OCaml for Industry

OCaml for Industry

[Jon Harrop]

Whilst considering future projects, it occurred to me that a management report 
along the lines of "OCaml for Industry" might be a viable product. Such a 
book could explain the utility of OCaml in overall terms (for managers and 
investors) as well as detailing the benefits, costs and various ways of 
adopting OCaml in industrial projects.

I'd like to hear from anyone who thinks their company might buy such a 
management report (for around 800UKP) and anyone who can provide information 
about case studies of OCaml having been used in industry.

-- 
Dr Jon D Harrop, Flying Frog Consultancy Ltd.
Objective CAML for Scientists
http://www.ffconsultancy.com/products/ocaml_for_scientists

Re: [Caml-list] OCaml for Industry

[Nicolas Cannasse]

>
> Whilst considering future projects, it occurred to me that a management
report
> along the lines of "OCaml for Industry" might be a viable product. Such a
> book could explain the utility of OCaml in overall terms (for managers and
> investors) as well as detailing the benefits, costs and various ways of
> adopting OCaml in industrial projects.
>
> I'd like to hear from anyone who thinks their company might buy such a
> management report (for around 800UKP) and anyone who can provide
information
> about case studies of OCaml having been used in industry.
>

That's bad news that nobody answered to this interesting topic within one
week.
As for buying a report, it might be nice to freely provide some results and
trends and then sell the "full report" if it's what you want to do.

My company is using OCaml is a lot of our projects, mainly for development
tools (compilers, level generators, resources builders... ). It's perfectly
suitable in the case where you need to manipulate a lot of data structures
and apply algorithms to them.

Nicolas

Re: [Caml-list] OCaml for Industry

[Jon Harrop]

On Monday 08 August 2005 18:14, Nicolas Cannasse wrote:
> That's bad news that nobody answered to this interesting topic within one
> week.

Yes. Back to the drawing board. :-)

> As for buying a report, it might be nice to freely provide some results and
> trends and then sell the "full report" if it's what you want to do.

Rather than writing a whole report (seeing as there's no interest in it), 
perhaps it would be better to have a news letter that people can subscribe 
to? That would be cheaper for customers and would require less initial 
investment on our part.

> My company is using OCaml is a lot of our projects, mainly for development
> tools (compilers, level generators, resources builders... ). It's perfectly
> suitable in the case where you need to manipulate a lot of data structures
> and apply algorithms to them.

Yes, that is exactly what we are finding. There is also a market for 
OCaml-generated C code, IMHO.

-- 
Dr Jon D Harrop, Flying Frog Consultancy Ltd.
Objective CAML for Scientists
http://www.ffconsultancy.com/products/ocaml_for_scientists

Re: [Caml-list] OCaml for Industry

[skaller]

[-- Attachment #1: Type: text/plain, Size: 975 bytes --]

On Mon, 2005-08-08 at 18:20 +0100, Jon Harrop wrote:
> On Monday 08 August 2005 18:14, Nicolas Cannasse wrote:
> > That's bad news that nobody answered to this interesting topic within one
> > week.
> 
> Yes. Back to the drawing board. :-)
> 
> > As for buying a report, it might be nice to freely provide some results and
> > trends and then sell the "full report" if it's what you want to do.
> 
> Rather than writing a whole report (seeing as there's no interest in it), 


That is a silly conclusion IMHO. Did you really expect existing
Ocaml experts to buy a report on use of Ocaml in industry targetted
at MANAGERS??

You're preaching to the converted here: your report is targeted
at people who are NOT subscribed to this list :)

> Yes, that is exactly what we are finding. There is also a market for 
> OCaml-generated C code, IMHO.

And C++ .. which is what Felix does .. :)

-- 
John Skaller <skaller at users dot sourceforge dot net>


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [Caml-list] OCaml for Industry

[Jon Harrop]

On Monday 08 August 2005 18:58, skaller wrote:
> > Rather than writing a whole report (seeing as there's no interest in it),
>
> That is a silly conclusion IMHO. Did you really expect existing
> Ocaml experts to buy a report on use of Ocaml in industry targetted
> at MANAGERS??

