caml trading

caml trading

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 483 bytes --]

For those who are interested, I just posted a video of a talk I gave at CMU
several weeks ago about Jane Street's use of OCaml.  This is an updated
version of the talk I gave at POPL last year.  Various people have expressed
an interest in showing the talk to their undergraduate classes, so I thought
it worth mentioning on the list.

Here's the link:

  http://ocaml.janestreet.com/?q=node/61

I can also provide the original quicktime file from which this is based on
request.

y

[-- Attachment #2: Type: text/html, Size: 577 bytes --]

Re: [Caml-list] caml trading

[Yoann Padioleau]

Yaron Minsky <yminsky@gmail.com> writes:

> For those who are interested, I just posted a video of a talk I gave at CMU
> several weeks ago about Jane Street's use of OCaml.  This is an updated version
> of the talk I gave at POPL last year.  Various people have expressed an
> interest in showing the talk to their undergraduate classes, so I thought it
> worth mentioning on the list.
>
> Here's the link:
>
>   http://ocaml.janestreet.com/?q=node/61
>
> I can also provide the original quicktime file from which this is based on
> request.

Great talk. I love that some people are doing business
with such a great language, and all the stuff Jane Street is doing
around OCaml is really cool (the open source library, the blog, 
the summer project). 

Having said that, about your company, Jane Street,
aren't you part of the people that put the countries in such trouble ? 
Do you create wealth ? It seems you are just a parasite ... that
exploits some loopholes in the system.

PS: I applied to your company 3 years ago, and got rejected, 
so I may be biased :) 

>
> y
> _______________________________________________
> Caml-list mailing list. Subscription management:
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
> Archives: http://caml.inria.fr
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs

Re: [Caml-list] caml trading

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 767 bytes --]

On Fri, Mar 13, 2009 at 8:56 PM, Yoann Padioleau <padator@wanadoo.fr> wrote:

>
> Having said that, about your company, Jane Street,
> aren't you part of the people that put the countries in such trouble ?
> Do you create wealth ? It seems you are just a parasite ... that
> exploits some loopholes in the system.
>

What Jane Street does is very far from the part of finance---and the ways of
thinking---that caused these troubles.  You can't just lump all of finance
together as "the guys who caused the trouble"; that would be like condemning
all engineers for

http://www.youtube.com/watch?v=P0Fi1VcbpAI

The beginning of the talk gives a flavor of role that Jane Street actually
does play in the markets, so that's a good thing to watch if you're
interested.

y

[-- Attachment #2: Type: text/html, Size: 1179 bytes --]

Re: [Caml-list] caml trading

[Yoann Padioleau]

Yaron Minsky <yminsky@gmail.com> writes:

> On Fri, Mar 13, 2009 at 8:56 PM, Yoann Padioleau <padator@wanadoo.fr> wrote:
>
>
>     Having said that, about your company, Jane Street,
>     aren't you part of the people that put the countries in such trouble ?
>     Do you create wealth ? It seems you are just a parasite ... that
>     exploits some loopholes in the system.
>
>
> What Jane Street does is very far from the part of finance---and the ways of
> thinking---that caused these troubles.  You can't just lump all of finance
> together as "the guys who caused the trouble"; that would be like condemning
> all engineers for
>
> http://www.youtube.com/watch?v=P0Fi1VcbpAI

Sure. 

>
> The beginning of the talk gives a flavor of role that Jane Street actually does
> play in the markets, so that's a good thing to watch if you're interested.

I did watch the video, that's why I asked the question. From the video
it sounds like your are buying and selling things on the market
at the micro-second level. I don't see any value, any wealth created
doing that. The only thing that come to my mind is that there should
be a law that forbid such micro-transactions. I can see the
value of banks, of investors, I don't understand the value
of "arbitrage". To make money flow better ? ...


>
> y

Re: [Caml-list] caml trading

[Jim Miller]

>>
>> The beginning of the talk gives a flavor of role that Jane Street actually does
>> play in the markets, so that's a good thing to watch if you're interested.
>
> I did watch the video, that's why I asked the question. From the video
> it sounds like your are buying and selling things on the market
> at the micro-second level. I don't see any value, any wealth created
> doing that. The only thing that come to my mind is that there should
> be a law that forbid such micro-transactions. I can see the
> value of banks, of investors, I don't understand the value
> of "arbitrage". To make money flow better ? ...
>
>

The Internet is a wonderful thing: http://en.wikipedia.org/wiki/Arbitrage

Now, attempting to keep this focused on what was a good talk.

