Path: news.cs.au.dk!not-for-mail From: Lars Balker Rasmussen Newsgroups: comp.lang.beta Subject: Re: Beta compiler for win32 Date: 29 May 2000 17:43:46 +0200 Organization: Mjolner Informatics ApS Lines: 146 Message-ID: <0fya4tpdjh.fsf@garm.mjolner.dk> References: <20000513221646.3288.qmail@noatun.mjolner.dk> NNTP-Posting-Host: garm.mjolner.dk Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: xinwen.cs.au.dk 959615027 16847976 255.255.255.255 (29 May 2000 15:43:47 GMT) X-Complaints-To: news@cs.au.dk NNTP-Posting-Date: 29 May 2000 15:43:47 GMT X-Newsreader: Gnus v5.5/Emacs 20.4 Xref: news.cs.au.dk comp.lang.beta:12360 Wow, that was confusing, Erik. Be careful with that copy-paste! :-) Here's what Erik probably intended to send: Erik Ernst writes: > >>>>> "Sascha" == Sascha Kimmel writes: > > Here I go, responding to a pretty old message - I guess that's me.. > > >> From: eernst@cs.auc.dk [mailto:eernst@cs.auc.dk] > >> > >> >>>>> "Sascha" == Sascha Kimmel writes: > Sascha> [..] the most important aspect of BETA is the platform > Sascha> [..] and [..] GUI independence. > >> > >> [..] it is a seriously cool language [..] few but powerful > >> abstraction mechanisms [..] > > Sascha> But the problem is that many ppl like to use languages > Sascha> which they've ever been working with? Why not those > Sascha> "shopping cart" languages like C(++)? It's been extend > Sascha> again and again, whereas BETA simplifies everything by > Sascha> abstract concepts. But for many users it's too difficult > Sascha> to understand the simplicity of the BETA structure > Sascha> therefore not using it... > > I believe that people have lots of very legitimate reasons to > choose, e.g., C++ (availability of programmers/tools/books, and > perhaps an OO assembler is _actually_ just the right tool for one or > two projects); but then again, they probably also have lots of less > enlightened reasons ("I will _never_ learn to write '(#' and '->'!", > "Who needs general block structure and true closures anyway?", and > "Virtual classes are hard to understand and even less useful!") > > I think the challenge is to spot the "bad" reasons whenever somebody > expresses them, and try to find out what they would really think about > life, death, and everything if certain misconceptions were just > slightly revised.. :-) > > >> However, it would be very nice if Sun/ > > Sascha> Sun? The million-dollar-marketing company that used its > Sascha> potential to bring Java (formerly known as Oak) to the > Sascha> world. > > Exactly. Many things are worse than Java, and programmers at large > seem to _insist_ on having syntax like > > while(*q++=*p++) { i+=(int)&j; ..} > > I'm actually pleased that the C-crowd is moving forward in history, it > does make another (incomparable) kind of difference if a few million > people take one step forward, compared with a few thousand people > taking a thousand steps forward. We need both! > > Still, it would be nice if more people were willing to try even more > "unusual" things. One of the promising options, as far as I can see, > is the collaborative, open source based development of alternatives to > the established behemoths. Because of the vast resources that might > potentially be put into such projects, it is actually possible to get > things started that would never be realistic if somebody had tried to > make money on it by hiring a lot of programmers. Then again, I'm not > sure such a topic as language design is the perfect example of such a > project. It is more likely to be tools and libraries, and perhaps > language design _adjustments_. Somehow, the holistic point of view > seems to indespensable in certain areas. Available source code for > compilers might also help, because a new compiler might reuse the > back-end (and perhaps much of the "middle-end") from an existing > compiler. VROOOMMM, and there's a compiler for your new language! > > Sascha> However Mjolner does not have such potential, but > Sascha> does this mean that BETA is worse than Java? I don't > Sascha> think so! > > I don't think so, either, but it does mean that Java has a level of > library/tool/book/course/community/conference/.. support which makes > Java a "better language than it is", whereas BETA has to lure people > based on a more purely language design oriented evaluation. Of > course, if good language design enables elegant libraries then the two > areas will interact, but in some respects it is still hard to > compensate for the lack of a billion dollar budget. > > >> IBM/ > > Sascha> Big Blue? Maybe - but do they even KNOW BETA? ;) > > I think they do. At international conferences such as ECOOP and > OOPSLA, people in general know something about BETA, even though they > tend to say "It's definitely interesting, but at some point it got so > weird that I gave up ... now, where did you put that CDROM?" I'm sure > there are numerous people in various parts of IBM with > more-than-buzzword knowledge of BETA. > > >> (or-even-)Microsoft would put > > Sascha> Visual BETA Enterprise Edition ;D > > "Where do you want to type '(#' and '->' today?" ;-) > > >> a few thousand man-years into the further development of tools and > >> libraries, too. > > Sascha> It's simply that problem that either no one is listening > Sascha> or that no one is interested. It's a fact. > > Oh, sometimes a single person _is_ listening. As long as "most > people" just keep out their antennas to ensure that they follow the > crowd as closely as possible (i.e., forever), it just does not make > sense to try to insist on capturing the world all at once.. > > >> And if a lot more people would use it.. :-) > > Sascha> IMO BETA should go mainstream. > > What would BETA be, then? BETA? > > In a way, it might be the same thing as I was saying, i.e., it would > be nice if some big player would show some concrete interest in the > development of and support for BETA programming (oh, yeah, and > gbeta, too :-) > > Sascha> As long as BETA is simply an "elité" language it won't be > Sascha> used much. In today's times you need such features as > Sascha> Internet (I don't mean only simple sockets!!), Multimedia > Sascha> and much more - it's not only programmers at some > Sascha> universities that want to implement their algorithms - > Sascha> this way BETA will not be used by the "usual" programmer. > > It's probably very important to work in the direction of support for > various standard component architectures, because the binary system > composition model makes it more convenient to use "special" languages > together with lots of shrink-wrapped, standard functionality. This > will also allow people to make gradual, graceful shifts from one > set of technologies to another one. Not that this is a new idea, > of course :) > > > regards, > > -- > Erik Ernst eernst@cs.auc.dk > Department of Computer Science, University of Aalborg, Denmark -- Lars Balker Rasmussen, Software Engineer, Mjolner Informatics ApS lbr@mjolner.dk