For some years now I have been saving J posts where I felt that "one day
I might want to use that". I now have so much 'stuff' that I will never
find what I need anyway. I am not even sure where to find my own
questions from five or so years ago.

The concept of tagging is music to my ears.

David

On Sat, 2008-02-02 at 18:34 -0500, Devon McCormick wrote:
> Skip - I think the way we mentioned _is_ extremely simple.  It consists of
> the following steps:
> 1) at the top of the wiki page, click on "Edit",
> 2) add keywords to the "#pragma keywords" statement at the very top of the
> page,
> 3) click "Save".
> 
> However, you make some good points about how this does not help us with
> non-wiki material.  Also, this method does not work for an immutable page;
> you'd have to create a page pointing to the immutable one and add the
> keywords there.
> 
> Could we agree on a more searchable phrase for J, like "J-language"?
> 
> By the way, I'm by no means a wiki expert unless that encompasses knowing
> only about seven things.  You can see these seven things on my "Wiki Tips"
> page (http://www.jsoftware.com/jwiki/DevonMcCormick/WikiTips)  which exists
> partly just to remind myself of the few things I've learned.
> 
> In any case, I'm glad to see others interested in this peristent problem of
> finding J-language information.
> 
> Regards,
> 
> Devon
> 
> On 2/2/08, Skip Cave <[EMAIL PROTECTED]> wrote:
> >
> > Oleg, Devon,
> >
> > In order for the keyword scheme to work it must be extremely simple, so
> > that anybody can add keywords, without having to read directions, write
> > scripts, or know how to create a wiki page. Keep in mind, users looking
> > for answers are not interested in building wikis, only finding the
> > answers to their problems. The keyword-add process should be so easy,
> > that even casual wiki or forum browsers would have no problem dropping
> > in a couple of keywords, if they felt it would help someone else have an
> > easier time finding their discovery.
> >
> > The basic rules that need to be in place to really make the keyword
> > scheme work, are as follows:
> >
> > 1. Every forum post, wiki page, or reference book chapter, or article on
> > J, should have two things at the bottom of that post or page:
> >     a. An active link which says "Add a new keyword" that, when clicked,
> > will bring up a dialog box that enables the user to enter a list of
> > keywords. Any keywords entered in the box will be added to the keyword
> > list for that page or post.
> >     b. The list of all keywords that have been added to that page or
> > post should be displayed at the bottom of the page. Perhaps the list
> > could be hidden by default, with a view/hide button/link to allow
> > viewing or hiding the list. We may have to convert some of the reference
> > docs to html to make this keyword-add scheme ubiquitous across all of
> > the J documentation, but I believe that it would be worth the effort.
> >
> > 2a. There should be one single search engine that encompasses all of the
> > forum posts, wiki pages, reference books, discussions, and articles on
> > J. That search engine should be prominently displayed on  the home page,
> > wiki home, reference book home, etc., and anywhere else that a user
> > might go to find information about J.   Actually, this has already been
> > done pretty well by the J software folks. The engine just doesn't find
> > the appropriate item often enough, which isn't the fault of the search
> > engine. The information is usually out there. The users just don't enter
> > in the right search terms. The keyword scheme can help fix this.
> >
> > 2b. A very brief explanation of the keyword scheme should be displayed
> > along with every search engine box, and also perhaps with every "add a
> > new keyword" link, so users will understand how they can help others
> > find the things that they discover, by adding their own keywords.
> >
> > Devon's #pragma scheme is fine for experienced wiki builders to add
> > keywords to wiki pages. IMHO, simple as it is, the average wiki-browsing
> > user won't take the time to learn how to add keywords to the wiki that
> > way. Also, it does nothing for the forum posts or reference docs, which
> > have much great info hidden in them. The keyword plan must be structured
> > simply enough so that as soon as someone finds the information they
> > need, they could add a keyword or two which would have helped them find
> > their information quicker.
> >
> > A little incentive could go a long way to help the cause. JSoftware
> > could have a raffle so that registered keyword "taggers" would get an
> > entry to the raffle every time that the added a keyword to the doc. The
> > taggers could be tracked using cookies. The prize shouldn't be too big,
> > to discourage inappropriate tagging, but big enough to encourage typing
> > in a couple of keywords that could help others find that useful post or
> > page.
> >
> > There are many "user-graded" posting schemes out there, (Digg, Slashdot,
> > etc) where users grade the importance or usefulness of an article or
> > blog, so this keyword concept is not too far out. However, the J
> > keyword mechanism is about adding information to a post, to improve its
> > chance of being found, instead of a grading scheme. Both schemes rely on
> > people's altruistic natures. Actually, the keyword mechanism is a
> > simple-minded way to back into Tim Berners-Lee's "semantic web" dream,
> > (http://en.wikipedia.org/wiki/Semantic_Web) without all of the
> > technical complexities Tim throws in.
> >
> > When you think about it, if every document, post or wiki page had a
> > unique keyword in it, such as jlanguage, or jay language, or jsoftware,
> > or Iverson, etc., then general searches on Google which included one of
> > those keywords, would find the same set of information that the
> > specialized J search engine would find...
> >
> > This discussion may be moot, as I am not sure that there is any easy way
> > to add the described mechanisms to the forum postings or reference docs.
> > If it could be done, it would be an interesting experiment that could
> > potentially revolutionize how web information is categorized and
> > discovered.
> >
> > Skip
> >
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> >
> 
> 
> 

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to