You discussed the issues that Jane Street has in "programming in the
large."  We have a small OCaml investment here and we are probably
going to significantly expand that.  While I know that time is a huge
factor, I'd be very interested in any blogs or writings you have on
what specific techniques you used to help with these issues.  In
particular, the build process that you use.  I've seen you mention in
the past on this list that you've had issues with some the existing
build tools.

Re: [Caml-list] caml trading

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 1047 bytes --]

On Mon, Mar 16, 2009 at 10:30 AM, Jim Miller <gordon.j.miller@gmail.com>wrote:

> You discussed the issues that Jane Street has in "programming in the
> large."  We have a small OCaml investment here and we are probably
> going to significantly expand that.


I'm curious, what are you guys doing with caml?


> While I know that time is a huge
> factor, I'd be very interested in any blogs or writings you have on
> what specific techniques you used to help with these issues.  In
> particular, the build process that you use.  I've seen you mention in
> the past on this list that you've had issues with some the existing
> build tools.
>

Our blog is here:

   http://ocaml.janestreet.com

And I wrote a post about our build system here:

   http://ocaml.janestreet.com/?q=node/36

Interestingly, it is the most popular post on the blog.  Turns out, there
are more people who care about version control than people who care about
functional programming.  (It's not that bad, though.  The second most
popular post is on type-indexed values).

y

[-- Attachment #2: Type: text/html, Size: 1651 bytes --]

Re: [Caml-list] caml trading

[Andres Varon]

[-- Attachment #1: Type: text/plain, Size: 1778 bytes --]


On Mar 16, 2009, at 2:35 PM, Yaron Minsky wrote:
>
>
> Our blog is here:
>
>    http://ocaml.janestreet.com
>
> And I wrote a post about our build system here:
>
>    http://ocaml.janestreet.com/?q=node/36
>
> Interestingly, it is the most popular post on the blog.  Turns out,  
> there are more people who care about version control than people who  
> care about functional programming.  (It's not that bad, though.  The  
> second most popular post is on type-indexed values).

Just to add the bit of our own experience here:

We have a relatively small group of people here (at most 5 working on  
the same project, some in separate locations). We also moved to  
mercurial more than a year ago (we started with subversion), and have  
automated tests before and after each commit using buildbot (http://buildbot.net/trac 
). Each developer can submit his local changes for test before pushing  
to the main repository, to avoid breaking the central repository that  
we all pull from. The tests run in numerous processors and OS'es  
(including parallel machines using MPI, Linux, Windows, and Mac OS X).

This setup works fantastic for our group (and is not very different  
from that of Jane Street). The main weakness was the Makefile (no  
wonder). To solve it, few weeks ago I moved all the build scripts to  
ocamlbuild, and it is working like a charm now. Our lives are a little  
bit better (once again), thanks to the french fellows :-).

Andres

>
> y
> _______________________________________________
> Caml-list mailing list. Subscription management:
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
> Archives: http://caml.inria.fr
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs


[-- Attachment #2: Type: text/html, Size: 2582 bytes --]

Re: [Caml-list] caml trading

[Markus Mottl]

On Mon, Mar 16, 2009 at 10:24, Yoann Padioleau <padator@wanadoo.fr> wrote:
> I did watch the video, that's why I asked the question. From the video
> it sounds like your are buying and selling things on the market
> at the micro-second level. I don't see any value, any wealth created
> doing that. The only thing that come to my mind is that there should
> be a law that forbid such micro-transactions. I can see the
> value of banks, of investors, I don't understand the value
> of "arbitrage". To make money flow better ? ...

This is clearly not on-topic for this list, but since it may be of
interest to OCaml-developers who want to work with us, let me quickly
dispel this unfortunately very common misconception concerning the way
markets work and what the actual service is that market makers /
traders like us (and in general) provide.

First of all, what do you think would be a "fair" average period that
traders should hold their positions?  Ten years?  Three months?  One
nanosecond?  It's clear that different people might have different
subjective opinions on fairness so any of the above would then just be
an arbitrary choice.

But there are clearly economic reasons for transaction speed.  E.g. a
trader in a grocery store surely won't buy and sell apples within
milliseconds.  Even if it were physically possible, we would certainly
expect that the required infrastructure would be too expensive.  It is
hard to imagine that the added benefit in more accurately meeting
demand for apples is big enough that customers would be willing to pay
the extra cost for that level of service.

Thus here we have it: cost and profit, these are the determinants of
how fast you should trade.  It is costly for us to be able to trade
fast: you may need to buy very expensive computer equipment, pay a lot
of money for realtime financial information, and, most importantly,
hire good OCaml hackers (a rare species!).  As long as the extra cost
of investment is at least compensated by extra profits (look up
"marginalism" on Wikipedia), it is economically advisable to trade
faster.

