On Mon, 24 Jun 2002, Pier Fumagalli wrote:
> >
> > Tomcat SHOULD be suited for reliability, scalabitity and performance,
> > and that's why I found Tomcat 5.0 proposal so important.
>
> Craig said the opposite...
>
Not true at all.
I have no problem with supporting the "large server / millions of hits"
use case being one of the goals. I adamantly reject your arrogant
assertion that nobody else should get what *they* need as well.
Same goes for Glenn's needs for supporting lots of different virtual
hosts. Makes perfect sense, and everyone has benefited from the security
manager support that he added. But that is not the only kind of
environment where Tomcat is useful. Kicking out features that other
people use is not appropriate.
Did the existence of the SSI or CGI servlets have the slightest impact on
whether Tomcat worked for either one of you? Of course not -- you used
what you needed and ignored the rest, or didn't use it at all. In the
former case, there's nothing wrong with other people using different
feature subsets in different ways. In the latter case, fix or add the
part you need (by the way, does mod_webapp *still* forward requests for
static webapp resources to Tomcat?).
Pier's "58% of the Internet" number is very nice -- but there are 10x -
1000x times that many applications running on intranets where you get more
like hundreds of hits per day instead of millions. Making that sort of
environment require installing a separate HTTPD server is ridiculous.
Note that nobody (including me) is arguing about whether some refactoring
is necessary -- it obviously is. I'll forward my list of favorite targets
in due course. But I care for Tomcat users of all types, who need
different sets of features; and I'm offended by people who think that only
their own needs are worth considering.
Craig
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
