An interesting blog by Matt Zimmerman touches on docs

[Kyle Nitzsche]

Hello Shaun and all.

On 07/07/2010 07:34 PM, Shaun McCance wrote:
> On Wed, 2010-07-07 at 17:13 -0400, Kyle Nitzsche wrote:
>    
>> http://mdzlog.alcor.net/2010/07/06/weve-packaged-all-of-the-free-software-what-now/
>>      
> [snip]
>
>    
>> With which I agree in general.
>> * on-disk docs might effectively be limited to only what is necessary to
>> get started and get connected to the web (localized, of course).
>> * run-time help links might instead display appropriate content in the
>> browser.
>>      
> [snip]
>
>    
>> Naturally, there are disadvantages, such as:
>> * no internet connection = no help (beyond the minimal on-disk help)
>> * umm.. any other disadvantages?
>>      
> A greater disconnect between applications and their help. Our
> traditional help consists of islands of documents that are
> largely separate from the applications they document.
>    
I don't quite understand this " greater disconnect". From the user's 
point of view (with an internet connection), I would expect that when 
they click the "Help" button, they don't care (and may not even know) if 
the content is served from the web or from the disk. In both cases, it 
displays in a separate application (browser vs. yelp). Indeed, since the 
web content is (I argue) more nimble, more able to be current, with more 
up-to-date translations, and includes (clearly labeled) user 
commentary/articles/forums/corrections, it is more likely to be 
(overall) correct and helpful, which means the connection is stronger, 
rather than weaker.
> One of my current projects is a library for deeply integrating
> help into applications. (It was Phil's idea, although he might
> not realize it.) Imagine help buttons and menus automatically
> populated with the most relevant content, searching for help
> directly in the help menu, and on-board help blurbs that come
> directly from the help and link into it for more information.
>
>    
I think context-sensitive help (minimal) is important in some cases. 
However, I don't see exactly why the content has to be on-disk. One can 
imaging that one clicks a help widget in an application, it opens a URL 
that has encoded in it: the user's locale, the OS, application, the 
application version, the currently displayed UI. These could be used on 
the server side to display context sensitive help.

Searching help is critical and currently underpowered. Again, I don't 
understand the requirement that this be performed on disk, given this 
^^^ sort of idea.

> These are the sorts of things that user assistance professionals
> are dreaming about, but most help tool vendors are still stuck
> in the 90s. We have the opportunity to blaze new trails with
> free software. Stop playing catchup and make UA professionals'
> mouths water.
>
> It's possible to have this sort of deep integration with cloud
> content, but it's harder. I have no doubt that help will move
> more and more to the web, but then, applications will move more
> and more to the web as well. If we jump there too early without
> thinking about how to really improve things, we'll lock ourselves
> into an outdated and inadequate help model.
>    

The following everyday apps from my lucid system (at least) have (only) 
web-based help:

  * Thunderbird
  * Firefox
  * Chromium-Browser
  * Google Docs

The lack of on-disk help for these does not seem to be impeding their 
paths through life. I would argue it *facilitates* their growth.

Overall, I am talking about a deep shift, and yes, we may consider it a 
"cloudy" shift.

Consider the transition of user expectations regarding "help". I assert 
that they now look to the web *first* for answers, and that this trend 
will only increase. They have come to feel, correctly, that they are 
more likely to get their actual problem answered quickly from the web 
than from Help. Why? Because experience has taught them that official 
help content is never complete, never up-to-date, never organized well 
enough to enable one to rapidly find the answer they need. (Well, 
'never' may be an overstatement, but you get the point.)

Embracing this idea means a conceptual transition on the part of those 
who produce help systems.
  * They don't write it all, realizing they can't do it, practically 
speaking
  * They write 'just enough'
  * They facilitate (through well-designed web portals) user commentary, 
user submissions, corrections, etc.
  * They empower "editors" to review user submissions and move them out 
of the fray and into prominence
  * They facilitate translation

It becomes primarily a role of setting things up, writing some required 
content, and facilitating the rest.

Cheers,
Kyle

> --
> Shaun
>
>
>