What is now the added benefit for society?  How do you personally
benefit from companies like ours?  This is easily explained, just
imagine we did not exist.  We can only be profitable by providing
better prices than competitors, otherwise nobody would trade with us
on a given market.  Therefore, if we didn't exist, you would have to
pay a higher price to somebody else.  Same if you want to sell again:
you would receive less for your investment from a competitor, thus
making you worse off.

Thus, besides providing wonderful jobs for OCaml-hackers, our company
makes it cheaper for people to invest their hard-earned money by
providing extra liquidity on financial markets.

Regards,
Markus

-- 
Markus Mottl        http://www.ocaml.info        markus.mottl@gmail.com

Re: [Caml-list] caml trading

[David Baelde]

Hi,

I'm not sure how much this is off-topic... Although this thread was
intended to be about an industrial use of OCaml, one cannot ignore
other aspects like advertising: it is rarely possible to separate
issues as clearly as we'd like. Anyway, I feel very much concerned
about this debate, and I'd like to add another non-Jane-St opinion.

Like many others, I've been feeling for a long time that many
financial products and practices don't make sense. I recognize that we
live in a complex world, and perhaps I should just learn more about
economics. However, the recent crisis showed that this feeling is not
entirely unfounded, even if it does not show that finance is harmful
as a whole.

I agree with Yoann's old-fashioned feeling: if you don't bring
anything to society, why should you make that much money? The point
against super-fast trading is not so much that speed is immoral in
itself, but that such fast transactions usually seem to fall in the
category of no-value-added.

A simple answer to this criticism is that the notion of usefulness to
the society is very controversial. Let me comment on the answer that
was given instead.

On Mon, Mar 16, 2009 at 4:31 PM, Markus Mottl <markus.mottl@gmail.com> wrote:
> Thus, besides providing wonderful jobs for OCaml-hackers, our company
> makes it cheaper for people to invest their hard-earned money by
> providing extra liquidity on financial markets.

All you're saying is that you believe in free trade. If my mobile
phone company is ripping me, they can justify it because other
companies might rip their customers even more? Notoriously, this
argument does not always work: in France, all mobile phone companies
practice similar insane prices and no "fair" one arises.

I'd like to understand finance, maybe I'll read some serious stuff
about it some day. But I feel that it's really too bad that such an
important matter is inaccessible to most citizens. More importantly,
one should not fool people (or oneself) with sloppy metaphors and
apparently flawless reasoning when it actually boils down to
non-trivial beliefs about the world economy.

In my opinion, a major sickness of our world is that it is ruled by
people who claim to understand it when they don't. Politicians always
have solutions, never doubts. Obviously, it's just a way to justify
why they're ruling those who don't understand. We should keep our eyes
open and don't let it happen unnoticed, nor reproduce it ourselves.

Despite my doubts, I enjoyed Yaron's talk and appreciate the Jane St
involvement in OCaml. I just wish it were that simple.

Cheers,
-- 
David

Re: [Caml-list] caml trading

[Alexy Khrabrov]

Ah, what the heck, let's put this into a good "Old Europe" vs "New  
World"-style, capitalism vs. socialism perspective!

On Mar 16, 2009, at 10:37 PM, David Baelde wrote:

> Like many others, I've been feeling for a long time that many
> financial products and practices don't make sense. I recognize that we
> live in a complex world, and perhaps I should just learn more about
> economics. However, the recent crisis showed that this feeling is not
> entirely unfounded, even if it does not show that finance is harmful
> as a whole.

Jane St is a market maker.  They have to enable liquidity in the  
markets.  They offer to buy or sell shares and they fulfill their  
promise -- exactly and fast.  The terms of their offers are crystal- 
clear to all participants and are enforced by the electronic market.   
The terms are known in advance, no hidden charges, no gimmicks.  Take  
it or leave it.  Very different from the "financial crisis."  Say a  
sewage tank explodes in Lyon and covers passers-by with its contents.   
Does it mean all sewage business is a scam?

The failed securities were derivatives, stemming from the rotten  
mortgages pushed onto unqualified American buyers, mixed with AAA  
bonds, rated highly by crooked agencies, and sold to the world  
investors.  Many European investors wanted to basically exploit hard- 
working American home owners, working with higher American  
productivity to uphold their higher American way of life.  The  
productivity of the US worker is higher than that of his European  
counterpart because EU is loaded with socialist obstacles to  
productivity, and EU capitalist investors know that and prefer to get  
a higher US rate of return.  This desire to live off of hard-working  
American mortgage payers was deftly exploited by smart US crooks, so  
the world and EU investors essentially got what they deserved.   
There's no easy money.