My theory was that programmers who appreciate better languages might want some 
evidence that they can present to their managers in order to help justify the 
use of non-mainstream languages. I was curious to see if anyone here was in 
such a position.

I get the impression that most people here (like me) write OCaml code without 
benefit of a manager. ;-)

> You're preaching to the converted here: your report is targeted
> at people who are NOT subscribed to this list :)
>
> > Yes, that is exactly what we are finding. There is also a market for
> > OCaml-generated C code, IMHO.
>
> And C++ .. which is what Felix does .. :)

Yes, and C++. :-)

-- 
Dr Jon D Harrop, Flying Frog Consultancy Ltd.
Objective CAML for Scientists
http://www.ffconsultancy.com/products/ocaml_for_scientists

Re: [Caml-list] OCaml for Industry

[Chris Campbell]

On 08/08/05, Jon Harrop <jon@ffconsultancy.com> wrote:
> On Monday 08 August 2005 18:58, skaller wrote:
> > > Rather than writing a whole report (seeing as there's no interest in it),
> >
> > That is a silly conclusion IMHO. Did you really expect existing
> > Ocaml experts to buy a report on use of Ocaml in industry targetted
> > at MANAGERS??
> 
> My theory was that programmers who appreciate better languages might want some
> evidence that they can present to their managers in order to help justify the
> use of non-mainstream languages. I was curious to see if anyone here was in
> such a position.

No one is going to pay £800 to buy a report containing evidence Ocaml
works in industry.  You may make a damn good case, but it ain't gonna
happen.

If you don't have a few million to spend on advertising, some free
t-shirts and big signs with words like "disruptive", "lower tco", and
all that jazz you're coming through the back door with word of mouth. 
That means you write applications in it, applications which are used!

This is one thing people fail to understand.  Ordinary users won't
care if your software is written in OCaml, Brainf**k or Java.  The
application is primary to them.  Developers otoh will care about the
underlying language.  To get their attention though you need to do
more than write some code.  You need to write code that gets used. 
That's the hard bit though.  What will people use?

> I get the impression that most people here (like me) write OCaml code without
> benefit of a manager. ;-)

Yep. :)

Re: [Caml-list] OCaml for Industry

[Erik de Castro Lopo]

Jon Harrop wrote:

> I get the impression that most people here (like me) write OCaml code without 
> benefit of a manager. ;-)

Well I tried to get Ocaml in the door where I work. I was tasked
with writing what was basically a compiler; programming language 
like syntax on the input, construct internal representation, 
manipulate internal representation, write out binary data which 
is basically a state transition table for a program running in an 
embedded device.

When this started I had already been programming in Ocaml for
about 9 months and felt really comfortable with it. Ocaml is
also ideal for compilers.

I suggested it but management said no and I didn't push it.
Ocaml is in their eyes a rather obsure language. I then suggested
Java (with ANTLR) but no, then C++, no. In the end it was written
in standard C with flex/bison. Every time I had to traverses a
damned linked list in C I though of how easy and foolproof this 
would have been in Ocaml.

So yes, the Ocaml code I write at home I do so without a manager
(not even my wife is allowed to tell me what programming language 
to use :-)).

Erik

-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo  nospam@mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
Good advice for everyone : stay away from churches, mosques and
synagouges.

Re: [Caml-list] OCaml for Industry

[skaller]

[-- Attachment #1: Type: text/plain, Size: 1079 bytes --]

On Tue, 2005-08-09 at 09:32 +1000, Erik de Castro Lopo wrote:

> I suggested it but management said no and I didn't push it.
> Ocaml is in their eyes a rather obsure language. I then suggested
> Java (with ANTLR) but no, then C++, no. In the end it was written
> in standard C with flex/bison. Every time I had to traverses a
> damned linked list in C I though of how easy and foolproof this 
> would have been in Ocaml.

