Why I don't use Dylan 
Author Message
 Why I don't use Dylan

I use Dolphin Smalltalk for Win32 programming, but have been browsing
in my spare time for something better.  Dolphin has been improving a
lot over the past couple of years, and I like it more and more, so I'm
spending less and less time browsing for something better.  Dylan
seemed very interesting, and I downloaded version 1 from Harlequin
sometime in the past.  But I was disappointed at how slowly it compiled
and how big the dll's needed by the executable were.  More recently, I
downloaded version 2 from F.O. and started to give it a try, by
following the examples and/or tutorials.  I started at the beginning,
where it suggested starting with the Reversi example.  But when I went
there, it said something to the effect that I had to pay money for
libraries before I could proceed with that.  But since I'm just
browsing for something better than Smalltalk, I'm not ready to pay
money for it yet.  Especially since Smalltalk keeps improving, and it
starts to seem like I will get the best by simply staying with
Smalltalk while it continues to improve.

And I think Dylan may gradually fade away, because it might not have
enough support to ever get critical mass.  Critical mass is what makes
Java useful in spite of how bad a language it is.  It causes the
vendors to compete against each other to provide the best tools at the
lowest prices, and the biggest and most functional libraries.

It's sad, because Dylan seemed almost like the ultimate programming
language, and I had high hopes for it.  It's too bad the best
programming language designers tend to suck so badly at promoting their
products.

But even though I have lost a lot of my interest in Dylan, I still do
discuss it sometimes, and sometimes suggest to other programmers that
it might be worth trying, especially if they're willing to pay money to
try it.  One thing I would like to know is how well it does in
benchmarks against LispWorks, because that question came up in a recent
discussion.

Sent via Deja.com http://www.*-*-*.com/
Before you buy.



Sun, 15 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:

> I use Dolphin Smalltalk for Win32 programming, but have been browsing
> in my spare time for something better.  Dolphin has been improving a
> lot over the past couple of years, and I like it more and more, so I'm
> spending less and less time browsing for something better.  Dylan
> seemed very interesting, and I downloaded version 1 from Harlequin
> sometime in the past.  But I was disappointed at how slowly it compiled
> and how big the dll's needed by the executable were.  More recently, I
> downloaded version 2 from F.O. and started to give it a try, by
> following the examples and/or tutorials.  I started at the beginning,
> where it suggested starting with the Reversi example.  But when I went
> there, it said something to the effect that I had to pay money for
> libraries before I could proceed with that.  But since I'm just
> browsing for something better than Smalltalk, I'm not ready to pay
> money for it yet.  Especially since Smalltalk keeps improving, and it
> starts to seem like I will get the best by simply staying with
> Smalltalk while it continues to improve.

> And I think Dylan may gradually fade away, because it might not have
> enough support to ever get critical mass.  Critical mass is what makes
> Java useful in spite of how bad a language it is.  It causes the
> vendors to compete against each other to provide the best tools at the
> lowest prices, and the biggest and most functional libraries.

> It's sad, because Dylan seemed almost like the ultimate programming
> language, and I had high hopes for it.  It's too bad the best
> programming language designers tend to suck so badly at promoting their
> products.

> But even though I have lost a lot of my interest in Dylan, I still do
> discuss it sometimes, and sometimes suggest to other programmers that
> it might be worth trying, especially if they're willing to pay money to
> try it.  One thing I would like to know is how well it does in
> benchmarks against LispWorks, because that question came up in a recent
> discussion.

You can try all the libraries for, like, 30 days without paying a cent.

And critical mass is created by people using the product.  Even Java
started from nothing.  There's probably a good quote about that by
someone famous...



Sun, 15 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan
I use Dolphin Smalltalk for Win32 programming also. I like Smalltalk
I think it is a cool language. That said, I followed the same course as you
did.
I downloaded Harlequin Dylan 1 and when I saw Functional Objects I
downloaded the free version. I was very disappointed when I tried to
compile some of the examples, but I was still interested in seeing what
it was all about so I went back to the website and did a little more
reading.
I downloaded the Professional Combo Pack and contacte Functional
Objects for an unlock code. (You'll have some issues if you do download
and install the free version and then get one of the 30-day evals).
I installed the whole set of software and docs.

