rebol crystals: [lucid rocks] => was {Re: Re: Browser gripe}
[1/4] from: jason:cunliffe:verizon at: 17-Mar-2002 14:33
Hi Steve, Gregg & list
Thanks.. Yes I think everyone here on the list feel that rebol 'rocks' in
multiple ways..
> Rebol takes the fragile-ness out of Internet programming, and provides
> safe solid and simple ways for developers and users to incorporate the
> Internet into their multimedia applications.
Q1: How+why does it take the "the fragile-ness out of Internet programming"
?
To answer that we must answer what makes it fragile to begin with..
> I have written some bots which do middleware applications which simply
> blow people away (And that "rocks" me!) so my wireless pieces are
Q2: What blows them away: concept, function, performance, speed of
development ??
> brought together at the server and it is so simple to do.
Q3: So it is ease of development and simplicity of execution. => easy to
make and run??
> And I have 6 Rebol clients connected to my embedded server which have
> run for 24 hours without a hitch, doing lots of messaging. I would hate
> to have tried to do that with plain old C/C++ or even plain old Java,
> and the scripts are _tiny_.
I want to know more :-)
Q4: Embedded server = IOS, Rugby, or your own ?
Q5: "24hours" you mean one whole day or "all day, every day.." If so, for
how long doing what? How much traffic, how many connections [sessions]?
> Hope that is "concrete" (lucid) enough for now! ;-)
NO, not yet. I really do want to hear more ..
Q6: What about REBOL is different architecturally, or tends towards
better/different Internet architecture. What distinguishes REBOL when
implementing a similar design written in Python or PHP JAVA or xyz...?
I know Rebol is fun, fast, small, cross-platform and includes Internet
protocols. But that is only a matter of degree, convenience and style.
Q7: Does Rebol make you all think differently about problems? How does your
approach differ when you use Rebol?
Q8: What do you think its greatest strengths and virtues are?
Q9: What do you think makes it unique?
Q10: What do think it is best for and why?
Q11: For any of the above questions, does IOS induce a different answer?
regards
./Jason
[2/4] from: steve:shireman:semaxwireless at: 18-Mar-2002 7:16
Jason Cunliffe wrote:
>Hi Steve, Gregg & list
>Thanks.. Yes I think everyone here on the list feel that rebol 'rocks' in
<<quoted lines omitted: 6>>
>?
>To answer that we must answer what makes it fragile to begin with..
Maybe one question at a time?
Software Engineering has long been an oxymoron. I suspect it was a term
made to glorify the practice of hacking,but give it better sound bytes
and a justification for higher salaries. The "Software Problem" was
highly publicized in the 1980's and 1990's, and Object-Oriented
Programming was touted to be the solution, but has grossly failed to
deliver.
In fact, I would be as bold to say that the "Software Problem" has
gotten worse, and Software is now even more fragile.
I note people being afraid to upgrade (let's call it upgrade-a-phobia,
the fear of the fragility of fresh newly written code) to the latest
versions, whether Netscape, Internet Exploder, or even GVIM,
The "Software Problem" was approached for a while with Application
Frameworks, trying to make it more difficult for programmers to lock-up
the system with errant code. These were kinda cool as they were
introduced, but their APIs or SDKs became mired in complexity, and
again, the cure was worse than the disease.
Part of the problem was that when C was invented (and I have worked with
C since Bell Labs introduced it in the 1970's) was that it was supposed
to be portable, but if you read the first chapter of the C books, you
will see (pun intended) that the OS was blatently ignored, and as with
the traditional over-specialization approach we inherited from the
Industrial Revolution
the language and the OS were carved apart,
destroying the fun part of programming, which early Basic rom
interpreters had (a holisting thing) where the language and OS were
married together, and made programming and experimentation easy. And
can't leave out Smalltalk.
Those early Basic and Smalltalk interpreters were pretty robust, and if
you crashed something, well, at least you could work your way out, and
the OS tried to give you some information about the crash.
Then Java tried again to marry the OS and language, but picked a
too-static language, but it was a good try, and brought some light to
the technical Software darkness of the late 1990's.
So Rebol is a holistic language plus environment which lets you crash
your code safely, even safer than JVM. And I find this less fragile for
programming. And the framework in it is much better than I could do
myself, (in fact, better than anyone in the world could do themselves,
except Carl Sassenrath, who knows how to pack those bytes in there with
maximum power) so when I write a Reblet, it is less fragile, because it
has an architecture beneath it which does the dirty work.
Not sure if that answers your Q1, but thought I would give it a stab.
Beats working...;-)
Steve Shireman
The Computer Revolution _Has_ Started...