This has noting to do with the clearly stated and honestly settled  
trades where the nature of stocks was known forever and never changed  
by any trading company.

> I agree with Yoann's old-fashioned feeling: if you don't bring
> anything to society, why should you make that much money? The point
> against super-fast trading is not so much that speed is immoral in
> itself, but that such fast transactions usually seem to fall in the
> category of no-value-added.

Why should stupid pseudo-art be sold at Sothebie's for millions?  Why  
do folks spend $15,000 for a brunch in a pseudo-French New York bistro:

http://www.nytimes.com/2009/03/15/nyregion/thecity/15part.html

Who decides what's value-added for society?  This is socialism and  
we'd have none of it in the US unless required by law.  No amount of  
bureaucrats in Brussels will ever tell us what to do.

> A simple answer to this criticism is that the notion of usefulness to
> the society is very controversial. Let me comment on the answer that
> was given instead.
>
> On Mon, Mar 16, 2009 at 4:31 PM, Markus Mottl  
> <markus.mottl@gmail.com> wrote:
>> Thus, besides providing wonderful jobs for OCaml-hackers, our company
>> makes it cheaper for people to invest their hard-earned money by
>> providing extra liquidity on financial markets.
>
> All you're saying is that you believe in free trade. If my mobile
> phone company is ripping me, they can justify it because other
> companies might rip their customers even more? Notoriously, this
> argument does not always work: in France, all mobile phone companies
> practice similar insane prices and no "fair" one arises.

Because socialism is bad, and US and Asian companies are more agile  
and their services are better priced than the EU ones.

> I'd like to understand finance, maybe I'll read some serious stuff
> about it some day. But I feel that it's really too bad that such an
> important matter is inaccessible to most citizens. More importantly,
> one should not fool people (or oneself) with sloppy metaphors and
> apparently flawless reasoning when it actually boils down to
> non-trivial beliefs about the world economy.

Trading is a highly specialized endeavor.  So is cutting diamonds and  
launching rockets.  Settling trades is as important as water and waste  
management.  Municipal waste requires a degree in civil engineering to  
do right, and trading is no different.  It's the blood of the economy.

> In my opinion, a major sickness of our world is that it is ruled by
> people who claim to understand it when they don't. Politicians always
> have solutions, never doubts. Obviously, it's just a way to justify
> why they're ruling those who don't understand. We should keep our eyes
> open and don't let it happen unnoticed, nor reproduce it ourselves.

The best way for smart people in such a situation is to unite and  
act.  Jane St is not a government or political organization.  It is a  
set of folks who believe they know how the world works and prove it  
daily.  That's the most honorable way for people to act in our times.   
They use their education, technical savvy, creativity and software  
skills to build an agile system which thrives.  This is one of the  
most beautiful things possible under the constraints you lament.  As  
we say here, stop whining, roll up your sleeves, raise some seed  
capital, and go change the world!

> Despite my doubts, I enjoyed Yaron's talk and appreciate the Jane St
> involvement in OCaml.

Here we finally agree.

> I just wish it were that simple.

It's simpler than you think!  "Just do it."

Cheers,
Alexy

Re: [Caml-list] caml trading

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 241 bytes --]

I would humbly propose that this thread has now gotten deeply off-topic, and
perhaps discussion on the list should turn back to programming languages,
rather than to deep philosophical questions about the nature of the
financial markets.

y

[-- Attachment #2: Type: text/html, Size: 252 bytes --]

Re: [Caml-list] caml trading

[Markus Mottl]

2009/3/16 Yaron Minsky <yminsky@gmail.com>:
> I would humbly propose that this thread has now gotten deeply off-topic, and
> perhaps discussion on the list should turn back to programming languages,
> rather than to deep philosophical questions about the nature of the
> financial markets.

I'd also rather keep this thread off the list now and will only reply
off the list in the future if it seems reasonable.  This thread has
become way too political for my taste.

As someone else here has already suggested, I encourage everyone that
wishes to learn more about different models of economy to pick up any
reputable, introductory textbook on microeconomics.  It will teach you
that markets, by and large, work well, that they can sometimes fail,
can also fail systematically in certain areas, and that it is
extremely hard to find good solutions for the latter that do not cause
more problems than they solve.  Ideological quick fixes do not exist,
just as the optimal type sytem doesn't so no more flames please...

