if oberon is so good, why isn't it more popular
Author Message
if oberon is so good, why isn't it more popular

Quote:
> so, if its so good, why isn't there any commercial interest?  is
> anyone doing commercial applications development with oberon.  why
> didn't steve jobs select oberon for the next machine instead of
> mach?  why is it so invisible?

I fear that this is entirely my fault.  I chose to learn the mc6809
instruction set rather than the 8088, and the IBM PC was introduced
thereafter.  I chose to learn Modula-2 rather than C, and Unix
became distributed to academic sites thereafter.  I chose to learn
and implement the UCSD P-system, and IBM quit bundling it with the
PC, and went straight for MS-DOS.

I chose to learn Oberon because it was small, nifty, and user data
could go from machine to machine without change, making a 15 Year
WORM archive attractive.  WORMs are out, CDs are in.

If you know of any really good technology that would benefit
mankind, perhaps it's best if I am not told about it.
--
6/27/94 Seasonal hyperbar in place.  Clear blue skies.  Highs near
100 (40).  No more beets.

Mon, 16 Dec 1996 01:43:25 GMT
if oberon is so good, why isn't it more popular

Quote:

>> so, if its so good, why isn't there any commercial interest?  is
>> anyone doing commercial applications development with oberon.  why
>> didn't steve jobs select oberon for the next machine instead of
>> mach?  why is it so invisible?

>I chose to learn Oberon because it was small, nifty, and user data
>could go from machine to machine without change, making a 15 Year
>WORM archive attractive.  WORMs are out, CDs are in.

>If you know of any really good technology that would benefit
>mankind, perhaps it's best if I am not told about it.

