Re: [xml-dev] XML aggregation question?

[peter murray-rust <pm286@xxxxxxxxx>]

At 02:22 27/08/2006, Robert Koberg wrote:


><winning-friends-and-influencing-people>
>A few years ago I really wanted to use an XML DB for a project. I 
>had narrowed it down to Tamino and Sleepycat's XMLDB. Tamino folk 
>were willing to give it away for something close to free (for 
>now...). Sleepycat's XMLDB offering was supposed to be open source 
>and free if you don't distribute. Since we were going to deploy on a 
>webserver (for a non-profit no less) and did not consider our webapp 
>as something that was going to distributed we started working with 
>the XMLDB product. After spending a few weeks working with it I 
>thought it would be a good idea to clarify with sleepycat. To my 
>surprise I would need to pay US$50k to license their product even if 
>I simply downloaded on the server machine (I would be developing 
>elsewhere and that is what constitutes distribution... now...). 
>That, you see was a deployment in their eyes. I argued. Their claim 
>probably would not hold up in court according to the existing 
>at-the-time GPL license (look at the main driver in the new GPL - it 
>was a sleepycat guy...).

I also looked at Tamino some years ago - about 6/7 I think. At that 
stage it seemed about all that was available without paying money. At 
that stage I wasn't "employed" so I couldn't even think of any 
cost-money system. I think things have moved on since then - not as 
much as I would have liked. That is how I got burnt with Xindice (if 
anyone that it is not dead it would be useful to know).

I don't understand the sleepycat story. The license should make it 
clear whether or not it is Open source. If it is, then there should 
be no cost and no restrictions other than the license. There *are* OS 
distributions which make it hard to use unless you pay someone and 
these include:
* no documentation; pay to know how to use it
* engine is open source but you benefit from a closed GUI


>Why everything NEEDS to be open is beyond me.

It doesn't. There need to be 2-3 systems which *are* open source. I'm 
an academic and I can only get my software used and developed by 
others if it's open. That means the components have to be open. I can 
put them all on a disk or server or whatever and others can clone the 
system. I need a Open, distributable,  XML DB. It doesn't have to be 
rocket science - it's not mission critical - it doesn't have to 
provide training courses and it will need work to develop the 
applications. If at any stage I need a gold-plated system I' will 
consider commercial offerings.

>(Why I would want to see another Open Source CMS is kind of like 
>needing to see another lawn mowing pattern.)

Unless we hear otherwise there are only two Open candidates - eXist and XML-DB.

>I did not want to deal with it and hired a RDB consultant...
>
>In other words, if you work on your laptop with a GPL product for 
>'some thing' with the latest license and you push that 'some thing' 
>to a server, you might be in trouble.

This is why we need a license which makes it clear that we can do this.

>...skipping forward to recent xml-dev thread ...
>
>OK... so I tried eXist XMLDB recently after some posts/comments to 
>this list. I did not go with it even though I had/have high hopes 
>for the product. I really hope it can be something worthwhile. When 
>trying it though, there always seemed to be some bug the popped up 
>that would 'always be fixed in the next subvesrion commit/update'. 
>That's great, but not something to base a current deliverable on. I 
>have been watching the list for eXist and it does not seem to change 
>- there is always something...
></winning-friends-and-influencing-people>

My impression is that eXist has had quite a lot of work recently and 
is continuing to improve.

I'll be interested to see a critical comparison of the systems here.

P.



Peter Murray-Rust
Unilever Centre for Molecular Sciences Informatics
University of Cambridge,
Lensfield Road,  Cambridge CB2 1EW, UK
+44-1223-763069