Date: Fri, 26 Mar 2004 07:56:57 +0100 From: Gabriel Ambuehl <gabriel_ambuehl@buz.ch> To: Chris Doherty <chris-freebsd@randomcamel.net> Cc: freebsd-java@freebsd.org Subject: Re[2]: WebLogic Message-ID: <94967001.20040326075657@buz.ch> In-Reply-To: <20040326060554.GZ8063@zot.electricrain.com> References: <OF8C0AA04E.34CB831F-ON86256E62.0051B4B1@transplace.com> <20040326060554.GZ8063@zot.electricrain.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello Chris, Friday, March 26, 2004, 7:05:54 AM, you wrote: > I find it fairly presumptuous and vaguely insulting to equate the use of > vendor-specific features with poor application design. I'm not sure about > the state of the art now, but not that long ago application servers like > ATG Dynamo had a wealth of features unavailable in free software that made > the loss of portability quite worthwhile. Didn't mean to insult anyone in first place (far and by large people around here are much better programmers/architects than what you usually come across) so I didn't equate it, I simply said "chances are" (which is conditional after all) and from my experience, many people jump at the replication features (especially session related) for scalability in clusters when their architecture itself doesn't really lead to scalability (using stateful session beans when they really should be using stateless ones). Caching layers are a nice concept but I've yet to see one that actually behaves like I would want it to (far and by large you can implement your own cache logic for the most critical points quickly anyhow). And I've come to the point where I think that one probably would be better off without EJB in most projects altogether (and it seems many people share my view these days). So if I stepped on anyone's toes, sorry for that. Best regards, Gabriel
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?94967001.20040326075657>