On 17.07-14:16, Jason Dixon wrote:
[ ... ]
> > ... but i was also making the point that if you "don't know anything
> > about Java client-side rendering" and are "not a flash or java developer"
> > perhaps you should refrain from spouting technical jargonese on the
> > subject. i know what you're getting but i don't think it's representative
> > of java or flash.
>
> I don't have to be a Java or Flash developer to engineer scalable
> web architectures (which is what I do). My original statement was that
> there _is_ a valid purpose for flash, particularly in client-side
> graphics generation. I've seen the good and bad of graphics generation,
> particularly with stubborn clients who insist on bad technology (GD,
> ImageMagick) for server-side rendering. This is where I'm coming from,
> hopefully you understand better the point I was making. Nowhere was I
> claiming that Java would be a bad tool (yes, I know it's a language)
> because I simply don't know what classes are available for it for
> performing client-side rendering.
>
> I *will* go out on a limb and suggest that it's highly unlikely there
> are any native classes for Java that optimize client-side graphics
> rendering like Flash. Prove me wrong, perhaps I can recommend it in a
> future project. ;)
jogl? but that is actually for client side rendering and not the
image manipulation that you appear to be talking about. there are
plenty of functions in the core java library that could also do the
image manipulaton. as i am sure you are well aware the trade-off
between transmitting the full image data to the client for manipulation
or transmitting a thumbnail and performing manipulations server-side
are application specific, not client specific.
of course, we are now almost completely lost because we have mixed
the client side (which was the initial discussion) with the server
side and integration issues of the two. suffice to say, there is no
real reason that a flash client couldn't be implemented in java.
[ ... ]
> > p.p.s: i expect the server side demands you are experiencing are from
> > a jvm on the server not the actual application demands that you are
> > correlating them to (although i do understand the overlap)
>
> This has nothing to do with Java or JVMs. Graphics rendering always
> requires high amounts of CPU, regardless of the
> language/solution/platform.
... and now i'm even more confused as this appears to contradict your
previous assertion that flash was more performant. my point was that
flash is compiled code and thus does these things faster than inside
a jvm. i don't believe your server performance issues relate, in
anyway, to the client side software.
unless, of course, your point was that flash makes it easier to move
the image processing (NOT rendering) off to the client. having
implemented this a couple of times in java client i'd have to
personally disagree whilst appreciating that common opinion may vary
(i.e. flash has more focused development kit for this stuff).
--
t
t
w
