> Personally, I think it should be supported since the https overhead
> is
> not huge and it's a good thing generally for people to be using https
> (I
> have mozilla-https-everywhere installed). And according to
> 
> http://stackoverflow.com/questions/149274/http-vs-https-performance
> 
> "Bottom line: making lots of short requests over HTTPS will be quite
> a
> bit slower than HTTP, but if you transfer a lot of data in a single
> request, the difference will be insignificant."
> 
> which is certainly the case for large ISO files.

Well zsync downloads just the differences using custom-range requests. I don't 
know whether they count as a single connection in HTTPS or as a large number of 
small data transfers.

If HTTPS is really a problem, then zsyncmake should either plainly refuse to 
create such metadata, or print a warning about it inefficiency and recommend 
using plain HTTP instead. It's very silly that zsyncmake allows to create it, 
but zsync can't then use it.

> 
> P.S. This reply is not threaded since your post didn't show up in
> Gmane.
> It's probably just slow, seems to take days sometimes to update.

I'm not sure what this means. I replied to the original test-announce@ message. 
But Zimbra, which I use, shows just one level of threading, so maybe I'm 
missing something.
-- 
test mailing list
test@lists.fedoraproject.org
To unsubscribe:
https://admin.fedoraproject.org/mailman/listinfo/test

Reply via email to