Beefy Boxes and Bandwidth Generously Provided by pair Networks
laziness, impatience, and hubris
 
PerlMonks  

Re^5: Perlmonks 'Not logged in Error' - Proposal for fix (cpu)

by tye (Cardinal)
on Jan 21, 2013 at 17:23 UTC ( #1014500=note: print w/ replies, xml ) Need Help??


in reply to Re^4: Perlmonks 'Not logged in Error' - Proposal for fix (load)
in thread Perlmonks 'Not logged in Error' - Proposal for fix

I'm sorry that you have no clue about CPU load. But a redirect is not cost-free. It means that there are two page loads (with an attempt to make the first one have low expense, of course). That's actually a ton more CPU than the tiny addition in the middle of the existing code for rendering user content (and the total CPU for rendering user content is a rather tiny part of the server load for rendering a page).

And that's even ignoring the extra CPU extravagantly wasted by CPU-guzzling bloated American-style browsers and all of the international network infrastructure that needs to be invoked for a second request! And some of that network gear will be located in poor countries that can barely afford such luxuries!

Though I also see value in having perlmonks.org redirect to www.perlmonks.org

Still you favor the solution which costs the most cpu

Actually, I wasn't proposing only one solution and I actually had no comment at all (yet) on the cookie domain stuff.

As for external links, there isn't a single external link on the internet that will result in me visiting PerlMonks "logged in". That is for security reasons. But... nevermind.

Everybody getting their panties in a wad about CPU usage that will really be completely lost in the noise if you tried to measure it has made this whole thing fall into the category of "unfun" and so I'll go extravagantly waste my CPU somewhere else, now.

- tye        


Comment on Re^5: Perlmonks 'Not logged in Error' - Proposal for fix (cpu)
Re^6: Perlmonks 'Not logged in Error' - Proposal for fix (cpu)
by tinita (Parson) on Jan 21, 2013 at 23:19 UTC
    I'm sorry that you have no clue about CPU load.
    oh yeah, sure.
    But a redirect is not cost-free. It means that there are two page loads
    It's an apache redirect which means there is no content coming back the first time and no perl code involved. and it's only for those links missing the 'www.' - the minority. And starting with the redirects, people will post less of the "wrong" links.
    That's actually a ton more CPU than the tiny addition in the middle of the existing code for rendering user content
    this would be happening for every request, also the ones with 'www.' - the majority.
    and I actually had no comment at all (yet) on the cookie domain stuff.
    I'm really looking forward to that now.

Log In?
Username:
Password:

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://1014500]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others exploiting the Monastery: (9)
As of 2014-12-29 14:54 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    Is guessing a good strategy for surviving in the IT business?





    Results (191 votes), past polls