What is the current preferred way to save user preferences in dotnet?

[Edward G. Nilges]

Edward, I think it was a mistake on your part to imply that VB is superior
to C#, for several reasons:

I didn't claim that "VB is superior to C#", because I started with
machine language and have used about 20 languages, all of which with
the exception of HTML are Turing equivalent. I made some points in
VB's favor.

1. This hints at the "language chauvinism" you have been saying you detest.
2. It won't be hard for most of the people on this newsgroup to make a
*longer* list of reasons why C# is better. Once that happens, are you going
to then change your position and base your book on C#? If not, you'll have
to explain why not, given that C# has been shown to be superior to VB by a
lot of intelligent people on this newsgroup (after all, you are the one that
initiated that topic).

The way in which a programming language is "superior" is never as
important as the science-behind. Consider the strange statement of
hero computer scientist Edsger Dijkstra: "computing science is no
more about computers than astronomy is about telescopes".

Dijkstra was trying to draw our attention to a shared reality which
isn't "reified" (in the sense of being-a-thing) which to him was the
true subject of discourse.

Paradoxically, Dijkstra was at one and the same time obsessed with the
need for an elegant and persipicuous notation, and contemptuous of
claims for notation magic.

That is, I think, because Dijsktra, unlike an American computer
scientist, was educated in a European intellectual tradition with its
own limitations, its own flaws, but one based on Kant.

If I "cudgel my brains" like Second Gravedigger in Hamlet, I almost
apprehend that Kant foregrounded the unbreakability of the link
between language and reality to such extent that Kantianism is the
refusal to consider a language as a thing-in-isolation.

Therefore Dijkstra didn't reify and reduce the solution to using the
right notation while at the same time realizing that notation is
important.

I conclude that what's important is elegant code in a Turing complete
language and not this or that language artifact. There are limitations
to this view. You can write elegant code in Fortran but the audience
would be vanishingly small (this isn't the case with VB as a Coast
Pidgin and lingua franca of commerce).

I am thinking of my Indian coworkers in Fiji who to get a book must
pay Amazon for shipping and who use VB and not C# and I would like to
send the message that it's possible to compile business rules in their
language.

I want developers to be empowered as an Army of One to write a quick
simple parser in the many cases where the user has a need for an
unbounded set of solutions. Military science has discovered since the
"stosstruppen" of WWI that you advance by empowering the unit and the
individual.

My inspiration remains the fact that FBI agent Colleen Rowley
testified before Congress in Dec 2001 that her team, investigating
Moussaui and the other bad guys prior to Sep 11, COULD NOT enter
simple queries of the form "terroristAssociation And
attendsFlightSchool", and 3000 people died...possibly as a result (we,
as ordinary citizens, still don't know the full story although we, as
ordinary citizens of the USA, have a right to know).

Sequestration of compiler algorithms in a language favored by
highly-paid specialists in compilers has the organizational effect of
forcing MIS teams to use little-understood "solutions", and,
apparently in the case of the FBI, to have to wait for solutions that
never arrive.

I am not privy to the FBI data base and its design. But let us suppose
that it was running (as it might well be) on a mainframe in Langley
running IBM's Conversational Monitor System.

This aging system is programmed at the user interface level by a
language called Rexx.

My guess is that this language prompted Colleen for queries and sent
them to a server. My guess is that it could have preprocessed the
queries using a lexical analyzer and recursive descent parser, and
sent a stream of simple queries.

This "compiler" could have been written, in this possible scenario,
quickly, and was not because organizational pressures meant it was a
nonstarter.

In my own MIS experience, it seems that every time developers approach
a sort of "sound barrier" at which the user's system chaotically
changes from a single solution to a
language-for-talking-about-solutions, the system's design has a
nervous breakdown.

Vague references to overly abstract entities appear such as "tags"
which "tag" a query and modify the query in ways we know not. The
modification is carried out in secret and in code, in ways that may or
may not correspond to user needs.

This is why I'd like to democratize compiler arcana.

3. C# is apparently Microsoft's language of choice for (much of) their next
generation operating system and framework. I'd be willing to bet that MS
favors C# over VB for compiler development. I assume it would be easy for

My personal experience is based on getting rejected from the Microsoft
compiler team last year. I was rejected because I haven't used C++
heavily. I think the plan is to use C++ going forward.

people more experience than me to argue that C# would be a better choice for
a book such as yours. Now that you have opened the discussion to the "best
choice", I don't see how VB can come out on top.

My compiler foregrounds a glass box GUI so you can see what makes it
tick and VB remains, marginally, the best choice.

 

[Mountain Bikn' Guy]

Edward, Thanks for your balanced comments. I was afraid I was going to get
blasted for my last post. I just hope all this discussion isn't impeding
your progress on the book!
Mountain