On Sat, Jul 24, 2004 at 09:06:14AM +0000, Johnny C. Lam wrote:
> Could you provide the reasoning that lead you to your opinion?

Maybe I misunderstood the patch (I don't pretend to grok most of the pkgsrc
makefile magic). If a transfer fails, while we still know the distfile is
incomplete, retrying and completing the file is ok. I don't even think we
need to check the last block of the old etc. since the sha1 hash will
prove the correct file later anyway. 

In my experience this will only succeed in a minority of the cases (aborted
transfers happen to me if my network connection drops for more than a few
seconds or the remote server goes down), but of course your milage may vary.

If we later detect a too short distfile with (of course) wrong checksum, I
think we should have the user fix it instead of magically trying to refetch.
If this is not what this change does, I apologize for my intrusion - just
ignore me.

Martin