[REBOL] Re: /View as a Product
From: gjones05:mail:orion at: 17-Feb-2001 9:40
There have been many interesting comments made in this thread. I've had
a lot of food for thought, and over the last day or so I find myself,
like a cow, regurgitating my cud and chewing on it again.
I've been convinced for years that communication is a very fragile
process. It is very easy to inadvertantly step on each other's toes
(much in the same way that is is very easy for the neophyte to overwrite
array space in C, or to inadvertantly reassign "system" words in Rebol,
and I've had my share of both). I suspect that most of us agree on two
points: 1) REBOL is a very exciting language that still has some room to
grow, and 2) Holy Wars don't tend to be very productive. I don't see
that any Holy War has been started. All I see is enthusiasm that is
pushing to make REBOL succeed.
I spent the 80's and the first half of the 90's in the MS-DOS/Windows
world as a computer enthusiast (and not necessarily a Microsoft
enthusiast). I cummulatively spent thousands of dollars buying
exciting
solutions, and then spent innumerable hours learning to
utilize those "solutions". As with all things in life, I had varying
levels of success, with limitations being more dependent on my available
time or abilities. The one thing that held constant in retrospect, is
that change in the industry required a continuous investment of time and
money (both of which become tough to justify at the advanced hobbyist
stage). I came to really resent the money part, for whatever reason
(maybe I'm practical or maybe I'm just stingy). From my perspective, it
wasn't so much that I ever minded an initial financial investment, it
was that the speed of change requiring frequent re-investments, and
usually at higher and higher amounts (unlike most of the rest of the
software industry, or, for that matter, the tech industry in general).
I could readily see and appreciate that the developers needed to pay
bills, but I felt that there was an increasingly diminishing return for
my additional investments of time and money. As a favor to myself, I
stopped buying prepackaged software development tools, and resigned
myself to the fact that my "advanced hobbyist" days were probably over.
In the latter 90's, easy access to the Internet made it easy to see that
there was a whole world beyond MS-DOS/Windows. I was thoroughly
intrigued to find that there were very powerful solutions available
beyond BASIC/Visual BASIC, TurboPASCAL/Delphi, and C/C++/Visual C++, and
I was amazed that they were "free." The investment of time remained to
ramp-up to even modest levels of productivity, and I seemed to be unable
to grasp a productive use of Tk (and its various incarnations. I
figured that I had become "intellectually spoiled" by the GUI IDE's of
days past.).
I stumbled across REBOL a couple of years ago when I was doing a brief
survey of languages with references on Yahoo. I was intrigued by the
simplicity of doing the simple things (the website did a great job of
highlighting these simplicities). I didn't have much time to spare, and
when I worked on my first real program, I quickly could see that the
paradigm shift for solving more complex would require more investment in
time than I had. I left it alone, but kept tabs on it every few months.
There's nothing like being laid off to free up time, so I have recently
revisited REBOL, now fresh with the /View beta. Perhaps the Tcl
paradigm eased my transition into REBOL, but, as I have said elsewhere,
I've been amazed at how quickly I have become productive in both the
utility aspect of the language and the GUI. For the first time in
years, I felt a gush of enthusiasm. But the sting of my past lessons in
has begun to temper my enthusiasm. Some of the discussions of late have
underscored my need to maintain perspective. The recent demise of
Scriptics has punctuated the need to maintain some objectivity.
(As a brief side bar, Scriptics was begun by Ousterhout, the founder of
Tcl. His goal was to foster the continued development of Tcl in open
source mode, while paying the bills by selling a more sophisticated
debugger and program wrapper, and by developing and promoting enterprise
level middleware to address the new era of commerce on the Internet. Of
course, this is a familiar theme that is not unique to Tcl/Scriptics nor
REBOL/RT. Recently practically the whole staff, including the founder,
was basically bought out to move their daytime energies to a different
environment. While it was pointedly denied that they were not able to
move their paying product to the point of sustainability, that they
merely decided to devote their energies elsewhere, I can't help but
wonder if they weren't seeing the writing on the wall given their
situation and capitalization.)
At this juncture, let me state clearly that I am **not** implying that I
think RT is headed down the same road.
