> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Karl Avedal
> Sent: 20 August 2000 10:20
> To: Orion-Interest
> Subject: Re: POLL: what are you missing in apache-like functionality?
>
>
> Hello Jeroen
>
> "J.T. Wenting" wrote:
>
> > I think such request merely indicate the need for more complete
> > documentation. Documentation (and you are doing a good job,
> but it should
> > not be needed, Evermind should do it for us)
>
> Yes, as I've said many times before, we are working on
> documentation but
> things are not always going as fast as we would want them to. J2EE and
> Orion is a vast area to cover and things are changing so rapidly that
> documentation needs to be updated extremely often. However, we are
> putting out more docs and will continue to do so, in a pace
> that will be
> increasing.
>
I know the problem. It is hard to just keep up, I can imagine the nightmare
having to update your problem every time someone at Sun has a bright idea ;)
> > is the greatest weakness Orion
> > has (not counting the bugs Swing causes in the tools, these
> are generic to
> > Java applications using Swing).
> > I like Orion, but without docs, I could never sell people
> on it (the people
> > who need to set up and maintain it are not programmers,
> they are Unix and NT
> > sysadmins...
>
> Yes, and we have focused a lot on this for Orion 1.2 with the
> graphical
> management console. However, it's more of a preview and it
> will improve
> much in the coming versions. Of course graphical tools can never be a
> replacement for quality documentation though, but you should
> definately
> not have to be a programmer or a Unix or NT admin to install
> or maintain
> Orion.
>
That is as it should be. good tools (and thet are quite good, apart from the
Swing-related trouble once in a while) can be a great way to help people
getting used to a product. Once they are a bit comfortable, it then becomes
easier to read the docs.
> Also, I think you are being somewhat harsh in saying "without docs",
> considering that there is a fair bit of documentation available. For
> example our taglib tutorial has become the default tag
> library tutorial
> for many people and you will find that for example Sun is
> linking to it
> from their site for people who want to learn them. But yes, the
> documentation isn't as complete as we would like and we'll continue to
> work on this until it is. If anyone wants a job as a technical writer,
> we are accepting applications, there is a lot of work to be done :)
>
It is good, but it is not documentation on installing, configuring and
maintaining the server itself (which is what is needed most, especially from
a comercial point of view).
> > And at the price, it is difficult to get management people
> > convinced anyway (the expensive == good syndrome is very
> strong here).
> >
>
> Of course we don't know all about this, but the expensive == good
> syndrome does not seem to be a problem for sales at all. People are
> getting more used to cheap or even free software every day and even
> though we still see people thinking like that, we are certain
> that it is
> a smaller problem than most people realize. Not a lot of people will
> suggest that Solaris 8 isn't a viable operating system anymore just
> because you can get it from Sun for the cost of media + shipping. More
> and more people start to realize that software isn't like other
> industries.
I know that, you know that, but I work at a large corporation where policies
are set by people who either have never worked in IT, or have quit a long
time ago. They do think that way (why else do iPlanet, Oracle and Weblogic
still sell strong?).
Not that you should increase price (though some do it) of course, your
client base will likely take a while longer to include the really large
companies.
>
> For a software purchase, you don't mainly pay for costs
> related to your
> license, but to the research, development and marketing behind the
> product. So if a company sells 10 licenses at $2,000, that's not much
> worse than selling 1 at $20,000 (of course it's worse, but not much).
> Whether you sell 10 cars for $2,000 or 1 car for $20,000 however is a
> huge difference, since 1 car may cost $15,000 to manufacture.
> Because of
> this, there's no reason why quality of a software product
> (unlike a car)
It can have an effect, if sales are expected to be low, and development cost
is high. In such a scenario, doing less rigorous testing can bring down
development cost (at the potential risk of didsgruntled customers because of
higher volume of bugs, thus lower quality).
Quality in software can indeed not be expressed in the quality of some fixed
component (especially with downloadable software), but must be expressed in
number and seriousness of bugs per KLOC (or MTBF in a more realistic
scenario). Lows MTBF (high # of bugs) will impact customer satisfaction, and
thus future sales (and with the current speed of information exchange, the
future can be tomorrow).
> affects the price of the product. What does affect the price most is
> where the companies think the optimal price lies to maximize the
> revenue. We are certain that there will be a continued and increased
> pressure on the larger vendors to lower prices and this will
> be good for
> the whole J2EE industry.
>
It will be good if prices come down, for imcome also. When the price of
office products like WordPerfect and MS Office dropped to 10% of their
previous prices, sales went up 100 times or more (probably all those people
with pirated copies finally thought the price right).
> Regards,
> Karl Avedal
>
>
Jeroen T Wenting
[EMAIL PROTECTED]
ICQ UIN #9191966
It's the end of the world as we know it and I feel fine (Michael Stipe)
