[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Y Store now C++



Sundar Narasimhan wrote:

>>Java was hardly the only language to provide important benefits over C/C++,
>>    
>>
>such as..
>
Most of the languages we discuss on this list.

>People need to make cashflow and payroll this month! What other
>languages were true contenders circa that time frame? (Remember: we
>already had/have production apps in Lisp -- so I'm looking for what
>were "new" solutions at that point in time frame that I might have
>missed :)
>
Why, at any given time, was anyone constrained to only use a "new" language?

>
>BTW, I just don't buy the "Java ran/rode with the WWW/Netscape
>coat-tail" argument at all -- this is one of those "press generated
>marketing half-truths".
>
It is what I observed myself.

> Coat-tails have a way of disappearing,
>
Yup.  This one has disappeared long ago. It only has to last long enough 
to get
the snowball rolling.  After that the snowball rolls by itself.

> and
>languages do fall out of favor.. remember AI and Lisp? 
>
AI never got the snowball rolling for Lisp.  AI was teensy in comparison 
to the
Internet/WWW phenomenon.  And Sun gave Java a huge push and many other
companies joined the bandwagon. Of course it helped that Java had a more 
familiar
syntax and stuff like that.

> Let me
>tell you - my empirical evidence proves otherwise
>
Well, you and I have evidently come to different conclusions.

>Sure, programmers are sometimes (mis)led this way or that by hype
>etc., but Dan -- you say PSE Pro (or was it the later Excelon XML
>stuff) was written in Java -- why did your company make that decision?
>Because of the hype wrt. applets/www/netscape?
>
Here's what happened.  People at Sun approached us and said that they 
were interested
in making something called the "Java Wallet". This would be a software 
module that would
be downloaded into the browser as an applet. It would remember your 
credit card numbers and
stuff like that, so that when you wanted to do e-commerce, all that 
information would be available.
So they needed a small-footprint transactional Java database for which 
very high performance
was not necessary. We thought that seemed reasonable. We didn't want to 
make a product
just for one customer (Sun). But by that time Java was starting to catch 
on as a general language
(for reasons that you and I differ on), so we thought there might be a 
broader market for such
a small-footprint Java data store. So, our decision was partly based on 
client-side Java and
partly based on the prevalence of Java and our guess (correct, as it 
turns out) that Java would
be a widely-used language.