Beefy Boxes and Bandwidth Generously Provided by pair Networks
Come for the quick hacks, stay for the epiphanies.
 
PerlMonks  

Re^4: Major update to perldoc.perl.org (functional)

by jdporter (Paladin)
on Jul 17, 2009 at 17:56 UTC ( [id://781129]=note: print w/replies, xml ) Need Help??


in reply to Re^3: Major update to perldoc.perl.org (functional)
in thread Major update to perldoc.perl.org

so I'll just use google w/ site:perldoc.perl.org when I want to find documents that mention "all of the following words".

Maybe we should change perldoc:// to use google. Then one can easily search for all these words "and this multi-word expression".

Between the mind which plans and the hands which build, there must be a mediator... and this mediator must be the heart.

Replies are listed 'Best First'.
Re^5: Major update to perldoc.perl.org (perldoc://)
by tye (Sage) on Jul 18, 2009 at 18:32 UTC

    I finally grok this suggestion and I think it is an excellent one. To be clear, perldoc:// is currently useless (and has always been) if javascript is disabled (for perldoc.perl.org) and having perldoc:// run a google search would be better.

    Update: perldoc:// now links to google searches: glob, if (doc:// still links to a "best guess": glob, if).

    - tye        

      I think
      http://www.google.com/search?q=...&sitesearch=perldoc.perl.org
      would be slightly better than
      http://www.google.com/search?q=...%20site%3Aperldoc.perl.org
      since it makes it slightly easier to refine or change the query.

        Thank you, that is a good suggestion. I am slightly concerned because "sitesearch" is the name of a "paid" service from google so I'm unsure of the ramifications of PM making use of that parameter. Given the google search form that perldoc.perl.org presents (as part of the "search results", at least in cases where there isn't an "obvious best match"), it might be that somebody has paid google for sitesearch against perldoc.perl.org (but I don't want to jump to that conclusion).

        But in either case, there are still portential risks associated with PM using that feature.

        If the feature has been paid for, there is a chance (though I consider it only a small chance) that PM could generate enough traffic against the feature that the price for the feature is pushed up due to volume, which wouldn't be fair to whoever is paying for it, unless they first give permission to PM (if this were to actually be a problem, I would expect that much of the volume of traffic would come from search engine spiders, since they are apt to follow any perldoc:// links over and over -- perhaps using sitesearch via GET links is a "no no" in itself for just this reason, since the use of sitesearch at perldoc.perl.org is via a FORM, something that a search engine spider won't navigate through).

        I also worry that the use of this feature is a "grey area" that may eventually rise to the point that google takes issue with it. Testing shows that sitesearch=perlmonks.org "works" but I'm pretty sure nobody is paying google for the "sitesearch" feature with perlmonks.org.

        Before I applied the changes to perldoc:// to use google, I searched for and scanned google "terms of service", hoping for hints about whether such use was strictly allowed, had certain restrictions, etc. I didn't find anything very informative, just a lack of any mention by google about such practices and some information about google offering a paid service called "sitesearch".

        It seems that "sitesearch", the paid feature, allows you to have the search results displayed "on your site" with your layout and without google's ads (but the documentation I found on sitesearch wasn't very explicit about this and certainly didn't say anything about "compared to X, which you can just use for free"). So there may be little in common between the paid feature and the sitesearch= parameter other than how they are spelled.

        So, my best guess is that having PM use sitesearch= in a ton of links probably isn't going to be a problem. But I would appreciate some confirmation of that impression (or even confirmation that it likely would be a problem) before moving forward with such a change.

        - tye        

Log In?
Username:
Password:

What's my password?
Create A New User
Domain Nodelet?
Node Status?
node history
Node Type: note [id://781129]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this?Last hourOther CB clients
Other Users?
Others wandering the Monastery: (7)
As of 2024-04-16 09:12 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    No recent polls found