You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tamaya.apache.org by Anatole Tresch <an...@apache.org> on 2014/12/24 00:20:14 UTC
Complexity Discussion
Dear all
to make further discussions hopefully less harmful and easier, I will
remove the clutter that is still in there, especially in core, but also in
API. The corresponding cleanup ticket is still in progress. Basically the
*PropertySource* interface will expect 1-2 methods be exact the same as in
Deltaspike, so we hopefully get a better starting point for further
evaluation ;)
Best and good night and merry X-Mas
Anatole
Re: Complexity Discussion
Posted by Anatole Tresch <at...@gmail.com>.
...also there is a clear verdict from David Blevins, I agreed with him:
first we care on Apache and the project, a JSR maybe possible after we had
done our job. But this is not a JSR here (and this is good so). If its good
(and I feel we have a chance to get there), we will probably quickly get
high adoption. Given that we will have broader acceptance and so be able to
start a JSR.
In the EC meeting during last JavaOne it was about 50-50 to start a SE
Configuration JSR. Discussions have shown that it might be too early now
and we should better shape our target, how it should look like abnd gain
adoption.
David also wants us to deliver quickly a first release candidate or
milestone release, so people start trying out things and (hopefully) get
involved. Given that we have a perfect position to start a JSR, where we
will rediscuss much of things again, but out of a position of strength
hopefully.
Best
Anatole
John D. Ament <jo...@apache.org> schrieb am Sun Dec 28 2014 at
16:16:49:
> On Sun Dec 28 2014 at 10:07:05 AM Mark Struberg <st...@yahoo.de> wrote:
>
> > John, the main intention of this podling _never_ was to polish an
> existing
> > code base but to aim for a JSR and provide a RI for it.
> >
> >
> Please double check the proposal.
>
> http://wiki.apache.org/incubator/TamayaProposal
>
> "At a later stage, if successful, standardizing it within a JSR also may be
> an option"
>
>
> >
> > This sometimes includes going back and forth a few times to review ideas
> > and get rid of dead ends.
> >
> > LieGrue,
> > strub
> >
> >
> >
> >
> >
> > > On Sunday, 28 December 2014, 14:10, John D. Ament <
> johndament@apache.org>
> > wrote:
> > > > So maybe two branches of the code base?
> > >
> > > The original sitting on a 0.1-prototype and maybe the new stuff on a
> > > 0.2-prototype branch (since I saw that Mark had set his code as
> > > 0.2-snapshot)? This way no one uses master for a little while until
> all
> > > this gets settled and if one branch wants stuff from the other they can
> > > rebase?
> > >
> > > We do have to remember that the original intention of this podling,
> like
> > > pretty much all other podlings, is to start from a donated code base
> w/ a
> > > set of initial committers on the project. The project's going to
> evolve
> > > over time based on who's working on it.
> > >
> > > John
> > >
> > > On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch <at...@gmail.com>
> > > wrote:
> > >
> > >> Hi, on the other side, we could also just let it be where it is and I
> > will
> > >> adapt things along our discussions
> > >> All we have discussed so far are no complete overhauls of everything.
> > The
> > >> API part is quite near already what we discuss, and the
> > >> core gets simplifies as well, when things are getting clearer.
> > >> This has the advantage that we have more or less rather quickly a
> > running
> > >> API/implementation, which does, what we have discussed. Similarly we
> > have a
> > >> POC that we really have understood everything correctly. When I see
> > that
> > >> something is getting completely differently, I would remove the
> > >> implementation parts and e.g. just add the interfaces as needed, so
> we
> > see
> > >> that we have work to be done in the corresponding areas.
> > >>
> > >> I think that way, we get both the advantage to have some kind of POC
> > and on
> > >> the other side the open discussions about how things should be...
> > >>
> > >>
> > >> 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rmannibucau@gmail.com
> >:
> > >>
> > >> > We rediscuss basis and startiong from an empty branch while we dont
> > > agree
> > >> > is sane IMO....also means we dont import anything while it is not
> the
> > >> case.
> > >> > Le 27 déc. 2014 15:15, "John D. Ament"
> > > <jo...@apache.org> a écrit :
> > >> >
> > >> > > No it's not a technical requirement. There's no reason
> > > to move stuff
> > >> for
> > >> > > the same of moving stuff.
