RE: [forwarded email: cygwin mirrors and installation]

2003-10-07 Thread Robert Collins
On Tue, 2003-10-07 at 13:02, Gary R. Van Sickle wrote: You'll have to forgive my ignorance here, but is there some way to query ftp/http servers for how much load they're under? Not as a standard. Thinking out loud, what we roughly want is: -users can choose any mirror explicitly. -on the

Re: [forwarded email: cygwin mirrors and installation]

2003-10-07 Thread Ronald Landheer-Cieslak
On Tue, Oct 07, 2003 at 05:35:24PM +1000, Robert Collins wrote: On Tue, 2003-10-07 at 13:02, Gary R. Van Sickle wrote: You'll have to forgive my ignorance here, but is there some way to query ftp/http servers for how much load they're under? Not as a standard. Thinking out loud, what

[forwarded email: cygwin mirrors and installation]

2003-10-06 Thread Christopher Faylor
- Forwarded message from Doug Porter - From: Doug Porter To: sourcemaster Subject: Cygwin mirrors and installation Date: Mon, 06 Oct 2003 15:17:08 -0500 Hi, I run archive.progeny.com, one of the Cygwin mirror sites. Our Cygwin mirror is heavily used; averaging around 140 GB per day. I

Re: [forwarded email: cygwin mirrors and installation]

2003-10-06 Thread Igor Pechtchanski
On Mon, 6 Oct 2003, Doug Porter wrote: Hi, I run archive.progeny.com, one of the Cygwin mirror sites. Our Cygwin mirror is heavily used; averaging around 140 GB per day. I was amazed by the demand for Cygwin, at least until I saw how mirrors were presented in the Cygwin installer. They

Re: [forwarded email: cygwin mirrors and installation]

2003-10-06 Thread Christopher Faylor
On Mon, Oct 06, 2003 at 09:18:54PM -0400, Igor Pechtchanski wrote: On Mon, 6 Oct 2003, Doug Porter wrote: I run archive.progeny.com, one of the Cygwin mirror sites. Our Cygwin mirror is heavily used; averaging around 140 GB per day. I was amazed by the demand for Cygwin, at least until I saw how

RE: [forwarded email: cygwin mirrors and installation]

2003-10-06 Thread Gary R. Van Sickle
Hmm, I'd guess this puts the mirror list prioritization by response time rather high on the TODO list... I never considered it before but that would tend to cause the mirrors with the best response time to be chosen almost exclusively, wouldn't it? That might end up having the opposite