I read the proposed project focus and think that the approach is wrong.  I think that we can draw parallels from documents used in very large applications, such as database, Mainframe software, and industrial processes.  From my experience these are the  documents that serve multiple users:

Installation and use  (just use the home appliance (TV, Air Conditioner, etc.) that you bought. You need to follow instructions describing the set up and first use.

User manual -- how to use the tool you installed (Oven, stove, bbq, car).

Support manual.  Given an error message, provide an answer as to the cause of the message and how to provide first and second level support.

Design Specs manual and source code. Messages and codes  needed to identify and to fix bugs

Q/A manual as part of Design Specs manual.

All the rest as outlined in the Prague proposal will not respond to the Linux Cloud and security users.

We must also discuss codifying all messages from software.  Kernel00o1I or Kernel0001W or Kernel0001E  For Informational, warning or Error.

Examples
Kernel0001I Linux version xxxxx is starting....
Kernel0023w Insufficient page space. Linux will work in degraded mode.
Kernel0099E Segment fault in xxxxxxxx. YYYYYYY cannot continue. Contact Support.

Its time to realize that Linux is in need of a documented detailed set of messages.

If the messages are coded, as in above example, the code will transgress translations to other languages.  Eventually, some messages will become obsolete, and after a time, reused with other text. a Kernel0023w would be the same informational message in English, French, Spanish, Russian, German, etc......
 
Regards

 Leslie
Mr. Leslie Satenstein
Montréal Québec, Canada





From: Pete Travis <me@petetravis.com>
To: For participants of the Documentation Project <docs@lists.fedoraproject.org>
Sent: Sunday, May 17, 2015 5:24 PM
Subject: Re: Personas for Fedora Docs




On May 17, 2015 6:37 AM, "Brian (bex) Exelbierd" <bex@pobox.com> wrote:
>
> I am restarting the discussion of Personas for Fedora Docs.
>
> Personas are critical for scoping the level and style of writing we will do on some/all topics. The current draft set of personas is here: https://fedoraproject.org/wiki/Docs_Project_Focus
>
> Before we dive into the various variant level personas that have been collected, we should think about whether we want to have more than a general set of personas.
>
> For example, we could just write to several different users in general, and allow the specifics to shake out.  I am personally in favor of a high level set of users who we can describe more fully relative to each variant/product situation.  I’ve put some examples on the wiki page above.
>
> I think this set of personas leaves us in a better position.  We can easily present the site different to each kind of user.  We can easily determine if we think a project requires a certain amount of knowledge, disclose that and then write only to personas that can handle it.  We can also do audits at every level and focus ourselves on the real areas that each type of user gets stuck in.
>
> This also gives our users a consistent person to “follow/shadow” in the documentation.  Project level personas may feel like reading a novel where every time you start a new chapter (look at a new part of Fedora) all of the characters suddenly change.
>
> I believe the best step forward is for us to decide if we want overall personas or project level personas and then to figure out who they are.
>
> regards,
>
> bex
> --
> docs mailing list
> docs@lists.fedoraproject.org
> To unsubscribe:
> https://admin.fedoraproject.org/mailman/listinfo/docs
I'm concerned that edition-specific personas would cause some unwanted inferences - writing only newbie docs for Workstation, only advanced docs for server, etc.
Thinking back to the metadata discussion - what do you think about rating the complexity of an article on, say, a 1-5 scale? Later, tools could use the complexity metadata value for sorting and filtering, or maybe link to our 'persona' for that experience level.

Now, I just realized you're thinking about literally writing narratives for the personas - I was thinking of it only as a composition / whiteboardind tool.  That could be interesting...
--Pete