Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search

[REBOL] Re: starting to be really late!?

From: ed:brittlestar at: 8-Jan-2004 1:23

To compete with Flash -- or any well-established commercial product or open-source language-- let's face it, it probably is too late. RT has limited resources and bandwidth to stake a claim against Macromedia, Adobe and numerous other public or privately funded companies. It's true that the "X-Internet" label was a marketing buzzword, but REBOL never appeared (to my eyes anyway) to be mainstream market-driven or customer-focused. Good languages and technologies aren't often the product of markets or concensus, so it probably makes sense. If popularity or money were truly central to the mission, consider the following choices: * platform-independence, instead of support for Win32/COM/CORBA (or Unix system calls) * support for common open network protocols/formats, no proprietary ones * free-form syntax, instead of C style syntax * language name choice is slightly awkward (COBOL, SNOBOL) instead of a hipster name * lightweight cloning instead of rigid OOP * ideas from Forth and Logo, instead of more popular C, VB or Perl * designed for "programming in the small", not for general programming or big-iron middleware * code interpreted instead of compiled * execution speed is respectable, but not a hallmark feature (such as with K, euphoria, erlang) * single execution thread * no modules/libraries, foreign functions, continuations, spawning of sub-processes (& other exotic MIT hacker profile) * interactive console and your favorite editor, instead of a 20 MB IDE * tiny disk footprint, instead of 5, 10, 20 MB or more * minimal install, instead of a 20 step Wizard-driven system infestation * think about your code instead of relying on rich, integrated debuggers * minimal GUI layer, not mature (missing html controls & formatting, no data-grid control, etc.) * no wysiwyg or GUI/forms design tools * /command db access not robust (i.e., count occurrences of #? in the SQL statement and bind a value to it) * BNF style grammars instead of a Regex engine * limited XML support, not mature * Latin characters instead of UTF-8 or Unicode * closed source instead of open source * REBOL is not available/mature on the number 2 desktop platform, Apple (2-3% of the market) * dialects instead of.... well, I've yet to see common problems that language dialects solve (but I hope to) Looking at the above list (and I appreciate many of the items as much as I dislike other ones), REBOL doesn't look like a commercial, market-focused language. Normally I would expect the data to form a profile of one or more marketable customer segments, i.e., scripter, hobbyist, corp. programmer, sysadmin, hacker, etc. The profile I arrive at from the list above is one of... well, a rebel. So rather than targetting a market opportunity, Carl created a language-- a platform, really-- that embodies his principles in a personal programming language (wasn't that what his original manifesto said?). I've been using REBOL off-and-on for over 6 years, and that's longer than I can say for most commercial products I've worked with. I believe that there are marketable & revolutionary products in REBOL, or a REBOL-like language. Given the limited resources of RT, the small user-base of REBOL, and the _heavily_ commoditized state of programming languages, I'd love to see an updated vision or roadmap for REBOL (forget delivery dates, naturally). Before I close this overly long message, I'll posit a parting thought. Tim Bray and other bloggers have asserted that there are 4 main kinds of apps: * information retrieval & consumption * database interaction * content creation * games/media players and that history implies that the web has the first 2 solidly covered. Will REBOL compete with the web, or will it mature enough to allow you and me to create robust apps in the latter 2 categories? Hey, despite the frustration, confusion and the passage of long periods of time, this is still mostly fun, right? Good luck with /View 1.3 all. Ed