Quick! Run out now with $400 and get Windows NT (don't waste anytime, or your$400 might literally turn into $300). Learn it. Love it. Acccept it. Prosetylize it! Quote: >-- >6/27/94 Seasonal hyperbar in place. Clear blue skies. Highs near >100 (40). No more beets. Ding dong the beets are dead! Taylor "Buy Tomorrow, They'll Be Cheaper Still" Hutt Pickled beets, pickled beeks. No matter what you call 'em -- I won't eat 'em. Mon, 16 Dec 1996 11:26:29 GMT if oberon is so good, why isn't it more popular ! i read a rather old book in the bookstore on oberon and i thought it was ! really cool. It seems like this could be a good, small soultion to the ! unix, next, windows, etc controversy. ! Glad you immediately saw some of the merits of Oberon. Most in this group are of the same mind as you. ! so, if its so good, why isn't there any commercial interest? is anyone ! doing commercial applications development with oberon. why didn't steve ! jobs select oberon for the next machine instead of mach? why is it so ! invisible? ! It is rather interesting that the under currents of another thread coincidentally have the same questions of interest and visibility of Oberon. The thread is called "Libraries for Oberon (was Brand New Oberon for Windows)" . I'll extract some of the stuff here. Wojtek Skulski points out a necessary feature for Oberon's success and also expresses your concern about the popularity of Oberon. ! > ! > Here, here, Wojtek! I'd like to carry that one step further. The need for ! >libraries ... math, string and others ... ! > ! In principle I agree. Libraries are essential for the general acceptance ! of the language. However, who is going to invest in developing libraries ! if the language is not being used ? And it is not being used (among ! physicists) due to a lack of libraries. I think this is a chicken-and-egg ! problem. Now Wojtek goes on to put the need for *some* libraries , as used in the Physics world, into proper perspective. Don't worry if you don't know what HBOOK or ZEBRA are ... you don't really want to know anyway! I offer this excerpt as background for my next point.. ! I also have mixed feelings about the need for said HBOOK and other ! CERN stuff. As you know, much of it was created to alleviate fortran ! deficiences. For example, ZEBRA is there to provide dynamic memory ! management which FORTRAN does not have. This is written explicitely ! in PAW manual. Similarly, HBOOK function is to implement complex ! data structures on top of FORTRAN arrays. They were needed for ! histogramming. Much of this stuff is disposable if one is using Pascal, not ! to mention Modula-2 or Oberon. Why people chose to use these "robust" ! libraries rather than use better language, is beyond me. Don't worry, I'm getting there. Wojtek's comment produced this response from Jonathan Shapiro. (Wojtek Quote: ! Skulski) writes: ! > Why people chose to use these "robust" ! > libraries rather than use better language, is beyond me. ! ! Perhaps because ! ! (a) their friends and colleagues use it and understand it, making it easier ! to share experience. ! ! (b) There are a few critical features that they need - e.g. a fortran ! compiler must honor parenthesis, which is essential if you are ! concerned about accumulated error. Many physics codes are. It's a ! continuing surprise to me that "modern" languages continue to be deaf ! to the most basic facts of numerical analysis. ! ! (c) They want to be able to take their code with them, and ! (your-language-here) is not widely enough available. The existence of ! the GNU compilers is eroding this impediment. ! ! Choosing a programming language is a social process, and very few language ! designers have taken the requisite social engineering into account as they ! design their languages. ! ! This is why Modula-II and Oberon, which are technically quite good, have ! never been widely accepted; while C++, which Bjarne acknowledges was a ! "compromised" language from the start, has taken off. ! Well that was certainly a long, but necessary, route to go. I guess you can see some of the reason for success of other languages if you read between the lines. If indeed Jonathan is correct and the great success of such languages as BASIC, FORTRAN, C, C++ have been due to "social engineering", it is a sad commentary about the programming discipline we programmers have ... or even care about. I remember that almost 10 years ago I would ask C programmers why they didn't use Modula-2 when they would gripe about getting caught by some type checking problems. They were quick to comment that LINT was available to them. I would ask if they used LINT and their response was NO. Why? Because it "got in the way" of getting the job done. Now, 10 years later, with C++ strong typing ability this is hailed as quite an advance to C! Never mind a "technically good" language had it 10 years ago (as well as Pascal before that). It was as if it was a brand new discovery because it came from the world of C! Go figure. It is interesting to note that such languages as ALGOL and Modula-2 are more popular and widely used in European countries. ! this is not an attack on the oberon folks out there. i just have these ! ideals that the good guy should win in the end and truth triumphs over ! deceit and i get upset when it seems to be going the other way. ! Let's all sail off to Switzerland and program in Oberon forever! (It certainly would solve the lack of manpower excuse ETHZ sometimes states!). As Wojtek points out:: ! The problem is to make other people see what you can see. This may ! be an uphill battle. And as Jonathan says: ! Personally, I believe that technical wins can be made to succeed in the ! market, but simply being a technical win is