On 7/26/05, John Henry Xu <[EMAIL PROTECTED]> wrote: > > Craig said:You can't properly measure a technology's overall success on > a single > > criteria like this. > > Craig, you are absolutely right. Maybe there is a better way to measure > technologies based on broader criteria. The problem was I (or public) > could not access proprietary networks as you said. I used search engines > a lot recently. Many topics I searched come up sites in PHP and cgi. For > example, I found www.javaworld.com, the site runs "Apache/1.3.26, Unix, > mod_gzip/1.3.19.1a , mod_perl/1.27". It likes cgi, isn't it? >
No ... that just means they have PERL installed on their installation of the app server, with no indication what else is going on. That has absolutely nothing to do with what technologies are actually used on that site (which happens to have a combination, so how are you going to count that? :-). > Again, I agree with you that maybe many hidden sites were Java but we > don't know. Then why are you trying to make any assumptions? You claim to be a journalist ... do some research and justify your findings :-) Of course, you're going to find it very difficult to find many meaningful metrics for open source projects. Let's consider one commonly used one ... download counts. On an Apache project, there is zero way to know what that count is, because good projects take advantage of the Apache mirror system, and there is no central accumulation of the overall stats (because the mirror sites consider that information to be proprietary). I can tell you that, a couple of years ago, before Struts started using the mirrors it would get about 70k downloads per month ... nowdays, it gets 15k-20k from the Apache site and a totally unknown number from the mirrors. So, has the total gone up or down? Darned if I know. But I do know a few other numbers .... consider downloads of the J2EE 1.4 implementation from Sun (or the 8.0 or later PE version of the app server, but can be downloaded for free) which come with JSF. You need seven digits (i.e. millions) to count these numbers ... yet Sun's overall penetration in the app server space is relatively small. Is Tomcat in the millions yet? (Answer: probably, but not as many :-). Tell me again how you come to the conclusion that Java is not a popular platform for web app deployments? (To say nothing of the fact that Microsoft might dispute the "PHP is king" rubric as well :-). > > John Henry Xu > Craig --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]