workshop

user:anonymous

_h1 ELS 2018

_ul [[ELS 2018|https://www.european-lisp-symposium.org/]]
_ul30 Notification of acceptance: 19 March 2018
_ul30 Early Registration Deadline: 25 March 2018
_ul30 Final papers: 02 April 2018
_ul30 Conference start: 16 April 2018
_ul30 Conference end: 17 April 2018
_ul [[easychair.org|https://www.easychair.org/my/roles.cgi?info=65850913.k6YkBmVz2dhUIbyc]]
_ul wiki page: [[els_2018]]
_ul paper submitted on 2018/02/18 (update 2018/02/24): [[els_2018_0224.pdf|http://lambdaway.free.fr/workshop/data/els_2018_0224.pdf]]

{hr}

_h6 The paper has been rejected.

{hr}

{pre {@ style="white-space:pre-wrap"}

{b Review 1}

-2: (reject)
The paper describes a Lisp dialect that is interpreted by a JavaScript program essentially by string substitution, and how it can be used to program web pages.

** General comments

I like the spirit of building up a minimalistic programming language
and taking it all the way to web-page programming. However, the paper has severe shortcomings: It's essentially a small variation of a paper from the previous year's ELS, without mentioning that fact.
Consequently, the "web part" of the paper has only very little room
and too little detail to be edifying.

** Detailed comments

*** Abstract

The abstract never mentions what the '{lambda way} is, or what the
motivation for the paper is.

*** Introduction

Mention the URL of '{lambda way} early. What I take from the text is that '{lambda way} is essentially a programmable wiki. Is that accurate?

1. '{lambda talk}'s foundations

- "unkwon" -> "unknown"

- I don't understand how the replacement engine works wrt words: It
seems to ignore word boundaries and replace inside words as well.
The paper fails to discuss the problems inherent in that.

- "SCHEME" -> "Scheme", also the citation should point to a primary
reference - one of the original Scheme papers or a standard, not
SICP.

- The enumeration is in tt, should be in roman font

2. '{lambda works}

The construction here is essentially standard, and could be reduced
greatly to make room for more web material.

The description of the evaluation engine is too brief to be
comprehensible. Instead, the reader should be referred to the
previous ELS paper.

3. '{lambda numbers}

This section also doesn't really do anything to bring the material
forward.

4. '{lambda web}

Now I could see a paper emerging from this section, but it's just no
there. The treatment of web programming total is only about a page.
The wiki aspect is hardly explained at all.

{b Review 2}

Overall evaluation: 	
0: (borderline paper)
This contribution reports about the '{lambda way} project which is about the interesting approach of obtaining a web language by implementing the Lambda calculus in Javascript running on a browser and then extending it in the usual way towards a more powerful language with Lisp-like syntax and semantics.

While I find the idea interesting, my main problem with this
contribution is that it is extremely similar to the contribution for ELS 2017.

Therefore, I would suggest to accept this contribution only after a modification is made which shortens it substantially by not repeating what was submitted to ELS 2017. I guess that the size should then be below 2 pages or at most 4 pages.

It should also address some or all of the following questions:

- How does '{lambda talk} differ from other tools? What is the
advantage?

- How can I install '{lambda talk}? (See also below)

- What are the future goals of the project?

Some smaller things:

* The use of boldface is not very consistent, e.g. "the making of" in the abstract.

* In Section 2: Is the usage '{HI} in HI, I just say '{HI}.
 really a function application in the sense of Section 1, (3.)?

* According to Section 2, we have

'{{lambda {fruit} The color of fruits ...} apple} ->
The color of apples ...

This looks a like a kludge, because it probably fails for nouns with
a more irregular plural like "cherry -> cherries".

* I had hoped that I could download a Javascript program from
somewhere and then play around with '{lambda way} myself running in my browser. However, I did not find such a link, and at the end the article mentions that also PHP on a server is needed. Maybe the author could clarify the setup a little bit more.

}

_p I have nothing to answer to such a reviewers. We are just in parallel worlds and I'm probably the alien. I'm just disapointed. Game over!

_p This is what I wrote today (2018/03/21) to some of my french friends.

{pre {@ style="white-space:pre-wrap"}
Bonjour,

Mon papier pour ELS_2018 a donc été rejeté, (http://lambdaway.free.fr/workshop/?view=ELS_2018). Les deux relecteurs me reprochent, entr'autre, la trop grande similitude entre ce papier pour ELS_2018, (http://lambdaway.free.fr/workshop/?view=els_2018) et celui que j'avais présenté pour ELS_2017, (http://lambdaway.free.fr/workshop/?view=brussels_paper). Et pourtant, s'ils les avaient vraiment lus, comparés et compris, ils auraient pu constater quelques clarifications importantes apportées dans le second justifiant une nouvelle présentation. Notamment :

1) Suivant une introduction compréhensible par un simple mortel, il est montré comment trois règles et un unique prédicat travaillant sur de simples mots suffisent pour introduire les concepts clés de structures de données et de récursion. Je ne connais aucune présentation de la récursion qui soit plus simple, reléguant le Y-combinateur au rang d'amusant exercice, en tout cas secondaire.

2) Le concept de nombre comme liste, remplaçant l'approche un peu ancienne proposée en son temps par Alonzo Church et donnant un éclairage plus simple et "symétrique" aux opérateurs classiques, [+,-,*,/,%].

3) L'utilisation des fonctionalités des navigateurs modernes pour construire à l'envie sur une base solide et unique des librairies de plus en plus sophistiquées conduisant à un vrai langage de programmation programmable.

Il n'y a donc vraiment rien à dire, à partager sur une aussi petite application web - 100kb - qui sait accompagner depuis les premières esquisses toute la création d'un document, son écriture, sa structuration, sa présentation, l'insertion de code vivant - testable en ligne dans une page wiki - et qui permet in fine la production directe de PDFs satisfaisant - de façon perfectible bien sûr - aux exigences académiques sans avoir à utiliser de monstrueuses usines à gaz comme LaTeX ou MSWord ?

Tout ceci ne mérite-t-il donc aucune considération ?
Suis-je perdu au fond d’une impasse à radoter sur d’inutiles questions ? 
Auriez-vous un peu de temps à m'accorder pour en discuter ?

Cordialement,
Alain Marty
}