history stats WAS Re: [pmwiki-users] Re: Redesigning the edit page - heirarchical menus?

[Radu]

Though I'm not sure scrapping that metadata is a good idea, I would indeed 
like to have the History in a more data-mining setting. That is go from 
less info to more info rather than the current setup where first you get 
all the changes, then you click to get less changes...

So a statistic idea would be fine. The first click on history would give 
counts per user (persistent ones, that is not influenced by the changes 
which 'fall out' of the history as they pass the $DiffKeepDays mark), then 
there would be links for major changes, all changes, and even changes by 
author, if it ever gets implemented.

Do other people feel like me that it's a good thing to have only the 
changes made by a specific user visible? My supervisor seems to think it 
would help to tell how well an author is doing without having to sift 
through loads of other changes.

At 10:49 AM 4/29/2005, Martin Fick wrote:
>On Thu, Apr 28, 2005 at 05:02:41PM -0400, Radu wrote:
> > Most of the time I am against allowing the computer to make decisions for
> > me. Computers are good at automating things, not at making abstract 
> choices
> > like the one between minor and major - which to begin with is fuzzy even
> > for people. Even if two people will agree in definition on what the
> > minor/major distinction means, in practice they'll most definitely diverge
> > (I can provide references for the HCI studies), and even if they did, it's
> > rather difficult to keep it consistent, especially if you also have to
> > remember to click a checkbox before saving. We're there to edit content,
> > not to provide metadata on our edits. The metadata should not require 
> extra
> > clicks.
>
>
>I think that your metadata point is very valid and wonder
>if we should not just question the idea of major vs. minor
>in the first place.  What is the purpose of major/minor?
>How is it actually used and is it actually helpful?
>
>I suggest that we scrap the major/minor distinction
>altogether and provide better statistics about a page for
>people who want them and allow simple page history
>filtering.
>
>Some simple statistics:
>
>1) # of lines changed
>2) # of authors
>3) # of edits per author
>4) Edit was a formatting only change
>...
>
>Every author (not just wiki admin) has a different concept
>of major/minor.  Authors make mistakes on submits which
>would be permanently marked as major/minor incorrectly, the
>resulting meta-data is useless...scrap it, make everyone's
>life easier!
>-Martin

Cheers,
Radu
(www.monicsoft.net)