Re: online help usability

[mpriestley -at- VNET -dot- IBM -dot- COM]

Vicki Rosenzweig, talking about applications in OS/2 which didn't have their
own help, said:

>but at least it's quick. In OS/2, you're dropped into a nice-loo
>multi-level help menu, which is _entirely_ about how to use
>OS/2 help. You still don't get the answer you're looking for,
>and it can take ten minutes to figure this out. Because the
>real joker is that you get all that stuff, _plus_ any actual
>application-specific information, if the information you want
>does exist, so there's no quick way to determine that what you
>want hasn't been written. My hunch is that anyone who has spent
>much time dealing with an OS/2 system is going to hesitate to
>press the "help" key, or select the help menu, again.

A quick note: if no help has been provided, OS/2 pops up a message: "Help
not found."  The situation you're seeing is probably the result of
a programmer hooking up help incorrectly or incompletely.  This ain't an
OS/2 problem, it's a dirty kludge by the programmer.

I'm curious - which applications didn't have their own help?
Now, personally, I don't think I've _ever_ seen a
Help menu without a General help choice (which would be help for the
current window).  Maybe I've been spoiled by working for the company; most
of the applications I've seen have been written by IBM, and they pretty much
follow their own rules.

If you've seen an application without help, however, then wouldn't the
lack of a General help choice be a clue in itself that your specific help
wasn't available?  Or am I misunderstanding something?

I've just done a quick random sampling of the applications on my system
(all ones that come with OS/2, since that's mostly all I use).  All of them
appear to have properly-implemented help: Help index, General help, Using help,
Keys help.  All the information is application-specific, except for the
applets in the Productivity folder, which share one index, and of course
Using help and (mostly) Keys help.

So: if you've seen an application that has no help, and only provides
system help (ie Using help and Keys help), the programmers should be shot
(okay, well, maybe only tortured a bit).  Like I said, I've never seen this.

In any case, though, I would have thought it would be preferable to still
have that help available, rather than none at all.  Maybe instead of taking
out the Help index and General help choices, they should have been hooked up
to a dummy panel that said: "application help available in a later release"
or some such.  Should OS/2 be blamed for that decision, though, or should
the programmers of that particular application be blamed?  The programmers
were the ones who hooked up the help wrong.

As a rule: the Help button, the General help menu choice, and the Help index
menu choice should _always_ provide application-specific help.  If the choices
aren't there, they should be; if they are but they're hooked up wrong, the
programmers did some funny things they shouldn't have.

I've been working with OS/2 help, as a user and then a writer, since OS/2 1.3
(about 2 years).  I don't think it's perfect, but I think it's a lot better
than your append suggests.  If an application that runs on OS/2 failed to
provide help, it broke every rule in the development guidelines (well, it
broke one anyway); it's not CUA-compliant either.  So I don't think it's fair
to blame the operating system for a failure on the programmer's part.

Anyway, pardon me for the defensiveness: I happen to like IPF (OS/2's help
system), and I think it's a powerful and flexible tool, multimedia-enabled,
with a whole lot of function that hasn't been used much yet.

I'll shut up now.

Michael Priestley
mpriestley -at- vnet -dot- ibm -dot- com
Disclaimer: I work for IBM, but I'm not speaking on their behalf.  I'm
speaking as an OS/2 user and an IPF help writer.