> > >> > >
> > >> > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg
> > > <st...@yahoo.de>
> > >> > wrote:
> > >> > >
> > >> > > > The reason for moving it into a separate directory is purely
> > >> technical.
> > >> > > > That way it is handled with GIT is that you can move files
> > > _without_
> > >> > > > loosing history of the content.
> > >> > > >
> > >> > > > So while we are moving bits back we still do have all the
> > > original
> > >> > > > contributions in place without loosing anything. It's
> > > just easier to
> > >> > move
> > >> > > > around files inside the same branch than to switch between
> > > multiple
> > >> of
> > >> > > > those.
> > >> > > >
> > >> > > >
> > >> > > > LieGrue,
> > >> > > > strub
> > >> > > >
> > >> > > >
> > >> > > >
> > >> > > >
> > >> > > >
> > >> > > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
> > >> > > werner.keil@gmail.com>
> > >> > > > wrote:
> > >> > > > > > E.g. DeviceMap emerged over time, even with
> > > multiple languages
> > >> > other
> > >> > > > than
> > >> > > > > Java, though you could call some of it
> > > "dormant".
> > >> > > > >
> > >> > > > > What we did there is a "contrib" section
> > > where new parts gradually
> > >> > got
> > >> > > > > refactored from, but as Tamaya already got a new
> > > structure,
> > >> > licensing,
> > >> > > > etc.
> > >> > > > > calling its root "contrib",
> > > "dormant" or
> > >> > > > > "stuffSomeDontLike" seems rather
> > >> > > > > useless.
> > >> > > > >
> > >> > > > > Better tag or even branch, then refactor what's
> > > worth changing and
> > >> > > merge
> > >> > > > > the results back if ALL who care about it agree
> > > it's good.
> > >> > > > > Not sure, if Apache uses e.g. Gerrit, but it feels some
> > > of us could
> > >> > use
> > >> > > > >
> > >> > > > > that;-)
> > >> > > > >
> > >> > > > > Werner
> > >> > > > >
> > >> > > >
> > >> > >
> > >> >
> > >>
> > >>
> > >>
> > >> --
> > >> *Anatole Tresch*
> > >> Java Engineer & Architect, JSR Spec Lead
> > >> Glärnischweg 10
> > >> CH - 8620 Wetzikon
> > >>
> > >> *Switzerland, Europe Zurich, GMT+1*
> > >> *Twitter: @atsticks*
> > >> *Blogs: **http://javaremarkables.blogspot.ch/
> > >> <http://javaremarkables.blogspot.ch/>*
> > >>
> > >> *Google: atsticksMobile +41-76 344 62 79*
> > >>
> > >
> >
>
Re: Complexity Discussion
Posted by "John D. Ament" <jo...@apache.org>.
On Sun Dec 28 2014 at 10:07:05 AM Mark Struberg <st...@yahoo.de> wrote:
> John, the main intention of this podling _never_ was to polish an existing
> code base but to aim for a JSR and provide a RI for it.
>
>
Please double check the proposal.
http://wiki.apache.org/incubator/TamayaProposal
"At a later stage, if successful, standardizing it within a JSR also may be
an option"
>
> This sometimes includes going back and forth a few times to review ideas
> and get rid of dead ends.
>
> LieGrue,
> strub
>
>
>
>
>
> > On Sunday, 28 December 2014, 14:10, John D. Ament <jo...@apache.org>
> wrote:
> > > So maybe two branches of the code base?
> >
> > The original sitting on a 0.1-prototype and maybe the new stuff on a
> > 0.2-prototype branch (since I saw that Mark had set his code as
> > 0.2-snapshot)? This way no one uses master for a little while until all
> > this gets settled and if one branch wants stuff from the other they can
> > rebase?
> >
> > We do have to remember that the original intention of this podling, like
> > pretty much all other podlings, is to start from a donated code base w/ a
> > set of initial committers on the project. The project's going to evolve
> > over time based on who's working on it.
