Will Johnson wrote:
> Ok let's hear some real numbers.
> Can a Pick database (on Windows) talk to the Web for
> say... under 5 grand?
Yes
> If so How ?
Write your own interface on time that has not been allocated to paying
customers.
Seriously, I have written many such interfaces using various technologies to
various MV back-ends, sometimes just for fun to see if it can be done. The
answer to "talking to the web" questions is in understanding the
technologies at each tier of your architecture - knowledge of MV alone is
not going to lead to the least expensive solutions.
Examples of free technologies and IDE's:
.NET (C#Builder), Perl (Komodo), Java (various)
InetD and WinInetD
UniObjects and InterCall
(Hint) Telnet is also an excellent protocol for raw data exchange
Web Services (see my articles in Spectrum Magazine)
If you want someone to provide more full-featured IDE's around these free
technologies then you may need to pay for them. Again, the more you can do
on your own, the less you need to pay someone else.
The cost for deployment depends greatly on whether you're running persistent
or non-persistent connections, and how many concurrent users you want to
connect via those connections. If you are not paying for N concurrent and
persistent licenses then the cost drastically drops. There are a variety of
free and commercial solutions that facilitate this.
The technologies you adopt may depend on how many clients you realistically
expect to connect at one time. I say "realistically" because most new
prospects will tell you they have thousands of people connecting to their
website at any given time. If you look at the website logs you can whittle
this down to real users if you don't count hits to static pages, bots,
images, menus, or anything else that isn't true access to real data. If you
have THAT many people who need your real data and you aren't covering your
development costs then you're either in the wrong business, you need to
change your architecture, or you need to change your business model and
start charging for access to your valuable resources.
Since you asked for real numbers, I recently mentioned that I was interested
in DesignBAIS. Please verify numbers with them, but here is my
understanding:
1) A single developer license costs $1000.
2) For deployment, each concurrent user costs $100 per year, but users are
non-persistent.
There are no other costs for this software. An MVDBMS license is currently
consumed during the period of connection, so "some number" of licenses are
required, again depending on how you architect the app and the services you
provide. For this software there is no "server license", so the only cost
for deployment is $100 per user at the end-user site. I'll rephrase: The
total added cost to an end-user for this software is $100 per user.
I also mentioned PDP.NET. Here is my understanding, I'll be happy to verify
the numbers for anyone who contacts me and is interested in a purchase:
1) The software development kit is $495.
2) Yearly maintenance for the SDK is $99.
3) End-user seats for U2 have a one-time licensing fee of $275.
4) Yearly maintenance per end-user seat is $55
5) A developer requires 1 client plug-in for testing so the real entry-level
price for a single U2 developer is $924.
6) PDP.NET integrates with Visual Studio .NET. While it's not required to
have VS.NET to develop with PDP, it would be silly to develop without it.
So add the cost of VS.NET to the developer cost.
7) Like DesignBAIS and other products, PDP consumes a U2 license during the
connection, but you can architect your solution to minimize the time per
connection and the number of connections.
To compare apples to apples between PDP.NET and DesignBAIS, there is also no
cost for the PDP server component (unlike mvDesigner and other products).
Therefore the added cost to deploy a single end-user for PDP.NET is $275
plus $55 per year.
When I say "added", I mean over OS costs (if any), DBMS licenses, and your
app license fees. But remember that if you now have, for example, 100 green
screen licenses at a given site, and you re-architect with a new UI, you may
be able to reduce the number of required DBMS licenses down to about 25 or
less. The cost savings from that alone justifies a large part of your
development.
The ideal way to fund such development is to have your end-users invest in
their own future savings. (I'm making up numbers here but...) If you have
an end-user paying $100 per year for DBMS maintenance and they know they can
save $5000 next year by using your new GUI, ((75 seats * 100 = 7500 ) - (25
seats * 100 for new UI = 2500) = 5000) they may be willing to fund your
development. Multiply that by multiple end-user sites and the cost of
developing a new UI is no longer important because all of your development
and deployment is funded. With this model, you may even be able to contract
out development; development costs are higher but you if you can deploy
sooner then you may be able to sell systems sooner. What's important is
picking the tool that satisfies your real business needs.
HTH,
Tony aka "Dudeus de Longus Windicus"
Nebula R&D
http://Nebula-RnD.com
(Always happy to develop systems using these technologies and others.)
-------
u2-users mailing list
[EMAIL PROTECTED]
http://www.u2ug.org/listinfo/u2-users
