cloning these repositories will give you access to the book in PDF format (suitable for both A4 and US Letter printing) and in EPUB format. Alternatively, the content can be downloaded directly from your browser without needing to clone the repositories. To ensure broader online availability and stability, I have also published the book on Zenodo, complete with a DOI (Digital Object Identifier). You can download both the A4 and 8.5 x 11 PDFs from there as well. The book is also available in print on Amazon and at most major bookstores in the US and Western regions. The ISBNs for both editions are included in this slide. Additionally, I have published this book in Iran through Jangal Publishers. I did not write this book for profit. My aim is to share my thoughts and encourage readers to engage with my views and ideas. Your feedback is welcome, and I am genuinely interested in hearing your perspectives. In Western markets, I have priced the print edition somewhat above production costs. If you find value in the book and the ByStar project, purchasing a copy will help support my work. Thanks in advance for your support. And here are the same links as a native Reveal slide. If instead of a video, you are viewing this presentation as a Reveal web page, you can just click on the pointers and URLs.
[00:12:52.640] Content processing - a ByStar/BISOS/Blee Capability Bundle (BCB)
Instead of the traditional model of giving you recipes in a DIY context towards the goal of creating content processing capabilities on top of what you may already have, I am doing the opposite. I am saying: take this whole BISOS and Blee thing, and in there you will also have the content processing capabilities that I am speaking of here. So, at the top level we have our own autonomy and privacy directed digital ecosystem, which in contrast to the center oriented American digital ecosystem, is edge oriented. We call it: "The Libre-Halaal ByStar Digital Ecosystem". All the systems in ByStar, run BISOS (By* Internet Services OS), which is a layer on top of Debian. The usage environment of ByStar and BISOS is Blee which is a layer on top of Emacs. With those in place, we then create a capability bundle called Blee-LCNT. So, when you buy into Blee and BISOS, you will naturally also get these content processing capabilities--- without a need for any recipies or DIY effort.
[00:14:23.120] ByStar containment hierarchy and ByStar capability bundles
If you were to look at the model that I introduced as containment hierarchies, it would look like this.
[00:14:31.280] Aggregated conviviality of ByStar capabilities
We love Emacs and we love Unix because their design is convivial. By convivial, I am referring to Ivan Illich's concept and terminology of "Tools for Conviviality". It was first published in 1973. It's a must read. A goal of the design of the ByStar Digital Ecosystem is to enlarge the aggregated conviviality of its capabilities. What distinguishes Blee-LCNT from other content processing tools and frameworks, is our emphasis on enhancing the aggregated conviviality. These tools let you express yourself. They let you be in charge.
[00:15:22.000] Parts list: integrated components
Here is our parts list. These are the components that we have chosen to bring together towards our goal of creating convivial tools. In this slide, we are using black ink to denote exisiting tools and we use violet ink to denote pieces that we have developed towards cohesive integration. [This] video,
[00:15:47.868] Resulting contents - output forms and formats
the video is just one of the outputs. There are other outputs as well. In this figure, the outputs are shown in the top layer. Using this video as an example, this presentation's output also include the "Presentation Form" and the "Article-Presentation Form". Let's look at these more closely. For Presentations, there are 3 different forms. The Video Form, the Presentation From and the Article-Presentation Form. The Presentation Form produces both a pdf output and Reveal output. Next we will walkthrough some of the benefits that availability of these forms and formats provide. The video presentation that you are watching is just one of the outputs of the Blee-LCNT machinery. There are two PDF format outputs and two HTML outputs that are also quite useful. The primary output of Beamer is a set of slides that people use to give their talks with. Typically that's done live. In my case I dissect the images of each frame and do a voiceover on it and then dispense it through reveal. In a second, you will see that as well. This PDF output is very useful. You get the table of contents, of course, and in addition to that, Beamer generates navigations for you where on any part you get a small table of content as well. This is heavily used amongst academics, and it's a good output on its own, and I'm augmenting it in a variety of ways. In addition to the presentation PDF format, there is also an article-presentation PDF format which gives you the same content, but it gives it to you in a textual form with the table of content and the rest. This is a good form to use when you are giving, for example, class lectures, and the students often prefer this format.