Regards,
Markus

-- 
Markus Mottl        http://www.ocaml.info        markus.mottl@gmail.com

Re: [Caml-list] caml trading

[Mike Lin]

[-- Attachment #1: Type: text/plain, Size: 1026 bytes --]

On Tue, Mar 17, 2009 at 9:56 AM, Markus Mottl <markus.mottl@gmail.com>wrote:

> 2009/3/16 Yaron Minsky <yminsky@gmail.com>:
> > I would humbly propose that this thread has now gotten deeply off-topic,
> and
> > perhaps discussion on the list should turn back to programming languages,
> > rather than to deep philosophical questions about the nature of the
> > financial markets.
>
> I'd also rather keep this thread off the list now and will only reply
> off the list in the future if it seems reasonable.  This thread has
> become way too political for my taste.


That's disappointing since you were quite willing to take the bait in the
first place :) But seriously, I'd suggest that your company could augment
its excellent technical outreach efforts to the
academic/engineering/free-software crowd with some education about why you
are proud of your business. It's good for recruitment! Especially
considering, however unjustified it may or may not be, how much embitterment
there is towards your field right now.

Mike

[-- Attachment #2: Type: text/html, Size: 1465 bytes --]

Re: [Caml-list] caml trading

[Xavier Leroy]

(I know I should let this thread die, but I can't resist!)

Mike Lin wrote:
> But seriously, I'd suggest that your company could augment its
> excellent technical outreach efforts to the
> academic/engineering/free-software crowd with some education about
> why you are proud of your business.

Wise words indeed.  <joke> May I also suggest some investment
counseling targeted to the working Caml programmer?  After all, it's
in Jane Street's best interest to make sure that key Caml developers
don't blow their little savings on mutual funds once presented to them
as "very safe, you can't lose with this one"... </joke>

- Xavier "but I didn't have any Madoff, or so my bank tells me" Leroy

Re: [Caml-list] caml trading

[Lawrence Austen]

[-- Attachment #1: Type: text/plain, Size: 2266 bytes --]

Hi Markus,

I agree that, empirically, market-based mechanisms tend to be the 
socially optimal solution to most real-world resource allocation 
problems.  However, this isn't really what is demonstrated in most 
introductory microeconomics textbooks, which focus on the analysis of 
systems in equilibrium. 

As is well known -- 
http://en.wikipedia.org/wiki/Theory_of_the_Second_Best -- once you're 
outside of equilibrium there is no particular reason why market-based 
mechanisms should result in social welfare improvements.  Producing a 
product that creates pollution (which is a form of market failure) is 
the classic example, making the manufacturing process more competitive 
might result in a net social loss if increased production results in 
increased pollution, but the problem is far broader than that.

I think what Jane St is doing is great and I've made my career in 
finance, but arguing that you're doing it to save the world doesn't seem 
like good chess.  At least I hope you don't actually believe that or 
you'll have to change the name to St Jane St :-)

Cheers,
Lawrence

PS Thanks for LACAML, very social welfare improving!


Markus Mottl wrote:
> 2009/3/16 Yaron Minsky <yminsky@gmail.com>:
>   
>> I would humbly propose that this thread has now gotten deeply off-topic, and
>> perhaps discussion on the list should turn back to programming languages,
>> rather than to deep philosophical questions about the nature of the
>> financial markets.
>>     
>
> I'd also rather keep this thread off the list now and will only reply
> off the list in the future if it seems reasonable.  This thread has
> become way too political for my taste.
>
> As someone else here has already suggested, I encourage everyone that
> wishes to learn more about different models of economy to pick up any
> reputable, introductory textbook on microeconomics.  It will teach you
> that markets, by and large, work well, that they can sometimes fail,
> can also fail systematically in certain areas, and that it is
> extremely hard to find good solutions for the latter that do not cause
> more problems than they solve.  Ideological quick fixes do not exist,
> just as the optimal type sytem doesn't so no more flames please...
>
> Regards,
> Markus
>
>   

[-- Attachment #2: Type: text/html, Size: 2894 bytes --]

Re: Re: [Caml-list] caml trading

[xahlee]

[-- Attachment #1: Type: text/plain, Size: 1084 bytes --]

On Mar 17, 2009 6:56am, Markus Mottl <markus.mottl@gmail.com> wrote:
> As someone else here has already suggested, I encourage everyone that
> wishes to learn more about different models of economy to pick up any
> reputable, introductory textbook on microeconomics. It will teach you
> that markets, by and large, work well, that they can sometimes fail,
> can also fail systematically in certain areas, and that it is
> extremely hard to find good solutions for the latter that do not cause
> more problems than they solve. Ideological quick fixes do not exist,
> just as the optimal type sytem doesn't so no more flames please...

well said.