> >
> > John
> >
> > On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch <at...@gmail.com>
> > wrote:
> >
> >> Hi, on the other side, we could also just let it be where it is and I
> will
> >> adapt things along our discussions
> >> All we have discussed so far are no complete overhauls of everything.
> The
> >> API part is quite near already what we discuss, and the
> >> core gets simplifies as well, when things are getting clearer.
> >> This has the advantage that we have more or less rather quickly a
> running
> >> API/implementation, which does, what we have discussed. Similarly we
> have a
> >> POC that we really have understood everything correctly. When I see
> that
> >> something is getting completely differently, I would remove the
> >> implementation parts and e.g. just add the interfaces as needed, so we
> see
> >> that we have work to be done in the corresponding areas.
> >>
> >> I think that way, we get both the advantage to have some kind of POC
> and on
> >> the other side the open discussions about how things should be...
> >>
> >>
> >> 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rm...@gmail.com>:
> >>
> >> > We rediscuss basis and startiong from an empty branch while we dont
> > agree
> >> > is sane IMO....also means we dont import anything while it is not the
> >> case.
> >> > Le 27 déc. 2014 15:15, "John D. Ament"
> > <jo...@apache.org> a écrit :
> >> >
> >> > > No it's not a technical requirement. There's no reason
> > to move stuff
> >> for
> >> > > the same of moving stuff.
> >> > >
> >> > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg
> > <st...@yahoo.de>
> >> > wrote:
> >> > >
> >> > > > The reason for moving it into a separate directory is purely
> >> technical.
> >> > > > That way it is handled with GIT is that you can move files
> > _without_
> >> > > > loosing history of the content.
> >> > > >
> >> > > > So while we are moving bits back we still do have all the
> > original
> >> > > > contributions in place without loosing anything. It's
> > just easier to
> >> > move
> >> > > > around files inside the same branch than to switch between
> > multiple
> >> of
> >> > > > those.
> >> > > >
> >> > > >
> >> > > > LieGrue,
> >> > > > strub
> >> > > >
> >> > > >
> >> > > >
> >> > > >
> >> > > >
> >> > > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
> >> > > werner.keil@gmail.com>
> >> > > > wrote:
> >> > > > > > E.g. DeviceMap emerged over time, even with
> > multiple languages
> >> > other
> >> > > > than
> >> > > > > Java, though you could call some of it
> > "dormant".
> >> > > > >
> >> > > > > What we did there is a "contrib" section
> > where new parts gradually
> >> > got
> >> > > > > refactored from, but as Tamaya already got a new
> > structure,
> >> > licensing,
> >> > > > etc.
> >> > > > > calling its root "contrib",
> > "dormant" or
> >> > > > > "stuffSomeDontLike" seems rather
> >> > > > > useless.
> >> > > > >
> >> > > > > Better tag or even branch, then refactor what's
> > worth changing and
> >> > > merge
> >> > > > > the results back if ALL who care about it agree
> > it's good.
> >> > > > > Not sure, if Apache uses e.g. Gerrit, but it feels some
> > of us could
> >> > use
> >> > > > >
> >> > > > > that;-)
> >> > > > >
> >> > > > > Werner
> >> > > > >
> >> > > >
> >> > >
> >> >
> >>
> >>
> >>
> >> --
> >> *Anatole Tresch*
> >> Java Engineer & Architect, JSR Spec Lead
> >> Glärnischweg 10
> >> CH - 8620 Wetzikon
> >>
> >> *Switzerland, Europe Zurich, GMT+1*
> >> *Twitter: @atsticks*
> >> *Blogs: **http://javaremarkables.blogspot.ch/
> >> <http://javaremarkables.blogspot.ch/>*
> >>
> >> *Google: atsticksMobile +41-76 344 62 79*
> >>
> >
>
Re: Complexity Discussion
Posted by Mark Struberg <st...@yahoo.de>.
John, the main intention of this podling _never_ was to polish an existing code base but to aim for a JSR and provide a RI for it.
This sometimes includes going back and forth a few times to review ideas and get rid of dead ends.
LieGrue,
strub
> On Sunday, 28 December 2014, 14:10, John D. Ament <jo...@apache.org> wrote:
> > So maybe two branches of the code base?
