rebelutionary / 2.0

Yeah, but you know, stuff like this becomes a de facto standard. Ant is one of the worst tools I've ever used, and is so badly designed, I can't even believe it's now at the APACHE level, yet everyone uses it, because it's XML and written in Java. The only way out of it is to come up with another tool, because make has such a stigma from all the trash-talk from the ant camp.

First off, I should point out that part of Atlassian's business is based on a Tomcat competitor (Orion). We use tomcat where I work (NASA) and it we have *never* had a problem. Oh wait...when we first got JIRA, it ran like dog crap on tomcat. The only question I have is it the application or the application server? Our tomcat web app has thousands of simultaneous users and flys. We have three developers that use JIRA to put in issues. Three simultaneous users at peak usage!

To be fair, this was a long time ago and you guys fixed the problem. We love JIRA, it is really a cool application. We also love tomcat too. We tried JBoss, Orion, Jrun, ServletExec. We liked tomcat the best. The benchmarks that I have seen show Orion blowing Tomcat away. Tomcat comes in at a measly 40-80 request per second. That is only a million requests a day. That is all the performance anyone should need.

Explain Jakarta marketing? The only marketing I have ever seen is word of mouth. I have never seen a newspaper, magazine, radio, or tv ad that promoted jakarta. The people who have promoted jakarta projects through discussion board postings, mailing lists, and newsgroups must all be ignorant slobs who like massaging apache people's egos.

The point is Apache has never claimed to be the "Greatest Open-Source Projects Assembled Under One Roof". People use jakarta apps because they like them. Not because they were tricked into believing that these are the best products out there. Some people don't use jakarta and jakarta people don't spam them into submission.

"All fired up". Sounds accurate. Trust me, I have way more plans for Tapestry than I've even been able to state on the lists. I can't wait to get the book done so I can concentrate on even more improvements to the framework.

I think there is a definate cachet with Jakarta and I can state for a fact that Tapestry's visibility has risen quite a bit since the ascendency at Jakarta. Moving Tapestry to Jakarta may make even more of an impression on project leaders and semi-technical managers ... once Tapestry gets a foothold, it sells itself. Perception is reality ... if it is percieved as popular, it won't be dismissed as a dead end. I thrill to see Tapestry mentioned in job requirements on monster.com.

But I like to think that Tapestry is raising the average quality level at Jakarta a bit, too.

Well, I don't care which is crap, which is not - just care it works for me - not talking about the available resources for students working a "commercializable" project.

Out there I can only try out Tomcat, JBoss, Jetty. Since my duty is mainly on developing webapps serving the Asian countries (yes I'm an Asian... will you hate me?), results showed that Jetty or JBoss/Jetty combo cannot show double-bytes correctly (even with UTF-8), sometimes.

That's the trick - I cannot figure out exactly where the problem goes. But Tomcat and JBoss/Tomcat will run them just fine.

However I think I have to regret to tell the fellows working at another group to use Struts... forgive me, I'm also new to Webwork1 at that time.

If you don't like what is available at Jakarta, then volunteer to help fix it and make it better. Duh.

There is no marketing engine...it is a bunch of people *like you* who decided to build cool tools and make them available. Nothing more. Nothing less. If something is unmaintained, maybe it just had its lifecycle of usefulness or just hasn't found the right people to pick it up.

I looked at JIRA's code and schema...I thought it had plenty of room for improvement. Honestly, I'm ashamed that you have created yet another issue tracker that is all fluff on the outside (which is why you are selling the product at all) with a terrible schema on the underside. What is the saying? Those who own glass houses shouldn't throw stones.

Ryan and Howard's posts are right on. To Howard, I agree, Jakarta always can use better software. =)

As a user of JIRA, I've been impressed. It's easy to install and easy to deal with. I've no idea of the schema and code underneath; as a user I don't care.

As someone who hangs around Jakarta a bit, I'm of two minds about the concept of the community. I think Apache had a good chance to 'market' the Jakarta brand and they've in fact taken the exact opposite direction, seeking to remind people that Apache is the brand, not a single project like Jakarta. It's taking time, but with the promotion of various projects the Jakarta brand is weakening. I suspect this is probably going to be a good thing, though I mourn the loss of a meaningful brand (Jakarta) for a relatively weak brand (Apache) in the Java community.

I like the open-ness of the Apache community, although this is harmed a bit by the battle to keep a good mailing list archive, and the mess which is the website. It feels more of a community than the huge mess of sourceforge, and more open than Tigris. The cross-language camps and the cross-philosophy (Avalon? Velocity? Tomcat?) issues at Apache make it seem like a chicken-shed sometimes (lots of squawking), but it also makes it feel far more like a bazaar than any other environment I've encounted. Which is meant to be a good thing in open source, if you believe what you read.