Return-Path: Delivered-To: apmail-mirrors-archive@apache.org Received: (qmail 40176 invoked by uid 500); 6 Jun 2001 07:48:36 -0000 Mailing-List: contact mirrors-help@apache.org; run by ezmlm Precedence: bulk Reply-To: mirrors@apache.org list-help: list-unsubscribe: list-post: Delivered-To: mailing list mirrors@apache.org Received: (qmail 40146 invoked from network); 6 Jun 2001 07:48:30 -0000 Date: Wed, 6 Jun 2001 17:48:33 +1000 (EST) From: jason andrade To: mirrors@apache.org Subject: Re: Apache mirror system - comments In-Reply-To: <3B1DB919.1FEECBEA@cc.kagu.sut.ac.jp> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Scanned-By: MIMEDefang 1.0 (http://www.roaringpenguin.com/mimedefang/) X-Spam-Rating: h31.sny.collab.net 1.6.2 0/1000/N On Wed, 6 Jun 2001, Devendra Narayan wrote: > I request you to look at it with more consideration and a historical perspective. > There probably are many sites ( including ours ) which started mirroring NCSA htttpd > and later Apache many years ago ( Apache since it's inception ). Now, these sites > have been 'official' mirror sites for many years. well, you did ask for comments. i'm just voicing my views - in particular based on what i think might actually produce a workable scaleable mirror system for a project. it's not gospel by any means. > In the meantime, however, Apache group has grown to develop a lot of > things besides the httpd server ( XML, Jakarta etc ). Just because the previously > 'official' mirror sites failed to mirror these new projects shouldn't be a reason to > 'demote' them from their official status. There could be many reasons why they are > not mirroring the new projects - lack of disk space, user interest etc. i respectfully disagree with this. i would love to have a situation where historical perspective is a factor (i'd qualify for a few archives on that basis..) but the reality is based more around having a mirror system that works - i.e users can always find things they're looking for at the mirror (e.g from your previous comments about having all the links relative etc), and that it requires minimal work to maintain on the part of the software project. the apache project at this stage is approximately 600M of disk space. if you think lack of diskspace is an issue, try and deal with the freebsd archive (50G ?) or redhat (80G ? 100G?). i would happily allocate 2G of disk to the apache project for 12 months if that guaranteed me a mirroring system that was consistent and therefore well used. i appreciate you have disk shortages (i have disk shortages myself!) but given the above, some perspective is needed. > Yes, if the facilities that they have permit, they should mirror the whole thing > but, those that have supported the distribution of the Apache httpd for so long, shouldn't > be punished for not being able to keep up with the rapid expansion of the Apache group's > projects. i don't wish to denigrate anyone's archive or efforts - please don't see it as such. i'd just like a workable mirror system. tomorrow would be great, thanks :-) IMHO, it won't work to have say 250 sites mirroring apache, with varying levels of up to dateness and completeness, i'd rather see 50 sites each with a complete archive. that doesn't stop the other 200 sites mirroring it and making it available to users. once again, that's just my opinion. regards, -jason