For those of you ignorant of the basics, i recommend:

Basic Economics: A Citizen's Guide to the Economy
by Thomas Sowell.
http://www.amazon.com/exec/obidos/ASIN/0465002609/

I read the first edition in the dot come age circa 2000, alone with other  
books that helped me understand stocks and related financial things. I've  
read Thomas's book 2 more times in the early 2000s.

Xah
∑ http://xahlee.org/

☄

[-- Attachment #2: Type: text/html, Size: 1218 bytes --]

Re: Re: [Caml-list] caml trading

[Philippe Veber]

[-- Attachment #1: Type: text/plain, Size: 2522 bytes --]

Suggestions to have this thread die quickly:
1. be more polite, and don't treat people with different opinions like lazy
ignorants that should read *your* reference textbook before talking.
Hopefully no old-school marxist (and I know there are some on this list ;o))
has bothered us with citations from Castioriadis, Che Guevara and Bob
Marley. Please don't let this happen : reply that you understand the
relevance of the initial question and that you'd be happy to debate on an
appropriate medium.
2. be more aggressive (for instance say that Marx should have taken care of
his beard instead of writing nonsense), create a mailing list
bored-camlist-free-market-argument@yquem.inria and announce a battle royale
on it. Settle it like real men do, but elsewhere ;o).

And then, maybe we could go back to "stuff that matters", like mixins, GADT,
ocamlbuild and camlp4 documentation, native opengl GUIs, LLVM compilation,
batteries etc ... all those topics for which this list (and its members) are
so valuable !

ph.




2009/3/17 <xahlee@gmail.com>

> On Mar 17, 2009 6:56am, Markus Mottl <markus.mottl@gmail.com> wrote:
> > As someone else here has already suggested, I encourage everyone that
> > wishes to learn more about different models of economy to pick up any
> > reputable, introductory textbook on microeconomics. It will teach you
> > that markets, by and large, work well, that they can sometimes fail,
> > can also fail systematically in certain areas, and that it is
> > extremely hard to find good solutions for the latter that do not cause
> > more problems than they solve. Ideological quick fixes do not exist,
> > just as the optimal type sytem doesn't so no more flames please...
>
> well said.
>
> For those of you ignorant of the basics, i recommend:
>
> Basic Economics: A Citizen's Guide to the Economy
> by Thomas Sowell.
> http://www.amazon.com/exec/obidos/ASIN/0465002609/
>
> I read the first edition in the dot come age circa 2000, alone with other
> books that helped me understand stocks and related financial things. I've
> read Thomas's book 2 more times in the early 2000s.
>
> Xah
> ∑ http://xahlee.org/
>
> ☄
> _______________________________________________
> Caml-list mailing list. Subscription management:
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
> Archives: http://caml.inria.fr
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs
>
>

[-- Attachment #2: Type: text/html, Size: 3374 bytes --]

Re: Re: [Caml-list] caml trading

[Loup Vaillant]

2009/3/18 Philippe Veber <philippe.veber@googlemail.com>:
> Suggestions to have this thread die quickly:
> 1. be more polite, and don't treat people with different opinions like lazy
> ignorants that should read *your* reference textbook before talking.

Err, Xah didn't sound offensive to me.
If I take "For those of you ignorant of the basics", literally (wich I
think was intended), the only way to be offended is to have an
*uninformed* different opinion.  I took this phrase personally because
I *am* Ignorant of the basics.  The book he proposed may convey a
particular ideology, but  I can't assume so.  And no, I have no right
to say anything about economics, until I have a clue beyond my salary.

Now, I may have misunderstood you. Sorry If I have.

> […]
> 2. be more aggressive […]

Anyway, your main point remains.

> And then, maybe we could go back to "stuff that matters", like mixins, GADT,
> ocamlbuild and camlp4 documentation, native opengl GUIs, LLVM compilation,
> batteries etc ... all those topics for which this list (and its members) are
> so valuable !

You're right. *That* is our business.

Cheers,
Loup

Re: [Caml-list] caml trading

[Kuba Ober]

> I'm not sure how much this is off-topic... Although this thread was
> intended to be about an industrial use of OCaml, one cannot ignore
> other aspects like advertising: it is rarely possible to separate
> issues as clearly as we'd like. Anyway, I feel very much concerned
> about this debate, and I'd like to add another non-Jane-St opinion.
>
> Like many others, I've been feeling for a long time that many
> financial products and practices don't make sense. I recognize that we
> live in a complex world, and perhaps I should just learn more about
> economics. However, the recent crisis showed that this feeling is not
> entirely unfounded, even if it does not show that finance is harmful
> as a whole.

As Feynman once said: if you cannot explain it clearly to an undergrad,
you likely don't really understand it yourself.

