* Mihai Popescu <mihai...@gmail.com> [2011-02-07 21:26]: > It looks like your answer is the most valuable for me. So, by my > understanding, the developer compiles and gets the *.tgz files, > uploads them to the ftp server but he doesn't compile the sha256 > checksum each times for those files. I don't know why, because i'm not > familiar with the entire distribution process, maybe they take to much > time. But from time to time, the developer makes it somehow that > SHA256 has the good checksums inside.
geez, how hard can it be. if you read the damn release manpage once and ran that process just ONCE you'd knew that the sha256 file is created as part of it. it is always "good". minus X, since that is built seperately, usually not even by the same person. you have been pointed to the reason for the mismatches more than once, how hard can it be... i try it one last time, and please pplz, let this useless thread die then. 1) sometimes a new kernel is thrown in, since we want something tested quickly 2) I repeat ONCE AGAIN: the mirroring process is not atomic. think about it for a second and it'll be obvious. ok, that doesn't work apparently, so super duper obvious: INSTALL.*, SHA256, base*, bsd* are already synced. cdXZ.iso is in the process of syncing and the rest is old, as in, from a previous snap. do i need to point out now that the checksums in SHA256 won't match for cd49.iso and everything after? and the same for the checksums in bsd.rd? -- Henning Brauer, h...@bsws.de, henn...@openbsd.org BS Web Services, http://bsws.de Full-Service ISP - Secure Hosting, Mail and DNS Services Dedicated Servers, Rootservers, Application Hosting