>
> The original sitting on a 0.1-prototype and maybe the new stuff on a
> 0.2-prototype branch (since I saw that Mark had set his code as
> 0.2-snapshot)? This way no one uses master for a little while until all
> this gets settled and if one branch wants stuff from the other they can
> rebase?
>
> We do have to remember that the original intention of this podling, like
> pretty much all other podlings, is to start from a donated code base w/ a
> set of initial committers on the project. The project's going to evolve
> over time based on who's working on it.
>
> John
>
> On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch <at...@gmail.com>
> wrote:
>
>> Hi, on the other side, we could also just let it be where it is and I will
>> adapt things along our discussions
>> All we have discussed so far are no complete overhauls of everything. The
>> API part is quite near already what we discuss, and the
>> core gets simplifies as well, when things are getting clearer.
>> This has the advantage that we have more or less rather quickly a running
>> API/implementation, which does, what we have discussed. Similarly we have a
>> POC that we really have understood everything correctly. When I see that
>> something is getting completely differently, I would remove the
>> implementation parts and e.g. just add the interfaces as needed, so we see
>> that we have work to be done in the corresponding areas.
>>
>> I think that way, we get both the advantage to have some kind of POC and on
>> the other side the open discussions about how things should be...
>>
>>
>> 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rm...@gmail.com>:
>>
>> > We rediscuss basis and startiong from an empty branch while we dont
> agree
>> > is sane IMO....also means we dont import anything while it is not the
>> case.
>> > Le 27 déc. 2014 15:15, "John D. Ament"
> <jo...@apache.org> a écrit :
>> >
>> > > No it's not a technical requirement. There's no reason
> to move stuff
>> for
>> > > the same of moving stuff.
>> > >
>> > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg
> <st...@yahoo.de>
>> > wrote:
>> > >
>> > > > The reason for moving it into a separate directory is purely
>> technical.
>> > > > That way it is handled with GIT is that you can move files
> _without_
>> > > > loosing history of the content.
>> > > >
>> > > > So while we are moving bits back we still do have all the
> original
>> > > > contributions in place without loosing anything. It's
> just easier to
>> > move
>> > > > around files inside the same branch than to switch between
> multiple
>> of
>> > > > those.
>> > > >
>> > > >
>> > > > LieGrue,
>> > > > strub
>> > > >
>> > > >
>> > > >
>> > > >
>> > > >
>> > > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
>> > > werner.keil@gmail.com>
>> > > > wrote:
>> > > > > > E.g. DeviceMap emerged over time, even with
> multiple languages
>> > other
>> > > > than
>> > > > > Java, though you could call some of it
> "dormant".
>> > > > >
>> > > > > What we did there is a "contrib" section
> where new parts gradually
>> > got
>> > > > > refactored from, but as Tamaya already got a new
> structure,
>> > licensing,
>> > > > etc.
>> > > > > calling its root "contrib",
> "dormant" or
>> > > > > "stuffSomeDontLike" seems rather
>> > > > > useless.
>> > > > >
>> > > > > Better tag or even branch, then refactor what's
> worth changing and
>> > > merge
>> > > > > the results back if ALL who care about it agree
> it's good.
>> > > > > Not sure, if Apache uses e.g. Gerrit, but it feels some
> of us could
>> > use
>> > > > >
>> > > > > that;-)
>> > > > >
>> > > > > Werner
>> > > > >
>> > > >
>> > >
>> >
>>
>>
>>
>> --
>> *Anatole Tresch*
>> Java Engineer & Architect, JSR Spec Lead
>> Glärnischweg 10
>> CH - 8620 Wetzikon
>>
>> *Switzerland, Europe Zurich, GMT+1*
>> *Twitter: @atsticks*
>> *Blogs: **http://javaremarkables.blogspot.ch/
>> <http://javaremarkables.blogspot.ch/>*
>>
>> *Google: atsticksMobile +41-76 344 62 79*
>>
>
Re: Complexity Discussion
Posted by "John D. Ament" <jo...@apache.org>.
On Sun Dec 28 2014 at 10:04:48 AM Mark Struberg <st...@yahoo.de> wrote:
> As I said in my original post: the goal of moving all to dormant for now
> was never to delete all the work which has been done!