insufficient. For years I called Oberon's predecessor, Modula-2, the appropriate blend of C and Pascal for it did indeed have what you need in C and wanted in Pascal. Yet to this day Modula-2 never became mainstream. Oh it has its niches here and there. So how does one make other people see what you can see? Evangelism perhaps. I leave you with a brief example to this tome. In the world of object oriented languages, other than C++, perhaps Eiffel comes to mind. Now Oberon-2 has roots in object oriented style programming but I don't think many would mention it if asked to name an object oriented programming language (sounds like a question for Family Feud). I have just gotten in the mail a brochure for Tools USA '94. Yea there are several Eiffel tutorials available. Well look at this ... the conference chair is none other than Bertrand Meyer himself! Have a nice 4th of July every one! ... You'll need it in order to read this novel! Zzzzzzzzzzz.... Alex "Oberon oblate" Waller ---------------------------------------------------------------------- Alex M. Waller M.S. 360 |"We want information ... Accelerator Controls Department | information ..." Fermi National Accelerator Laboratory | Batavia, Illinois 60510 | No. 2 from the Prisoner ---------------------------------------------------------------------- Tue, 17 Dec 1996 03:12:22 GMT if oberon is so good, why isn't it more popular Because: 1. A simple, efficient, language and system does not lend itself to empire-building so well as bloated, obscure, obfuscated, unmaintainable monsters such as C++, Windows, never mind Windows NT 2. You cannot get, at least here, in Melbourne, any of the necessary literature on Oberon. I went to the Royal Melbourne Institute of Technology bookshop, found nothing. Went to the technical bookshop two streets away, found nothing. In a nutshell: Oberon is not popular because mediocrity and humbug reign supreme. Tue, 17 Dec 1996 12:34:27 GMT if oberon is so good, why isn't it more popular Quote: > Let's all sail off to Switzerland and program in Oberon forever! > (It certainly would solve the lack of manpower excuse ETHZ > sometimes states!). Hmm . . . What kind of student financial aid does Switzerland offer? How quickly would I have to learn German? -- Every so often, I like to go to the window, look up, and smile for a satellite picture. -- Stephen Wright Wed, 18 Dec 1996 08:45:32 GMT if oberon is so good, why isn't it more popular Quote: > Because: [...] > 2. You cannot get, at least here, in Melbourne, any of the > necessary literature on Oberon. I went to the Royal > Melbourne Institute of Technology bookshop, found nothing. > Went to the technical bookshop two streets away, found > nothing. Not entirely true, but mostly. I got "Project Oberon" and "OOP in Oberon-2" at the other technical bookshop *four* streets away, and saw "Programming in Oberon" there as well. The last time I looked a couple of weeks ago, there were no more on the shelves. It seems they got in small quantities and didn't re-order when they sold out. Even when they were there, it was away in the back corner, among the 'obscure' academic stuff, past the "Windows Control Panel for Dummies" books. Another possibility for Oberon's lack of popularity: Pascal. I don't understand it myself, because Modula-2 and then Oberon blew Pascal out of the water, but no-one told the rest of the world it seems. Frank Copeland -- MODULE Sig; IMPORT StdDisclaimer, CleverQuote, AdvocateOberon; END Sig. Wed, 18 Dec 1996 07:05:59 GMT if oberon is so good, why isn't it more popular The vast majority of programmers work on x86 based PC's. Modula-2 never really caught on because there never was a highly polished integrated development system. Right or wrong, it looks like the people who control the choice of programming language are highly influenced by lots of pretty windows and tons of documentation. The only way Oberon stands a chance is someone could put together a package with 15,000+ pages of documentation, (this should include lots of graphics and many even sound so a soft copy would take up an entire CD), then another CD of just code examples (another 600 Meg). The package would have to have lots of bells and whistles to impress people with. The system would require a minimum of 16 Meg of ram and take about 150 meg of hard disk space to install. Jeffrey M\kern-.05em\raise.5ex\hbox{\b c}\kern-.05emArthur a.k.a. Jeffrey McArthur Thu, 19 Dec 1996 04:28:52 GMT if oberon is so good, why isn't it more popular Quote: >The vast majority of programmers work on x86 based PC's. >Modula-2 never really caught on because there never was a >highly polished integrated development system. >Right or wrong, it looks like the people who control the choice >of programming language are highly influenced by lots of >pretty windows and tons of documentation. >The only way Oberon stands a chance is someone could put >together a package with 15,000+ pages of documentation, >(this should include lots of graphics and many even sound >so a soft copy would take up an entire CD), then another >CD of just code examples (another 600 Meg). The package >would have to have lots of bells and whistles to impress >people with. The system would require a minimum of 16 Meg >of ram and take about 150 meg of hard disk space to install. If you put together all .ps files from neptune, you will get pretty impressive documentation. There are also lots of code examples, esp. in recent System-3 release. Then, there are four Oberon books which together total 1500 pages and about$200.
Since these are really needed for any serious work, one cannot
There is also something to impress people with: Gadgets.
So the situation is not that bad. Still no answer to the
title question.

