I'm not sure how GUEST 14 May 06 - 05:06 PM related to other guests, but that post contains some points that are worth discussing in detail, and I think have relevance to other comments in this thread. For what it is worth, I am been involved in developing computer systems for more years that I care to recall and my current and last jobs have both involved developing some fairly hairy multi-platform, distributed web applications. That is not to claim my thoughts have any special merit; more that I've made a goodly share of wrong decisions in my time.

So let's take a few of the points. You can download and burn to CD damn near anything on the internet.. Yes, you certainly can burn a single item. Burning inter-related items so that the relations are maintained is much more difficult and not a task for the faint-hearted. Sure, there are tools that can help, but don't just assume they can sort it all out. A trivial example: on Unix platforms, the case of filenames is significant, on Windows it is not. Unless you have played that game, you would be surprised how many times that is at the root of real-world problems. Similarly, the set of permitted characters in filenames differs - the most obvious being the forward-slash, backward-slash issue, but there are plenty more. And another problem is deciding what exactly gets 'burnt'. Do you burn everything that is within 1 link of the page you are dealing with? Two? Have some interactive dialogue saying I want this subtree not not that? It's much more tricky than it sounds.

Then there is the question of how the 'next-generation' DT is to be used. For historical reasons, the DT evolved as a stand-alone system. It was, and is, fully usable on a laptop away from any network or wireless hotspots. To build a Wiki-like system where everything can link to MP3s all over the net is great, but it comes with the need for either a good network link - and don't forget there are still a lot of 56Kbit modem users around - or with a very well considered way in which it can 'degrade' when not connected to the net. I'm not saying one requirement is superior to the other, necessarily, but you need to weigh up the pros and cons before you leap into it.

Then there is the software used: If you go to the bottom of the page & click on About Folktunes it shows the tech support used to create their new site. A lot of the problems with the DT come from the decision to use AskSAM. That was probably the best decision at the time, but over the years has come back to roost. How do you ensure that whatever decisions you make about the technology minimise these issues in the future? There's much more to it than simply saying use open source. I've spend about 15 minutes browsing the example site and it makes use of at least five different emerging standards. In all probability, at least one of these will fail by the wayside in a year or two. What happens then? It's one thing to say, as Jon did, that he was considering Java and HSQL: that's a closed and controllable set of constraints. It's another when you start throwing all sorts of things like WebJay into the equation. Nothing wrong with WebJay as such, but is that going to be the dominant solution in a few years time? Do you remember Veronica and Gopher? Or how Netscape was going to make push technology a dominant part of the web experience?

Next contributions: Not so complicated, now is it? You don't need programming knowledge to contribute, or at least not much.. This is the greatest strength of a web-based approach. ANY system that needs a manual intervention to get from a contributer providing a piece of information to letting others see it will inevitably lead to delays because it takes time and effort for the people concerned to move things across. That's the real world for the DT as it currently exists. And its also the same for moderated sites like some the BBC message boards, where they have people employed full time to keep the transfers going, but want to monitor each post for legal reasons.

Trust of strangers, ... , of the competence and integrity of those who wish to contribute... (I left out 'of the unknown' because I think that's a different issue) This again goes back to what the DT actually IS. It has never been a reliable source of lyrics and tunes or history in the sense that an academic would mean that term. So I don't think there is too great a risk in giving a more immediate facility to update the database, were this possible.

But the biggest danger, in a way, comes from the idea of multiple copies. As the DT stands, we have essentially major releases, the last being Spring 2002. That's an age, I know - as do Susan and Dick. But if you have lots of copies all being maintained and updated separately you will not necessarily have something better; what you will have is a number of disparate sources that grow ever further apart. Again, you may feel that the benefit outways the cost. All I say is think carefully before you choose.