I must admit that it is a complete package. The price is reasonable.
The language seems complete and well thought out. It takes a little
getting used to because instead of providing many different ways
to get something done, Dylan gives you one process that is used to
do many different things. The libraries seem to have alot of functionality
included. Object orientated programming is done a little different but it
kind of grows on you after a while.

It won't replace Smalltalk in the near future, but it is certainly worth
a look. Free evaluations allow you to try before you buy and if you
don't like it, remove it and move on.

Give it another look.

Tim

Quote:

> I use Dolphin Smalltalk for Win32 programming, but have been browsing
> in my spare time for something better.  Dolphin has been improving a
> lot over the past couple of years, and I like it more and more, so I'm
> spending less and less time browsing for something better.  Dylan
> seemed very interesting, and I downloaded version 1 from Harlequin
> sometime in the past.  But I was disappointed at how slowly it compiled
> and how big the dll's needed by the executable were.  More recently, I
> downloaded version 2 from F.O. and started to give it a try, by
> following the examples and/or tutorials.  I started at the beginning,
> where it suggested starting with the Reversi example.  But when I went
> there, it said something to the effect that I had to pay money for
> libraries before I could proceed with that.  But since I'm just
> browsing for something better than Smalltalk, I'm not ready to pay
> money for it yet.  Especially since Smalltalk keeps improving, and it
> starts to seem like I will get the best by simply staying with
> Smalltalk while it continues to improve.

> And I think Dylan may gradually fade away, because it might not have
> enough support to ever get critical mass.  Critical mass is what makes
> Java useful in spite of how bad a language it is.  It causes the
> vendors to compete against each other to provide the best tools at the
> lowest prices, and the biggest and most functional libraries.

> It's sad, because Dylan seemed almost like the ultimate programming
> language, and I had high hopes for it.  It's too bad the best
> programming language designers tend to suck so badly at promoting their
> products.

> But even though I have lost a lot of my interest in Dylan, I still do
> discuss it sometimes, and sometimes suggest to other programmers that
> it might be worth trying, especially if they're willing to pay money to
> try it.  One thing I would like to know is how well it does in
> benchmarks against LispWorks, because that question came up in a recent
> discussion.

> Sent via Deja.com http://www.deja.com/
> Before you buy.



Sun, 15 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan
You can try every available component in Functional Developer free for
30 days.  Installing is pain-free... unlike with some other free
evaluations.  They don't even really make you register.

Functional Developer has improved a lot since the Harlequin Dylan 1.0
days.  Give it a try.

Don't forget to install the Microsoft linker which you can download from
an SDK at Microsoft's web site.  It speeds up the building process by
about twice because the GNU linker is so slow.

Anyway, the cost of Functional Developer for many will be trivial
compared to the productivity gain.

Sent via Deja.com http://www.deja.com/
Before you buy.



Mon, 16 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan
So there are 30-day free evaluations for everything from Fun-O.

I also can't fathom for the life of me why someone would simultaneously
think that Dylan might almost be the ultimate programming language,
but then won't even take the trouble to see that there is a free evaluation,
and furthermore spreads misinformation to other people about them
having to pay money even to try it.

I don't work for Functional Objects, so don't hold my sharp tongue against
them.

Quote:

>I use Dolphin Smalltalk for Win32 programming, but have been browsing
>in my spare time for something better.  Dolphin has been improving a
>lot over the past couple of years, and I like it more and more, so I'm
>spending less and less time browsing for something better.  Dylan
>seemed very interesting, and I downloaded version 1 from Harlequin
>sometime in the past.  But I was disappointed at how slowly it compiled
>and how big the dll's needed by the executable were.  More recently, I
>downloaded version 2 from F.O. and started to give it a try, by
>following the examples and/or tutorials.  I started at the beginning,
>where it suggested starting with the Reversi example.  But when I went
>there, it said something to the effect that I had to pay money for
>libraries before I could proceed with that.  But since I'm just
>browsing for something better than Smalltalk, I'm not ready to pay
>money for it yet.  Especially since Smalltalk keeps improving, and it
>starts to seem like I will get the best by simply staying with
>Smalltalk while it continues to improve.

