Seaside (Smalltalk web development framework)

[Robert Holmes]

Check out R. It's got good interpolation and plotting routines, it has a
good package for time series analysis, it's free and has a huge user base.

http://www.r-project.org/

Robert

On 9/14/06, Phil Henshaw <sy at synapse9.com> wrote:

>
> Owen types:
> > I'd like to get back to the task at hand -- evaluating rapid
> > prototyping environments.
> ... clip
>
> > BTW: I'm starting to think answer is that rapid prototyping has
> > splintered into pieces:
> >    Excel: great for fast exploration of data
> >    R: great for fast statistical evaluation
> >    Gnuplot: great for fast and simple plotting of data
> >    NetLogo: sorta smalltalk for simulation
> >    Processing: Killer 2D/3D visualizations
> >    Sh/Bash: Lightning fast trivial hacks for unix folks
> >    Python: Scripting++ and more
> >    JavaScript: Great access to browsers for both users & AJAX
> >    .. and so on.
> > So thus there really is an explosion of application specific
> > environments rather than the old smalltalk which was pretty universal.
>
> Probably off topic, but has any of them solved my problem?   I can only
> limp along with my live systems data analysis using home made lisp
> routines on a graphics platform because of the standard grid model of
> data that seems to be used elsewhere?    All my data streams come with
> different, and often irregular, time periods and a vector graphics
> program is the only one I can find that allows direct interrelationships
> between sequences with different periods.  I'm also doing analysis of
> various ways to reconstruct the points in-between the points.   Excel
> and others have no place to put them.   Is there an affordable and
> flexible stat package that treats all scales as continuities?
>
> Phil
>
>
>
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> lectures, archives, unsubscribe, maps at http://www.friam.org
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/friam_redfish.com/attachments/20060915/bfe66fdb/attachment.html

[Phil Henshaw-2]

marvelous!   but now I've got all sorts of stuff to transfer to a system
I don't understand...  know any tutors in NY?
 
 

Phil Henshaw                       ????.?? ? `?.????
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
680 Ft. Washington Ave
NY NY 10040                      
tel: [phone]                
e-mail: pfh at synapse9.com          
explorations: www.synapse9.com <http://www.synapse9.com/>    

-----Original Message-----
From: [hidden email] [mailto:[hidden email]] On
Behalf Of Robert Holmes
Sent: Friday, September 15, 2006 1:31 AM
To: The Friday Morning Applied Complexity Coffee Group
Subject: Re: [FRIAM] Seaside (Smalltalk web development framework)


Check out R. It's got good interpolation and plotting routines, it has a
good package for time series analysis, it's free and has a huge user
base.

http://www.r-project.org/

Robert


On 9/14/06, Phil Henshaw <sy at synapse9.com> wrote:

Owen types:
> I'd like to get back to the task at hand -- evaluating rapid
> prototyping environments.
... clip

> BTW: I'm starting to think answer is that rapid prototyping has
> splintered into pieces:
>    Excel: great for fast exploration of data
>    R: great for fast statistical evaluation
>    Gnuplot: great for fast and simple plotting of data
>    NetLogo: sorta smalltalk for simulation
>    Processing: Killer 2D/3D visualizations
>    Sh/Bash: Lightning fast trivial hacks for unix folks
>    Python: Scripting++ and more
>    JavaScript: Great access to browsers for both users & AJAX
>    .. and so on.
> So thus there really is an explosion of application specific
> environments rather than the old smalltalk which was pretty universal.

Probably off topic, but has any of them solved my problem?   I can only
limp along with my live systems data analysis using home made lisp
routines on a graphics platform because of the standard grid model of
data that seems to be used elsewhere?    All my data streams come with
different, and often irregular, time periods and a vector graphics
program is the only one I can find that allows direct interrelationships
between sequences with different periods.  I'm also doing analysis of
various ways to reconstruct the points in-between the points.   Excel
and others have no place to put them.   Is there an affordable and
flexible stat package that treats all scales as continuities?

