Is Cocoon obsolete?


That is the question on our lists today.

Here's what I feel to be the most important excerpt from my reply:

For me Cocoon today is:

- An ├╝bercool integration platform: combine widely varying data sources easily and safely while keeping your code and system structure clean.

- A great help in structuring and implementing complex apps which include some form of web interaction (but the web part might not be the most important).

- The best tool that I know today to build long-lived systems, which are meant to be actively developed for years and as such need a very clean, modular and open structure, to be maintained at a reasonable cost and stay manageable.

- A collection of components written and assembled by some of the brightest programmers around, and it shows, although as with all bazaar developments finding your way is not always easy.

- A framework to build powerful tools: Forrest, Lenya, Daisy...

All this has little to do with the client-side technology being used, so no, I don't think Cocoon is obsolete, not at all.

In my opinion Cocoon is and will stay a niche tool, but for a fairly big niche, which includes many silent feeling is that many companies don't want to talk about the kind of strategic systems where Cocoon is involved.

See also comments from Berin, Ugo and Andrew, and Sylvain.

Update: Stefano has more to say on his weblog, in his usual enlightening way. A great read, and I guess the PiggyBank code must be worth looking at!