The question is .. why did you ask management?
This should be your decision not theirs. They're paying you because
YOU are the expert in software development.

I used Ocaml to develop a translator for a telco:
I didn't ask .. I *told* them how I was doing it.
I myself raised the issue of maintainability, and even
suggested a teacher be paid to train other programmers.

I made it quite clear that in the available time-frame
with the available resources .. there was no possible
alternative. Haskell could have done it .. but I
don't know Haskell. C++? Java? NO WAY!

-- 
John Skaller <skaller at users dot sourceforge dot net>


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [Caml-list] OCaml for Industry

[Erik de Castro Lopo]

skaller wrote:

> The question is .. why did you ask management?

I didn't really ask. It came up in conversation and I suggested
Ocaml which was immediately shot down.

> This should be your decision not theirs. They're paying you because
> YOU are the expert in software development.

I agree, but he pays my salary and if wants to make those decisions
thats his call.

Erik
-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo  nospam@mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
"Or here's an even simpler indicator of how much C++ sucks: Print out
the C++ Public Review Document. Have someone hold it about three feet
above your head and then drop it. Thus you will be enlightened."
 -- Thant Tessman

Re: [Caml-list] OCaml for Industry

[skaller]

[-- Attachment #1: Type: text/plain, Size: 1789 bytes --]

On Tue, 2005-08-09 at 13:18 +1000, Erik de Castro Lopo wrote:
> skaller wrote:
> 
> > The question is .. why did you ask management?
> 
> I didn't really ask. It came up in conversation and I suggested
> Ocaml which was immediately shot down.
> 
> > This should be your decision not theirs. They're paying you because
> > YOU are the expert in software development.
> 
> I agree, but he pays my salary and if wants to make those decisions
> thats his call.

Well for me that would be an ethical problem leading to
my resignation. 

I mean, if someone says "Write a function
to calculate this in C" its disappointing that they
want to waste their money and some of my time --
I'd be looking for a better job, but its their money.

But if they say "We need to solve this problem" and then
they say "And you have to use these tools" .. I'm out
the door because they're simultaneously asking you
to take responsibility -- and then denying the freedom
of choice that is required to actually be responsible.

I've worked with technically aware bosses, but generally
its a discussion leading to a reasonable division of
responsibility -- eg I can say "Ocaml has these advantages
and disadvantages" and the boss can say "Well, this could
be pretty serious for us" and it is his job to weight
technical factors in a business context. 

But it is mine, exclusively, to make the final technical decision 
about how to achieve the outcome required. I have hardly ever
written C/C++ code by hand in the last 2 decades: invariably
I write a code generator to help meet requirements effectively.
Hehe .. some people got suspicious in one job at the perfectly
formatted code i was producing  .. :)

-- 
John Skaller <skaller at users dot sourceforge dot net>


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [Caml-list] OCaml for Industry

[Nicolas Cannasse]

> Well for me that would be an ethical problem leading to
> my resignation.

Well your thinking is correct as long as you're the only programmer
involved. If you work with teams, unless you're the project manager you
can't really force other people learning a new language even if it's better
suitable. There's important human factors.

BTW when I think that writing applications is a good way to get people
involved. I got some of interest about OCaml when I presented MTASC at OSCON
last week, and PUGS (Perl 6 compiler in Haskell) is also bringing people to
learn the language.

Nicolas

Re: [Caml-list] OCaml for Industry

[Yaron Minsky]

[-- Attachment #1: Type: text/plain, Size: 961 bytes --]

On 8/8/05, Nicolas Cannasse <warplayer@free.fr> wrote:

> That's bad news that nobody answered to this interesting topic within one
> week.
> As for buying a report, it might be nice to freely provide some results 
> and
> trends and then sell the "full report" if it's what you want to do.


I'm not sure it's such bad news. The company I work is making pretty 
succesful use of OCaml at the moment, and there are a few others out there 
doing the same. I just didn't see much interest in the kind of document Jon 
described. We're already past the part where we had to convince people to 
take a gamble on developing in OCaml. So far, it's gone pretty well...