Phil



============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org



-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/friam_redfish.com/attachments/20060916/21253e6c/attachment.html

[Marcus G. Daniels-3]

Phil Henshaw wrote:

> I can only
> limp along with my live systems data analysis using home made lisp
> routines on a graphics platform because of the standard grid model of
> data that seems to be used elsewhere?
[..]
> marvelous!   but now I've got all sorts of stuff to transfer to a
> system I don't understand...  know any tutors in NY?
If you like Lisp, you'll find R natural.    Witness lambda expressions
and lexical scope:

test <- function (l) {
  sum(sapply (l, function (x) { x + l }))
}

 > test (c(1,2,3,4))
[1] 80

It's also has consed lists (called pair lists), hash tables (called
environments), etc.

There are tutorials of various sorts on the R website.  

R is a great package.   It has a plug-in interface for native code, and
hundreds of contributed packages.
Folks that are mainly interested in presentation graphics or
visualization might not find it is what they want as that's not it's
greatest strength.  It's greatest strength is that it is good
full-fledged dynamially typed programming language with about every
statistical test ever conceived.  But visualization is improving.  
There's a very usable OpenGL package that works fine, for example.  Also
some will gripe that it is slow.  It's true there is no compiler for R,
yet, but given the native code plug-in interface and the fact that most
operations are vectorized with intensive numerics done in R's native
code libraries, I think that's a pretty bogus complaint.

Marcus

[Phil Henshaw-2]

Thanks, found some help on the R help forum, though now I'm buried in
resources and haven't gotten to square 1 with understanding the syntax.
What I really want is to communicate my own purposes, how to watch the
masterful artwork of nature evolve by cross connecting real events and
real dynamics...   What the skillful use of the right tools shows is
that *every* event is emergent as an original development, not accident,
of its time!


Phil Henshaw                       ????.?? ? `?.????
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
680 Ft. Washington Ave
NY NY 10040                      
tel: [phone]                
e-mail: pfh at synapse9.com          
explorations: www.synapse9.com    

> -----Original Message-----
> From: friam-bounces at redfish.com
> [mailto:friam-bounces at redfish.com] On Behalf Of Marcus G. Daniels
> Sent: Saturday, September 16, 2006 11:22 AM
> To: The Friday Morning Applied Complexity Coffee Group
> Subject: Re: [FRIAM] Seaside (Smalltalk web development framework)
>
>
> Phil Henshaw wrote:
>
> > I can only
> > limp along with my live systems data analysis using home made lisp
> > routines on a graphics platform because of the standard
> grid model of
> > data that seems to be used elsewhere?
> [..]
> > marvelous!   but now I've got all sorts of stuff to transfer to a
> > system I don't understand...  know any tutors in NY?
> If you like Lisp, you'll find R natural.    Witness lambda
> expressions
> and lexical scope:
>
> test <- function (l) {
>   sum(sapply (l, function (x) { x + l }))
> }
>
>  > test (c(1,2,3,4))
> [1] 80
>
> It's also has consed lists (called pair lists), hash tables (called
> environments), etc.
>
> There are tutorials of various sorts on the R website.  
>
> R is a great package.   It has a plug-in interface for native
> code, and
> hundreds of contributed packages.
> Folks that are mainly interested in presentation graphics or
> visualization might not find it is what they want as that's not it's
> greatest strength.  It's greatest strength is that it is good
> full-fledged dynamially typed programming language with about every
> statistical test ever conceived.  But visualization is improving.  
> There's a very usable OpenGL package that works fine, for
> example.  Also
> some will gripe that it is slow.  It's true there is no
> compiler for R,
> yet, but given the native code plug-in interface and the fact
> that most
> operations are vectorized with intensive numerics done in R's native
> code libraries, I think that's a pretty bogus complaint.
>
> Marcus
>
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> lectures, archives, unsubscribe, maps at http://www.friam.org
>
>

[Giles Bowkett]

> Alright, I give up - why do fun languages like Python or more fun
> languages like Squeak get passed over in the market compared to
> rather annoying languages like Java? How come they haven't been as
> competitive as say Linux as a server OS platform? Why is C++ vs.
> Java still our fate in 2006? Is there no God? Have we been bad?

