Howard Ship has released Tapestry 5.2.4, what he calls the first stable release since April 2009. Enhancements: no more page pooling, live service reloading, query parameter support, bean validation, documentation, and a new motto!

The motto is "code less, deliver more," and is accompanied by a new logo and new committers.

Does anyone still use Tapestry? I know it's got some fanatical supporters who puke at the thought of using anything else, but Wicket looks like its stolen Tapestry's thunder; plus, Tapestry's tendency to be incompatible between revisions looks like it's been a problem?

Tapestry just looks like it updated itself into being an also-ran, even though it seems to have been pretty influential at one point.

As a core Wicket contributor I'd like to congratulate the Tapestry community with their newest release.

Just as JSF users want the world to forget about the past (any JSF release prior to JSF 2.0) I'm sure TSS community can do without the obligatory rehashing of old flamefests. Take a look at T5's progression instead of digging up 3 year old grudges. T5 has enough merit on its own: compare T5 current with Wicket 1.4 instead. Or JSF2, Play Framework, or GWT 2.x.

Stripes is very flexible and powerful.  And very easy to use and understand.

there I fixed that for you

Tapestry is shit.

Tapestry is shit.

 

Care to elaborate, or was that all you had?

Everybody with half a brain know's there's only framework that actually works: loom

 

http://loom.extrema-sistemas.org/

As a core Wicket contributor I'd like to congratulate the Tapestry community with their newest release.

Just as JSF users want the world to forget about the past (any JSF release prior to JSF 2.0)

Indeed, very true words. Are you a Wicket committer?

There is one very big difference between JSF and Tapestry here though (and actually EJB, even though this isn't a web framework). JSF and EJB started out as utter crap and were each taken over by completely different teams and reshaped into something much better.

The same happened with Tapestry basically, but this time it was the same guy who rewrote his own crap a couple of times over.

I'm not saying one is better than the other, but it's a striking difference.

I was involved a while back in a large Tapestry-based projekt, a large, commercial, global website based on Tapestry 5.1. I liked a lot of the ideas and thinking behind Tapestry (5.1). The core patterns are good and I liked the component lifecycle. The block/delegate structure I found very powerful as well.

One of the main issues was the documentation. Especially the convention-based configuration structure was very hard to penetrate and the documentation was more or less abscent on that issue. What I understand, the Tapestry team has done a tremendous effort on producing documentation on the 5.2 release and that should definitly increase the potential of the framework for larger, commercial applications.

I wonder who, in their right senses, would choose Tapestry for any serious project in this modern times. Tapestry is a failed project. Go take a look at their users mailing list and you would quickly realize that that over-engineerd piece of software is being used by only students and newbies on their school projects or hobby projects. I know no serious enterprise using Tapestry for serious stuffs other than POCs.

Also all the major comitters have left Tapestry. Even Jesse Kuhnert, Tapestry's second man has long abandoned Tapestry. Mr. Howard Lewis Ship once took up a payroll job because nobody was using Tapestry seriously enough to pay consulting fees for it.

I have used Tapestry in the past. I have used all the versions, namely 3.x, 4.x, 5.x, so I know what I'm talking about. Tapestry is simply dead. Fortunately Mr. Howard Lewis Ship is still alive and may come up with something new and better than that ugly and hairy beast. Though I had wished Mr. Howard Lewis Ship was a little bit more beautiful.

R.I.P. Tapestry. Long live all others.

Jan

 

Hey, you've been slinging mud on Tapestry for couple of years or more, so when I see your question 'who, in their right senses, would choose Tapestry for any serious project' side by side with your statement 'I have used all the versions, namely 3.x, 4.x, 5.x' I feel tempted to ask, are you in your right senses?

I've used 3.x,4.x,5.x and the only real real pita was lack of documentation. I do believe that Howard should first be made to write book for the next version and only after be allowed to start coding it :)

Hey, you've been slinging mud on Tapestry for couple of years or more, so when I see your question 'who, in their right senses, would choose Tapestry for any serious project' side by side with your statement 'I have used all the versions, namely 3.x, 4.x, 5.x' I feel tempted to ask, are you in your right senses?

I've used 3.x,4.x,5.x and the only real real pita was lack of documentation. I do believe that Howard should first be made to write book for the next version and only after be allowed to start coding it :)

I've been only telling the truth from my own experience, and many, many others who have also used Tapestry in the past and present. And believe me, Tapestry has a very bad record based on these experiences. Tell me, are you one of those using Tapestry for the "Hello world" kind of projects or on a hobby project? If your answer is yes, then I don't think you are not in your right senses so don't feel offended.

Jan

Well, you are absolutely right, I have not used it for high load web-apps but just for administrator web UI-s (i.e. limited number of admin users) and I found that after overcoming that well recognized steep learning curve of Tapestry, development gets much more productive/enjoyable compared to, say, JSF.

Having said that, I also have to admit that I do not think using MySQL or MS Sql for heavy load database apps is very sensible (I had to resort to timesten) but I do not post comments questioning developers' credibility at TSS or MySQL forums.

