I really want answers from everybody, if at all possible, but even Stefan's
responses are enlightening. A +0 at best, almost a -1. Interesting, I wanted
a sanity check, maybe I'll get one here.

I guess I need to examine why would a rewite (forcing change) be wanted? I
think it needs to bring some things radically new. I (personally) have made
the effort to learn Python, and it took a while. The Python Gump code is
pretty simple, but still takes some time. I can understand that this is a
barrier to entry for a lot of Java programers.

> > *Python Gump (as a future replacement for traditional Gump)
>
> +0 as long as it can do everything that the "traditional" Gump can.
> -1 otherwise.

As it stands it does the core (as I understand it) yet it fails to cope with
a few things, and I was hoping to make it

1) It doesn't process outputs (copying jars somewhere, publishing javadocs,
publishing junit). This is towards the top of my TODO.
2) It doesn't process projects on workspaces only on modules.
3) It doesn't handle depends on ant tasks (keep meanign to ask you Stefan,
'cos I've seen you use it, what is that?)
4) I am sure there are some other subtle differences.

Otherwise, I feel it does everything and more. (1) I will fix, but I'm not
sure about (2) or (3) [partly 'cos I don't see understand them.]

The more:

1) [IMNSHO] Easier development entry (even for Java programmers, XSLT isn't
Java).
1) A GUI (there are lots of benefits here, some really nice ws/profile
visualizations.)
2) Better "check tools", better debugging, and other tools.
3) Statistics, the sky is the limit. (especially if we write to some
database, I'm using DBM right now.)
4) Cleaner extension points can be added (for users, like me, who use Gump
to do more than just build, but publish/other from inside gump's projects).
5) Easier personalized installation ... daft, but true IMHO. The pain is in
the ws/profile configuration, and Python Gump helps the users get inside it
easier.
6) More OO (if far from perfect) and open for extension/customization.

> I won't find the time to learn enough Python to jump into coding so I
> can't be +1.  I've already had my share of code contributions to the
> "traditional" Gump and expect to keep supporting it.

This I'll leave until the future.

> > *Python Gump using Forrest
>
> I don't care at all, honestly.

Do you run Gump personally/locally, or do you run local custom Gumps? In
short -- do you run it remotely and have it communicate to you via
HTML(/RSS)? Do you want HTML output (whether it is forrest or XYZ that helps
produce it?)

> > *Gump Statistics (as a tool for automatically communicating
> > reuse/reliability/robustness, etc.)
>
> Depends on the kind of statistics.  I'm really concerned about
> reliability/robustness values computed from Gump failures.

Ah, really really? Cool, do I have something to entice you? I think
statistics (along with easier debugging) are what drove me to want a new
Gump (or one I could understand/maintain). I agree 100% that
reliability/robustness values are needed.

Are you game to help me with thoughts on what these ought be, and how to
determine them, and algorythms for them? I consider this a key constribution
Gump can make to OSS, a dispassionate view of a project for folks to
consider in order to make "build verse buy" (i.e. re-use) decisions. If
there were two choices, perhaps to help them pick one that has reached
greater maturity.

I'd also like statistics to show "activity" (at CVS level I guess). I'd like
statistics to show "responsiveness" (to problems). FOG Factor is simple
"successes - (failures + prereq failures)", but it might be ok (over time).
I recall the archives having a richer algorythm, but I think this one is an
adequate start.

BTW: What bugs me is if product X changes it's API (or fails to run unit
tests) and hence breaks innocent product Y, then Y gets the bad mark. If X
is then stagnant/unresponsive, Y gets slammed. I don't see a way that Gump
can detect it was X that broke Y, and not Z. Now I realize that Y will
eventually move off X, but that fact won't be captured in stats, which is a
shame. Still, I feel better some information than no information, we can
improve/mature it over time.

And, oh yeah, I consider statistics as subversive marketing for good
products. ;-) ;-)

> > *Gump and Ruper (using Ruper to download packages)
>
> I'm not sure whether Ruper can help a lot when many of the installed
> packages require you to click through a license.

Good point, noted.
But what of the outputs of Gump that are redistributable.

> > *Distributed Gump Trees (also using Ruper to download upstream Gump
> > outputs)

I run Gump on my stuff, but I have to build much of Jakarta first, and that
is tough on my limited resources. If I could automatically (via Ruper)
download last night's Gump'd outputs from Jakarta, then I'd be "in pretty
good shape" for less. Say I did this, and produced my own outputs. Say
others fed of my repository and/or Jakartas, and on. A tree of nested gumps.

> > *Gump Documenter (as a tool for communicating the OSS map, and
> > promoting)

Mapping out the dependencies in the OSS world, to document it. Showing
dependencies, dependees ... perhaps showing history of this. I think there
is extreme value in the project descriptors (I'm sure you agree here, you
invest so much time in them) and I think we need to slice/dice that metadata
to present it to humans for perrusal.

> I don't think I understand these 8-)
>
> > *Promoting more gump projects in public Gump (do resources exist,
> > will the 7+++ hours grate on Stefan?)
> > *Promoting more public/private/personal Gumps.
> > *Promoting more OSS Gumps (e.g. on SF.net. on Java.net, etc.)
>
> Marketing has never been my business and will never be.  +0.

Ok, so the flip side. What say folks did this, and the main gump became a
15+ hour build. How would you be affected? Would you stop running Gump, and
use rsynch/Ruper/other to get jars from public? Would this be bad for you?
(I guess I don't know what do you use Gump for. Do you gump your own private
projects?)

Thanks for your time/insights Stefan.

regards

Adam


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

Reply via email to