[Mono-list] .NET version 1.1 and 2.0

[Douglas and Elena Husemann]

>
> > > Also, as far as the database (ADO.NET) classes go they're,
> > not part of
> > > the ECMA class libraries. Therefore they're Microsoft proprietary
> > > class libraries. Anyone trying to emulate their
> > functionality is doing
> > > so at their own risk, right? Same goes for windows forms, asp.net,
> > > enterprise services, etc.
> >
> > The fact that they are Microsoft-built does not make them any
> > less broken.
>
> I wasn't trying to say they weren't "broken", but I was trying to point out
> that it's really Microsoft's problem if they are, not other .NET
> implementation's.

Actually, ADO.net, is a good thing once they get it fixed and all the Libraries
are the same.  And perhaps work out a few issue of it loading the entire data
into memory instead of a local File to work with. (but that is an issue with XML
in general)

ADO is a Database Abstraction device, it allows for the data to be downloaded
and
worked on and updated.  There are a few issues of Syncing. (and if many users
have updated the same data, which sync is correct).
I feel that this is something that MS is keeping for themselves and should have
been
implemented in the BCL.

The advanatages are that those that Implement a version of it can fix the
inconsistancies.
and perhaps make it work more efficiently.  But having worked a bit with
databases
(not a guru for sure) and played a bit with ADO.net.  It would be a big
disappointment
if it wasn't implemented in some way.  (even more so than Win forms IMO)

> > And indeed, the article is about .NET, and not
> > about the ECMA specs.
>
> Where are you drawing the line then? The ECMA specs are what every runtime
> needs to provide to be considered .NET standards compliant. Any other class
> libraries written outside of the ECMA BCL are proprietary, right? Whether
> John Developer writes it or Microsoft writes it certainly makes no
> difference. If John Developer makes poor design decisions or ties himself to
> a single platform using P/Invoke or COM interop, it's the same thing...
> right?
>
> So, from your point of view, is it that .NET == Microsoft in the sense that
> .NET is the brand name of Microsoft's implementation of the ECMA
> specification. That's just not the way I looked at it.
I can't speak for Migual. but from my point of view .NET is exactly that.
an Implementation of the ECMA standard with several extentions.

A true implementation of the ECMA standard that has MS blessings is Rotor.
and in My opinion is fairly useless except for Command Line Programs. Because
it only implements the BCL.  so you don't have winforms, ado.net and asp.net
and the other libraries extentions that MS chose not to include in the standard.
(at this time anyway).

Although it is very interesting from an educational point of view to look at the
code
and see how most of the CLI is implemented. which is a round about way to
to get peer review on code in a closed sourced company.

The main thing I think would be for the CLI to standup to security attacts (even
with
the stupid idea of still have Unmanaged code in the system. (but some thing
still need
the ability to run closer to the machine)  Since that is one of the primary
reasones of
imlementing this. and moving it in as the primary development platform for
windows.
so having the peer review will help/hinder in that regard as the platform is
exposed
at the source level. attacks can be seen and generated that take advantage or
benevalent
programmers can suggest fixes Although unlike OS. probably won't get the credit.

Besides cleaning up the Horrendous Development platforms related to visual c
etc.

Douglas