Ken McNeil wrote:
>
> >I'll second you on the bugs issue and the problem with Suns 1001 API's
> >approach. However to be fair you probably picked a bad eg. with
> >parametric types as Sun has just put up a proposal for them.
>
> I thank you for pointing this out, but they should have responded earlier. By not
>implementing >this in Java 2 but instead creating the new collections API (which is
>an outstanding API based on >the tools available in the language) they have done a
>great injustice to the developers. Many have >built designs paritally around this new
>framework and would have to revise existing code >significanlty because of the large
>differences between the two solutions to the need for a >collections library.
I have mixed feelings (to say the least) about parametric/generic stuff
in java so I will not comment further on this.
> There isn't any need for a Java
> >alternative because there is _nothing_ fundementally wrong with Java -
> >only _implementations_ of it. However this is slowly being overcome - I
> >would suggest that if you want to see a good quality Java implementation
> >that you contribute to one of the free alternatives
> >kaffe/japhar+classpath/etc. Even if you have been crippled by signing
> >Suns licensing - you can at least run the programs maybe send in bugs
> >fixes, write test scripts, docs, whatever.
> >
> >IMHO you're suggestion of an alternative to Java (I guess you mean the
> >whole platform - ie. langauge,VM & class lib API) is on the same level
> >as M$'s latest announcement of their alternative to Java (I forget its
> >name - "Magic" or something like that).
>
> The alternative is code named Cool, but I think there is a fundamental difference.
>The idea is to >seperate Java from Sun and there corporate goals not to kill Java in
>general. I believe without >seperating the two we will always be fighting this
>conflict of interest. And since many believe >(like I do) that Java will evolve into
>*the* platform it seems wise to fight this sooner than >later. If removing the profit
>motive from Java requires creating a new platform so be it.
Cool it is (I don't know why I was thinking Magic). Anyway I totally
agree with you about the need to seperate Sun from Java - "nothing good
will come from this" is the phrase that comes to my mind. I to hope (if
not yet believe) that Java will become _the_ next platform and it would
be best to settle this issue as early as possible. Removing the profit
motive I think puts it to simply though - C/C++ vendors have a rpofit
motive - but with a published ANSI/ISO standard whatever "clever
add-ons" that mightcome up with, they are just that and it's the profit
motive (ie. "the market") that makes them all try to meet the standard.
>
> >I think I (and probably alot of other people) have spent alot of time
> >and effort learning java, writing apps in java and are happy with java.
> >I for one am more then happy with java and would rather work on
> >improving its implementations then expending what would be need to be
> >_huge_ amounts of resources on the of chance of coming up with something
> >comparable. Such an effort would not be an answer to any of my prayers.
>
> I see this as just another example of GPL'ing a commercial product and thus the
>small pains of >incompatablity would be warranted. Anyways, there is no need to
>change the fundamentals of the >language to the point that there would be any kind of
>significant learning curve.
I may have misunderstood you here - if so I apologise. From reading
your 1st post I thought you meant creating a new language/platform like
java but different (to java, as java is to c++ for instance). If on the
other hand you mean simply a java that isn't Java(c)(tm) by Sun them we
are in total agreement.
> >Well having critised, its only fair for me to put forward an alternative
> >suggestion: Move ahead with STANDARDISATION WITHOUT Sun. Surely its
> >possible for ANSI/ISO to produce a standard without Sun. The API's,JVM
> >& langauge specs are all published - only Java(tm)(c)(whatever) really
> >is Suns property - so why not have the ANSI/ISO "J" standard, which
> >just happens to be based on Suns Java.
> >
> >Now I admit I'm to young to know first hand what happened with C, but
> >from what I've read K&R invented C, lots of people used/implemented it,
> >ANSI after alot of work made a Standard for C that was a bit different
> >from the original that everyone now uses (admittedly this is very
> >simplified view of history).
>
> I'm also to young to know the history of that event, but I wonder if this sort of
>action could be >contested by Sun on legal grounds. K&R wasn't a commercial product
>(or was it?). If this was to be >legal it would be a great option to avoid the
>problems I spoke of in my original post and I would >support it.
Well hopefully one of the "elder statesmen" will correct me if I'm wrong
on this, but my understanding is that there were lots of vendors all
selling C compilers+libs and each implemented K&R C plus what ever
"extensions" they felt like. Now Borland/M$/Watcomm/gcc/etc all made a
name for themselves selling products that were at least mostly complaint
to a spec, even if they all did have their own wierd propreitery
extensions. Even if the vendors weren't terribly happy about
standardisation (again I'm not very knowledgeable on the history of
these events - though I can't imagine all they vendors were thrilled by
standardisation) they all put out compliant products eventually - full
stop.
Now today we have the same thing with java dev tools, but since with
java they are only half the game, the same thing needs to occur with the
java (jvm+class libs) platform.
Maksim Lin.
---
Remember Darwin; building a better mousetrap merely results in smarter
mice.
----------------------------------------------------------------------
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
