[java] Cool Java video on youtube

That was cool, although it's not like Java invented the idea of thread priorities/scheduling. :)

Next time I write Java, I am going to imagine that my program is a pair of giant blue hands that pushes stuff around.

It may well have soft real-time capabilities but i doubt the scheduling is deterministic and therefore i doubt it is remotely capable of hard real-time applications.

Yeah, but at least you you can keep a pendulum up :P

Pretty cool video, but the actual information content is really low. It's more of a marketing spin for Java RTS. For those who want to know what this is actually about: link. At $6500 just for the development kit, this is unfortunately not something that most of us can easily get our hands on to play with.

Quote:

Original post by lightbringer Pretty cool video, but the actual information content is really low. It's more of a marketing spin for Java RTS. For those who want to know what this is actually about: link. At $6500 just for the development kit, this is unfortunately not something that most of us can easily get our hands on to play with.

This isn't something you play with.

Writing real-time system is trivial. At least to play with.

A real-time system would need to make certain guarantees with regard to reliability and accuracy, given their system requirements, it likely is backed by hardware.

For almost all problems, a soft-real-time solution will do. But there are certain problems which require 100% timing reliability, and there you need some form of company guarantee that you have it.

$6500 isn't even that much (although it depends on how many systems it covers). And even if it's just used to get a piece of mind in some data acquisition scenario, it's not really that much money.

Contrast this price to the stocks pushed by real-time system, and suddenly it doesn't matter anymore, especially considering it's Java, and as such, reduces the cost of training somewhat (much more than $6500 for a small team).

But yes, this is one of those "serious(tm)" technologies. You don't gain anything from playing with it. You use it when you need it, exactly because you need it.

Oh, come on, that's no fun :) You could say the same about so many other things. When it's about a hobby, getting concrete results is not directly related to playing with something, which can be fun in of itself.

Obviously this is peanuts from a business perspective, although actual licensing costs would be higher since running something internally and especially on one CPU is not likely to be the goal of anyone licensing it.