38 comments

[ 4.0 ms ] story [ 64.2 ms ] thread
never submit a free.fr here, it won't be able to handle the traffic
Thanks, but could you elaborate on that, I don't understand why.
Free is a French ISP, and *.free.fr is free hosting (ie. not made for high traffic).
Usually nobody visits this personal site. After the short avalanche of access will return the calm of loneliness :)

And maybe I will have had a few useful feedbacks on this work...

freecodyx is saying that there are high chances that the site will go down due to the amount of traffic coming from this source regardless of its average daily traffic.

We've experienced many issues on HN about content that was hosted on Free.

EDIT: as expected, website's is down.

martyalain is saying that (s)he won't go through the hurdle of scaling a website for a 2-days buzz.
« EDIT: as expected, website's is down. »

No website is still alive, even if the "save" function has been disabled for a while. I was stupid enough to believe in people respecting a defenseless site. Destroying such a site is in no way an exploit. The exploit would be to understand how such a small language could be funny to use in a wiki.

I'm afraid it's hopeless.

some feedback about the landing page:

- it looks like a mobile design (text way to big on desktop) - needs more room to breathe - the picture takes up the whole screen (while still displaying the h1 below the fold)

Thanks. I do agree with you, the design must be improved. It's a summer workshop. What do you think of this one http://lambdaway.free.fr/workshop/ ?
On my my mobile device I get some imagery (a face and a building) and text all stacked on top of each other, making the text very hard to read.

It would be good to have a single clear statement describing the aim and purpose (what and why) of lambda speech, followed by a description of how to use it. Otherwise people will give up in confusion

The first time I saw it, I was suprised to learn that the evaluator is based on text transform: evaluation is a fixpoint where basically each match of "{[^}]*}" is replaced by its evaluation. The regex ensures values at the deepest level of nesting are evaluated first: http://b2b3.free.fr/lambdaspeech/meca/JS.js
You got it. Thanks for your attention.
In short: something very much like s-expressions is applied to create markup (and likely any computation) on a web page; The pages are combined in a wiki-like structure.

The implementation is ~20kb gzipped js.

I thought this was some kind of extension to make wikipedia or others online encyclopedia more smart. Like visualization of math stuff that can be processed and played in browser or something like that.
Am I the only one who is utterly confused by what I am looking at?
You're not alone
Did reading the text help you any? Did the word "wiki" make sense in the context? (I suppose "lambda", "lisp", and "s-expression" are not very familiar to you?)

No snark here, I just try to understand how do different people read and understand a copy like this.

On mobile I see no reference to any of those words or help. I also have no idea what's going on.
Not sure where you're reading this. All I can see is this page http://b2b3.free.fr/lambdaspeech/?view=start
Ah, the page I saw initially was pretty self-explanatory, a wiki page describing things like `{b{i{Hello World}}}` and all the way to introducing basic Lisp constructs.

It is now taken offline due to too many "stupid people", as the last line says, apparently due to them vandalizing it :(

The menu which is shown now is much less helpful indeed.

Thanks for your help. The site is still alive.
tldr; - Think of Wikipedia. Even though it’s a web site with html, people who make edits do not create html. Instead they edit this nasty quasi-text document bearing a vague resemblance to markdown. This seems to be intended for a similar purpose in wikis.

post tldr - The terrible thing was this was painful to use and early on wikis had almost no structured data. So you have incredible amounts of useful information that mostly cannot be leveraged or processed efficiently because it’s a living nightmare to parse.

Over time things have improved. They added a wysiwyg editor so you don’t have to look at it. They’ve incrementally added ways to allow for documents and data to be more structured and parsable.

Not sure how far along everything is at present.

I see. Would be nice if there was some kind of about page explaining this
OP if not on website could you please post a short description here.
This project seems very interesting, yet unfortunately I can’t think of how it could ever be the best solution for a wiki.

It’s said on the site that html and coding are often too complex of an abstraction for a wiki because so many less technically oriented folks have use for editing them.

This audience I would guess doesn’t want to hear lambda anything, I doubt they could ever fully understand the concepts. Not because they are dumb, but because they have other specialties and by definition are not investing time to learn the theory.

It’s probably been a while for many people here since they first learned about lambda calculus, functional programming, or other concepts. But try to think back to when it first hit your brain. You may have started to do some things quickly but even for many in the field it only really sinks in over time and with some continued investment.

I love how there are some undergrad theorems that someone with a middle school understanding of math can grok and learn to prove and derive the result themselves in a matter of hours. The simplicity can be so beautiful for what someone might have thought was scary.

But that doesn’t apply here. This stuff would be scary to lots of people, it’s made my brain hurt before. Now after however many years I have the luxury of reading the paper, appreciating the elegant aspects, and even enjoying thinking off all the connections to other cool concepts. But even for me is this really the most efficient abstraction?

So I like the design and theoretical aspects, nice work on all of it. I just have trouble seeing any mainstream use cases as described on the site that don’t have better potential solutions.

I think there's no one best solution for a wiki, because audiences differ.

Some people prefer a feature-laden word processor, some, a complex language like LaTeX, and some, the distraction-free notepad.exe. For each group, for their subject tasks and mental habits, their chosen solution feels best.

I don't see why this approach can't be best for some niche.

Maybe, the problem is they didn’t define the niche just talked about people who need a higher abstraction. If they did maybe it would be easier to imagine how a particular audience could be most productive with it.
When I click the link nothing renders and after I closed the window and tried to go back the link now redirects me to 4chan? I guess I wont browse HN at work anymore.
Beware: The site seems to have been compromised. I'm getting re-directed to 4chan when visiting.
Sorry. I knew that there are stupid people on the net but I didn't think so many. So {lambda speech} is locked for a while. See you later. Meanwhile, you could have a look at this one: http://lambdaway.free.fr/ .
(comment deleted)
Website just redirects to Rick Astley's "Never Gonna Give You Up" on YouTube, although it stopped doing even that now.
Sorry. I knew that there are stupid people on the net but I didn't think so many. So {lambda speech} is locked for a while. See you later. Meanwhile, you could have a look at this one: http://lambdaway.free.fr/ .
Sorry. I knew that there are stupid people on the net but I didn't think so many. So {lambda speech} is locked for a while. See you later. Meanwhile, you could have a look at this one: http://lambdaway.free.fr/ .
Lambda Talk is a language that looks like Lisp, but is based on character string processing. The code is literally the text you see, with the parentheses (braces, actually) and all. Evaluation proceeds by term rewriting using textual substitution.

Similar work:

* James Gosling of Java made something called "Mock Lisp" in the 1980's to support an Emacs editor: Now deleted Wikipedia page: https://en.wikipedia.org/wiki/Mocklisp

* Lisp in Sed: https://github.com/shinh/sedlisp