Note: There was some technical problem since yesterday to cypherpunks postings, so that my posts to the list didn't result in copies back to me. This morning I reposted these, again without success. An hour ago Igor Chudov informed me that the problem was solved and I have just verified this with a test message. I am therefore reposting once again one of my posts of yesterday here (not exactly identical to the original, being reconstructed from memory). It is a response to Enzo Michelangeli. Please ignore it if you have already seen it. Enzo Michelangeli wrote:

It would be helpful if the list included the approximate size in Mb of each archive. I'm making arrangements for some mirrors here in Hong Kong, and I need an estimate of the space needed.

I don't know. But a number of years ago I accessed some archive sites of software for processing of Asian languages. From that and the fact that modern disks have become much bigger in capacity and cheaper, I suppose that storage shouldn't be a bottleneck. I like to take the opportunity to mention a non-trivial issue that happens to have refreshed my attention by a reader's letter in a computer journal, namely that software at public archives could be infected by virus. I suppose one should give this potential risk some consideration, since in the present context there could be some third party particularly interested in disrupting the goals of crypto archives. Another point I like to raise is what to archive, i.e. whether only program texts (which are small) or also executable files, and how to let the archive contents be up-to-date. Related to this is whether the archive owner has to fetch everything himself or there is also a directory for uploading by contributors (this is the case with the archives I mentioned above and is sensible for new software and for updates.) M. K. Shen