Re: [SQU] refresh_pattern, aka dealing with an arrogant consultant

Does this affect in any way the caching of pages without explicit
'content-length'?? We would like to cache these (and control their
expiration like other pages) but our squid refuses to cache them...

------Original Message------
From: Bert Driehuis <bert_driehuis@nl.compuware.com>
To: raarts@office.netland.nl
Sent: October 13, 2000 10:19:52 AM GMT
Subject: Re: [SQU] refresh_pattern, aka dealing with an arrogant consultant

Hi Ron,

> I have a problem with a consultant accusing me of misconfiguring squid.

The world is full of assholes. I'd suggest talking to his manager, and
if there is no resolution, consider moving to a different company. I've
been in this situation before (with the same issue, even), and have had
my authority on this issue reconfirmed by higher up management.

> I told him this should be solved on the webserver side (they made this
site
> themselves), and send them a macromedia advisory.

And rightfully so. There is nothing wrong with trying to cater to the
bugs out there in the real world, but users who do not accept the fact
that the underlying problem is not yours are totally unacceptable.

> He tells me the following:
>
> (I censored the part where he was bragging about have been working
> at a lot of fortune 500 companies, and where he was trying to make fun of
the
> fact that we put three hours of work in it: "this took you THREE hours?")

Snicker... This part of the response is the only part I'd discuss with
his manager. It sort of brings home his attitude. I appreciate users who
bring their concerns to me and are willing to discuss things, but this
kind of attitude will eventually bring your company down. If his manager
doesn't see that, it's time to jump ship.

> It's a fact that many webdesigners do not handle dynamic pages correctly
> so every proxy has an option to always check if a new page is available
> before serving the page from the cache. It generates more traffic, and
> slows things down, but everybody uses this. MSproxy has that, Winproxy
has that.
> You configured squid, it's a proxy misconfiguration, so we won't pay you
for
> repairing this.

> I originally did not put refresh_pattern in the squid.conf because I
thought
> the default would suffice, now I did.
>
> But, is he right about proxy servers having this option, and about the
fact
> that everybody has this switched on?

Well... Squid has the option (which is switched on by default) to cache
content without explicit cacheability information. You can tune this
with refresh_pattern, as you found out.

Other proxies may not have this feature or have it switched off by
default. The refresh_pattern is not very important for saving bandwidth;
it's impact comes mostly from reducing latency.

Seeing how I hate Flash, I wouldn't mind sabotaging it by setting the
refresh_pattern to always check: he gets what he wants but it may or may
not be usable, and good luck to him.

> Is this the reload_into_ims option? The docs say 'Enabling this feature
could
> make you liable for problems which it causes.'. What problems are meant
here?

If you enable reload_into_ims, a "Pragma: no-cache" gets translated into
an if-modified-since check. That wise ass consultant thinks he is
sending a reload, but he is only reloading the HTML page that frames the
..SWF file. If you don't have reload_into_ims set, the clients "pragma:
no-cache" is honored and the object is refetched irregardless of its
freshness, so obviously he is not sending a "pragma: no-cache".

So, reload_into_ims makes things worse by disallowing a true reload (not
that I care -- if the web server is so misconfigured as to fail an IMS
check I don't care about it anyway).

Good luck in dealing with this issue. I know how you feel.

Cheers,

-- Bert

--
Bert Driehuis, MIS -- bert_driehuis@nl.compuware.com -- +31-20-3116119
Dihydrogen Monoxide kills! Join the campaign at http://www.dhmo.org/
--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html
______________________________________________
FREE Personalized Email at Mail.com
Sign up at http://www.mail.com/?sr=signup
--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html