After this lengthy introduction, I'll finally get to the reason I am
writing. People who program for a living will always need to invest
significant time and possibly money into the skills and tools necessary
to succeed. I personally believe that what we consider to be intuitive
is probably subconsciously learned. If a language matches the intuitive
sense of the person, then the language will be more acceptable to learn
provided the the person is initially exposed to its existence. (By way
of example, I offer that a highly engineered language like Eiffel that
is linguistically restricted will seem very "intuitively" different to a
person who finds Ada to be intuitive than a person who finds Lisp to be
intuitive, and yet no one would likely doubt that Ada and Lisp were
successful in their own relms.) Although the number of "professional"
programmers is growing, so to does the landscape of available languages
and paradigms. I don't *know* this to be true, but I suspect that the
percentage of market growth of REBOL amongst professional programmers
may always be limited unless there is either a clear and compelling
advantage that is brought to the attention of the audience and/or there
is a virtual monopolistically-limited growth of alternatives (I suspect
that the growth of Visual BASIC, for example, benefited from both).
Rapid application development, cross platform support, and true
glue-ability
of pre-existing components have been some of the features
that have aided the acceptance of the more widely known script languages
to date. Having a "killer application" can also help to bring a
language to the attention of others. (Being first, or at least early,
in the *scripted* cgi processor helped Perl, and, in my opinion, Zope
may have done this for Python, which had previously remainded fairly
obscure.)
I suspect that the true, untapped, growth market is the "programming"
language for non-programmers. Making the computer more accessible to
the techno-phobe is what has driven PC growth ("technophiles" and
gamers
adopted computers much earlier). This concept has been
featured more prominently in the on-line information for REBOL in the
past, but frankly one has to dig a lot more to find "those pages" these
days. Given that bills need to be paid, I can readily appreciate "why"
the enterprise-focused, revenue-generating aspects are prominently
displayed. But I am really disappointed to see that the message that
REBOL might be a language-of-choice for non-programmers given its more
natural
linguistic appearance. And this is the juncture where my
inability to understand the developing product line becomes a problem: I
would imagine that these enterprise-level programs will allow for
non-programmers to take more appropriate control of doing the little
things, in the same way that the macro languages helped the
non-programmers take more control of their word-processors, etc. And
here is where I think the real growth potential is for the language at
large: quicker and more seemless development for the enterprise by the
developer with easier access to the good stuff by the non-programmer. A
marriage made in "heaven"; possibly the killer application for REBOL.
VBScript does NOT do this for Microsoft's enterprise-level software in
my opinion, and is hardly approachable by the mere mortal, and there is
*no way* that Java does this, nor the Java-like C# (sorry Redmond).
Microsoft's enterprise-level software will continue to only be
approachable by way of "real" development by "real" programmers for the
foreseeable future. That is not necessarily true for REBOL, in my eyes
as a non-professional "hobbyist programmer."
I truly don't "know" what is right for REBOL. I like it; it speaks to
me in a way few languages have. I don't understand the currently
proposed direction for the various evolving products, and I think Petr
spoke for me when he said that this prospect makes people nervous. I
also don't really know what is the best marketing approach is for REBOL,
but I do know that in considering basic principles of human existence,
that as the product is being marketed currently, REBOL will need to
somehow pay the bills for RT, or both may die on the vine. This
prospect would understandably make developers and RT staffers both
nervous. I believe that this is the reason that a few toes were
inadvertantly stepped on of late. It is understandable.
Here are my two suggestions. First, I believe that RT in its leadership
role likely "owes" its list-members/language-followers a little more
idea of where it is heading. In the medical profession I learned long
ago that honest communication goes a long way toward improving
understanding. Being evasive or being silent only breds mistrust,
misunderstanding, and ultimately disloyalty.
Second, I believe it can really help to occasionally take a huge step
back and try to see the situation from a different perspective. I
sometimes do this by using the Internet as my vehicle by doing a sort of
stream of consciousness websurfing (nothing metaphysical really being
implied). This morning I did this and ultimately stumbled across an
interview with Larry Wall (Perl founder) from last year. I believe most
would agree that he is articulate, humorous and insightful (far more
than I am and in all three categories). Without further commentary I
offer the following link as additional food for thought on these issues
presented in this thread:
http://www2.linuxjournal.com/cgi-bin/frames.pl/lj-issues/issue61/3394.ht
ml
Have a nice weekend, all.
--Scott