It seems You have a hobby of using every release of Tapestry in high profile projects (you said that, not me) only to get frustrated and write posts like that year after year. Sorry maybe I just should not have replied, but believe me, I dont feel offended, just having fun :)

I'e actually been pretty happy with Tapestry--particularly in comparison with some of the other frameworks.  I'm not saying that the others are bad, but the Tapestry paradigm works very well for me.  Jan seems bent out of shape because Tapestry broke compatibility between different versions. Honestly I think this was a good thing and I'm glad to see that Tapestry 5 started with a clean slate. There are always new frameworks coming out to take advantage of different things that have been learned on previous frameworks.  Perhaps Jan would be happier if Tapestry 5 was called something else.

I've seen a lot of people start trying to use Tapestry and get frustrated.  Some of this was because of a lack of documentation (which has mostly been fixed--the current docs are great).  A lot of it is just that that Tapestry take a different approach to doing things and a lot of programmers have a hard time getting their head around the way things work.  I feel it is worth acquiring the different mindset because it is an extremely productive programing environment, but for people who have a lot of difficulty understanding new concepts it might not be worth the effort.

I use tapestry 5 every day at my company. We work on large complex software projects. I can't imagine doing what we do without T5. IMHO It's an amazing framework that is a pleasure to use and makes it incredibly easy to build complex software. Also, worth noting that the IOC in T5 is by itself worth using - very sophisticated. Have any of you who are bashing T5 actually used it professionally (no, not T4 which is a completely different framework)? I strongly recommend to any java developer who's passionate about discovering and using cool and powerful new tools to check it out! It's interesting that people complain that T5 wasn't backwards compatible with T4 - maybe T5 should have really been called something else because it is essentially is a new framework and is it's own thing.

I strongly recommend to any java developer who's passionate about discovering and using cool and powerful new tools to check it out! It's interesting that people complain that T5 wasn't backwards compatible with T4 - maybe T5 should have really been called something else because it is essentially is a new framework and is it's own thing.

This is a bold but irresponsible advice you are throwing into the air. Have you thought about when T6 comes out? You would then have to throw away your T5 code and start all over again if you want to reap the benefits of whats on offer there. Mr. Ships claims there won't be T6. That is also what he said during the T4 era- there won't be T5. Your company has made a huge mistake by choosing Tapestry. And they would find out when T6 is released, believe me.

Jan

I completely agree with you! I chose Tapestry 5.0.X for a IT project during the last year and I must confess that compared with other frameworks, it's amazing. It was great, because we needed to create two  eGoverment apps, and we chose Tapestry for one of them and Wicket for the other.

I must say that, at least in that times, the Tapestry support was better than Wicket's.

I must say that, at least in that times, the Tapestry support was better than Wicket's.

You may be right about the support, though from my own experience I think Wicket has a very friendly, vibrant, enthousiastic and highly supportive community. Having said that, I must also confess that Wicket, in terms of techinicalities and productivity, is far, far superior to Tapestry regardless of what the web framework religious fanatics may claim.

Jan

I must also confess that Wicket, in terms of techinicalities and productivity, is far, far superior to Tapestry regardless of what the web framework religious fanatics may claim.

Jan, the only "fanatic" in this thread is you. Dial it down a bit. You dont like Tapestry, we get it. Some people disagree with you, get over it.

Wow this looks like a real bashing contest.

If you like something and are passionate about it, good. If you want to evangelise, fine. But no need to try to say something is bad without arguments.

As a long time user of Tapestry I must say that Tapestry is a nice framework. I like it a lot. From my feeling, the framework has made a lot of progress in the last years and the community seems to be growing. In the last year or two the active community seems to have grown (no hard figures, just my impression from following the project mailing list), and development is now done by an active group of committers.

I would recommend people to have a look at it and form their own opinion.

Happy coding.

I was using T4.0 (the one that was running on java 1.4 as it was a limitation of our project at that time) and I found a lot of good ideas in Tapestry. It is true, that each version of Tapestry is actually a new framowork with it own codebase. The only thinks in common are ideas, but they are evolutioning too. For example, Hivemind IOC, which was used in T4 is similar in ideas to Tapestry-IOC which is used in T5, but Tapestry-IOC also contains a lots of improvements, like Guice style syntax, etc. It is also important to mension that Tapestry-IOC is official part of Tapestry, when Hivemind was standalone project (It was an issue for Howard to keep both frameworks in synch - at least, that was one of the main reasons as I remember). BTW, no matter if you like Tapestry itself or not, Tapestry-IOC is worth attension even as a separate element. It is fully independant maven artifact and can be used without any reference to Tapestry. And some aspects of Tapesty-IOC are unique (like configurations) and can't be found in other IOCs (Spring, Guice). So even if you already have your favorite framework for web, Tapestry still has something to offer, at least for widening your horizonts. The only problem I found with this "improved" documentation, is that I was not able to find IOC specific docs, as they were created fot T-5.0 or T-5.1, old organization of docs was more intuitive for me.

In a conclusion: If you like new ideas and enjoy discovering something new then you need at least to play with Tapestry. Maybe you will find out that it not what you need, but I am sure you will find it interesting anyway. Choosing Tapestry for your business is completely different question and this is a problem of all this thread: Jan blames Tapestry because of political reasons, other people hails it for it technical side.

P.S. Today, For RIA applications I would use GWT, not Tapestry or anything else. But for page based I will take a look on a Tapestry once again.