Re: requirements for an OFBiz IDE

[David E Jones]

Such a tool could get pretty fancy, but the most useful thing I think  
we've discussed in the past is a fairly simple IDE plugin (for  
Ecipse, NetBeans, IDEA, etc) that could display some sort of graph of  
relates resources given a specific resource.

For example, you could right click on a screen name in a screens XML  
file and have it pop up a graph showing all related forms, services,  
entities, controller views & requests, etc.

To make things more fun these graphs could vary for different common  
patterns in sets of screens, services, even sets of entities.

To make things yet again more fun the tool could identify missing  
elements that would be expected, and link to a code template of some  
sort (perhaps using a convenient and well known templating language  
like FTL...), to generate service defs and implementations based on  
entity definitions, and screen and form definitions based on service  
defs, and so on...

-David


On Dec 27, 2006, at 10:10 AM, Anders Hessellund wrote:

Hi Torsten,

Would you be looking at building one? Are you thinking of Eclipse  
as the
framework for that as well? (I understand SAP NetWeaver is based on
Eclipse, isn't it?)

I am currently doing research on how to improve the development  
experience
for enterprise systems such as Navision, Axapta and SAP. OFBiz is
interesting in that respect since it is open source and supported by a
large and enthusiastic community. I was actually considering  
prototyping a
small eclipse plugin for OFBiz. In order to do that, though, I need  
some
realistic requirements from the OFBiz community. I am not sure that  
the
requirements that I bring from looking at the commercial  
counterparts are
relevant here.

Now the interesting question would be: Would an IDE be used only  
by the
groups 1-3 or by group 4 as well?

I agree that end-users should also be taken into account when  
discussing
tool support. Nevertheless, I believe that offering tool support  
for group
1-2 and especially group 3 is most interesting. In the commercial  
tools
that I have been looking at, the support for group 3 (often  
partners and
3rd party customizers) is extremely weak. It would really add value  
if one
could come up with relevant tool support for the customization  
process.
Hence my request for a list of requirements ;)

Nevertheless, this triggers the question wether an IDE should be
something that lives on someone's PC (as Eclipse would) or if all the
nice new web technologies (AJAX and the like) are mature enough in  
the
meanwhile to provide customizing through a web interface.

I am not sure I think this is very important. Whether tool support is
offered as a desktop app or in an AJAX-powered webapp, the business
requirements should still be the focal point.

-- Anders


Hi Anders,

I've been doing a little reading of previous posts in this forum in
order to determine the requirements for a specialized OFBiz IDE
(similar to NetWeaver for SAP).

Would you be looking at building one? Are you thinking of Eclipse  
as the
framework for that as well? (I understand SAP NetWeaver is based on
Eclipse, isn't it?)

However, as I see it, there are three main
user groups in the OFBiz community: 1) framework developers, 2)
application developers, and 3) application "customizers".

I would add a forth group here:

End-users of the application.

Believe it or not, but this is ultimately all about "normal" (=  
non-IT)
people who sit in front of OFBiz and get real work done that makes  
the
company that's USING (not developing, not consulting, not training) a
particular OFBiz installation some bucks. I think this aspect often
get's lost in discussions.

Now the interesting question would be: Would an IDE be used only  
by the
groups 1-3 or by group 4 as well?

My initial feeling was that probably some kind of IDE would be  
something
for "us, the IT people" and should not be of any interest to business
users. But after giving this some more thought, I am not sure I would
stick with that view.

I think that customizing work can be done by both IT people as  
well as
business users, depending on complexity. If you think of  
customizing the
layout of an invoice or a report for example, this is definitely
someting a user would want to do. But of course, there are aspects to
customizing that would be beyong the scope of a non-programmer.

Nevertheless, this triggers the question wether an IDE should be
something that lives on someone's PC (as Eclipse would) or if all the
nice new web technologies (AJAX and the like) are mature enough in  
the
meanwhile to provide customizing through a web interface.

Regards,
Torsten


Anders Hessellund schrieb:
Hi,

I've been doing a little reading of previous posts in this forum in
order
to determine the requirements for a specialized OFBiz IDE  
(similar to
NetWeaver for SAP). Since I am still new to this community, my
observations might be wrong. However, as I see it, there are  
three main
user groups in the OFBiz community: 1) framework developers, 2)
application developers, and 3) application "customizers". The  
last group
is probably the largest, because the typical use case of OFBiz  
seems to
be
to download, customize and deploy the standard OFBiz  
distribution. The
two
first groups seems to have a significant overlap since application
development provides feature requests for the framework development.

I would appreciate some feedback on concrete tool requirements  
for the
three groups:

1) framework developers
   - e.g., better profiling and performance measurements of base
components
2) application developers
   - e.g., better editors for different artifacts, analysis tools to
ensure consistency among XML files (such as checking whether a
referenced entity in a minilang file actually exists), navigation
tools, generators for boilerplate code (similar to Neogia)
3) application "customizers"
   - e.g., visual editors for frontend customization, easy
configuration,
simple mapping from user requirements to actual ofbiz components

Please contribute to this list, if you have any ideas.

-- Anders