I agree with everybody :-) However I also think it is important to get something up quickly. Is it possible to get some simple server going. Whether it be based on PG USA PG EU, Google, I don't think it matters. Something with very little work and quick. Then at the same time we can work on putting together the all singing, all dancing machine which will cook your breakfast in the morning as well. Let's continue to debate what the ideal final solution is, but at the same time lets start a new thread that discusses our options for a quick solution as too. Andrew Sly wrote:
On Tue, 9 Nov 2004, James Linden wrote:
Regarding backend: I'm very opposed to basing PG Canada on the current PG system, which does not meet even minimal requirements for cataloging and does not rely on a master XML document.
This is the issue that draws my attention the most. Actually, I believe it's a brief summation of three areas: 1)cataloging information 2)file format 3)directory structure All of these are proposed by James to be done differently from what has been done at PG. This is a valid choice, however, it could be quite an uphill struggle to work out enough bugs to have them at a usable point.
If my ideas are too much for the PG Canada group that is coming together here, I'll step aside and not get in the way. I'd rather learn the appropriate lessons from PG's experiences, and do something better, which, I must point out, it not hard.
This is somewhere I am unsure where to draw the dividing line. At one extreme, we could just barge ahead and start posting whatever files people want to submit, working things out as we go. At the other extreme we could try to work out a detailed plan for how to deal with various exceptions that will come up in the areas mentioned above before going any further.
The first option has the benefit of drawing in volunteers. PG's growth can be attributed to its historical openness to accepting files prepared in different ways. A drawback is that it leads to many inconsistencies in the files. The second option has the benefit of being very organized, which can end up making the collection easier to use, search, etc. The drawback is that working out the details, without basing them on an already working model, would be very time consuming.
So is there some middle ground?
Andrew
_______________________________________________ Project Gutenberg of Canada Website: http://www.projectgutenberg.ca/ List: pgcanada@lists.pglaf.org Archives: http://lists.pglaf.org/private.cgi/pgcanada/