GNUe is quite OK (was Re: [GNUe] Re: GNUe)

[Joost Helberg]

Darryl,

>Darryl Caldwell <address@hidden> wrote about `[GNUe] Re: GNUe (was 'What's 
>up?')' on Date: Mon, 15 Mar 2004 18:02:17 -0600
> On Sun, Mar 14, 2004 at 10:33:30AM +0000, Peter Sullivan wrote:
> > In article <address@hidden>, Derek
> > Neighbors <address@hidden> writes
> > >Yes we have a very real problem with people thinking this project is not
> > >highly viable when it is.  The interesting thing is we actually were at
> > >the core of a study on how to communicate in a distributed world[0].
> > >
> > 
> 
> 
> Hi all,
> 
> Meaning no disrespect to the hard volunteer work you have all done, I will
> continue to bypass GNUe until there is up-to-date docs, more
> traffic on this list, case studies, and screen shots of working apps
> in the field. Is anyone besides Derek and Jason using this stuff? Show
> me.

I'm using [GNUe] for editing the content of my datamodel which drives
the technical part of a Dutch IT-services operation which employs 50
people.

This includes management of all CV's (core business), education
budgets and Balanced Score Card (limited) reporting on over 80 Key
Performance Indicators.

GNUe matches my requirements quite good. I do reporting using a
self-developed SQL-resolver, which allows mixing of SQL with any
content (text, tex, xml, html etc.).

Furthermore, I don't like the UI-software to impose too much on
users. I'd rather implement reporting to find anomalies in the data,
instead of doing extensive UI-validation. 

I was able to migrate my datamodel from proprietary Java software to
GNUe without touching it. Writing applications is quite easy, although
there's still too little documentation (I'm working on that).

With a good application-construction set like GNUe (although pretty
old-fashioned in architecture), it's easy to use any datamodel and
business model for implementing applications.

The business logic can be implemented using triggers and reporting of
anomalies. I don't think the business logic should be present too much
in the end-user application.

Even the problem of `missing items in a form', should not be solved in
the UI, but should be reported as `things to do' to the particular
user.

In my old proprietary Java product, I used various queries for
detecting to-do-items, and presented these in a `ToDo'-menu.

I'm sure there is a way to implement this in GNUe.

> I have complained before that all new developments are regulated to IRC
> discussions, which is not a good log for developer serious about
> researching solutions. I don't want to spend anymore time grepping IRC
> logs....

On IRC, I agree. I joined once and was not amused at all: too much
noise. Didn't join again and probably will not in the near future. 


> Sorry for kicking up dust, but I suspect this is what is holding up
> the horde.

Don't apologize, although I'm not sure this is true.

GNUe is quite OK; i.e. the toolset is. The applications are not
available yet, and I don't think they will ever be.

I don't think I can use a bit of GNUe HR code for my Human Resource
process, it's not universal enough. What I can use are business rules
and their mappings to datamodels, views and constraints! Also
anomaly-spotting (using e.g. views) can be extremely universal and
useful.

Writing down these for a particular solution (CRM, finance, HR etc.)
is tough, but valuable. You need specialists on the subject or that,
these don't come easy.

Having dropped the automatic anomaly-->todo-list feature of my
previous generation of office applications, I'll think of how to
implement this in GNUe now.
 
regards,

Joost Helberg

-- 
Joost Helberg
Snow B.V.        http://snow.nl Tel 0418-653333 Fax 0418-653666
Voorzitter VOSN  http://www.vosn.nl Tel 0418-653336 Fax 0418-653666