>And I think Dylan may gradually fade away, because it might not have
>enough support to ever get critical mass.  Critical mass is what makes
>Java useful in spite of how bad a language it is.  It causes the
>vendors to compete against each other to provide the best tools at the
>lowest prices, and the biggest and most functional libraries.

>It's sad, because Dylan seemed almost like the ultimate programming
>language, and I had high hopes for it.  It's too bad the best
>programming language designers tend to suck so badly at promoting their
>products.

>But even though I have lost a lot of my interest in Dylan, I still do
>discuss it sometimes, and sometimes suggest to other programmers that
>it might be worth trying, especially if they're willing to pay money to
>try it.  One thing I would like to know is how well it does in
>benchmarks against LispWorks, because that question came up in a recent
>discussion.

>Sent via Deja.com http://www.deja.com/
>Before you buy.



Mon, 16 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan


Quote:
> a look. Free evaluations allow you to try before you buy and if you
> don't like it, remove it and move on.

> Give it another look.

The problem is the 30-day limit.  I only have a small amount of spare
time per week.  I have to wait till sometime in the future when I have
30 days of free time.

I also want to evalutate other languages, such as OCaml.  I will
probably start that soon, but have no idea how long I will continue
with it.  I have a hunch that I will probably end up still using
Smalltalk for the foreseeable future, because it keeps improving and
becoming more popular.

When ParcPlace/ObjectShare had financial problems, they sold their main
Smalltalk product, VisualWorks, for less than $1 million.  If the same
thing happens to Functional Objects, I doubt they could even get that
much, because VisualWorks was a proven product with a well-established
market.  What if some company such as Sun buys F.D. for a bargain
price, and uses it for internal research only, like they did with Self?

I've already invested a lot of time into Smalltalk, which seems like a
good investment.  But common sense tells me Dylan is not the same kind
of investment at all.  I do want to get involved with Dylan, when I
have time, but I don't think I should delude myself into thinking it
will be anything but a hobby for me.  Meanwhile Smalltalk is how I earn
my living, and I have to give it a lot more respect and attention than
it probably deserves, at the expense of other possibly more interesting
languages such as OCaml and Dylan.

Sent via Deja.com http://www.deja.com/
Before you buy.



Tue, 17 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:



> > a look. Free evaluations allow you to try before you buy and if you
> > don't like it, remove it and move on.

> > Give it another look.

> The problem is the 30-day limit.  I only have a small amount of spare
> time per week.  I have to wait till sometime in the future when I have
> 30 days of free time.

Second! Demo versions should be limited by the number of uses, not the
number of days.  I keep never seeing that I will have enough free time in
the next period to do a proper evaluation, so I never get around to
installing the downloads.

The last time I bought Dylan I eventually gave up on the macro syntax.  I
never did get the dialog to display, though I followed the example
exactly.  Presumably things have been fixed since then, but now I'm mainly
looking at other languages (and Harlequin has "spun off" it's Dylan
compiler).  Python, Ruby, Ada95 ...  If I could blend Ruby and Ada95....
Dylan remains very interesting, but FO Dylan is limited to Windows, and I
would prefer to move to Linux.  I don't know how compatible FO Dylan is to
gwydion (or what state gwydion is currently in).  Etc.

Limited time.  Lots of choices.  A lot of what I need to do involves screen
design, databases, and printing reports.  If it weren't for MS and VB, then
Access would be the obvious choice, but Basic is still a horrible language,
and MS changes the file formats incompatibly between every version.  And
the errors!  So I really want out of the entire MS system.  (Not to mention
UCITA, and what *THAT* implies, or that MS is one of its big supporters,
and what THAT implies.)

  charleshixsn.vcf
< 1K Download


Mon, 23 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:


> > The problem is the 30-day limit.  I only have a small amount of spare
> > time per week.  I have to wait till sometime in the future when I have
> > 30 days of free time.

