Re: XML databases [ was: S.O.D.A. database Query API - call for comments ]

Philip Lijnzaad wrote:
[Query a file without reading it completely]
> Carl> No, not impossible, if indices are included where needed.
>
> but then you'd have to have standard for indices as well,

True.

> and still you can't query complex things untill you
> have read all of the index(es).

All the relevant indices. Also true.
Still this might only be a very small part of a file.

> Carl> I have been researching CDR a year ago and I found it quite amusing.
>
> as in: 'useless' ?

No, not that bad, just strange, if you reed my comments. It was useless for us.

> Carl> An unofficial remainder of a document is still left over here:
> Carl> http://www.db4o.com/db4o/yap.html#Why Some links might be dead.
>
> I couldn't make much of this. Have you guys investigated ASN.1 ?

No, I have never heard of it. Thanks for the pointer. We won't change our format now though. It works quite well performant and stable. We have uncomparatively worse troubles in other areas.

> Carl> CDR includes the necessity to specify the byte order over-and-over
 again
> Carl> with every little bit of data.
>
> one byte per request, I believe.

One byte per variable, if I remember this right.

> The reason not to standardize the endianness
> is so you don't have to do byte-swapping if the machines on either end
> already have the same (I think this is a bit silly, though, with i/o
> practically always being network-bound). Cheers,

...well I also thought it was silly.

Cheers,
Carl Received on Sun Jul 22 2001 - 01:24:47 CEST