Bowerbird wrote:
jon said:
Most of the PNGs are in the 70-80k range
yes, that was my mistake, sorry. i was misled by the .djvu version, where most of the pages are <10k. (i've _got_ to see how to use that!)
DjVu is cool, but the "openness" and long-term viability of the format is still open to question. The Internet Archive uses it, so they must feel it is open enough to use. There's a dearth of free tools using DjVu, but from what I understand there's no impediment to open source DjVu compile tools.
distributed proofreaders keeps _saying_ that they are going to, but as of yet, they haven't done it.
Yes, if this is the case, it is mysterious since IA will gladly host them once the etext version is out the door. I do know that some scan sets are encumbered (they are "loaned" to DP under some sort of arrangement, but cannot be made public -- this is somewhat troubling, but hopefully the scans will be made available elsewhere at a future time, such as through IA's scanning activities. One thing I do know is that DP does keep full source metadata for each text they produce, even if that data is not turned over to PG.)
d.p. still seems to be constrained by disk-space, even with generous help from ibiblio and internet archive.
I know that for production purposes they want to use their own servers -- IA is not reliable enough. IA's focus is on archiving and storing, so 24-7 with full-throttle availability is a lower priority to IA, while DP *must* have 24-7 availability and sufficient speed to not keep volunteers waiting. Thus, disk space is an issue for the DP production process, especially in that DP is still a shoestring operation. Anyway, this is my interpretation of what Juliet told me a few months ago. Maybe someone from DP will reply to this... Jon