Now the deal is that I wish the pro-trading argument was presented by  
someone
who truly understands (in the Feynman sense).

I don't know if there's any single person out there who has as much  
understanding
of trading as Feynman had of physics. There's no "Feynman Lectures on  
Trading" :(

Cheers, Kuba

Re: [Caml-list] caml trading

[Mike Lin]

[-- Attachment #1: Type: text/plain, Size: 717 bytes --]

2009/3/14 Yaron Minsky <yminsky@gmail.com>

> On Fri, Mar 13, 2009 at 8:56 PM, Yoann Padioleau <padator@wanadoo.fr>wrote:
>
>>
>> Having said that, about your company, Jane Street,
>> aren't you part of the people that put the countries in such trouble ?
>> Do you create wealth ? It seems you are just a parasite ... that
>> exploits some loopholes in the system.
>>
>
> What Jane Street does is very far from the part of finance---and the ways
> of thinking---that caused these troubles.  You can't just lump all of
> finance together as "the guys who caused the trouble"; that would be like
> condemning all engineers for
>
> http://www.youtube.com/watch?v=P0Fi1VcbpAI


It's OK, that bridge was insured by AIG :)

[-- Attachment #2: Type: text/html, Size: 1404 bytes --]

Re: [Caml-list] caml trading

[Jon Harrop]

On Friday 13 March 2009 19:52:14 Yaron Minsky wrote:
> For those who are interested, I just posted a video of a talk I gave at CMU
> several weeks ago about Jane Street's use of OCaml.  This is an updated
> version of the talk I gave at POPL last year.  Various people have
> expressed an interest in showing the talk to their undergraduate classes,
> so I thought it worth mentioning on the list.

Very interesting. I do have a couple of points:

. If you want to see shared-memory parallelism done right look at Cilk. It is 
both painless and incredibly efficient.

. I don't know how F# handles nulls but it has never caused me a problem and 
I've used many .NET libraries from F# including MDX, WPF, XNA and, of course, 
the framework itself.

-- 
Dr Jon Harrop, Flying Frog Consultancy Ltd.
http://www.ffconsultancy.com/?e

Re: [Caml-list] caml trading

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 1310 bytes --]

On Sat, Mar 14, 2009 at 3:25 AM, Jon Harrop <jon@ffconsultancy.com> wrote:

> On Friday 13 March 2009 19:52:14 Yaron Minsky wrote:
> > For those who are interested, I just posted a video of a talk I gave at
> CMU
> > several weeks ago about Jane Street's use of OCaml.  This is an updated
> > version of the talk I gave at POPL last year.  Various people have
> > expressed an interest in showing the talk to their undergraduate classes,
> > so I thought it worth mentioning on the list.
>
> Very interesting. I do have a couple of points:
>
> . If you want to see shared-memory parallelism done right look at Cilk. It
> is
> both painless and incredibly efficient.


I know about Cilk, but I'll go back and look at it again.  Thanks for the
suggestion.


> . I don't know how F# handles nulls but it has never caused me a problem
> and
> I've used many .NET libraries from F# including MDX, WPF, XNA and, of
> course,
> the framework itself.


Interesting.  I haven't done a ton of F# hacking, so I don't know the
details, but several people who work at Jane Street have stubbed their toe
on this.  It's quite possible that it mostly just doesn't come up, if you
code in the appropriate style.

And as I tried to say in the talk, F# definitely has a lot to recommend it.
It's just not a good match for us.

y

[-- Attachment #2: Type: text/html, Size: 1886 bytes --]

Re: [Caml-list] caml trading

[Jim Miller]

2009/3/13 Yaron Minsky <yminsky@gmail.com>:
> For those who are interested, I just posted a video of a talk I gave at CMU
> several weeks ago about Jane Street's use of OCaml.  This is an updated
> version of the talk I gave at POPL last year.  Various people have expressed
> an interest in showing the talk to their undergraduate classes, so I thought
> it worth mentioning on the list.
>
> Here's the link:
>
>   http://ocaml.janestreet.com/?q=node/61
>
> I can also provide the original quicktime file from which this is based on
> request.
>

Thoroughly enjoyed the talk.  You did an excellent job of describing
the environment and business case for the language.  Your discussion
of the reasons why you use OCaml, its benefits, and its drawbacks was
very balanced and makes it more credible.  Well done and thanks for
sharing.

