I worry this model could lead to a "wall of text" rather than forcing the presenter to be concise, speak to the concepts, and rely more heavily on images. The format is good for academic lecture scenarios but I'd probably just use Jupyter, as noted by others.
I get that it’s usually seen as bad practice to write down all the points and read from the slides, but honestly, I like that style sometimes. My monkey brain gets distracted easily, and I often lose track. Having detailed agendas displayed on the screen helps me follow along better.
I especially like it for reading the slides without the video. Oftentimes, the slides for a presentation will be available but a video recording will not be, and having all the information on the slides makes it easier to learn from those presentations.
Of course, the "correct" way to do this would be with "speakers notes", but those seem to often be stripped off of archived presentations for whatever reason.
The timing is super important. Writing on a chalk/whiteboard or overhead takes time, which is required to absorb the content. Hitting the right timing and cues with this approach seems like it would take a lot of practice, which isn't a deal breaker, though I believe the 2 biggest problems with most presentations is too much/too fast content, and not enough practice and this approach might make those both harder, not easier.
This immediately reminded me of the "infinite" blackboards we had in some university lecture theatres such that you could just revolve the board surface and keep claiming new writing space (you could also write at a consistent level - no bending / stretching).
I thought you were talking about the lecture theaters where there were 6-8 giant boards arranged on the wall with pulley systems so the lecturer can raise/lower them and access them independently, kind of like in [0]. Some of the setups are have enough writable area that it's not even really practical to fill them all up in a 2-hour lecture period. Using the available space in a smart way is definitely a learned skill; it's far too easy to start in the "wrong" place and end up needing to write "board via"s / "goto"s.
Do companies still do whiteboard interviews anymore for engineers? Whiteboard management in that context was also a learnable skill, though easier than managing multiple movable boards.
When I was in college they would use a transparency projector, but instead of transparency pages, there would be a roll with a crank that you would turn to advanced the roll. You would get the same effect as slipshow.
I did an old experiment on a scrollable whiteboard with replay that I built after watching a khan academy style video and wanting to scroll to back to a formula without pausing the audio. This makes me want to dig it back ^^
Prezi lost me a PhD co-advisor, this person only liked traditional slide formats. In retrospect, this was a good thing to not have this person oy board. I used it pretty close to traditional slides, but they could not be made happy unless it was powerpoint... Separately, I did have a lot of fun making a Prezi to explain how GPUs worked for another class, which went over swimmingly for a different audience.
I ended up using reveal.js for my defense slides because the 2D slide grid allows you to go deep on a subject and keep the main flow clean.
The markdown to presentation approach is great. You can manage your slide (or slips) as code giving you history, offline collaboration, pull requests, etc. I don't think you can do that with most other presentation tools. I've used Marp [1] for traditional slides, and wrote a GitHub template repo that outputs the Marp HTML to GitHub Pages [2]. Similar workflows should be possible for Slipshow.
I see a huge markdown fan! I'm on the same page! We're building an entire ecosystem for markdown slides for the same reason.
The same design for all slides in the same workspace, a hosted version for non-developers, each slide has its own public url for easy slide playback, and more ...
Try if you are interested in slidepicker.com (beta)
I've wanted to love MARP for years but the documentation is so limited (and tutorials so sparse) that building a presentation with any complexity beyond the basics is almost impossible unless you have a solid understanding of CSS and YAML front matter.
It's a shame, because I think the idea of MARP is revolutionary, but in practice PowerPoint (or even free alternatives like Google Slides) is easier to use.
Having modifiable source inside the html is a great idea. I'll still need to have a compile workflow to add image while keeping a standalone produced html file (and some other feature such as live-preview), but I'll definitely add this as a possible workflow.
Thanks for the link, and the idea! Definitely a github action to automatically publish your Slipshow presentation on push would be a great workflow! (I'll open an issue for that!)
Also, Quarto to reveal.js is an amazing pipeline, and ties in well with the rest of the Quarto ecosystem meaning that you can reuse content from other mediums more easily.
I like it. Slides are the right approach for a lot of things but writing on blackboards and overheads had their own benefits that just scrolling a document doesn't really replicate.
As others have noted, you probably need to be cautious about just creating a wall of text but I can definitely see its uses.
> When using traditional slides, you are given a rectangle of white space to express your thought. When this rectangle is full, you have no other choice than erasing everything, and start again with a new white rectangle.
It's not entirely solving the issue though because you are animating everything all the time, you barely have much time to see the content
>> When using traditional slides, you are given a rectangle of white space to express your thought. When this rectangle is full, you have no other choice than erasing everything, and start again with a new white rectangle.
But if you scroll to far you have the same result and if you scroll less you may just use a slide where just move the bottom content to the top and add new content at the bottom.
I myself often still remember the position of the content even if I can't remember the content itself.
So I can skip through the slides and only watch certain parts of it without the need to read everything.
I think it's a very nice concept, but sometimes when you present, you have to use other people's computers, and I am wondering if it's possible to generate a PDF from this and use it for presentations?
My second thought is that something like this could be made as a LaTeX package, similar to Beamer. For scientists, there are many benefits to using LaTeX for presentations.
If I'm understanding it correctly, I think it renders your presentation to an html file that you can just open in a browser. I assume the effects and transitions are all powered by embedded JS. So it should be portable.
Reminds me of Prezi, a lot more than it reminds me of Jupyter notebooks.
However, as someone who prefers to create PDF slides using LaTeX instead of PowerPoint, I completely prefer a scriptable tool instead of a web-based tool.
This. It's notebooks as presentations and I'll argue there is nothing about slides that explicitly prevents you from taking this visual approach, but for the case where you want less flexibility it could make sense.
This is a cool idea! A few years ago I did a (short) presentation out of a Notion doc for some radio-related prototyping and the format was a hit - scrolling works really well for certain topics.
Nice! I do something similar with Dropbox Paper in presentation mode and also with Jupyter notebooks with the presentation feature. Of course, without all the extra nice features this has.
Good question :) Many reasons, I could make a blog post about it!
First, let me start with the biggest drawback of using OCaml: it might prevent potential contributors from making contributions. That is very sad!
However, OCaml is the language I know the most. I use it at work and I'm very very satisfied with it. It's main strengths for me are its type system for maintainability, the tools around it (the lsp server, the build system dune, the autoformatter ocamlformat, ...) and the fact that it can compiles to Javascript! It also has some libraries of very high quality.
With a "single" codebase, leveraging the javascript ecosystem, I could make the compiler work as a statically linked binary, as a node script published on npm, inside a Tauri app, and inside a VSCode plugin. ("single" codebase in quote since I needed some specific code for each application, but the core logic is shared.)
For sure, the same could have been made with another language, but I knew OCaml, and knew I would have a pleasant experience using it for that.
(I even plan to rewrite the engine in OCaml. The engine was written in javascript quite quickly, at a time when I had no experience with largish projects. It is now very difficult to maintain and extend.)
I don't know Miro or Figjam very well, but I would say that:
- You can write a Slipshow presentation in Markdown, which can be more convenient than in those collaborative whiteboards (depending on the person): the source file is plain text.
- Slipshow is made for presentations, so it is themed for that: the usual ratio for the screen, blocks such as example/theorem/definition, titles, ...
- It is easy in Slipshow to reveal new content/go to a new position by simply pressing the "right arrow" key.
- The output from Slipshow is a single file that you can view offline, send to your audience, ...
There might be more differences. For sure, those tools are different and adapted to a purpose, or a style of presenting. For some kind of presentations, Miro and Figjam might be much better than Slipshow!