> The reason was just to have a clean sheet and move back parts we agree to
> be in a good shape. But really only the parts we really need. No bells and
> whistles, just really the important stuff.
>
Yes, and the ideal way to do that is via a vote. What actually happened
was you said (paraphrased) "let's move the source around" and then a couple
of hours later started doing it. This clearly goes against the ASF notion
of consensus - ( for those unfamiliar; The Apache Way:
http://www.apache.org/foundation/how-it-works.html )
In the interim, Anatole, Werner and myself all said (paraphrased) "probably
not a good idea."
> LieGrue,
> strub
>
>
>
>
>
> > On Sunday, 28 December 2014, 16:00, Anatole Tresch <at...@gmail.com>
> wrote:
> > > Well, should be .[..but at least have now good discussions]. Small but
> > important point...
> >
> > Anatole Tresch <at...@gmail.com> schrieb am Sun Dec 28 2014 at
> > 15:20:15:
> >
> >> I think that would be better than move all to dormant. Nevertheless
> still
> >> not satisfied with the way it was done by Mark, but at least have no
> good
> >> discussions, so I will not create elephants from flies... ;)
> >>
> >>
> >> John D. Ament <jo...@apache.org> schrieb am Sun Dec 28 2014 at
> >> 14:10:21:
> >>
> >> So maybe two branches of the code base?
> >>>
> >>> The original sitting on a 0.1-prototype and maybe the new stuff on a
> >>> 0.2-prototype branch (since I saw that Mark had set his code as
> >>> 0.2-snapshot)? This way no one uses master for a little while until
> > all
> >>> this gets settled and if one branch wants stuff from the other they
> can
> >>> rebase?
> >>>
> >>> We do have to remember that the original intention of this podling,
> > like
> >>> pretty much all other podlings, is to start from a donated code base
> w/
> > a
> >>> set of initial committers on the project. The project's going to
> > evolve
> >>> over time based on who's working on it.
> >>>
> >>> John
> >>>
> >>> On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch
> > <at...@gmail.com>
> >>> wrote:
> >>>
> >>> > Hi, on the other side, we could also just let it be where it is
> > and I
> >>> will
> >>> > adapt things along our discussions
> >>> > All we have discussed so far are no complete overhauls of
> > everything.
> >>> The
> >>> > API part is quite near already what we discuss, and the
> >>> > core gets simplifies as well, when things are getting clearer.
> >>> > This has the advantage that we have more or less rather quickly a
> >>> running
> >>> > API/implementation, which does, what we have discussed. Similarly
> > we
> >>> have a
> >>> > POC that we really have understood everything correctly. When I
> > see that
> >>> > something is getting completely differently, I would remove the
> >>> > implementation parts and e.g. just add the interfaces as needed,
> > so we
> >>> see
> >>> > that we have work to be done in the corresponding areas.
> >>> >
> >>> > I think that way, we get both the advantage to have some kind of
> > POC
> >>> and on
> >>> > the other side the open discussions about how things should be...
> >>> >
> >>> >
> >>> > 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau
> > <rm...@gmail.com>:
> >>> >
> >>> > > We rediscuss basis and startiong from an empty branch while
> > we dont
> >>> agree
> >>> > > is sane IMO....also means we dont import anything while it is
> > not the
> >>> > case.
> >>> > > Le 27 déc. 2014 15:15, "John D. Ament"
> > <jo...@apache.org> a
> >>> écrit :
> >>> > >
> >>> > > > No it's not a technical requirement. There's no
> > reason to move
> >>> stuff
> >>> > for
> >>> > > > the same of moving stuff.
> >>> > > >
> >>> > > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg
> > <st...@yahoo.de>
> >>> > > wrote:
> >>> > > >
> >>> > > > > The reason for moving it into a separate directory
> > is purely
> >>> > technical.
> >>> > > > > That way it is handled with GIT is that you can
> > move files
> >>> _without_
> >>> > > > > loosing history of the content.
> >>> > > > >
> >>> > > > > So while we are moving bits back we still do have
> > all the original
> >>> > > > > contributions in place without loosing anything.
