[erlang-questions] Lets build a stock exchange!
Mon Nov 12 18:37:58 CET 2007
We need many books. Enough to drown the Java books. Yes we do!
I imagine the stock exchange to be an example of something that
needs throughput and high service levels. Lessons learned will be
applicable for other projects. A system related to the financial sector
can also help to popularize Erlang more there.
2007/11/12, Valentin Micic <>:
> Would not like to sound negative, but how many people would actually have a
> chance to do something similar (i.e. build a stock exchange system) in their
> careers? To me, this sounds like teaching someone to build affordable
> housing by explaining how to build a pyramid. If the main goal of writing a
> book is to popularize a specific subject, the stock exchange system, IMHO,
> would not cut. If you want a popular support, write a book about popular
> subject. For example, I think that erlang's implementation of jabber
> protocol (XMPP, etc), did a great deal to erlang's popularity...But then
> again, how many books do we need?
> ----- Original Message -----
> From: "Joe Armstrong" <>
> To: "Torben Hoffmann" <>
> Cc: "Erlang Questions" <>
> Sent: Sunday, November 11, 2007 11:45 PM
> Subject: Re: [erlang-questions] Lets build a stock exchange!
> > This sounds like a great idea.
> > I have no idea whether it would succeed or fail - if it succeeds then
> > (gulp)
> > If it fails the interesting thing is not that it failed but WHY it failed.
> > If things fail and we know why we can do research and fix things. It
> > is the existence of things
> > that don't work that stimulates software research and development.
> > Interesting problems are very difficult to think up - asking the right
> > question is
> > often the really difficult bit - once asked lots of brains can try and
> > find the solution.
> > Joels suggesting poses a difficult question - great - let's try and answer
> > it.
> > If Joel can in any way structure the problem then there are loads of
> > us that can try and
> > write the individual bits.
> > Breaking the system is great - a program that breaks the compiler is
> > far better than 100 that work.
> > A program that breaks mnesia is far better than 1000 that work.
> > If this breaks Erlang great - we can fix it and make an even better
> > system.
> > Breaking stuff is easy - the interesting point is *where* does the system
> > break!
> > /Joe Armstrong
> > On Nov 11, 2007 10:10 PM, Torben Hoffmann <>
> > wrote:
> >> On Nov 11, 2007 9:43 PM, Christian S <> wrote:
> >> > Anyone know a book on technical requirements that brokers put on stock
> >> > exchanges? Order matching procedures, what information do tax offices
> >> > want, typical features as in stop-loss rules...
> >> >
> >> > Oh, and yes. A stock exchange is cooler than poker.
> >> It may very well be the case that a stock exchange is cooler than poker,
> >> but
> >> I think that the poker server has some interesting aspects that could
> >> inspire others. E.g., the dynamic allocation of a server to run a new
> >> table
> >> as well as the handling of the messages between the players is similar in
> >> nature to some of the challenges there is in telecom with regards to
> >> group
> >> calls and where to "host" the call.
> >> Furthermore, I would rather see a well-written book on hard-core Erlang
> >> hit
> >> its window of opportunity than see the whole thing turn into "let's make
> >> the
> >> coolest application that will never run in real life just in order to
> >> spice
> >> up a programming book"-kind of projects.
> >> A well written book with a clear example of usage including discussions
> >> about where the general ideas can be applied is what I would bet my money
> >> on. If a potential buyer of the hard-core Erlang book cannot abstract
> >> from
> >> the specific case study used to teach the principles then it is time to
> >> utter "Joe, We have a problem!" ;-)
> >> Cheers,
> >> Torben
> >> P.S. My first reaction to the stock exchange idea was "Wow!! Cool!!!",
> >> but
> >> then my analytical side issued a kill to that process for the reason
> >> listed
> >> above.
> >> >
> >> > _______________________________________________
> >> > erlang-questions mailing list
> >> >
> >> > http://www.erlang.org/mailman/listinfo/erlang-questions
> >> >
> >> _______________________________________________
> >> erlang-questions mailing list
> >> http://www.erlang.org/mailman/listinfo/erlang-questions
> > _______________________________________________
> > erlang-questions mailing list
> > http://www.erlang.org/mailman/listinfo/erlang-questions
> erlang-questions mailing list
More information about the erlang-questions