[00:18:45.720] reveal.js
Now for the HTML format output, the most relevant, of course, is the reveal itself. If you have not used reveal before, in my view, it's a HTML slide dispenser. I don't look at it as a presentation framework. I use, as you are seeing, we use Beamer to feed into it and we use it to dispense the information. It has all the typical navigation capabilities that you would expect, and most of what I have as slides are images, but occasionally, particularly when there is a need to provide pointers, HTML pointers, I then also include a textual output. This is also produced from the Beamer LaTeX source, but it's HTML through textual HTML, through HeVeA, not the image. You can... you get a table of contents. You can navigate and there are a whole lot of other features that reveal also provides.
[00:20:31.980] Generating the video
So to generate the video, what I do is I come to the very beginning of the presentation. I turn on the screen capture recorder, and then I start playing the voiceover for each slide and at the very end, you get a video, but what you just did is you dispensed every frame, one at a time, through reveal. In addition to this HTML form, you also get an article presentation form of it, with a full table of contents and the videos are there, and the notes are there, and this is also quite useful.
[00:21:33.480] A unified single input -- a sequencef of frames
Now, let's look at the one single input file that produced all of the outputs that we just saw. I have put both the input file and some of the output files for this presentation on Github. Here are some links to these repos and files. And here are the same links as a native Reveal slide. This figure gives us an overview of how one set of inputs encapsulted in a single file can produce all of the outputs that we saw. The main TeX file shown at the bottom is processed by both XeLaTeX and by HeVeA. That main TeX file, in addition to LaTeX syntax, also include org-mode constructs that facilitate addition of audio and video files. Later, I'll walkthrough the bodyPresArtEnFa.tex file that generated this very presentation with you.
[00:22:39.180] Abstractions to keep in mind
When you construct that primary TeX file, there are several abstractions that you need to keep in mind. Is my presentation going to go from Left-To-Right or from Right-To-Left? Perso-Arabic presentations go from Right-To-Left. Another consideration is the types of forms of results that you want. Just the presentation or Article-Presentation as well? With those choices in place you can produce condition based text for each of your desired outputs.
[00:23:16.200] Frame control types
Think of this video presentation as a sequence of frames. Each frame is controlled by an org-mode dynamic block. This table lists available dblocks from which you can choose. For example, this particular frame that we are watching is controlled by b:lcnt:pres:frame/derivedImage. Beamer creates a pdf file that includes the image of this slide. That image is then injected into Reveal. And in the end, a video of that image is produced with the narrations that I am uttering right now. All of this has similarly been applied to each and every frame that you have been watching. Similar to Frame Controls, there are org-mode dynamic blocks for "Frame Body Types". You can easily insert an image which is typically created by OpenOffice Draw into a frame. Same with say a screen capture video.
[00:24:24.360] How outputs are generate from the inputs
Now that we have looked at the "Outputs" and the "Inputs", let's look at how the Outputs are generated from the Inputs. Let's bootstrap Raw-BISOS and Raw-Blee. Starting from scratch, get yourself a fresh copy of Debian 12. Then go to https://github.com/bxGenesis/start . The README.org file of that github repo is same as Chapter 18, "Engineering Adoption of BISOS and ByStar" of the book. We will next run "raw-bisos.sh", but prior to that, let's take a quick look. This bootstrap scripts will do a lot as root on your Fresh-Debian. It is best to first try it on a disposable VM. raw-bisos.sh adds the current debian user to sudoers. Then it installs pipx. And then with pipx it installs from PyPI bisos.provision. bisos.provision includes additional bash scripts that are then executed. Full installation involves setting up various accounts, groups, various directory hierarchies, lots of apt packages and lots of python packages from the bisos namespace. If you are ready, copy and paste this line and run it. You will be prompted for the root password. Then be patient. Full installation can take 15 minutes or so. The logs of this script are also captured in ~/raw-bisos-${dateTag}-log.org
[00:26:25.200] Context for unified source walkthrough
Now that we have Raw-BISOS and Raw-Blee installed, we are ready to walk through the unified source of the very presentation that you are watching. The "bodyPresArtEnFa.tex" file that we will visit is in COMEEGA-LaTeX syntax with lots of org-mode dblocks which generate Beamer-LaTeX frames and conditioned LaTeX bodies. After the walkthrough, I'll describe dblocks and COMEEGA in more detail. At the tail end of the walkthrough, we will also go through the generation process which runs XeLaTeX and HeVeA and a lot more. Let's look at our input file. It's a LaTeX file in LaTeX mode, and it has org syntax org-mode included in it, and I can toggle between LaTeX and org-mode. So, now I'm gonna be in org-mode, and org-mode gives me everything that org has to offer, including a very convenient navigation framework.