> > It's just easier
> >>> to
> >>> > > move
> >>> > > > > around files inside the same branch than to switch
> > between
> >>> multiple
> >>> > of
> >>> > > > > those.
> >>> > > > >
> >>> > > > >
> >>> > > > > LieGrue,
> >>> > > > > strub
> >>> > > > >
> >>> > > > >
> >>> > > > >
> >>> > > > >
> >>> > > > >
> >>> > > > > > On Saturday, 27 December 2014, 14:28, Werner
> > Keil <
> >>> > > > werner.keil@gmail.com>
> >>> > > > > wrote:
> >>> > > > > > > E.g. DeviceMap emerged over time, even
> > with multiple languages
> >>> > > other
> >>> > > > > than
> >>> > > > > > Java, though you could call some of it
> > "dormant".
> >>> > > > > >
> >>> > > > > > What we did there is a "contrib"
> > section where new parts
> >>> gradually
> >>> > > got
> >>> > > > > > refactored from, but as Tamaya already got a
> > new structure,
> >>> > > licensing,
> >>> > > > > etc.
> >>> > > > > > calling its root "contrib",
> > "dormant" or
> >>> > > > > > "stuffSomeDontLike" seems rather
> >>> > > > > > useless.
> >>> > > > > >
> >>> > > > > > Better tag or even branch, then refactor
> > what's worth changing
> >>> and
> >>> > > > merge
> >>> > > > > > the results back if ALL who care about it
> > agree it's good.
> >>> > > > > > Not sure, if Apache uses e.g. Gerrit, but it
> > feels some of us
> >>> could
> >>> > > use
> >>> > > > > >
> >>> > > > > > that;-)
> >>> > > > > >
> >>> > > > > > Werner
> >>> > > > > >
> >>> > > > >
> >>> > > >
> >>> > >
> >>> >
> >>> >
> >>> >
> >>> > --
> >>> > *Anatole Tresch*
> >>> > Java Engineer & Architect, JSR Spec Lead
> >>> > Glärnischweg 10
> >>> > CH - 8620 Wetzikon
> >>> >
> >>> > *Switzerland, Europe Zurich, GMT+1*
> >>> > *Twitter: @atsticks*
> >>> > *Blogs: **http://javaremarkables.blogspot.ch/
> >>> > <http://javaremarkables.blogspot.ch/>*
> >>> >
> >>> > *Google: atsticksMobile +41-76 344 62 79*
> >>> >
> >>>
> >>
> >
>
Re: Complexity Discussion
Posted by Mark Struberg <st...@yahoo.de>.
As I said in my original post: the goal of moving all to dormant for now was never to delete all the work which has been done!
The reason was just to have a clean sheet and move back parts we agree to be in a good shape. But really only the parts we really need. No bells and whistles, just really the important stuff.
LieGrue,
strub
> On Sunday, 28 December 2014, 16:00, Anatole Tresch <at...@gmail.com> wrote:
> > Well, should be .[..but at least have now good discussions]. Small but
> important point...
>
> Anatole Tresch <at...@gmail.com> schrieb am Sun Dec 28 2014 at
> 15:20:15:
>
>> I think that would be better than move all to dormant. Nevertheless still
>> not satisfied with the way it was done by Mark, but at least have no good
>> discussions, so I will not create elephants from flies... ;)
>>
>>
>> John D. Ament <jo...@apache.org> schrieb am Sun Dec 28 2014 at
>> 14:10:21:
>>
>> So maybe two branches of the code base?
>>>
>>> The original sitting on a 0.1-prototype and maybe the new stuff on a
>>> 0.2-prototype branch (since I saw that Mark had set his code as
>>> 0.2-snapshot)? This way no one uses master for a little while until
> all
>>> this gets settled and if one branch wants stuff from the other they can
>>> rebase?
>>>
>>> We do have to remember that the original intention of this podling,
> like
>>> pretty much all other podlings, is to start from a donated code base w/
> a
>>> set of initial committers on the project. The project's going to
> evolve
>>> over time based on who's working on it.
>>>
>>> John
>>>
>>> On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch
> <at...@gmail.com>
>>> wrote:
>>>
>>> > Hi, on the other side, we could also just let it be where it is
> and I
>>> will
>>> > adapt things along our discussions
>>> > All we have discussed so far are no complete overhauls of
> everything.
