17 Apr 2012 00:10
Re: PyPI mirrors are all up to date
Donald Stufft <donald.stufft <at> gmail.com>
2012-04-16 22:10:32 GMT
2012-04-16 22:10:32 GMT
On Monday, April 16, 2012 at 5:57 PM, "Martin v. Löwis" wrote:
Maybe a better checksum would be a global hash calculated differently ?Define a protocol, and I present you with an implementation thatconforms to the protocol, and still has inconsistent data, and notin a malicious manner, but due to bugs/race conditions/unexpectedevents. It's pointless. Ultimately, clients will need to verify thedata that they receive (if they suspect issues), and fall back gracefully.I can definitely see a mirroring implementation where thelast-modified field is updated at the end while some packages are notcopied over at the end for whatever network issue.That mirroring implementation would violate the principle thatlast-modified should only be updated when the mirroring run wascompleted successfully.
Is this documented anywhere? I don't see it in PEP381. I agree that
last-modified should only be updated when mirroring was completed
successfully but I don't see that specified as part of the mirroring protocol
which could lead to inconsistent definitions of what last-modified is.
Regards,Martin_______________________________________________Catalog-SIG mailing list
_______________________________________________ Catalog-SIG mailing list Catalog-SIG <at> python.org http://mail.python.org/mailman/listinfo/catalog-sig
RSS Feed