On Sat, 19 Aug 2017 14:44:09 +0200, Jochen Wiedmann wrote:
On Tue, Aug 15, 2017 at 4:26 PM, Gilles
<gil...@harfang.homelinux.org> wrote:
How about creating "Commons Geometry"?
Honestly: There are other subprojects (Vfs comes to mind), which are
perfectly able to produce a set of jar file without adding to the
list
of jar files for every one.
I don't understand the purpose of that remark.
Why do you require a new subproject?
The reasons were on the original post.
Haven't I been proven right that CM would become unmaintained?
I am in favour of attempting to salvage what can be (prioritizing
according to the given criteria).
Given the amount of noise, that numbers, RNG, etc. have produced over
the last year, I am more than hesitant to have more of that.
1. Isn't this list supposed to hold discussions? [If it bothers you
(as it does me, on subjects which I'm not interested in), please
support the creation of separate MLs. The "noise" is not IMO a
reasonable argument to oppose other's work.]
2. It's quite unfair to cite "RNG" in this regard. Do you refer to
the "history" (i.e. the dispute I had with our now Apache chairman)?
Or to something else? I sure hope that you are not criticizing the
fact that I developed Commons RNG in full view by commenting every
aspect on this list and on JIRA.
Whatever, wasn't it released? Did it incur any "noise" since then?
3. What noise does the development (mostly porting work) of "Numbers"
entail? [I certainly wish there would be more noise if it could
help making the release imminent; remaining issues depend on other
people (see JIRA).]
4. It's quite unlikely that a "Geometry" component will create any
noise at this point: cf. my suggestion about a "beta" release.
If it does afterwards, that it will mean that the goal of
reviving interest in that codebase (and, hopefully, attracting
maintainers and contributors) will have succeeded.
(In
particular, when considering the rather limited amount of releases,
which have grown out of that.
The first request (partly spin-off, mostly improvements and new
features) was about package "o.a.c.math4.random". As a result,
"Commons RNG" v1.0 has indeed been released.
I'm willing to release v1.1 (but waiting for a feature taken on
by someone else).
The second spin-off was "Numbers". No release yet, that's true,
but didn't I indicate that a "Geometry" installment would wait
for that to happen?
What else? Yes, a "Commons SigProc" idea was floated, but it's
hardly a spin-off from Commons Math: rather it is new donated
code that has a few CM dependencies (which I guess would mostly
become dependencies on "Numbers"). [And not much noise entailed
by that one either, because the donor (and expected main
contributor) hasn't started the porting work yet.]
The "Downloads" section for numbers is
still pointing to RNG.)
It's a copy/paste bug. Hardly a reason to stop initiatives.
As far, as I am concerned, I am clearly -1
Obviously, I'm at a loss understanding why...
Regards,
Gilles
Jochen
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