[00:27:46.480] One slide
Let's take one slide and take a look at how it was done. So I would come to this scope slide and while I am there, I'm going to click on N. N takes me to the native LaTeX form back, so that I'll be looking at it not in org, but in LaTeX. So we're back in LaTeX, and as you can see it uses a dynamic block starting with the comments and the BEGIN, and it uses a dynamic block named a framedDrive image, which means the content of this frame will be dispensed as an image, not as text, and it also automatically creates for me a name, a label, that can be used for voiceover augmentation. So a file in the audio directory called ScopeOfBleeLcnt.mp3 is this audio that will come on top of this slide and then the rest is the LaTeX itself.
[00:29:24.080] Dynamic blocks
The concept of "Org Dynamic Blocks" is very powerful. I think of them as universal visible macros. But, why should they be primarily used in just Org-Mode? I say, let's generalize them to "Emacs Dynamic Blocks". Have defaults for org-dblock-start-re in every relevant mode and use them everywhere. Blee does that. In COMEEGA-LaTeX, Dynamic Blocks create Frame Controls and insert Image and Video contents. Much of Blee and BISOS are implemented in COMEEGA. Almost all of our Elisp, Python, Bash and LaTeX work uses COMEEGA. COMEEGA stands for Collaborative Org-Mode Enhanced Emacs Generalized Authorship. It is the inverse of org-babel. COMEEGA adds org-mode to your programming mode. Full and proper use of COMEEGA, requires Polymode. Let's call that Poly-COMEEGA. But Emacs's Polymode is work-in-progress, particularly now with the new tree-sitter. So, in the interim, my usage of COMEEGA has been in the form of Toggle-COMEEGA. Where I manually switch between the programming-mode and org-mode. For me this has proved to be a fine interim solution.
[00:31:05.800] Internationalization - a non-Americanist perspective
Naturally, content processing should be multi-lingual and internationalized. Let's look at that dimension. I am Iranian and much of what I write is in Farsi. Getting Perso-Arabic text right is often a challenge, as it involves Bi-Directional text (BIDI) and shaping of characters. In the context of our content generation these need to span all relevant tools, not just emacs. For emacs, I have created my own input method called farsi-transliterate-banan. My EmacsConf 2021 talk was about that. Now let's look at some examples and spice it up a bit with semantics. As an example of proper BIDI text, here is the orignal Farsi text along with English translation of Imam Khomeini's text with respect to invalidity of Western Intellectual Proprty Rights regime. And as another example of proper BIDI text, here is Ayatollah Mothari's take on Western IPR not being private property. Note that these predate by more than half a century Jack Dorsey and Elon Musk's tweets of April 11, 2025 saying "Delete all IP law". This topic is too important and too sensitive to be left to American billionaires and their tweets. Let me again refer you to the logic of polyexistentials in my book. Chapter 14 of the book is dedicated to Ethics and ownership in Religions. With respect to my preference for Ethics over Freedom, let me refer you to Section 12.4 "A Cynical Perspective on Freedom Orientation of Americans" in which I describe where the FOSS labels and the likes of Stallman, Raymond, Moglen and Lessig have gone wrong. If you are one of their followers, perhaps Chapter 12 is for you. My emphasis thus far has been on content generation.
[00:33:42.280] Autonomous self-publication and federated re-publications
Let's very briefly also look at Autonomous Self-Publication and Federated Re-Publications of our content. From the very beginning the Debian folks understood the importance of "Universality" and coined the "Universal Debian" label. This means that we can base our entire digital ecosystem on just the Libre-Halaal Debian distro. And that is what we have done with ByStar. In ByStar, everything is based on just the Universal Debian everywhere. This has made our Usage Environment totally harmonious with our Service Environment allowing for very powerful software-service continuums. Of course, all of this is immediately applicable to our ByStar Content Bundle as well. Some have asked, why don't you also include Ubuntu? I think the opposite makes more sense. Ubuntu should converge with Debian. I tried to explain this to Mark Shuttleworth in an email a while back. I have included that email in Section 12.1.5.
