not marketing /people/ so much as a marketing /mindset/
if you want people to use stuff, they have to be able to understand
it, and you just cannot demand that they devote a month of their
lives to using your thing... cuz your thing might be a big deal to
you, but it's a small piece of every adopter's world, and they don't
have time to become an expert like you.
This is where I often am flamed, for complaining that "RTFM" is not a
fair answer to a potential adopter of your stuff...but I've seen many
legitimate questions shot down with that or much more vicious attack
language, in forums, mailing lists, elsewhere, leading to a
reputation for many projects that are otherwise very nicely
engineered and just shy of completely usable.
As for the docs, they seem to run the gamut from completely absent,
to way way way too wordy with not too much in the sweet middle spot.
I've been reading the docs for a popular, rather nice open source
project that is proud to finally have proper documentation...
.. and the writer is so happy about the new docs that he spends FOUR
CHAPTERS talking about why the project is so nice...
... while leaving huge gaps in the small, essential details around
using the thing, with much redundancy, and many missing details...
The documentation is lovely. The code examples are nice. But they are
tripped up by internal contradictions and overlaps with the author's
preferred way of doing things (in a world where this project is the
only thing he works with) and they way coders would use it as a small
component of some other project.
The why of "why things aren't better" is easy... it's because these
things are quite hard:
- writing good technical documentation
- writing good installers
- documenting installation and use
- factoring code libraries, even if just defining the single class
and some methods
It's hard to be too annoyed with someone who is giving away code.
On the other hand, it is disappointing and rather sad to see projects
that are so overwhelmed with their own wonderfulness that they are
all but impossible to use (hello Apache Axis).
Maybe a difference with commercial code is that the developer's own
weirdness and personality tend to be squeezed out of commercial
releases, whereas those features may strongly dominate a badly- or
entirely unedited open source project.
I like the craft aspect of open source code, so maybe if 80% of the
author's idiosyncrasies could be removed from the project and
documentation, that'd leave enough to make it special, while keeping
the project from getting in its own way.
The bottom line for code is the same as for anything else: an
excellent editor is the most important thing of all. As you will
agree, even this note could have used one! (A writer of code or words
cannot be his/her own editor).
On Jan 26, 2007, at 2:31 PM, Raj Singh wrote:
(changing the subject line to be more relevant)
I'll throw a radical idea into the mix. The biggest thing missing
from open source projects is marketing people. Yes, I agree
marketers are 90% evil, but that last 10% is important. You need
the voice saying things like... "What will people's 1st impression
of this software be?" "Will using this application make me feel
good?" And even things like, "orange is a hot color right now. Get
that into the UI."
Steve's email was right on target. ArcGIS looks and feels great
compared to the open source competition, whatever else you may
think of it. OK, ready to be flamed now.
---
Raj
On Jan 26, 2007, at 12:39 PM, Landon Blake wrote:
I'm currently trying to figure out ways to tackle this problem on
my own
open source project. I'd love to hear some ideas and suggestions
on that
topic. I'll start another thread so that I don't hijack this
discussion.
_______________________________________________
Geowanking mailing list
[email protected]
http://lists.burri.to/mailman/listinfo/geowanking
