Dougal Stanton wrote:

On 22/11/2007, Richard Kelsall wrote:

...

Did I do something wrong when searching haskell.org?

You didn't use Google first? ;-)

Seriously though, using the search box at haskell.org seems to be a dead loss. I'm sure this has come up in the past.

In general, I find *most* search functions to be fairly unhelpful. Google is the shining exception to this rule; it almost always seems to figure out what you're after. I guess doing text searching is just a fundamentally difficult problem, and the guys at Google have spent a hell of a long time on it.

Tillmann Rendel wrote:

Andrew Coppin wrote:

...

In general, I find *most* search functions to be fairly unhelpful. Google is the shining exception to this rule; it almost always seems to figure out what you're after.

I guess doing text searching is just a fundamentally difficult problem, and the guys at Google have spent a hell of a long time on it.

text searching is a well-known problem. ranking search results by relevance is the key to google's success. read the paper about google to learn more:

Sergey Brin and Lawrence Page, "The Anatomy of a Large-Scale Hypertextual Web Search Engine" in: Proceedings of the 7th International WWW Conference, 1998, Brisbane

http://infolab.stanford.edu/~backrub/google.html http://infolab.stanford.edu/pub/papers/google.pdf

Thank you for that link. An interesting paper. I hadn't seen it before. I've added a link to a Google site search to the haskell.org front page. I can't find a way to link the existing search box at the top directly to it and can't create a 'form' element, so I've created a link labelled 'Search haskell.org' to this intermediate page on my site http://www.millstream.com/haskellorgsearch.html If someone can point the haskell.org search box directly at the Google site search please do so. It currently searches *.haskell.org/* which may be too broad? I can adjust as required. Richard.

Mads Lindstrøm wrote: ...

...

Did I do something wrong when searching haskell.org? Properly not. I think the problem is that haskell.org do not index words, that have length <= 3. MediaWiki (which I think haskell.org uses) do not by default index short words (length <= 3 or length <= 4 - can't remember which).

If you search for yhc you also get zero results, which does not make sense either. ...

Yes, it does look like MediaWiki and so I guess from the number of configuration options (if this is the right software) http://www.mediawiki.org/wiki/Manual:Configuration_settings there ought to be some clever setting somewhere rather than it needing any programming. Maybe the full search has been switched off to save resources? Experimenting with the search it seems to miss some longer words too, but I can't see a pattern to it. Richard.

Thomas Schilling wrote:

On Thu, 2007-11-22 at 13:23 +0000, Richard Kelsall wrote:

...

I was reading the 'Problems with do notation' thread and Thomas Schilling suggested reading about mdo. Not knowing mdo I thought that sounds interesting and went to

Gah, I was too lazy to add the proper references:

"A Recursive do for Haskell" by Erkök and Launchbury http://www.cse.ogi.edu/PacSoft/projects/rmb/recdo.pdf

"A Recursive do for Haskell: Design and Implementation" by Erkök and Launchbury http://www.cse.ogi.edu/PacSoft/projects/rmb/mdo.pdf

And here's a nice use case for it:

"Assembly: Circular Programming with Recursive do" by O'Connor http://www.haskell.org/sitewiki/images/1/14/TMR-Issue6.pdf

Thank you. I'll have a read of those. I didn't mean to suggest you should have given all the details, just that haskell.org confused me by saying mdo didn't exist. I've just tried some other words from that page in the Search box and couldn't see any pattern to whether the page appears in the search results. Very strange. Maybe a message along the lines of "We are hoping some generous person will improve this search feature at some point. The source code is here ... The search currently produces incomplete results. Please try Google if you do not find an answer." would be useful? Richard.