[00:35:07.720] Ingredients of BISOS platforms and their progression
In this presentation, we have stopped at the "Raw-BISOS" stage. We can further evolve Raw-BISOS and make it be "Sited" and provide autonomous publication services. But here by going through EmacsConf and youtube we are using the "Federated Re-Publications" model. Something this large, should be well documented. In Emacs, the way that we have been dealing with documentation and information retrieval is archaic. Man-pages, TeXInfo, Helpful-Mode and convention based Doc-Strings are old and limited. In BISOS and Blee, we use Blee-Panels for all kinds of documentation. Let me show you some examples.
[00:36:02.560] Moving forward
So, what next? If Blee, BISOS, ByStar, Libre-Halaal, Polyexistentials and these Content Processing capabilities have piqued your interest, please feel welcome to contact me. These Emacs Conferences have proven to be very useful and productive. I look forward to your thoughts, feedback and questions. I want to thank all the EmacsConf 2025 Organizers for their great work, and Sacha in particular.

Captioner: mohsen

Q&A transcript (unedited)

Greetings. Salaam again. I'm delighted to be here and happy to answer your questions. I'm not seeing anything yet, but let me take advantage of this time to speak about one question that I saw come in earlier on the pad,
[00:00:22.880] Q: Where do we find all the inputs and outputs you mentioned?
which was all the outputs and the inputs that you mentioned, where are they? So they are on GitHub, and this is in one of my slides. I mentioned the URL for it. I'll show you that as well. So the URL for it is https://github.com/bxplpc/180068, which is the handle for this talk. In there, you have all the PDFs and the HTMLs, a citation, a bib input, and also the sources. So if you were to go to the PDF, you will see um, both the article presentation and the beamer, let's take a quick look at the beamer, which is what you have seen. So. And as far as the sources are, there are two primary files. This presentation, left to right, is the one that includes all the LaTeX packages. We might as well take a quick look. So what's in there is primarily the use packages. And then it dispatches to bodyPresArt, and this is where the code is. And I walked through this briefly. So, notice here again that this is a mixture of LaTeX and Org. Each of the presentation slides are here. For example, my introduction is just a video that gets included. And then the notes that I use, the voiceover, is also included in the LaTeX file. Let me... It'll probably be easy to take those voiceover notes and then align them with a tool like Aeneas to make subtitles for your videos. Exactly, and that is what I do. So there is a way to gather them all as P-notes. And so all the P-notes get together in a single file, and then you feed that to Aeneas, and it will align them. And then there is the work of using your subed to just get the right sort of line length on them. But you did all of that for me this year, Sacha. Thank you very much. It was just a matter of not having time. Otherwise, I planned to do it myself. It's all right. It was very easy since he provided the full narration. I still need to tweak it sometimes, so I often use the waveforms in subed to find the right starting time and ending time for things. But it is so nice to have a presentation where you can experience it in different forms, as an article, as a video, as a post with links and everything. Very handy. Right, and in case a teacher uses this for class lectures, then the student profits from all sorts. The article presentation format is very useful for a student to add their own notes to it and the rest. Exactly as you said, having multiple forms is great. Video has its place, reveal has its place, PDF has its place, article has its place. All of them work together.
[00:04:48.400] Making presentations easier to distribute
I've been having a hard time figuring out how to make the reveal.js version of a presentation more easily distributable, though. This is something we've had a hard time with in the past, too. You have these lovely EmacsConf presentations that could be reveal presentations, but hosting them doesn't quite fit in the usual assumptions people have. Exactly, exactly. So if we were to have a reveal server, then we could upload our reveal inputs to it. But you're right, we should look for some sort of a packaging that is more plug and distribute. Yeah, so I'd be all game if we wanted to do it for next year, I'd be happy to provide all the reveal outputs.
