A little more about metabasep.

Recently, an interesting topic habrahabr.ru/blogs/i_am_clever/53170 appeared in Habré, in which it was suggested to backup not the data itself, but only information about the location of files + their identifiers for later finding in peering networks.
The author wrote a script to collect this information, and then, when it was needed, merged everything in DC ++ back to the hard disk.

It occurred to me that for a backup and then merging through DC it is not at all necessary to write any script, because DC ++ itself stores the “own file list” with TTH file identifiers. Ie, it is enough to backup only the filelist, then open it in DC through the "open own filelist" and put the files on download. The file list itself is stored in a convenient XML form along with TTH.

Also touched upon the problem that not all files can subsequently be found. It would be interesting to know your opinion on the use of Reed-Salomon recovery methods, when you, having scanned all the files, create a file with redundant information, for example 10% percent of the original, according to which it will be possible to recover unworn files, in our case, if their number is will not exceed 10% of the total.