Re: taxonomy for data-structure modules?

2007-05-27 Thread David Nicol

On 5/26/07, Ken Williams [EMAIL PROTECTED] wrote:


I think that's part of it.  But another problem is that many people
don't know the official names for the data structures they're
thinking of, e.g. trie or skiplist or whatever, so they can't
search very effectively for it.

It does seem like a user-applied keyword tagging scheme on CPAN (e.g.
search.cpan.org) would be very fruitful and would go a long way
toward solving a lot of the organization-related problems people
bring up on this list.

  -Ken


Does perlmonks have a wiki with a data structures section?  Wikipedia
has loads of data structures documents on it -- for instance,
http://en.wikipedia.org/wiki/Skiplist has links to SkipDB and an implementation
in C#.  Creating a norm of linking data structure modules from the wikipedia
article on the structure in question would do the trick using existing
infrastructure,
as would other proposals that do not demand tiers of new services appear from
nowhere.


Re: taxonomy for data-structure modules?

2007-05-26 Thread Ken Williams


On May 23, 2007, at 4:15 PM, Eric Wilhelm wrote:


Would a sort of taggable search scheme and/or collaborative
categorization help?  That is, implements functionality modules  
(e.g.

Net::FTP, LWP::UserAgent) seem easier to find and navigate than data
structure modules.  Maybe that has something to do with the subtle
differences in data structures and the number of possible names for  
any

given structure/variation?


I think that's part of it.  But another problem is that many people  
don't know the official names for the data structures they're  
thinking of, e.g. trie or skiplist or whatever, so they can't  
search very effectively for it.


It does seem like a user-applied keyword tagging scheme on CPAN (e.g.  
search.cpan.org) would be very fruitful and would go a long way  
toward solving a lot of the organization-related problems people  
bring up on this list.


 -Ken



taxonomy for data-structure modules?

2007-05-23 Thread Eric Wilhelm
That hour of searching-and-not-finding got me to thinking...

Is it just me, or is the collection of data-structure modules on CPAN 
somewhat difficult to discover/navigate?

Maybe the frustration is just exaggerated by the fact that it takes just 
as long to not-find a module for a simple data structure as it takes to 
not-find a module for some big chunk of functionality?  You do the 
legwork, find nothing, look at the clock and think I could have 
written this and uploaded it by now.  This is of course slightly 
optimistic or even false economy because we all know the burden of 
testing and maintenance (and yetanotheritis.)  But being empty-handed 
after an hour of searching and reading will always feel less rewarding 
than having created something.  This then contributes to even more 
yetanotheritis as we continually search through and read more 
not-quite-it modules.

Would a sort of taggable search scheme and/or collaborative
categorization help?  That is, implements functionality modules (e.g.
Net::FTP, LWP::UserAgent) seem easier to find and navigate than data
structure modules.  Maybe that has something to do with the subtle
differences in data structures and the number of possible names for any
given structure/variation?

Or, is this just yet-another-variation on searching 45k+ modules is 
hard?

--Eric
-- 
Matter will be damaged in direct proportion to its value.
--Murphy's Constant
---
http://scratchcomputing.com
---