Yaron 

My company is using OCaml is a lot of our projects, mainly for development
> tools (compilers, level generators, resources builders... ). It's 
> perfectly
> suitable in the case where you need to manipulate a lot of data structures
> and apply algorithms to them.
>

[-- Attachment #2: Type: text/html, Size: 1425 bytes --]

Re: [Caml-list] OCaml for Industry

[Jeremy O'Donoghue]

On 08/08/05, Nicolas Cannasse <warplayer@free.fr> wrote:
> >
> > Whilst considering future projects, it occurred to me that a management
> report
> > along the lines of "OCaml for Industry" might be a viable product. Such a
> > book could explain the utility of OCaml in overall terms (for managers and
> > investors) as well as detailing the benefits, costs and various ways of
> > adopting OCaml in industrial projects.
> >
> > I'd like to hear from anyone who thinks their company might buy such a
> > management report (for around 800UKP) and anyone who can provide
> information
> > about case studies of OCaml having been used in industry.
> >
> 
> That's bad news that nobody answered to this interesting topic within one
> week.
> As for buying a report, it might be nice to freely provide some results and
> trends and then sell the "full report" if it's what you want to do.

One of the problems with the idea of 'OCaml for Industry' is that
Industry can mean many things. In my case, 'industry' means the
development of complex real-time embedded software, generally
targetted at mobile telecomms.

With my manager's hat on, I don't want to know that OCaml is good for
(say) commercial games design. I need to make a case for it in my
industry (sadly this is hard in my case - no ARM target). I probably
can't get the cash to pay for a report, to assist in decision making.

However, many managers do have the ability to recommend the use of
consultants to train and mentor their teams, once the decision to do a
pilot has been made. This is often surprisingly easy, but it usually
takes an enthusiastic evangelist to get things moving.

This person probably knows 80% of the issues and pitfalls already. The
other 20% - well, they probably can't get funding to pay for a report
to help in this area, and that's what the pilot is there to find out.
In my opinion you'd be better off giving the report away (probably via
mailing list) with the expectation that if you make a good case you'll
see the return via consultancy, training and support contracts.

> My company is using OCaml is a lot of our projects, mainly for development
> tools (compilers, level generators, resources builders... ). It's perfectly
> suitable in the case where you need to manipulate a lot of data structures
> and apply algorithms to them.

Much the same for me. The advantage, in this field, is that the writer
of the tool generally 'owns' it forever, and the company doesn't much
care how the tool was written as long as it works. Same for utilities
which generate C/C++.

Moving slightly OT, if you want to penetrate industry, what is needed
is good, clear documentation, tutorials, good quality training,
support. OCaml lacks many of these. Serious indistrial adoptors will
pay for them. My employer spends many thousands of pounds per software
engineer per year on training, and pretty vast amounts on support as
well. We already play well with some leading-edge Open Source teams
(although we don't make a fuss about whom), but it's generally been in
the areas of training and support that the real money is spent.

Jeremy

> Nicolas
> 
> _______________________________________________
> 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] OCaml for Industry

[Eric Stokes]

What you might do is try to get your report published in an industry  
magazine aimed at managment.

On Aug 1, 2005, at 10:26 AM, Jon Harrop wrote:


>
> Whilst considering future projects, it occurred to me that a  
> management report
> along the lines of "OCaml for Industry" might be a viable product.  
> Such a
> book could explain the utility of OCaml in overall terms (for  
> managers and
> investors) as well as detailing the benefits, costs and various  
> ways of
> adopting OCaml in industrial projects.
>
> I'd like to hear from anyone who thinks their company might buy such a
> management report (for around 800UKP) and anyone who can provide  
> information
> about case studies of OCaml having been used in industry.
>
> -- 
> Dr Jon D Harrop, Flying Frog Consultancy Ltd.
> Objective CAML for Scientists
> http://www.ffconsultancy.com/products/ocaml_for_scientists
>
> _______________________________________________
> 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
>
>