Wojtek

Wojtek Skulski, Physics Div., Oak Ridge Nat'l Lab, Oak Ridge, TN 37831

Thu, 19 Dec 1996 20:11:00 GMT
if oberon is so good, why isn't it more popular

Quote:
>Another possibility for Oberon's lack of popularity: Pascal. I don't
>understand it myself, because Modula-2 and then Oberon blew Pascal out of
>the water, but no-one told the rest of the world it seems.

I think PASCAL's popularity is largely there because of commercial support
for PASCAL-like thingies from Borland and Apple. TP looked at M2 and tried
adding what they thought were its best "features" into Pascal - very
successfully in many cases. Borland et al know what sells in the market
place much better than the language theorists who couldnt care less and
thence lies the problem.
Phil

Fri, 20 Dec 1996 05:06:14 GMT
if oberon is so good, why isn't it more popular
: >Another possibility for Oberon's lack of popularity: Pascal. I don't
: >understand it myself, because Modula-2 and then Oberon blew Pascal out of
: >the water, but no-one told the rest of the world it seems.

: I think PASCAL's popularity is largely there because of commercial support
: for PASCAL-like thingies from Borland and Apple. TP looked at M2 and tried
: adding what they thought were its best "features" into Pascal - very
: successfully in many cases. Borland et al know what sells in the market
: place much better than the language theorists who couldnt care less and
: thence lies the problem.

Sure. Turbo has nice products. By far in the way their greatest
strength is their libraries. Borland as put a lot of effort into
this. The libraries work well with both the DOS and Windows
environments. As with most application frameworks they are clumsy
to work with and overly rigid - but this is the state of the art
for commerical based products. In the mean time it would be nice if
they "fixed" some of the language problems such the lack of dynamic or open
arrays. It would save a lot of time and aggravation for
Turbo-Pascal programmers and eliminate a lot of unneccessary questions
on  comp.lang.pascal. I don't understand why Borland remains
deaf to the pleas for some of these basic features and yet
introduces useless features such as pure pointers. Don't the folks

Oberon is a university based project. It used to demonstrate
software techniques without the baggage required to market
a commercial package. As such it is hardly surprizing
that they put little effort into the completely backwards technologies
of 386 chips, 16-bit Windows or real mode DOS -- even
the slow moving market place is turning away from these products.

Some universities have setup or are setting up means to transfer
technologies from universities to industry. Perhaps one day there
will be a smooth flow of ideas from academia to practice and we'll
will see the good ideas of the world implemented in commercial
products..... I won't hold my breath.

Whitney

Sat, 21 Dec 1996 03:05:15 GMT
if oberon is so good, why isn't it more popular

Quote:
>As such it is hardly surprizing
>that they put little effort into the completely backwards technologies
>of 386 chips, 16-bit Windows or real mode DOS -- even

I agree its hardly surprising but also makes the effort pretty pointless.

Quote:
>the slow moving market place is turning away from these products.

