On 2011-10-22 13:22 , Daniel J. Luke wrote:
> On Oct 21, 2011, at 8:39 PM, Ryan Schmidt wrote:
>> Does MacPorts output any debug/info when redirects happen? If so, then yes, 
>> that would be a possibility. But we don't want to fire off yet another email 
>> to the committer/maintainer. It would be nice if it were part of the lint 
>> mail.
> 
> I don't think it's a big deal to have committers/maintainers get a different 
> email (only if there's an issue)... We all probably deal with enough volume 
> that it doesn't really matter.
> 
>> Not all http redirects are necessarily a bad thing or an indication of a 
>> problem. The SourceForge redirects are particularly wrong though because 
>> they negate the benefit of our list of SourceForge mirror servers. We ping 
>> them and try to direct users to a nearby server, but if all it's going to do 
>> is redirect the user back to the main download site because the URL was 
>> wrong, then it's a waste of time. We decided to keep our list of sf mirrors, 
>> so we should make sure they're getting used properly by rewriting sf 
>> master_sites that do redirects. My earlier suggestion about just checking 
>> the syntax of sourceforge master_sites should suffice for this.
> 
> 
> We actually should probably just let the sourceforge redirect do it's 'magic' 
> and fallback to pinging sites and selecting a mirror from the list only if 
> there's a problem with the main site (it's very likely that it can do a 
> better job than we can of directing people to an appropriate server).

I find it does a considerably worse job. I'll often get a mirror on the
other side of the country and relatively poor download speeds when going
through the redirector, whereas pinging selects a local server and I get
3-4 times the throughput.

- Josh
_______________________________________________
macports-dev mailing list
[email protected]
http://lists.macosforge.org/mailman/listinfo.cgi/macports-dev

Reply via email to