You all know I do love Node.js
, and some of you have noticed me one of its competitors (maybe the only one) Vert.x
is getting quite a momentum. You can get in the video a personal opinion about the comparison by yourself.
- Vert.x is a great (huge) piece of SW. It covers most of the XXI century demanded programming paradigms.
- Maybe it covers too much. Some things it offers are not going to be used ever by average programs (including large corporations ones).
- It claims you have to deal with the low level detail to do hard things in Node.js but, I assume it as adding black magic layers
to the framework: If you don't know how it's working you'll misuse it for sure. It reminds me of EJBs which they were a good thing in theory. Extending the event bus
to a wider scope than the thread (multi-host, client side...) amazes and scares me a little bit at once.
- Even though it's Open Source, there are two corporations behind it: VMware and Oracle (in the Java side). I'm thinking of the Oracle/Google legal dispute and the Vert.x author project withdrawal when he left VMware: http://www.theregister.co.uk/2013/01/14/opensource_ownership/
just a little bit disturbing.
- Finally, Vert.x is prepared for a share-nothing asynchronous-only programming model, BUT
it natively offers some scape tools to do other kind of programs. Maybe you'll end up using a technically great new toy to do old tricks. I've never missed what they're offering me to compete Node.js
Anyway, we're playing around Vert.x because it's a engineering masterpiece and the benchmarks look great (we're doing our own). Maybe I'm wrong but it smells like an overkill for most uses. Once we test it I'll change my mind if it worths it. Coherency is for weak people
, you know ;-).