[OSM-talk] Why isn't any XAPI server available ?

[Frederik Ramm]

Hi,

David Murn wrote:
> Am I missing something here...?  People are complaining about how bogged
> down and slow the current service is, so its being re-written in java?
> Is there any language slower or more resource intensive than java?

We have no performance comparisons between the old XAPI which was 
written in GT.M (formerly MUMPS), and the new XAPI written in Java. It 
is very well possible that the new version is actually slower than the 
old one on comparable hardware - we simply don't know.

The problem with the old version was that there is practically only one 
person in the whole OSM universe who can work with it (80n) because the 
programming language is so unusual, and everyone else would need quite 
some time to reach 80n's level of proficiency. Which has led people to 
sit back and complain (or politely request features) instead of having a 
go themselves.

It's not that 80n has stopped working on XAPI or refused to cooperate 
but one man can only do so much. (Indeed the old XAPI has, often in 
response to user requests, received a few really innovative features, 
like finding out which versions of nodes match a historic version of a 
way, or listing all past contributors for an object - stuff that the new 
XAPI doesn't even support.)

The hope is that by implementing XAPI in a technology that more people 
are familiar with than with GT.M, more people can be encouraged to help 
XAPI development, and also more people can simply set up their own local 
XAPI, perhaps with a filtered data set, to satisfy their needs.

> If the service isnt designed to be portable (it only runs on one system
> currently, in the world),

There used to be three XAPI instances, so an effort at distribution has 
been made already and will likely be made in the future.

> then who cares about java, why isnt it written
> in optimized C or some other similarly lowish level language, rather
> than java?

I share your anti-Java sentiment (and would add to it an anti-database 
sentiment). However, in an Open Source project it is often the 
availability of programmers, and not the availability of technologies, 
that dictates how something is done. What good is knowing that 
everything would be faster in C if we don't have anyone to write it?

Plus, there would be some danger to again maneouvre us into a niche 
where nobody but the original programmer dares to touch the code.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"