[3/4] from: sunandadh:aol at: 18-Mar-2002 9:24
Steve:
> Software Engineering has long been an oxymoron.
Like country music!?
> I suspect it was a term
> made to glorify the practice of hacking,but give it better sound bytes
> and a justification for higher salaries. The "Software Problem" was
> highly publicized in the 1980's and 1990's, and Object-Oriented
> Programming was touted to be the solution, but has grossly failed to
> deliver.
You can go back even further -- the term was invented (as far as I know) for
a NATO conference in 1968 -- a bit before my working life started -- but they
had all the same problems we do and much the same snake oil. Just one quote
from the conference:
In aiming at too many objectives the higher-level languages have, perhaps,
proved to be useless to the layman, too complex for the novice and too
restricted for the expert. I maintain that high-level programming languages
have, to this extent, failed.
(d=E2=80=99Agapeyeff)
> In fact, I would be as bold to say that the "Software Problem" has
> gotten worse, and Software is now even more fragile.
Yes indeed. Environments are far too complicated and change too fast. If it
wasn't for the fact that Microsoft have brainwashed an entire generation into
lowered expectations ("the new version of MS-xxx only falls over five times=20a
day when running on an otherwise empty machine") most developers would have=20a
tarnished reputation.
The traditional approach to solving the software engineering problem has been
to create new languages. It's not worked to date but -- hey I wouldn't be a
developer if I wasn't hopelessly optimistic -- maybe Rebol will be the One.
Sunanda.
[4/4] from: pwoodward:cncdsl at: 18-Mar-2002 10:16
Jason -
> Q6: What about REBOL is different architecturally, or tends towards
> better/different Internet architecture. What distinguishes REBOL when
> implementing a similar design written in Python or PHP JAVA or xyz...?
>
> I know Rebol is fun, fast, small, cross-platform and includes Internet
> protocols. But that is only a matter of degree, convenience and style.
True - but that's one heck of a difference. Java really raised the bar a
few years ago by including (what at the time seemed like) a very rich set of
classes to build applications on top of. Sure, you could use other class
heirarchies (MFC anyone? Lotus Notes?) - but by including and assuring some
basic level of functionality in the SDK, they helped make it so Developers
could concentrate on the software that needed to be built, instead of doing
yet another memory manager, and linked list implementation.
REBOL is taking those steps as well. By making it trivial to access the
Internet from within an application, one can start to build applications
that synergistically take advantage of the Internet. In part it's an
alternative to the "Web Services" stuff you keep hearing about. REBOL (with
add-ons like Rugby - or IOS) makes it easy to build applications that
leverage remote services. Whether it's fetching stock quotes, or the
current weather. IOS or Rugby can be used to build relatively powerful
Intranet
applications for organizations.
So the matters of degree, convenience, and style really do matter. They are
the barriers to developer adoption of features.
> Q7: Does Rebol make you all think differently about problems? How does
your
> approach differ when you use Rebol?
Yes and no. It can really depend on what kind of problem you're solving.
Much like Python - you can write quick one-off scripts - with little thought
to their object orientatedness, or any real planning. On the otherhand, it
is possible to build very reusable bits and pieces. My approach hasn't
really changed when using REBOL. I have the tendency to approach things
proceduraly when working with it though. Except when writing /View scripts,
I don't find myself doing much event-driven or remotely object-oriented
things.
> Q8: What do you think its greatest strengths and virtues are?
REBOL, like Java brought a lot of the joy back into programming for me.
Working in C/C++ (and Pascal before that) was a real trial. Even simple
things required a lot of work, and setup code. With REBOL - it's easier to
dive in, and accomplish something. Instead of writing two pages of setup
code to open a video screen, and setup the display raster. Plus, when I
want to automate internet based activities, it's really easy.
> Q9: What do you think makes it unique?
It is what it is. It has similarities to things like Lisp or Scheme - but
it is it's own language.
> Q10: What do think it is best for and why?
It's best for the newer breed of applications that are on their way. Ones
which are going to leverage the interconnectedness of systems. I can
imagine a "dash board" style application that pulls data from all sorts of
sources, aggregating and comingling it to present a coherent, single source
interface to that information. That way you're not opening your browser,
and going to 5 or 10 different sites/pages each with their own navigation
system.
> Q11: For any of the above questions, does IOS induce a different answer?
Not really, other than it raises the bar again on what features and services
you can rely on when developing an application. Sharing data securely with
people becomes pretty trivial.
- Porter Woodward
Notes
- Quoted lines have been omitted from some messages.
View the message alone to see the lines that have been omitted