>>> The
>>> > API part is quite near already what we discuss, and the
>>> > core gets simplifies as well, when things are getting clearer.
>>> > This has the advantage that we have more or less rather quickly a
>>> running
>>> > API/implementation, which does, what we have discussed. Similarly
> we
>>> have a
>>> > POC that we really have understood everything correctly. When I
> see that
>>> > something is getting completely differently, I would remove the
>>> > implementation parts and e.g. just add the interfaces as needed,
> so we
>>> see
>>> > that we have work to be done in the corresponding areas.
>>> >
>>> > I think that way, we get both the advantage to have some kind of
> POC
>>> and on
>>> > the other side the open discussions about how things should be...
>>> >
>>> >
>>> > 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau
> <rm...@gmail.com>:
>>> >
>>> > > We rediscuss basis and startiong from an empty branch while
> we dont
>>> agree
>>> > > is sane IMO....also means we dont import anything while it is
> not the
>>> > case.
>>> > > Le 27 déc. 2014 15:15, "John D. Ament"
> <jo...@apache.org> a
>>> écrit :
>>> > >
>>> > > > No it's not a technical requirement. There's no
> reason to move
>>> stuff
>>> > for
>>> > > > the same of moving stuff.
>>> > > >
>>> > > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg
> <st...@yahoo.de>
>>> > > wrote:
>>> > > >
>>> > > > > The reason for moving it into a separate directory
> is purely
>>> > technical.
>>> > > > > That way it is handled with GIT is that you can
> move files
>>> _without_
>>> > > > > loosing history of the content.
>>> > > > >
>>> > > > > So while we are moving bits back we still do have
> all the original
>>> > > > > contributions in place without loosing anything.
> It's just easier
>>> to
>>> > > move
>>> > > > > around files inside the same branch than to switch
> between
>>> multiple
>>> > of
>>> > > > > those.
>>> > > > >
>>> > > > >
>>> > > > > LieGrue,
>>> > > > > strub
>>> > > > >
>>> > > > >
>>> > > > >
>>> > > > >
>>> > > > >
>>> > > > > > On Saturday, 27 December 2014, 14:28, Werner
> Keil <
>>> > > > werner.keil@gmail.com>
>>> > > > > wrote:
>>> > > > > > > E.g. DeviceMap emerged over time, even
> with multiple languages
>>> > > other
>>> > > > > than
>>> > > > > > Java, though you could call some of it
> "dormant".
>>> > > > > >
>>> > > > > > What we did there is a "contrib"
> section where new parts
>>> gradually
>>> > > got
>>> > > > > > refactored from, but as Tamaya already got a
> new structure,
>>> > > licensing,
>>> > > > > etc.
>>> > > > > > calling its root "contrib",
> "dormant" or
>>> > > > > > "stuffSomeDontLike" seems rather
>>> > > > > > useless.
>>> > > > > >
>>> > > > > > Better tag or even branch, then refactor
> what's worth changing
>>> and
>>> > > > merge
>>> > > > > > the results back if ALL who care about it
> agree it's good.
>>> > > > > > Not sure, if Apache uses e.g. Gerrit, but it
> feels some of us
>>> could
>>> > > use
>>> > > > > >
>>> > > > > > that;-)
>>> > > > > >
>>> > > > > > Werner
>>> > > > > >
>>> > > > >
>>> > > >
>>> > >
>>> >
>>> >
>>> >
>>> > --
>>> > *Anatole Tresch*
>>> > Java Engineer & Architect, JSR Spec Lead
>>> > Glärnischweg 10
>>> > CH - 8620 Wetzikon
>>> >
>>> > *Switzerland, Europe Zurich, GMT+1*
>>> > *Twitter: @atsticks*
>>> > *Blogs: **http://javaremarkables.blogspot.ch/
>>> > <http://javaremarkables.blogspot.ch/>*
>>> >
>>> > *Google: atsticksMobile +41-76 344 62 79*
>>> >
>>>
>>
>
Re: Complexity Discussion
Posted by Anatole Tresch <at...@gmail.com>.
Well, should be .[..but at least have now good discussions]. Small but
important point...
