whoa! rhinola rocks! – at least if you have a reliable and up-to-date operating system...

whoa! rhinola rocks! – at least if you have a reliable and up-to-date operating system at hand that lets you easily install ant or the necessary apxs libraries or even more sophisticated stuff like the latest gnu compiler collection in a snap.

furthermore, either some expertise or intuition on how to fix a Makefile is inevitable: i got an "apxs:Error no config variable PREFIX" on make install and thus, simply removed the corresponding -S argument in modgjc/src/Makefile (l. 626).

after that, everything compiled and installed smoothly, and both, modgcj test as well as rhinola's hello world example worked fine.

now i am really curious where this thang will develop to... how performant is it in comparison to helma? will missing helma features be added (as javascript libs) and the result lead to helma version 2.0? or will it become the basis for a more client-focused rails-like re-incarnation? by all means exciting times for server-side javascript.

hns, 2005.08.15, 12:48

thanks. good to know it builds on ubuntu. I don't know what the prefix option to apxs is supposed to do, so i'm going to remove it.

A few notes: rhinola is not confined to apache/mod_gcj. with a few minor changes, It could run just as well inside any servlet container (e.g. jetty, mod_jk etc.). possibly, rhinola might become bigger than mod_gcj, so it will eventually start a life of its own and beyond mod_gcj. Regarding performance: basic performance is very fast. res.writer.write is currently pretty slow, the cvs version already fixes this with a native rhino res.write() function which doesn't incur introspection overhead and is about 4 times faster. For the next release, there will be native request, response and cookie objects replacing the mod_gcj object wrappers. and: I don't think rhinola will ever become helma 2.0, but it would be a nice frontend controller to helma 2.0.