> Second! Demo versions should be limited by the number of uses, not the
> number of days.  

It is a common practice in the software industry, though. If you don't
like it, email Fun-O and tell them how a different system would
encourage you to evaluate their products.

Quote:
> The last time I bought Dylan I eventually gave up on the macro syntax.

I can relate to that. :-) It looks so ugly and unlike the rest of Dylan
at first.
If you look through the Dylan library open sources at gwydiondylan.org,
the examples can help you make the leap much more quickly (I'd have been
lost without them). I can give you the urls if you're interested.

Quote:
> Limited time.  Lots of choices.  A lot of what I need to do involves screen
> design, databases, and printing reports.  

You can do this in Functional Developer, I think. There was a
testimonial from a database report project done in Functional Developer
on this newsgroup a few months back.

Quote:
> If it weren't for MS and VB, then Access would be the obvious choice

Functional developer has ODBC, its own object database, the ability to
call Win32 APIs, its own GUI system and lots of other features it sounds
like you are after. I don't know enough about printing or screen
painting to comment on these, though.

- Rob.

--
Rob Myers - http://www.robmyers.org/   H2G2 - http://www.h2g2.com/
MacOS wonderfulness for The Hitch Hiker's Guide to the Galaxy Game.
"Don't talk to sociologists. Social practice has no sociological
content." - Art & Language.



Mon, 23 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan


Quote:
> So there are 30-day free evaluations for everything from Fun-O.

> I also can't fathom for the life of me why someone would simultaneously
> think that Dylan might almost be the ultimate programming language,
> but then won't even take the trouble to see that there is a free evaluation,
> and furthermore spreads misinformation to other people about them
> having to pay money even to try it.
[...]

[...]
> >downloaded version 2 from F.O. and started to give it a try, by
> >following the examples and/or tutorials.  I started at the beginning,
> >where it suggested starting with the Reversi example.  But when I went
> >there, it said something to the effect that I had to pay money for
> >libraries before I could proceed with that

Okay, so tejavu missed that the libraries could be tried free. But apart
from that, I think the point is valid, because that same thing tripped me
up too. The point is that this is *the main example* at the *beginning*
of the help for the free product. If I download a free C compiler, I
don't expect to have to download a separate, time-limited trial version
of some other thing in order to try "hello, world", especially if the
help hasn't mentioned it. And yes, I know there's a "hello world" program
included, but it isn't what the Quick Start topic in the help files uses.

Another gotcha, while we're on the subject, is the way the example
programs are paused so that you can read their output. If the dialog box
that tells you they are paused was to say "and you can un-pause them by
clicking the "start" button again", that would be an improvement in
usability out of all proportion to the effort involved.

Peter.
--

quin.co.uk       Barrington, Cambs CB2 5RG, England.



Tue, 24 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan


Quote:
> I must admit that it is a complete package. The price is reasonable.

But it's hard for most programmers to take it seriously.  Functional
Objects does not inspire confidence that the programmer would be making
a wise investment of time and money.  The company is run by former
Harlequin employees who give the vague impression they might be doing
it as a part time hobby.  There have been so many such products in the
past which have gradually faded into obscurity, when their proprietors
found their time occupied by other pursuits and put the product on the
back burner, till they gradually reached the point where they could no
longer even do any more maintenance on it.

One thing Functional Objects could do to improve their credibility a
lot would be to put their source code in escrow with SourceForge to
become free open source if Functional Objects ever drops the ball.  A
clear guarantee, that the product will definitely have a future no
matter what, is exactly what people need to be able to take it
seriously enough to invest the time and money needed to get started
with it.

What is the development situation at Functional Objects?  I get the
vague impression some contract programmers working for other companies
might be doing the Functional Developer work between contracts and/or
in their spare time.  Or does Functional Objects have venture capital
to pay the salaries of full time programmers?

Sent via Deja.com http://www.deja.com/
Before you buy.



Thu, 26 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:

> Subject: Re: Why I don't use Dylan


>> I must admit that it is a complete package. The price is reasonable.

