AW: [Auth]Re - [Auth} A simple server-side authentication scheme

[Carsten Kuckuk]

Norbert,

I still have to get aquainted to the ideas and thoughts of the others on
this list.

>> Because the ecommerce market is the one market that really
>> matters, and Microsoft is targeting it aggressively.

I disagree here. You are right in that this is where the money is, and that
this is the field that MS is targeting, but money is irrelevant for a GNU
project, and I don't think people should blindly follow others. By looking
too much at Microsoft you allow them to control your thoughts and take away
your time. When you enter the realm of VES, CIL, C#, etc. you have to spend
weeks and weeks reading their specs, learning their language and
abbreviations and in the end you see that all they did was just to copy
Java, and make it incompatible to Java. I have implemented several compilers
in the past decade, and reading their specs with this background makes it
pretty obvious. So in the end they have lured you into wasting valuable
weeks of your life by just reading their docs.

Please take the time and visit
http://www.edge.org/3rd_culture/gelernter/gelernter_p1.html and read David
Gelernter's "The Second Coming - A Manifesto". In this text you'll learn
about a different concept of virtual identity -- cyberbodies -- that has
almost nothing to do with e-commerce but has the power to transform the
Internet into something really useful. And this is what I was thinking of
when I made my posts on the list here. It is immediately useful to all
online users, implementable, and non-critical in such a way that initial
errors would not lead to lawsuits.

>> > If you stick with LAMP architecture on the server side, the system
>> > can be rolled out right now on millions of servers.
>> Please explain LAMP architecture, and how we can "roll out a
>> system right now on millions of servers."  That sounds exciting.

LAMP means Linux, Apache, MySQL and PHP. This is the standard web server
architecture that hobyist can rent from ISPs for prices starting at USD 10
per month. All my friends have one, I have one for myself, one for my wife,
and one for the toastmasters club I'm a member of. They're ubiquitious. So
if we can implement part of the system using the relational database MySQL
for data storage, and write the access code in PHP we would have a small
package that could be installed on almost every hobyists web server. The
instructions would read: "Please log into your web server. Start mysql.
Execute the following SQL statement: CREATE TABLE PROFILES( uid int, account
varchar(100), pwd varchar(100), data varchar(200), primary key(uid)); Then
create a directory /dotgnu under you HTML root. Copy the files
putprofile.php and getprofile.php there." Every hobbyist can do this. And
suddenly you have thousands, possibly millions of identity servers on the
web. The clients making use of this data could be implemented as server
plug-ins or as standalone Java applications, whatever people like. That's
what I was thinking about. It would put an infrastructure into place very
quickly, and more important: It would make the project known and immediately
useful.

Am I too far away from the goals of the other members of this list?

Carsten Kuckuk <-- on digest