Subject:
|
Re: "jump.cgi" considered harmful ? (1)
|
Newsgroups:
|
lugnet.off-topic.geek
|
Date:
|
Sat, 25 Nov 2000 16:12:03 GMT
|
Viewed:
|
504 times
|
| |
|
|
In lugnet.admin.general, Larry Pieniazek writes:
> In lugnet.admin.general, Mike Stanley writes:
> > Last time I looked into this (and I think posted about it) it seemed fairly
> > obvious the biggest delays were had by IE users. I recall the delay being
> > minimal when using Nutscrape. That doesn't help me (or the overwhelming
> > majority of the browsing public) though, because I'm not about to switch to
> > Navigator or Mozilla or whatever its called this year just to read LUGNET
> > posts.
> >
> > I don't care if its IE's fault, or this script's fault, or anyone else's
> > fault - the delay is real and it is incredibly annoying.
>
> It seems that slow DNS is the culprit if Todd is right.
Well, not quite...I'm not ready to posit a cause-effect chain. I was merely
trying to say that I couldn't fathom how the script itself could add anything
more than a small fraction of a second. Slow DNS is one possible answer, but
if it happens to MSIE users and not to NN users, then it's more likely (IMHO)
that it's a bugaboo of some kind in MSIE.
> Why does cutting and
> pasting make it go faster? If Todd is right, you would have to pay the same
> DNS price either way. Or is it just a perceived speedup?
Well, Mike said that when he cuts and pastes, he removes the prefix -- which
means he only goes through one DNS lookup (the second one -- the only "real"
one; the first one (doubly moot if the prefix is stripped) should be cached
anyway).
> So what can be done about making IE do DNS faster? (other than changing to a
> different product) Or at least changing the timeout before it gives up. I'd
> be happy to wait it out, the error screen is the biggest part of the
> annoyance. I usually run with many windows so I just flip away and come
> back, I'd never notice the delay.
>
> I have just (again) looked at all the options I could think to look at and I
> didn't see anything that controls the timeout delay. Win 2000, IE5.5
Let's try an experiment. Let's encode the URL that gets passed to jump.cgi
two different ways: one with an unencoded colon and one with an encoded
colon. The two URL versions appear here on a test page:
http://www.lugnet.com/test/jump/colontest.html
--Todd
|
|
Message has 5 Replies:
 | | Re: "jump.cgi" considered harmful ? (1)
|
| (...) exact URL and hadn't flushed any caches. Don't forget, though, that this may not be the best test destination, since www.yahoo.com isn't a single DNS location, it's many, due to that technology whose name I can't remember... Akatomi? Doing a (...) (24 years ago, 25-Nov-00, to lugnet.off-topic.geek)
| | | Re: "jump.cgi" considered harmful ? (1)
|
| Is it possible there's any firewall issues? I know at work I regularly see situations where it takes ages for a page to come up, during which time, I can't do anything else in Netscape because the whole thing is locked up (it won't even re-paint (...) (24 years ago, 27-Nov-00, to lugnet.off-topic.geek)
| | | Re: "jump.cgi" considered harmful ? (1)
|
| (...) at a solid 1.2mb over my ADSL connection, with most websites popping up near instantly, the way they used to at the office before Napster killed our DS-3. Win98 SE IE 5.5 Colon about 4 seconds to get to the bricksmiths page encoded colon about (...) (24 years ago, 30-Nov-00, to lugnet.off-topic.geek)
|
Message is in Reply To:
36 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
