Dear Gabriel,


Thanks for all your comments. I will take them into account.


> the Caml Light error messages used to be
> localized, so there are french translations available for Caml Light.


Yes, I tried to reuse a lot of things from there.

Actually, I first wrote a patch for Spanish, but less complete.


> You should definitely share your patch in case someone is interested.



Thomas Gazagnaire offered me to put it in OPAM. Duly noted.





> I heard that there is some work ongoing to evaluate the feasability of
> using OCaml in the French junior high school [...]



If you have some reference, I would be interested.



> On a strictly personal level, I'm not terribly interested in such a
> feature: I have only interacted with OCaml beginners that, if maybe
> not completely comfortable with English, accepted to read technical
> english as part of their programming activity.



In a "perfect" world, made of young graduated polyglot engineers, you are totally right.


But in a world like planet Earth, where 70% of people never learnt English,

only 7% have a computer and 1% have an academic diploma,
things are a bit complicated.
There is the problem of aged people and children, too.


> I would be interested in more information on the profile of your
> students, that made such a change an important part of the teaching
> process.

I have been professor at the National University of Bogota (Colombia)
where I taught Mathematics and Computer science for 5 years.
There, the students have very diverse origins.
Some are teachers from other universities that come to learn more, 
but some are young people that come from very little villages and have
difficulties to even correctly speak Spanish.
Their parents sometimes barely knew how to read and write,
they have no idea how to behave in a classroom, how to redact an exam, etc.
But their good willing is not in cause... they just don't know.
They are doing a big cultural effort to come to study in a big city.

If you, the "gringo" (with all that it means), show that you are doing
a cultural effort too, you gain some respect.
This effort must be at least linguistic.

If you are involved in the development of what you are teaching
(if only adapting, debugging, deploying, documenting it) and not just only copying
someone else's book to the blackboard, you gain some credibility.
If you present something modern, tailor-made and experimental, you gain some interest.

All of that is indispensable if you want to sincerely convince the students
to play the game of intellectual rigour, what is really difficult, taking into account
the state of the educational system (especially public) of the country.

CAML is interesting from that point of view, because it helps to have clear
ratiocination, it automates almost all that can be automated in programming
and the error explanation is really instructive and sufficient to solve the problem.
(no "segmentation fault " but "this pattern-matching is not exhaustive.
Here is an example of a value that is not matched: 0" )
If everything goes well, the student is autonomous, and can progress alone.

But if a student really can't understand the sentences, the magic is not working. 
And all you get is a polite but definitive "Sir, it doesn't work !"

My experience shows that most of the time, in a classroom, when a student
is blocked by a CAML error, reading the error slowly and carefully
with words easy to understand by him, is sufficient to solve the problem.

And curiously, this fact is independent of the country... :-)

As a consequence, a big part of your job is... translating!
Here comes the idea of translating once and for all.


Moreover, IMHO, in a normal process, one obtains every error just once
(well, let's say... "a little number of times") and then learns how not
to do the same error again, until having problems with another error, etc.
Until, at last, just obtaining typos or very uncommon errors.

So, for the student, investing time learning the precise translation of an error,
thinking he will use that knowledge a lot in his whole life, is not the reality.
The reality is that he is learning how not to obtain it again, and he needs
to concentrate on that fact, the rest of the process being immediate and transparent.
The only one that will obtain the same basic errors, again and again, in his whole life is... the teacher.
Second reason for translating those errors once and for all.


That's why I wrote a Spanish patch for my Colombian students.
(I hope to be able to finalize it in a few weeks)
I also wrote a more complete and precise one for French, which is my language.
(That one is ready and I will share it soon)

Hope this answers your question.

> If such a localization
> feature was available I'm not sure I would encourage people to use it
> -- if only because a single format for error message makes it easier
> to ask for help and search the web about it.

You are right.
But what do you actually find on the web?
An very explicit clarification, from people like you, who speak with your words,
of a message in English, that was very explicit too (as usual in CAML).
Finally, some kind of... translation!
(Third reason?)


> I have more reserves about encouraging people to

> participate to the discussion in French [...]


That wasn't my intention ;-)

I was just trying not to exclude people who have big difficulties writing

in English, because the topic is all about that. Nothing more.

Sorry for that.



Alexis