next meeting
jdelgado
Nice meeting past wednesday!
On the next one. Johan Montelius accepted to give a talk (between
30'-1h) on Erlang, and I suggested tuesday April 13th (it's difficult
for me to get to the meeting before 20:00h on wednesdays).
Since I was appointed "chair of the meeting" I should ask whether
there are any volunteers to give a lightning talk (provided there is
enough time left). I will prepare my lightning talk on the Futamura
projections (or, who needs a compiler?).
Have a nice holidays!
Bests,
Jordi
Andy Wingo
On Mon 29 Mar 2010 08:51, jdelgado <jdelga...@gmail.com> writes:
> On the next one. Johan Montelius accepted to give a talk (between
> 30'-1h) on Erlang, and I suggested tuesday April 13th (it's difficult
> for me to get to the meeting before 20:00h on wednesdays).
Excellent!
Johan Montelius: http://web.it.kth.se/~johanmon/
Andy
--
http://wingolog.org/
Raimon Grau
> Hi,
>
> Nice meeting past wednesday!
>
come to the meeting.
> On the next one. Johan Montelius accepted to give a talk (between
> 30'-1h) on Erlang, and I suggested tuesday April 13th (it's difficult
> for me to get to the meeting before 20:00h on wednesdays).
>
Great! For me it's ok on Tuesday.
Btw, I've seen there's a variant of Scheme intended for distributed
computing that mimics erlang's concurrency model called termite.
It runs on top of Gambit scheme and its site is here:
http://code.google.com/p/termite/
If noone takes it, maybe I can prepare a lightning talk for the next one (may)
and see if I can come up with something interesting. In the worst case,
I'll be showing the termite syntax to do message passing and so
(which can be found in the docs), but...
> Have a nice holidays!
>
> Bests,
Regards,
Rai
--
blog -> http://puntoblogspot.blogspot.com
jdelgado
asked again for confirmation, so let's see...
We keep, for now, next tuesday as the date of the next meeting,
starring Johan Montelius and the guest, lightning, stars
would be myself and... anyone else?
Bests,
Jordi
On 29 mar, 12:47, Raimon Grau <raimons...@gmail.com> wrote:
Xavier Noria
Sent from my iPhone
> --
> To unsubscribe, reply using "remove me" as the subject.
jdelgado
place as usual.
We should not give a poor impression of the group, so:
- I think this is the first time we have an external speaker in our
group. So, please, come to the meeting (I am not very polite, and even
if I were, I do not know how to be polite in english, so let me talk
imperative).
- I will give a lightning talk. We *need* at least one more. Xavier,
five minutes of metaprogramming in Ruby? Rai, five minutes of Erlang
FSMs generated from Perl? Andy, five minutes of hygienic macros in
Scheme? Andy, five minutes of how nice is haskell? Jao, five minutes
of your experiments with Termite? Come on!, it's not that we have no
topics, isn't it?
See you tomorrow!
Bests,
Jordi
On 12 abr, 00:46, Xavier Noria <f...@hashref.com> wrote:
> Would like to attend if Johan confirms
>
> Sent from my iPhone
>
Colin Fleming
I'll definitely try to make it, I've missed the last couple and I'm
very interested in Erlang. However I twisted my neck on Friday and I'm
still on the couch, so I'll see how I'm feeling tomorrow.
Cheers,
Colin
Xavier Noria
> Ok, Johan confirmed, so we will have Erlang tomorrow. At 19:30h, same
> place as usual.
>
> We should not give a poor impression of the group, so:
>
> - I think this is the first time we have an external speaker in our
> group. So, please, come to the meeting (I am not very polite, and even
> if I were, I do not know how to be polite in english, so let me talk
> imperative).
>
> - I will give a lightning talk. We *need* at least one more. Xavier,
> five minutes of metaprogramming in Ruby? Rai, five minutes of Erlang
> FSMs generated from Perl? Andy, five minutes of hygienic macros in
> Scheme? Andy, five minutes of how nice is haskell? Jao, five minutes
> of your experiments with Termite? Come on!, it's not that we have no
> topics, isn't it?
Can't promise.
I don't think that a ton of talks is a goal to aim for anyway. If I
was the guest, gave my talk, and then sat down with all of you around
that big table with some beers I wouldn't have a poor impression of
the group at all. I am the guest! It is fine that I am the one that
gives the (only) talk.
But that's the way I like meetings anyway, a talk (perhaps) and
informal chat. I see from the mails that flibug is more oriented to
talks and I am not a regular, just expressing my viewpoint in case it
can relax a bit things for tomorrow :).
Colin Fleming
> that big table with some beers I wouldn't have a poor impression of
> the group at all. I am the guest!
+1. I like the talks, but I like the discussions as well.
Marc-André Lureau
> place as usual.
>
> We should not give a poor impression of the group, so:
>
> - I think this is the first time we have an external speaker in our
> group. So, please, come to the meeting (I am not very polite, and even
> if I were, I do not know how to be polite in english, so let me talk
> imperative).
>
Sure, I'll be coming!
--
Marc-André Lureau
Colin Fleming
Thanks for the meeting last night, it was very interesting. I'd wanted
to learn more about Erlang, it was a nice introduction to it.
For those that are interested, the LWT/Actor project on the JVM I
talked about was Kilim (http://www.malhar.net/sriram/kilim/). The two
papers there are a very interesting read, and Sriram's PhD is pretty
interesting too
(http://www.malhar.net/sriram/sriram-diss-final-jan28.pdf). Kilim
doesn't provide everything Erlang does, in particular process
monitoring isn't present which IMO is one of the more interesting
aspects of Erlang.
Curiously enough given the discussion around independent
implementations, there's actually a project to run Erlang on the JVM
with Kilim under the scenes
(http://wiki.github.com/krestenkrab/erjang/). It's pretty impressive,
looks like they can actually boot OTP now, although it still doesn't
run anything particularly complex. Preliminary results are that it's
actually faster than Erlang, but loses some of the benefits. Processes
aren't as tightly isolated and in particular don't have independent
heaps so you'll still get stop-the-world GCs. That said the Java GC is
an unbelievable piece of work, we run production JVMs with 4GB heaps
and full GCs are pretty rare if you tune things right. If they manage
to build Erjang into a production project it could be a serious
competitor for the runtime, although it wouldn't help with language
evolution since they're still running the same Erlang bytecode.
Interesting stuff!
Cheers,
Colin