If I may contribute a modest opinion : this whole thread started, I
believe, because someone wrote that Amazon may move away from perl and
may go in the direction of Java.
How many of us, honestly, will some day have to create or run a website
that sees even 10% of the traffic and load of the current Amazon site ?
And to the practical knowledge of this regular Amazon customer, their
site is working pretty well today, isn't it ? What I mean is that
Amazon might very well have taken a decision to move to other tools in
the future, but it does not look as if they did it because their current
one isn't working. And we also don't know if their future website will
work better.
In my humble opinion thus, if today Amazon is (still) using perl, to me
that is recommendation enough. I will happily and humbly follow in
their tracks, and hope that my website some day reaches the level of the
current Amazon website, and that I might have to think about
investigating something else. I would consider that quite an achievement.