Re: java serialization alternative

Bob Lee wrote:
> You can't accuse me of not knowing serialization. I know the spec inside and
> out.
Alright, then justify your opinion with solid facts.  What is it in
Serialization that you can point to that makes it such that ti has to be
either bloated or slow?


> I'm certainly not the only one who thinks this way about serialization--take
> a poll.
Theres a whole lot of misinformation and mythology about serialization
out there. Common wisdom is seldom either.
> I'm just the only one stupid enough to argue with Jeff.
>
Naw, lots of people argue with me.  Occasionally, when its proved, I'll
even admit Im wrong.

So, go ahead and prove it.

>
>> But watching you two bicker over
>> something trite that essentially reflects the fact that each of you has
>> drunk different Kool-Ade is really quite an inefficient use of bandwidth
>> and
>>
Pardon me?  I'll also take issue with the Koolaide comment.  I've seen
Serialization used well. We USE it well in Project Darkstar and it isn't
a problem for us.

This has ntohing to do with cool-aide, this has rto do with the fact
that I spent my first 2.5 years at Sun treaching people to performance
tune their code and., whenever I saw a serializationre lated problem, it
was inevitably because they were trying to serliaze way more data thne
they really had to or wanted to.

Serialization is not intended to be a general data persistance mechanism
and  in general I recommend *against* using it as such and poitn people
at real data technologies like RDBMSs.  As it happens, the kind of
persistence we do in Darkstar maps to its capabilities well, but its a
limited domain problem.

JK

===================================
This list is hosted by DevelopMentor®  http://www.develop.com

View archives and manage your subscription(s) at http://discuss.develop.com