[ot] commercial containers (was RE: localhost:8080 vs localhost?? ?)

["Fernandez Martinez, Alejandro" <a....@ibermatica.com>]

Hi Drew,

> -----Mensaje original-----
> De: Drew Davidson [mailto:drew@ognl.org]
> Enviado el: martes 23 de julio de 2002 4:39
> Para: Jakarta General List
> Asunto: Re: localhost:8080 vs localhost???
> 
> I have to disagree with you about "supporting" commercial containers. 
>  First of all, I think that the consensus about Tomcat is that it is 
> _not_ a commercially viable option when running a site.  The 
> price for 
> Resin is very reasonable given the performance, features and awesome 
> technical support.

Not that this conversation looks like getting anywhere, but... if you come
to general@jakarta saying these things, the least you can expect is a little
bashing :)

> The purpose of Tomcat is not to be a commercially competitive product 
> but a _reference implementation_.  To treat it otherwise is 
> contrary to 
> common sense; using it to check correctness of behaviour of an 
> application is sensible and smart; using it to run a 
> commercial site is 
> foolhardy.

How can you know this? Are you a Tomcat developer? Have they told you that
their purpose is to build a reference implementation? C'mon, Tomcat 3.3 and
Tomcat 4 are attempts at making the old codebase faster and more robust.

Tomcat *is* a reference implementation, but its purpose is to make quality
software.

> I _really_ disagree with your anti-capitalist stance on this (if I'm 
> reading wrongly into your post please tell me).  Commercial 
> containers 
> (and commercial software in general) help spur acceptance of the 
> technology in which we are all making a living; the more 
> people use Java 
> the better right?  If there are commerically supported platforms then 
> corporations will feel at ease using the technology and they are the 
> ones with the money that drives technology spending.

There is no anti-capitalism or (shiver) communism or anything like that
implied here. And your argument makes much more sense when reversed: It is
free containers (and free software in general) that helps spur acceptance of
a technology. It is because millions of people have downloaded Tomcat for
free, that servlets have gained acceptance. It is because other people see
the successes and mistakes made in the Tomcat source code, that commercial
software has got better, as Jon explained in a separate mail.

It is also a matter of ROI -- using Tomcat, apart from costing nil, ensures
that your application will be compatible with most other containers, as long
as they are compliant with the spec. You don't have to import weird
weblogic.* classes to make it work. Free software encourages compliance to
specs, while competition tends to lead to lock-in behaviors (i.e., to
eliminate that competition). If you want choice, try free software(TM).

> Those of use who have neither the time nor inclination to 
> contribute to 
> Tomcat generally prefer to have the problems with a container 
> be solved 
> by someone who is monetarily incented to - if their container 
> sucks they 
> can't make a living.  Whereas if Tomcat sucks no one can 
> really complain 
> because it may or may not get improved, depending on the whim of 
> non-paid contributors.

This is really an odd argument, wrong on all points. Please just reverse all
the sentences, and it will look straight :)

[snip]

Un saludo,

Alex Fernández.