Hi,
Just proofed your bug report. Very well documented and complete.
Thanks.
How do new features get discovered by documentation writers?
By mining the wiki? Or is there a new feature list as things are being
corrected and added?
Currently developers collect development notes and write release notes,
blog about the implementation progress ...
So while there are some information around, this is currently not
satisfying for the people who need to know most of the new features.
This is not only documentation writers - these are also localizers,
quality assurance, people giving user support etc.
This was one of the reasons why I tried how a small specification can be
put to the wiki. Once more people use this approach,we could just add
categories (like NewIn3.5) and would have a good list.
But (i wrote this already at the design and at the German lists):
although I personally think, that a (basic and not over-formalized)
specification is welcome to most of the developers, hardly any developer
will write a detailed specification on his own. This is not because
developers are lazy - it is just that ~1/4th of the specification can be
written by non-developers. E.g.
- you do not need a developer to find the correct place for a new option,
- you do not need a developer to know what the new label should read,
- you do not need a developer to search in the OOo usage-tracking data
and identify how often a (supposed to be removed) UI item has been used.
- you do not need a developer to identify side-effects for the user's
workflow
- you do not need a developer to write the help
Some of these tasks even *should* not be done by developers, because
they (naturally) think more in terms of code (e.g. a developer's
decision where to put a new UI element might be heavily influnced by the
internal code structures - what is in many cases not the best for the
end user, not knowing the implementation). But the developer needs to be
involved, as she needs to give input about what works and what not (and
developers have ego's, so even their opinion about user interaction
should be considered ).
To give you an impression about the work that a non-developer could do:
I worked about 2 days on the spec (discussion at mailinglists, do some
mockups / prototyping, search for usage data, document what I found...).
The implementation was done in one and a half day - and it took only
that long, because I never ever did contribute new functionality to
LibreOffice before - during this one and a half day I learned how to
implement new UI elements and store options. I even did almost three
full implementations (the not used alternatives had been implemented for
learning purposes). After all - the real code-related work was less than
1/3rd of all the work.
We need to come to a point, where non-developers help with all these
things and collaborate with developers (code hackers). For this, we need
to make people aware (what I did was just a test of what people are
already aware).
regards,
André