[techtalk] RE: [issues] Standards?? PHOOEY!!!

[Fan, Laurel]

Lothan, lothan at newsguy.com, said:
> Sent: Friday, June 30, 2000 11:44 PM
> The American National Starndards Insititute (ANSI) and the International
> Standards Organization (ISO) have several standards (and authority) on
> computer devices and languages (QIC tape hardware and protocols, and the C
> and C++ languages are examples of ANSI standards). The Internet
Engineering
> Task Force (IETF) also has authority to set standards for Internet
protocols
> (IPv4). Like it or not, Microsoft is the standards authority for Windows
> protocols (OLE, ActiveX controls).

Exactly:

If you want to write C code or a C compiler, and want to use this
code/compiler on a C compiler/code, you must do this.

If you want to talk to ipv4 hosts, you must do this.

If you want to talk to Windows, you must do this.

Nobody's saying "You must write C code" or "You must use ipv4" or
"You must use Windows" (ok, maybe someone is in this case.  But it's
not a standard). There's nothing to stop me from 

> > The lack of agreements results in gratuitous incompatibility, creeping
> > featurism, and data loss.  Standards arise when the participants in the
> > creation of the standard feel the need.
> 
> Unfortunately there is no guarantee that standards won't prevent
> incompatibility, creeping featurism or data loss. Neither do standards
> guarantee acceptance by the community it intends to serve.

True.  When standards work exactly as intended and the standards-makers have
no ulterior motives, everything is perfect.  Unfortunately, this is not
usually the case, so things are somewhat less that perfect.

> AOL owns both instant messengers... Unfortunately I can't think of
anything
> nice to say about AOHell so I'll keep quiet.

Yes.  Perhaps demonstrating that without standards, no company will
even make the tiniest effort toward interoperability? :) (However,
AOL has indicated that they will accept the possibly forthcoming
IM standard, perhaps to avoid antitrust)
 
> Unfortunately it can also make your life more miserable sometimes. I still
> have and use an old version of Borland C++ because (at least at the time
of
> its release) it followed the ANSI C++ standard (for both the compiler and
> the library). Of course, at that time Microsoft C++ blatantly thumbed its
> nose at the ANSI standard. Needless to say, to this day that ancient copy
of
> Borland C++ is by far more compatible with UNIX source code.

No, it's not standards making your life miserable.  It's a certain
company's attempt to "embrace and extend" a certain standard.
Unfortunately,
the same things that allow innovation allow this. (One attempt to prevent
stuff like this was Sun's Java standard; basically, you're not allowed to
call your stuff Java unless it conforms to (Sun's) Java standard.  Microsoft
attempted to "extend" Java, Sun sued them, and Microsoft has now essentially
dropped Java and are replacing it with some strange "C#" thing.  It's left
as
an exercise to the reader to determine if this is a good thing or not.)