Re: GNUstep version number(s) (was: Re: GNUstep article (was: Re:gnustep compared to other toolkits))

[Dennis Leeuw]

<marketingmode>
What I am experiencing in the Linux market is that people are ready for
Linux
on the desktop, but there is nothing that they feel comfortable with.
The news
flashes in the press about difficult installations, GNOME and no office
suites
are dominant.
We need a counter attack. And that is what this reply is based on


> Anyway - here's my opinion - I think part of our current `marketing'
> effort should be to communicate to users and developers the concept that
> gnustep *is* composed of different parts, which *are* at different stage
> of development.

No problem with that, it is just that GNUstep is one project. The
website is
one project. It is viewed and percieved as one project. The problem is
more
WHAT is GNUstep.


> people should not think that because Gorm running on DGS is not ready for
> the end user, then gnustep as a whole is not ready for usage.

But WHAT are we offering a user. I think we are offering two things
1- GNUstep the development environment
2- GNUstep the NeXT look-and-feel desktop

That means we have two audiences to address:
1- Developers
2- Users


> the fact this idea is so widely spread is a problem - because gnustep-make
> and gnustep-base, which are at least half of the whole story - are
> perfectly ready for the end developer.
>
> this is our current major `marketing' problem - we have probably the best
> available non-graphical OO framework for GNU/Linux,

But to be synical, and please don't start a flame war, who is waiting
for that.
WHAT does the GNU/Linux User or Developer want?


> we have it running and
> working, we have used, fixed, debugged, optimized most of it, we are at
> release 1.0.2, and nobody out there knows about it !  because when they
> think of gnustep they think of it only as the full thing, or if not, they
> only think about it as the graphical stuff.

They don't think of GNUstep at all! I work daily with Linux users and
developers and until they met me nobody knew of GNUstep. Now slowly
people are
getting convinced that it might be something, but none of them have ever
tried
it. While compiles of programs go on every day none are GNUstep.


> I am for having the full thing - I am for having the whole of gnustep
> finished - but it's also very important that we sell efficiently what we
> already have - that we bring our products to the users.
>
> gnustep-make/gnustep-base provides you with highly advanced, state of the
> art, easy to use stuff for non-graphical OO stuff which nearly ridicolizes
> the competitors in the GNU/Linux market.  And it's done - stable - ready
> for use - a finished product !  We need to find a way to tell that to the
> world.

Good point, but who is the audience? Not the entire world, who do you
want to
address with a non-graphical OO framework ?


>  IMO the marketing approach to that is to try to sell it as a
> separate product from the less advanced gui stuff.  we don't want the idea
> that the gui is not yet complete shadow the fact that the base libraries
> are done!  this is one of the reasons why we have made the `GNUstep
> LaunchPad' as a separate product I think - to mark the difference between
> that product, and the gui product (which is not at all bad anyway).

Exact and THAT is what I am missing. The GUI is quite good actually.


> in brief - the concept for the masses is -
>
> the gnustep gui stuff will rock and kick asses when finished; but the
> gnustep non-gui stuff rocks and kicks asses *ALREADY* *NOW*.
>
> to bring this concept forward, we need to have clearly different release
> numbers between gui and non-gui stuff.

Okay I have been very negative in the reply. Next I have provided more
questions than answers. But this is where we are. I agree with Nicola
that
GNUstep LaunchPad is rock solid, but I don't think that is a selling
point. The
current press hypes are GNU/Linux on the desktop for the average user
and
GNU/Linux on the PDA for the developer.

And THAT is what we have to address. Marketing for GNUstep is IMHO two
fold.
You have to convince developers that time to market with GNUstep is
better then
with any other toolkit and you have to convince users that GNUstep is
the
future for their desktop.

What needs to be done is to first get people to compile GNUstep. Take
them by
the hand and help them out, and that is a big cheer to everyone on this
list.
The support is great and quick.
Next tell people what GNUstep actually is and that means write
documentation.

For developers the big things are what are the advantages of Objective-C
over
C++, Java and any other language. A clear document that explains the
differences and benefits. This document should give me (a
non-programmer) the
ability to explain why people should use Objective-C (probably be part
of the
GNUstep System Overview).
Then a second document that explains the relationship between
Objective-C, the
runtime and GNUstep this is probably a highly technical document.
Volunteers ?

This should be published to a wide audience through LWN, slashdot etc.

The there has to be a document for the end-user, the desktop user.
That should explain the simple things like, simple configuration stuff:
setting
up time-zones, default languages etc.
It should show how to add new software packages, compile tips and a
directory
explanation.

</marketingmode>

And now back again to the subject: version numbers. I don't see any
difficulty
with different version numbers, but I do see difficulty with all the
different
naming conventions. Imagine you are new. You know linux a bit. You have
seen
GORM and KDE and you take a look at GNUstep, here we go:
Foundation Kit, AppKit, LaunchPad, gnustep-base, gnustep-make,
gnustep-gui,
gnustep-xdps, gnustep-xgps and some are part of one and some are not.

Let's clear things. I think we should have two and only two parts a
on-gui part
and a gui-part. Meaning you have GNUstep LaunchPad, which contains
gnustep-make
and gnustep-base. It is one download and one version.
Then we have GNUstep ApplicationPad that contains gnustep-gui and all
the
backend bundles. One download one version.

This is I hope maintainable. This can be documented and explained.

Sorry for the rage, but I think GNUstep needs more marketing. GNUstep is
ready
for use as was GNOME 1.0 it lacks a lot, but it works (don't call it 1.0
as
GNOME did ;)

Greetings

Dennis

dleeuw.vcf
Description: Card for Dennis Leeuw