[00:05:42.040] Reveal output
In fact, they are here, let me touch on that quickly. Good point you brought up. So, this is the Reveal output. So all of these, you see the images and the audio in my own environment are SIM links to where I keep them. So those need to come in, but the reveal output really comes right here. And this is how it looks. If that's a link you can add to the pad, then I can add you to the top page afterwards. I think as a, as a speaker, I tend to just self host the thing. And that way also, if I find a bug, which often happens, I can go and quietly update it. Exactly. Exactly. Sure. Sure. Yeah. This right now, what we are looking at is from GitHub. And the audio and the images, of course, are large, and I did not upload them. But I can send them, sure. So, a few other things maybe I can elaborate on. This is the bootstrapping page for BISOS, also on there. And if somebody wanted to actually dig deeper and go through this, a good starting point would be my GitHub page. But my GitHub page is not done and organized the usual way. So I only have three repositories and the repositories basically say, here is where you need to go. So all of my work is organized as organizations. So, for example, for Emacs, if you were to go to BxBlee, you will see all the relevant repos for that purpose. So, for example, if you wanted to come and take a look at mail template, templating, distribution, and tracking, you would get a Emacs package ready to go over here.
[00:08:15.000] GitHub organizations
But again, all of this is through the use of GitHub organizations. So my repos are by subject and the BISOS itself has 69 repositories where different sections of it are packaged as PIP packages. So, for example, for LCNT, we can go there. So, these are the packages that let you, let me go there, that let you dissect. Sorry, you have your slides shared at the moment, so I don't know if you're in a different tab. Oh, am I in a different tab? I thought that that would follow me. Yes, I am in a different tab. Let me see. This is the one that you are seeing, so let me go over here. Perfect. Now you're seeing it. So literally on the same page. Awesome. Right. So you were not actually seeing what I was saying. So, so what I was saying that is that if you go to my primary website, GitHub page, you will see that there are only three repos there. And those repos just give information about where the real repos are. And the real repos are organized in various organizations. So, for example, for Emacs packages, I use the bxblee And in there, there are 40 repositories. And as an example, you can choose, for example, AI plus is just a few additional libraries for menus, for Aidermacs. And the rest of BLEE is done that way as well. So if you were to let me also show more relevant stuff to the content generation. And everything related to BISOS in Python is in BISOS-PIP. These are packages that are ready to be exported to pypi and For example, for LCNT, if you go here, let's go to the bin directory. These are the utilities that dissect the PDF output of Beamer convert it to images so that you can insert them in Reveal. And then again, if you were to just wanted to dig deeper in any of these as components, you would simply start from this top level page and explore the organizations. So in total, maybe there is 300 repos, but they are organized by subject matter within GitHub organizations. Yeah. One additional general comment. If you were to look at my presentation, I'd say I touched on five different topics. So, one topic was this content generation in general.
[00:12:24.040] Challenge of DIY model and recipes
The second one was this challenge of DIY model and recipes versus building something large and including everything in it. And that is, that has been the motivation for BISOS and BLEE. And I'm interested in getting feedback on it. In general, the open source culture has been focusing on components and large American corporations have focused in integration of these components. So my short message here is that we should start thinking as providing solutions as opposed to minor pieces and packages and put them all together and claim them as our own digital ecosystem. And this is the concept behind ByStar, BISOS and BLEE. The third point I was making throughout is this concept of dynamic blocks everywhere, and Comeega, which is the inverse of Babel. And I'd also be interested in feedback on those.
[00:13:57.480] Dblocks everywhere
I think the dblocks everywhere concept, I can very quickly show that. it would be in the LaTeX file, as an example. I didn't really do a good job in digging deeper into that. So if you go to the sources and you look at any of the slides, All the slides are in here inside of a dynamic block. This is an org dblock, but you would invoke it in LaTeX mode. And then from this begin to this end, everything is auto-generated through the Elisp function, body:mm/video. And the parameters that it takes are the video path. So all of this code is repeated all over the place. And all you need in it is the video path. So it's very easy to think of this as a macro capability, except that the macro is visible. And it has one additional benefit on top of general macro capabilities, and that is that it's open. In this particular case, it closes, but let me show you another one where it is open. So if you look at, this is latex section, this one is good. This is a derived image. And by saying that it's open is that you see my begin verbatim and my begin frame. They end and there is no closing for them. So the extra text that is outside of the D block and closes it is down here. And none of the existing macro capabilities gives you this feature. This even comes handy in Lisp. So that is a proposal saying that let's make D blocks, dynamic blocks, generalized to all of Emacs, as opposed to just org mode. And that's relatively simple. And the Emacs source team could easily decide that this is worthwhile doing. And then COMEEGA, of course, I've gone over it through the presentation.
[00:17:09.960] Q: What changes have you seen in the culture while developing all these things like libre-halal system and now blee-lcnt?