OK -- this is a huge topic, but, for what it's worth, here's my theory.

Short answer: marketing.

Long answer:

Most language choices in the programming world come down to
programmers selling stuff to management. Selling is a people skill and
programmers often lack it. Every so often a programmer manages to use
some obscure language they prefer, and the results are impressive --
Yahoo! Stores was originally written in Lisp, there were virtually
zero Ruby Web programmers before one Ruby Web programmer created Ruby
on Rails, Seaside is written in Smalltalk and has productivity that
makes even Rails look sluggish, etc. -- but generally that's not
because somebody discovered a language's heretofore-overlooked power,
it's because somebody who was already aware of that power managed to
convince some business person that the power actually existed, or
learned enough business skills that these decisions became their own
to make.

More frequently, what happens is that the language choice actually
comes down to a management decision, and most people who manage
programmers are in the business of making decisions they do not
understand. So they follow the dictates of fashion. This is why the
industry is so prone to buzzwords and trends -- the people making the
decisions do not actually have any other metric at their disposal
besides popularity.

Anyway, apologies Owen, but I have to disagree with the "languages are
hard" thing. Languages are easy, at least, for me personally, I've
never found syntax or even idioms difficult to get the hang of. As
arrogant as I might sound, I don't think I'm actually unique in this.
The sheer number of languages, and the fact that most programmers know
several, demonstrates that they aren't **that** difficult. But the
points about interoperability and deployability seem much more valid.

I think programmers who enjoy obscure languages tend to be better
programmers, but the causality works the other way around: enjoying
obscure languages doesn't make you a better programmer, being a good
programmer gives you an increased appreciation for obscure languages.
That being said, there's a lot of interest in learning new languages
these days, and I think it's gotten out of hand.

A lot of the big surprise success stories in the last decade or so
have come out of left field in terms of the languages they were
written in, but this is almost tautological -- they wouldn't have been
surprises if they had been written in the languages everyone else was
using at the time. It does seem, though, that if you study languages
strictly to learn more languages, you'll find a lot of totally
careerist programmers doing the same thing. About a year and a half
ago I was going to Python users' group meetings at Google, and the
slickness, desperation, and indifference which marks a room full of
careerist programmers was so thick in the air that I could hardly
breathe. Maybe a year or so before that, it had become common
knowledge that Google preferred to see Python on resumes when hiring
for Java positions, and so a lot of people had started learning Python
purely to get better Java jobs, in much the same way that a guy who
has no interest in actually managing a project might still seek to be
officially named the project's manager.

I hope I don't sound like a bitter Silicon Valley refugee... ;-)

--
Giles Bowkett
http://www.gilesgoatboy.org

[Robert Holmes]

Phil,

Of the various tutorials out there, I'd recommend "simpleR - using R for
introductory statistics" by Verzani.
www.cran.r-project.org/doc/contrib/*Verzani*-*SimpleR*.pdf

Robert

On 9/16/06, Phil Henshaw <sy at synapse9.com> wrote:

>
> Thanks, found some help on the R help forum, though now I'm buried in
> resources and haven't gotten to square 1 with understanding the syntax.
> What I really want is to communicate my own purposes, how to watch the
> masterful artwork of nature evolve by cross connecting real events and
> real dynamics...   What the skillful use of the right tools shows is
> that *every* event is emergent as an original development, not accident,
> of its time!
>
>
> Phil Henshaw                       ????.?? ? `?.????
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/friam_redfish.com/attachments/20060917/60575aac/attachment.html