On Sun, 10 Apr 2011, Erik Christiansen wrote:
On Sat, Apr 09, 2011 at 02:24:23PM +0200, Christian Brabandt wrote:
On Fr, 08 Apr 2011, donothing successfully wrote:
A user may well assume the site is down if the url doesn't resolve.
Seconded. This had bugged me many many times with various browsers.
"Thirded".
"Fourthed".
In response to the domain manager's list-quoted response from the DNS
adminstrator in the thread referenced by John:
Yes. This is by design. Its a bad habit to omit the "www." part if you
intend to talk to a webserver. Maybe call me an old fart, but you are
connecting to a webserver, not a domain, and the URL you use should
reflect that.
You're an old fart.
Kidding aside, both the 'http://' and 'www.' in 'http://www.vim.org/'
are redundant when typed for the intent of navigating to a website. It
initially made me really mad when Chrome started hiding the
'http://'[1], but times change, and now I appreciate the less visual
clutter. Redundant is redundant. Let the user save a few keystrokes.
Particularly, if after trying all those http applications, the user
decides to have a casual look at DNS:
$ dig vim.org
...
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0
^^^^^^^^^
and sure enough, there's no "ANSWER SECTION:"
You're ignoring the other problem, though: what should the A record for
vim.org. be?
www.vim.org. is a CNAME to vhost.sourceforge.net.
The bare domain *can't* be a CNAME[2]. And there's no canonical IP
address returned for vhost.sourceforge.net. As a large site, they
geotarget their DNS responses, the way a CDN does. And they also use IP
lists for load balancing.
Compare the result for google.com, where there are five entries in the
ANSWER SECTION.
As an even larger site with an even more complex delivery
infrastructure, google.com. resolves in an even more complicated way
than vhost.sourceforge.net. The six IP address responses from my home
PC (all under 74.125.113.0/24) don't even match the six returned from my
neighbor's wireless router (all under 74.125.93.0/24) less than 20 feet
from my house. Nor do they match the six returned to my web hosting
provider (all under 209.85.225.0/24) located less than a few miles from
my house.
It seems decidedly dufus for vim.org not to resolve.
Agreed, but it's still not clear what the correct alternative is. The
simplest setup would probably be an Apache box somewhere with
mod_rewrite set up to redirect every http://vim.org(rest-of-url) to
http://www.vim.org(rest-of-url):
RewriteEngine On
RewriteCond %{HTTP_HOST} =vim.org
RewriteRule - http://www.vim.org${REQUEST_URI} [L,R=permanent]
This:
$ whois vim.org
gives telephone and email contacts to ask for proper nameserver set up
for vim.org.
Those contacts don't match "Stefan `Sec` Zehl", whose off-list response
in the role of DNS administrator was quoted in the other thread. Sven
Guckes, listed in the WHOIS info is a regular here, so I won't put words
in his mouth. But pushing for telephone/email contact for this seems
excessive.
It's a registered domain (at least until 14th september)
(at which point it will undoubtedly auto-renew)
, so it's only crappy nameserver config which prevents it resolving.
[See above] I agree that vim.org should work, too. But, it's not clear
the easiest setup to allow that.
--
Best,
Ben
[1] http://code.google.com/p/chromium/issues/detail?id=41467
[2] http://tools.ietf.org/html/rfc1034#section-3.6.2 RFC 1034 §3.6.2 ¶3:
If a CNAME RR is present at a node, no other data should be present
[and vim.org. *must* have other data, e.g. SOA and NS records.]
--
You received this message from the "vim_use" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
