[REBOL] /View As A Product
From: robbo1mark:aol at: 19-Feb-2001 8:38
The more I read over Holger & Jeff's recent comments
regarding some of my post to this list, the more angry
and annoyed Iam becoming.
Somehow Iam being painted as the BAD GUY here, as if
Iam highly unethical, unscrupulous and down right wrong!
My opinions might be wrong for some people, I can accept that, but that is surely what
debate and constructive criticism is all about.
I have never once criticised REBOL Technologies or any
person who works there. I may disagree with their strategic directions for REBOL but
I've never once said they are wrong or urged them to make REBOL open source.
I fully respect their rights to pursue whatever strategy they feel is best for them and
their company.
Iam not the ONLY person who would prefer that REBOL source code was openly available
and freely distributable, I wish this was the case but I would never say REBOL Technologies
Inc. should or shouldn't do this or that.
I have never criticised REBOL products on their technical merits, I do feel that /Command
is not good value for money and that thesuggested pricing levels for /Express may be
too much for some people and discourage rather than encourage some people from
using REBOL, but that is a commercial decision for the peopleat REBOL, I can only make
my choice in the market place and decidewhether or not to buy these products. I have
never and would never try to tell Carl or anybody else how to run their business.
However people on this list DO have some concerns and frustrationsregards some of REBOL's
capabilities or future product directions.
let's look at three which have been discussed here recently.
1. /Apache
2. REBOL-CGI
3. /Shell features in /Core
/Apache was a product that was in development and was under BETAtesting to selected developers.
However it has recently been statedby REBOL Technologies Inc. that although /Apache remains
as a "Strategic" future product for REBOL it is not under active development just now.
So people who would like to do server side
scripting with APACHE server and REBOL now have to wait indefinitely for this to be developed
or pay something like $2500 US Dollars for an /Express Server licensing fee.
This decision is almost certainly financially oriented by REBOL Technologies Inc. need
to get revenues, which is perfectly fine, you are a business entity after all.
But what about the people who need or would like to do REBOL & Apache,YOU are not addressing
their needs, that is rightfully your decision.Is it unethical to suggest an alternative
strategy, that rather than waiting for REBOL to do something someday, that these people
might
take matters into their own hands, and help themselves by considering an open source
REBOL which they can adapt to suit the needs of server side scripting and integration
with APACHE webserver.
Next up REBOL-CGI, various people have made comments to this list about REBOL's effectiveness
and speed in doing CGI. The speed issue is affected by REBOL loading and decompressing
it's internal source for each instance of CGI. I think it was [chris--starforge--co--uk] who
recently commented that this is on the margins of what is acceptable performance for
CGI operations. This was in respect to not having a /Core product
and only having /View. Now this is probably unlikely, but in any event REBOL still suffers
in this respect in that /Core is a very good but still a general purpose interpreter.
Correct me if Iam wrong but I don't think REBOL is optimised in any way for CGI.
REBOL has a deficiency here in that a one-size-fits-all-needs interpreter cannot be stripped
down to suit the needs of CGI.
Python, Perl, Ruby & TCL etc. in particular have this advantage because they are open
source. A minimal interpreter can be produced specifically for CGI processing. This mapping
of the solution to the
problem is not possible because REBOL is only available in binary executable format.
I can't remember anyone from REBOL ever saying that a CGI specific interpreter is on
the top of the priorities list, so again what do these people do if they want to use
REBOL for CGI but with improved
performance?
Also REBOL/Core cannot be easily be used as an embedded interpreter in applications although
it is potentially ideal for this purpose.
TCL, Perl & Python all have capabilities to embed C programs in scripts and also themslves
be embedded within other C applications & programs. REBOL doesn't do this yet, and also
will it be possible to do this in REBOL for free?
Which brings us to /Shell features being available in /Core.Petr Krenzelok has argued
the case for this for nearly two years as well as a clear roadmap for componentization
in REBOL. As far as I can see his words have fallen of deaf ears.
What if you want /Command features with /View or /Shell availablein /Core, this has never
been clearly answered.
/Command to me doesn't strike me as good value for money for it'sextra feature set, but
that is a personal opinion.
The /Shell command 'CALL which passes a message string! to the operating system shell
seems to be not much more that a REBOL implementation of the ANSI C EXECL () function.
Sure /Command has /Database, /Library & Encapsulation functionality as well but if you
only want /Shell then $250 dollars is a high price to pay.
With an open source REBOL you could add this functionality yourself.
And that is the whole crux of my posts. OSCAR: :REBOL is not vaporware, I have never
claimed it can do this or that. Right now it cannot do anything as we are only at the
reverse engineering, design & specification phase.
However OPEN SOURCE provides a mechanism for people to address these areas of "missing"
functionality in REBOL. If for whateverreason you choose not to provide or prioritise
these "needs" people have then you what is unscrupulous and untactful tosuggest people
help COMPLEMENT the existing REBOL offerings by developing an open source REBOL in these
directions.
OSCAR by definition can only ever have whatever functionality people are willing to put
the time and effort into producing.
However because it is open source, it can in theory be pushed in any direction people
want to take it.
I realise the commercial pressures REBOL Technologies Inc. operate within and you literally
do not have the resources to DO & prioritise everything.
You have chosen your strategic space and direction for REBOL.Is it really so wrong for
me to want to help these people provide complementary improvements to REBOL rather than
simply asking them to have faith, patience & please wait.
These are real needs that real people have NOW!
What is wrong with OSCAR trying to fill this space if you won't or can't?
Iam only trying to improve REBOL as a technology, does this make me a bad guy?
Mark Dickson