> But it's hard for most programmers to take it seriously.  Functional
> Objects does not inspire confidence that the programmer would be making
> a wise investment of time and money.  The company is run by former
> Harlequin employees who give the vague impression they might be doing
> it as a part time hobby.

Having communicated with Functional Objects personnel in both public and
private email, and having read through the business plans and white papers
on their web site, I can assure you that I have not got this impression.
They seem to be a hard working group of very capable people dedicated to
moving Functional Developer forward.
Ex-Harlequin personnel not with Functional Objects do answer questions on
this list as far as I can tell. I'd take that as a vote of confidence rather
than anything else.

Quote:
> There have been so many such products in the
> past which have gradually faded into obscurity, when their proprietors
> found their time occupied by other pursuits and put the product on the
> back burner, till they gradually reached the point where they could no
> longer even do any more maintenance on it.

I can't see evidence of that happening here. Fun-O have already finalized
and released the 2.0 version of Functional Developer and have announced
plans for a UNIX version.

Quote:
> One thing Functional Objects could do to improve their credibility a
> lot would be to put their source code in escrow with SourceForge to
> become free open source if Functional Objects ever drops the ball.  A
> clear guarantee, that the product will definitely have a future no
> matter what, is exactly what people need to be able to take it
> seriously enough to invest the time and money needed to get started
> with it.

SourceForge has an air of "amateur project" about it that I'm not sure would
help. The public sources are freely available with the Functional Developer
downloads, so if a neutron bomb does get dropped near their offices, you'll
still have them. Some are even on the Gwydion site and CVS server.

Quote:
> What is the development situation at Functional Objects?  I get the
> vague impression some contract programmers working for other companies
> might be doing the Functional Developer work between contracts and/or
> in their spare time.  Or does Functional Objects have venture capital
> to pay the salaries of full time programmers?

This is company confidential stuff for a start-up, so I can't imagine them
letting us know.

I was worried when Harlequin got split up, but Functional Objects have come
through with flying colours. My only small concern is that their website
could do with making more boring-corporate, but it has grown on me :-)

- Rob.



Fri, 27 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:

> I was worried when Harlequin got split up, but Functional Objects have come
> through with flying colours. My only small concern is that their website
> could do with making more boring-corporate, but it has grown on me :-)

I'm neither a customer nor am I known to be very sympathetic
to Dylan in general. But I really like how the product is being
sold in pieces by "Functional Objects, Inc." over their web site.
This really makes sense to me in the Internet age.
Make libraries affordable and let customers buy libraries
based on their real demand - this is really a good thing.

--
Rainer Joswig, BU Partner,
ISION Internet AG, Steinh?ft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999



Fri, 27 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:
> Having communicated with Functional Objects personnel in both public and
> private email, and having read through the business plans and white papers
> on their web site, I can assure you that I have not got this impression.
> They seem to be a hard working group of very capable people dedicated to
> moving Functional Developer forward.

  I hope you don't mind me jumping in here.

  For some reason that I don't understand, computer programming is based on
a popularity contest.  The faintest whiff of suspicion on the future of any
development system is enough to send it into the trash by the vast majority
of the user it's targetted at. Sometimes this is deserved, consider
ProGraph's almost-comical series of "we're back!" followed by "sorry, dead
again".

  However, consider Dylan at Apple.  Seems to me that the vast majority of
Mac users are still better off on Dylan than they are on, say, C++ for the
"average" project, and yet it's usage is effectively zero.  There's a number
of good reasons for this, but an equal number of bad ones that are likely
more often the case.

  My point is that I'm not sure Ex-Harlequin personnel mean anything one way
or the other.  They do not have the market clout to have a serious name, and
as a result I'd suspect they'll shortly have "day jobs" like Pictorius does
today, if not already.  And that just makes the matter worse.

  Too bad really, if only BillJ wanted a new Dylan rather than a new C++...

Quote:
> SourceForge has an air of "amateur project" about it that I'm not sure
would
> help. The public sources are freely available with the Functional
Developer
> downloads, so if a neutron bomb does get dropped near their offices,
you'll
> still have them. Some are even on the Gwydion site and CVS server.

