Thanks for taking the time to review and provide feedback.  Comments inline
below.

Linus said:

> this is an attempt to join the release cycles in order to get
non-conflicting priorities.

I think it's premature to try to do that.  The ArgoUML project is over 10
years old.  The ArgoEclipse project is only a couple of months old.  I
expect the ArgoEclipse releases to happen every 1-2 months initially as
contrasted with 1-2 releases per year for ArgoUML.

Even after ArgoEclipse matures to the point where it's feasible to
synchronize release schedules, I think there's still value in it having a
separate identity.  It's pretty clear from the name that it's closely
related to both Eclipse and ArgoUML, but it serves a different customer base
than the standalone ArgoUML users.

> If you have concerns about the time spent and the value of a certain
> non-eclipse-related change in ArgoUML (as the change in file save format),
> then this list is not the right place. It is better if you discuss that on
> the argouml dev list or with the developer that made the change.

Agreed.  As you know, I had already done this before I wrote my note.  There
may be instances however where what's right for the ArgoUML users is
different
than what's right for the ArgoEclipse users.
 
> The purpose of integrating the changes into ArgoUML ... is to avoid 
> the need for having to maintain a separate version of the ArgoUML code.

I probably should have described what we are doing in my first reply, both
for the benefit of others who are less familiar with the structure of the
projects and also to make sure that there were no misconceptions.

There is *not* a separate version of the ArgoUML code.  For a small number
of classes where the Swing UI code is too tightly entwined with the non-UI
code, we've created separate copies for ArgoEclipse.  This is effectively
a very small "overlay" on the ArgoUML code.

This is also a temporary measure which became necessary when the ArgoUML
project slipped and it's release code freeze coincided with a period of
active ArgoEclipse development.  It's certainly the goal to make this
"overlay" as small and as temporary as possible.

> A headless build is not an absolute requirement. 

Actually, it is.  I didn't mean to imply that it was a requirement inherited
from ArgoUML.  It's a requirement for ArgoEclipse to be able to set up a
continuous integration build infrastructure.

> I think the ArgoEclipse+ArgoUML should be called ArgoUML when seen from
within Eclipse.
...
> This will make ArgoUML be the market name and ArgoEclipse is the project

>From a user perspective, I think it would be more confusing to download and
install one thing and then have it called something different in the UI.

Tom

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to