So, I see 1 question coming in. What changes have you seen? the culture while developing all these things like Libre-halaal system and now Blee-LCNT? Well, it's a work in progress, I would say. We learn from one another. And what I'm doing may be considered just a stepwise increment, but the cultural input is that we really should start thinking about providing solutions as opposed to packages. The FOSS culture is really limited in its scope to packages or even if when you think something very large like Debian, which is a collection of packages. And it is still choice oriented, as opposed to solution oriented. Yeah. Are there any additional topics or questions? Otherwise, I'll just add a few additional concepts.
[00:19:11.160] Intellectual property rights
So the two other points made throughout the presentations are that this statement about clear invalidity of the Western IPR regime. So throughout the FOSS movement, we have been focusing on providing alternative licenses which coexist with the IPR system. And that is the practical thing to do. We are doing a jujitsu on IPR. We are saying that this is our license to it. But conceptually, there has been little discussion and also positioning on this basic question as to whether or not copyright and patents are valid or invalid. And what I am saying is that it is clear that they are invalid, particularly once you start looking at them with the lens of polyexistentials. And that the FOSS movement really needs to combine these two, this notion of free software and open source licenses, and combined with the belief system that we are completely and utterly against the validity of the Western intellectual property rights regime. And I use the Western and the American occasionally, and the reason behind that is that in fact it is, they are Western. If you go to any other language, if you go to Farsi, if you go to Arabic, if you go to Chinese, if you go to Japanese, which are non-Western cultures, the concept of intellectual property, the words, the vocabulary of those combination of intellectual and property, did not exist anywhere in those cultures. It's only in the past, maybe 60, 70 years that they have been translated from the Western world and brought into it. So there is an inherent root into the intellectual property rights system, which goes to the Western culture. The second point that I have brought into this presentation and the previous ones is this question of, if we go with free software, if we go with open software, are we really creating the right labels? And my point is that no, neither of them, neither free software, nor open source are capturing really the essence of what we are trying to do. And I claim that that is in fact ethics and morality. And it is societal belief that if we reject intellectual property rights regime, what do we replace it with? And in my thinking is that a software developer does not get to choose what license goes with his or her software, and that the equivalent of a Affero GPL is the default correct license to use for all of your software because it is the one that reflects the belief system that all software should be ethical software. Yeah, I'm looking at the Etherpad again and
[00:23:43.560] Q: Given that large AI companies are openly stealing IP and copyright, thereby eroding the authority of such law (and eroding truth itself as well), can you see a future where IP & copyright flaw become untenable and what sort of onwards effect might that have?
The question is about IP and AI. So yeah, over the past two years, something huge has happened. And what I am seeing in there as a solution is essentially comes down to a talk that was given maybe two years ago by someone at EmacsConf, and its label was attribution-based economics. In my thinking, intellectual property as a whole is invalid. But that means that through something like a Affero GPL, you focus on attribution basing, proper attribution basing. If somebody has done some work, it should be clear, no matter what, that that work is his. And that we already, even prior to AI, we were seeing this. We were seeing large GitHub repos with hundreds of authors. And it was utterly unclear as to who would own this whole thing. And any piece of it is not of significance. What is of significance is the whole thing. So moving towards that attribution based economics is key. And then once we do that, and then we accept AI as a reality. AI should still take very seriously and conform to attribution-based economics. In other words, what is generated by the machine should not be claimed to be no one's or the machine owners, the AI owners. It should still clearly be attributed to the people who contributed in its creation. This all becomes very muddy, very clear, and I don't have a simple or clear answer to it. But the perimeters of the solution lie in rejection of intellectual property, replacement of the intellectual property with attribution-based economics, and restrictions on AI use of not properly attributed content. Yeah, I'd say that would be, it's a complicated topic and I would simply say I haven't figured it out at all. I just have a perimeter set of concepts that can be used to drive it. Are there any other questions? If there aren't any, I thank everybody again, and particularly the EmacsConf organizers and Sacha. And I look forward to continuing all of this next year. Unless there is any objection, I'll leave the session and close it. Thank you.

Questions or comments? Please e-mail emacsconf-org-private@gnu.org

Back to the talks Previous by track: Basic Calc functionality for engineering or electronics Next by track: GNU Emacs Greader (Gnamù Reader) mode is the best Emacs mode in existence Track: General