Re: [Caml-list] caml trading

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 970 bytes --]

On Sat, Mar 14, 2009 at 3:30 PM, Jim Miller <gordon.j.miller@gmail.com>wrote:

> 2009/3/13 Yaron Minsky <yminsky@gmail.com>:
> > For those who are interested, I just posted a video of a talk I gave at
> CMU
> > several weeks ago about Jane Street's use of OCaml.  This is an updated
> > version of the talk I gave at POPL last year.  Various people have
> expressed
> > an interest in showing the talk to their undergraduate classes, so I
> thought
> > it worth mentioning on the list.
> >
> > Here's the link:
> >
> >   http://ocaml.janestreet.com/?q=node/61
> >
> > I can also provide the original quicktime file from which this is based
> on
> > request.
> >
>
> Thoroughly enjoyed the talk.  You did an excellent job of describing
> the environment and business case for the language.  Your discussion
> of the reasons why you use OCaml, its benefits, and its drawbacks was
> very balanced and makes it more credible.  Well done and thanks for
> sharing.
>

Thanks.

[-- Attachment #2: Type: text/html, Size: 1488 bytes --]

Re: [Caml-list] caml trading

[aditya siram]

[-- Attachment #1: Type: text/plain, Size: 1618 bytes --]

Thanks for this awesome video! I was unaware that the Caml compiler did not
optimize source code before generating native code. Now I am a little less
afraid to study the actual compiler!

Your point about a lack of  GUI libraries is great - however it seems as
though they are not as important as they once were. The amazing javascript
toolkits [1] that have sprung up over the last year (along with the Ocsigen
and OBrower project [2]) have filled most of that vaccum.

As long as there are libraries that provide an easy way to hook up with a
web client is there still a compelling reason to develop native GUI
toolkits?

Regards
-Aditya Siram



[1] http://www.sproutcore.com/ , http://qooxdoo.org/ ,
http://cappuccino.org/
[2] http://ocsigen.org/

2009/3/13 Yaron Minsky <yminsky@gmail.com>

> For those who are interested, I just posted a video of a talk I gave at CMU
> several weeks ago about Jane Street's use of OCaml.  This is an updated
> version of the talk I gave at POPL last year.  Various people have expressed
> an interest in showing the talk to their undergraduate classes, so I thought
> it worth mentioning on the list.
>
> Here's the link:
>
>   http://ocaml.janestreet.com/?q=node/61
>
> I can also provide the original quicktime file from which this is based on
> request.
>
> y
>
> _______________________________________________
> Caml-list mailing list. Subscription management:
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
> Archives: http://caml.inria.fr
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs
>
>

[-- Attachment #2: Type: text/html, Size: 2483 bytes --]

Re: [Caml-list] caml trading

[Jon Harrop]

On Monday 16 March 2009 17:01:15 aditya siram wrote:
> Thanks for this awesome video! I was unaware that the Caml compiler did not
> optimize source code before generating native code. Now I am a little less
> afraid to study the actual compiler!

The OCaml compilers are remarkably comprehensible. I attribute that to the 
OCaml developers at INRIA discarding implementations of ideas that turned out 
to be rubbish rather than accumulating such cruft. So there is very little 
irrelevant baggage in there compared to many other compilers.

> Your point about a lack of  GUI libraries is great - however it seems as
> though they are not as important as they once were. The amazing javascript
> toolkits [1] that have sprung up over the last year (along with the Ocsigen
> and OBrower project [2]) have filled most of that vaccum.
>
> As long as there are libraries that provide an easy way to hook up with a
> web client is there still a compelling reason to develop native GUI
> toolkits?

Performance. If you want to do OpenGL-based visualization then you at least 
need a fast route from the high-level data you are visualizing through to a 
low-level representation (e.g. tesselations) down to the C FFI to the OGL 
driver. Javascript is, of course, not up to the task. OCaml is much better 
but there is still room for improvement.

-- 
Dr Jon Harrop, Flying Frog Consultancy Ltd.
http://www.ffconsultancy.com/?e

Re: [Caml-list] caml trading

[Philip]

On Mon, 2009-03-16 at 12:01 -0500, aditya siram wrote:
> Thanks for this awesome video! I was unaware that the Caml compiler
> did not optimize source code before generating native code. Now I am a
> little less afraid to study the actual compiler!
> 
> Your point about a lack of  GUI libraries is great - however it seems
> as though they are not as important as they once were. The amazing
> javascript toolkits [1] that have sprung up over the last year (along
> with the Ocsigen and OBrower project [2]) have filled most of that
> vaccum.
> 
> As long as there are libraries that provide an easy way to hook up
> with a web client is there still a compelling reason to develop native
> GUI toolkits?
> 
> Regards
> -Aditya Siram
> 
> 

It is very easy to interface ocaml with an gui-toolkit like qt, living
in its own process. The communication is done via localhost/ http.

-Philip