[MUD-Dev] (no subject)

[John Buehler]

>From: "Joe Andrieu" <joe at andrieu.net>
>Date: Sun, 19 Nov 2000 21:52:35 -0800

>Are there any particular online resources you'd recommend to learn more
>about component languages of the kind you are talking about?  Is it just
>COM and CORBA?
>
>Any pointers would be appreciated.

The context in which I was taught this stuff was entirely COM-centric.
There were academic mentions of Eiffel, CORBA, DCE, COM+, Oberon, C++ and
other languages and technologies, but we were doing everything from the
ground up.  So the net result is that all the documentation that I relied
upon was either walking around in the form of Tony, Clemens and company, or
it was online but internal to Microsoft.

Try the following web location to find out about the research group I'm was
with:

http://research.microsoft.com/comapps/

There are pointers to some online documents, and I suggest that you follow
Clemens' link to his list of publications.  Or just search for his name on
the web.  It's thoroughly unique and will give you good hits (Clemens
Szyperski).

I noticed that the "COM Programmer's Cookbook" by Crispin Goswell is
pointing at dead space.  I'll ask Crispin to straighten that out.  His
'cookbook' deals with some important issues that relate to COM component
constructions.

I guess the bottom line is that I can't offer pointers to significant online
resources about languages.  Clemens' book is great material, and he is a
primary author of the Oberon language.

I just tripped over a column that it seems Clemens writes for Software
Development Online magazine.  His October 2000 article can be found at:

http://www.sdmagazine.com/articles/2000/0010/0010k/0010k.htm?topic=uml

Note that there are links to other articles there as well.  Listen to
Clemens.  He's a smart guy.  A very nice guy too (sometimes a little too
nice - you should see his list of obligations for the next year or so).

I hope this is of some value.  The stuff we were working on was intended to
be tight, fast, production code.  Essentially, doing software the way every
software engineer always wanted to.  And because we were looking at software
from the ground up, we just didn't bring in any outside technologies - thus
my ignorance of those outside technologies.  As a product engineer at
Microsoft, I worked in C++, but building production code at Microsoft is not
exactly an academic effort.  The research work was all C code and assembly.
Half the project was to build production components and the other half of it
was to figure out how to do it rock solid.  I spent a LONG time building the
equivalent of a templated vector class that required something like 22K of
code.  Every instruction and every arc covered in testing (not full path
coverage).  I think a total of one bug has been found since we signed off on
it.  It was a pretty minor one at that.

JB


_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev