Re: Some Laws

"Marshall Spight" <mspight_at_dnai.com> wrote in message news:<XWO4d.98398$MQ5.24347_at_attbi_s52>...

> "robert" <gnuoytr_at_rcn.com> wrote in message news:da3c2186.0409231518.29479ecd_at_posting.google.com...
> > "Marshall Spight" <mspight_at_dnai.com> wrote in message news:<I5k4d.91137$D%.72678_at_attbi_s51>...

> >
> > let's count:
>  
> > XML data,
> 
> This abomination is the work of Tim Berners-Lee and the w3c,
> who have set the field of user interface back 15 years and the
> field of data management back 30. This has nothing to do with
> Java.

if not in your neck of the woods, then be thankful. lots o folk are stocking up on o'reilly java/xml books, and diving off the cliff. i see it every day. no joke.

> 
> > semantic web,
> 
> Ditto. Utterly unrelated.
> 
> 
> > java beans (Holub, Arnold, and Stroustrup, at least, are on
> > record that the get/set paradigm is not
> > OO),
> 
> A citation, please? Although I'm not going to count Stroustroup
> as an authority on OO. Get/set is painfully verbose, but necessary
> in the face of a lack of syntactic sugar if you want to have
> encapsulation, which is mandatory if you lack declarative
> integrity constraints. I note that languages designed *since*
> Java deal with this issue simply by having syntactic sugar
> for get/set.

having no luck finding arnold. holub: his javaworld articles stroustrup: artima interviews; and he is an OO guy.

> 
> 
> > EJB,
> 
> EJB 1.0 sucks ass, and every Java programmer I know thinks the
> same. If you're going to judge the platform by the worst thing
> ever done with it, then you have to accept as valid any critique
> of the relational model based on an evaluation of MySQL.
> 
> 
> > Swing.
> 
> Swing is as good a cross platform window system as you're likely
> to find. It's MVC to its core. It's certainly popular to bash it, but
> I've never heard a criticism of swing that didn't make it clear
> that the critic hadn't worked with swing long enough to know
> what he was talking about.
> 
> 
> > there's a great quote, may have it in cubeland, that java is perfect for
> > corporate development since it fits the "acretion of code" mode of
> > operations.
> 
> Indeed. Which is another way of saying that if you've got a lot
> of work to do, Java is an excellent platform choice because it
> scales up to hundreds of programmers, something that few
> programming languages do well. (I can't think of another
> that succeeds at this as well as Java does.)

in the original context: a nice way of saying it makes plates of linguine

> 
> 
> >  it fits very nicely with COBOL folk who see code and data
> > as separate things:  in java they call 'em data objects and action
> > objects.
> 
> All the prominent OO writers I have read stress objects as being
> the union of code and data, (they say "state and behavior") so
> I have no idea where this criticism is coming from. Nor have
> I ever even encountered the terms "data objects" or "action objects."

then you've must not have seen struts and other frameworks. they're built on the bean paradigm, which is definitely *not* state and behaviour. just confabulated structs.

> 
> 
> > in all, i like java.
> 
> I find this this statement hard to reconcile with you earlier one
> about the twinks and the lemmings.

java should stay out of larger data management. just as COBOL learned.

> 
> 
> > what irritates me is the drive to turn it into
> > COBOL.  java folk i've met see the redundancy in their language:  the
> > relational database is really an object store without the froo-froo.
> > it is data and behaviour rolled into one.  with a standardized, which
> > is to say language and application independent, access.  this is not
> > a good thing if you're desperate to lock in clients.
> 
> Again I find a criticism that makes no sense to me. The vast majority
> of the Java programmers I know are sufficiently ignorant about
> database fundamentals that they could not possibly perceive a threat
> coming from the dbms corner. It's not just that they don't understand
> data independence, they are not even conscious of it as a concept,
> and so they certainly can't be feeling threatened by it.

again, if not in your neck of the woods.... but spend some time at the publisher sites. see where the wind is blowing. language and application centric data stores. this is regressive. if one is a language maven (well, au courant language), then the idea that data concurrency belongs under your hegemony is neat. not so if one believes that the data store protects the data.

pace,
BobTheDataBaseBoy

> 
> 
> Marshall