  Methinks you doth protest too much.  Let me assure you that escrow on
SourceForce would help credibility far more than it would hurt! This is an
excellent suggestion, you shouldn't write it off so easily.

Quote:
> This is company confidential stuff for a start-up, so I can't imagine them
> letting us know.

  Sounds to me like you're all too willing to be their appologist.  The
issue here is credibility, I'd say that the company would be happy to tell
you if you asked in order to support just that.

Maury



Fri, 27 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:

> For some reason that I don't understand, computer programming
> is based on a popularity contest.  The faintest whiff of suspicion
> on the future of any development system is enough to send it into
> the trash by the vast majority of the user it's targetted at.
> Sometimes this is deserved, consider ProGraph's almost-comical
> series of "we're back!" followed by "sorry,  dead again".

Hard to avoid, I think.  It takes a large investment of time to learn a
new programming language, and once you build up your own library of
useful stuff you really don't want to find that the company that
produced your programming language is dead and that a new OS version is
incompatable with what you're using, or that show-stopping bug will
*never* be fixed.

It's amazing how long THINK Pascal kept on keeping on years and years
after it was no longer developed, sold or supported, but something in
about MacOS 8.1 or so finally killed it and it's dead, dead, dead today,
even thought it's *still* probably the best ever Pascal or C learning
environment.

Which is why I'm so e{*filter*}d about Gwydion Dylan.  Right now it's less
cosmetically polished than either the Apple Dylan or Functional
Developer products, but it's also very bug-free and produces good code
and produces better code with every month that goes by.  And because I
have the source code and it's a simple command-line application I can
make it work on any system that I want to use in the future, and can fix
any bugs myself.

Quote:
> However, consider Dylan at Apple.  Seems to me that the vast
> majority of Mac users are still better off on Dylan than they
> are on, say, C++ for the "average" project,

I certainly think they would be.

Quote:
> and yet it's usage is effectively zero.  There's a  number
> of good reasons for this, but an equal number of bad ones that are likely
> more often the case.

Where would we be now if Apple has included the source code on the Dylan
Technology Release CD?

-- Bruce



Sat, 28 Dec 2002 03:00:00 GMT  
 Why I don't use Dylan

Quote:



> > SourceForge has an air of "amateur project" about it that I'm not sure
> > would help.

>   Methinks you doth protest too much.  Let me assure you that escrow on
> SourceForce would help credibility far more than it would hurt! This is an
> excellent suggestion, you shouldn't write it off so easily.

OK, I'll take a look at SourceForge again. I'm possibly stuck in the
early web-era snobbery of everything having to have its own domain name.
What happens if SourceForge's business model is broken and they go down? :-)

Quote:

> > This is company confidential stuff for a start-up, so I can't imagine them
> > letting us know.

>   Sounds to me like you're all too willing to be their appologist.  

Eep! I wouldn't want to be an apologist for anything! The background to
my comments here is that I've been involved in a couple of startups now
and you tend to keep your cards prety close to your chest unless there's
a very good reason not to. Responding to concerns about credibility can
harm credibility.:-)

- Rob.

--
Rob Myers - http://www.robmyers.org/   H2G2 - http://www.h2g2.com/
MacOS wonderfulness for The Hitch Hiker's Guide to the Galaxy Game.
"Don't talk to sociologists. Social practice has no sociological
content." - Art & Language.



Sat, 28 Dec 2002 03:00:00 GMT  
 
 [ 20 post ]  Go to page: [1] [2]

 Relevant Pages 

1. Dylan Tip #101: don't use keyword init-values with class allocated slots

2. Why patterns don't suck

3. why awk don't display the results?

4. Why I don't Like HTML Help...

5. Animated Gifs Some work some don't why

6. Why don't Memo Entry Fields wrap?????

7. I don't understand why my haskell program works

8. Why collected languages don't need a free operation

9. Why don't large companies use Ada?

10. Why don't my textures map correctly?

11. Why don't Paragraph reply?

 

 
Powered by phpBB® Forum Software