Where do you get this idea?
Software Publishing Assoc figures on application sales in $(millions) 92 Q1-Q3 93 Q1-Q3 Windows$1500                   $2600 Dos$1650                   $1250 Mac$550                    $600 All others$100                    <$100 including Unix and OS/2 Note that Windows for 93 not only out-sold Dos, but outsold all other other combined. No questions about windows being shelf-ware - this is figures for software applications, not OS. Like or hate it, making a living in programming means living with Windows for a vast no. of programmers. Quote: >will see the good ideas of the world implemented in commercial >products..... I won't hold my breath. Nor will I. Another terrible reality programmers in the market place - I like many other C haters are having to seriously consider moving to C++. Reason, simply being able to track the changes in the Win32 API. Borland just arent tracking fast enough. Clients are already starting pressure for OLE 2.0 and Chicago isnt even released yet. I'm holding and hoping but it certainly has put a dampener on plans to buy an Oberon-2 compiler. Phil Sun, 22 Dec 1996 06:33:19 GMT if oberon is so good, why isn't it more popular Quote: >The vast majority of programmers work on x86 based PC's. >Modula-2 never really caught on because there never was a >highly polished integrated development system. Well I certainly still have a lot of affection for the TopSpeed M2 system. It was great. Trouble was that a windows version never appeared let alone tracking the changes in the Win API. I am sadly thinking of moving TS off the disk simply because I no longer do any DOS programming. Quote: >Right or wrong, it looks like the people who control the choice >of programming language are highly influenced by lots of >pretty windows and tons of documentation. I think the ability to produce programs for their market is criteria that most use. A good IDE and tons of documentation help but that is still only the start. Phil Sun, 22 Dec 1996 06:38:34 GMT if oberon is so good, why isn't it more popular : >As such it is hardly surprizing : >that they put little effort into the completely backwards technologies : >of 386 chips, 16-bit Windows or real mode DOS -- even : I agree its hardly surprising but also makes the effort pretty pointless. Hardly. Oberon is fun ! [ 16-bit Windows ] : >the slow moving market place is turning away from these products. : Where do you get this idea? : Software Publishing Assoc figures on application sales in$(millions)
[ snip - lots of \$ ]
: Note that Windows for 93 not only out-sold Dos, but outsold all other
: other combined. No questions about windows being shelf-ware - this is figures

You misunderstood me. I wrote that few ( developers ) have intrest
in Microsofts 16-bit systems. Microsofts has made it very clear
that its flagship product will be  32-bit "real soon now". The
folks at the ETH would be wasting their time developing a 16-bit
system this late in the game. I think that they did the right
thing by implementing their system on top of Win32s - even though
Win32s is premature technology that does not fit very well Microsofts
16-bit products.

: for software applications, not OS. Like or hate it, making a living in
: programming means living with Windows for a vast no. of programmers.

Sure. I don't see what that has to do with Oberon. The ETH has
an operating system/environment called Oberon. It is fun system
to work with and great way of demonstrating the versatility
this system available for people to experiment with and thankful
that it is robust enough to be practical to work with. More important
to many is that Oberon is a general purpose language. I can not see any
limitations in the language for using it for developing code for
Windows, or any other operating system. What is lacking and what
makes Oberon painful to use on other OS's is lack of module interfaces
to libraries of the host OS. ( Some commerical implementors supply POSIX
ANSI-C interfaces but these are relatively useless for Windows/NT
or OS/2 programming. ) Borland did a great job of implementing
an API for Turbo Pascal on top of Windows - surely the same could
be done for Oberon.

: >will see the good ideas of the world implemented in commercial
: >products..... I won't hold my breath.

: Nor will I. Another terrible reality programmers in the market place - I
: like many other C haters are having to seriously consider moving to C++.
: Reason, simply being able to track the changes in the Win32 API. Borland
: just arent tracking fast enough. Clients are already starting pressure for
: OLE 2.0 and Chicago isnt even released yet.

Intresting... in our lab one of the most frustrating aspects of
Windows is not fast things are moving but rather how slowly things
are moving from 16 to  32-bit products.

Who knows ... maybe the US government will break up Microsoft the
way they broke up Bell. If they botch it up as badly as they
botched up the telephone system we may soon be moaning about
how slowly things move...

: I'm holding and hoping but it
: certainly has put a dampener on plans to buy an Oberon-2 compiler.

Ahh... but C/C++ is devoid of fun that ..... you'll be back
or at least looking at something comparable.

Whitney

Sun, 22 Dec 1996 11:37:07 GMT
if oberon is so good, why isn't it more popular

Quote:
>Well I certainly still have a lot of affection for the TopSpeed
>M2 system. It was great. Trouble was that a windows version never appeared
>let alone tracking the changes in the Win API.

It's been in alpha/beta for some while now. If you want a copy you should be
able to get on the list because they were offering just that a few months
ago. Money is a bit tight at the moment so I declined. (want to live in a
country that deliberately hurts itself? madness)

Mon, 23 Dec 1996 05:48:00 GMT

 Page 1 of 1 [ 14 post ]

Relevant Pages