Anatole Tresch <at...@gmail.com> schrieb am Sun Dec 28 2014 at 15:20:15:
> I think that would be better than move all to dormant. Nevertheless still
> not satisfied with the way it was done by Mark, but at least have no good
> discussions, so I will not create elephants from flies... ;)
>
>
> John D. Ament <jo...@apache.org> schrieb am Sun Dec 28 2014 at
> 14:10:21:
>
> So maybe two branches of the code base?
>>
>> The original sitting on a 0.1-prototype and maybe the new stuff on a
>> 0.2-prototype branch (since I saw that Mark had set his code as
>> 0.2-snapshot)? This way no one uses master for a little while until all
>> this gets settled and if one branch wants stuff from the other they can
>> rebase?
>>
>> We do have to remember that the original intention of this podling, like
>> pretty much all other podlings, is to start from a donated code base w/ a
>> set of initial committers on the project. The project's going to evolve
>> over time based on who's working on it.
>>
>> John
>>
>> On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch <at...@gmail.com>
>> wrote:
>>
>> > Hi, on the other side, we could also just let it be where it is and I
>> will
>> > adapt things along our discussions
>> > All we have discussed so far are no complete overhauls of everything.
>> The
>> > API part is quite near already what we discuss, and the
>> > core gets simplifies as well, when things are getting clearer.
>> > This has the advantage that we have more or less rather quickly a
>> running
>> > API/implementation, which does, what we have discussed. Similarly we
>> have a
>> > POC that we really have understood everything correctly. When I see that
>> > something is getting completely differently, I would remove the
>> > implementation parts and e.g. just add the interfaces as needed, so we
>> see
>> > that we have work to be done in the corresponding areas.
>> >
>> > I think that way, we get both the advantage to have some kind of POC
>> and on
>> > the other side the open discussions about how things should be...
>> >
>> >
>> > 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rm...@gmail.com>:
>> >
>> > > We rediscuss basis and startiong from an empty branch while we dont
>> agree
>> > > is sane IMO....also means we dont import anything while it is not the
>> > case.
>> > > Le 27 déc. 2014 15:15, "John D. Ament" <jo...@apache.org> a
>> écrit :
>> > >
>> > > > No it's not a technical requirement. There's no reason to move
>> stuff
>> > for
>> > > > the same of moving stuff.
>> > > >
>> > > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg <st...@yahoo.de>
>> > > wrote:
>> > > >
>> > > > > The reason for moving it into a separate directory is purely
>> > technical.
>> > > > > That way it is handled with GIT is that you can move files
>> _without_
>> > > > > loosing history of the content.
>> > > > >
>> > > > > So while we are moving bits back we still do have all the original
>> > > > > contributions in place without loosing anything. It's just easier
>> to
>> > > move
>> > > > > around files inside the same branch than to switch between
>> multiple
>> > of
>> > > > > those.
>> > > > >
>> > > > >
>> > > > > LieGrue,
>> > > > > strub
>> > > > >
>> > > > >
>> > > > >
>> > > > >
>> > > > >
>> > > > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
>> > > > werner.keil@gmail.com>
>> > > > > wrote:
>> > > > > > > E.g. DeviceMap emerged over time, even with multiple languages
>> > > other
>> > > > > than
>> > > > > > Java, though you could call some of it "dormant".
>> > > > > >
>> > > > > > What we did there is a "contrib" section where new parts
>> gradually
>> > > got
>> > > > > > refactored from, but as Tamaya already got a new structure,
>> > > licensing,
>> > > > > etc.
>> > > > > > calling its root "contrib", "dormant" or
>> > > > > > "stuffSomeDontLike" seems rather
>> > > > > > useless.
>> > > > > >
>> > > > > > Better tag or even branch, then refactor what's worth changing
>> and
>> > > > merge
>> > > > > > the results back if ALL who care about it agree it's good.
>> > > > > > Not sure, if Apache uses e.g. Gerrit, but it feels some of us
>> could
>> > > use
>> > > > > >
>> > > > > > that;-)
>> > > > > >
>> > > > > > Werner
>> > > > > >
>> > > > >
>> > > >
>> > >
>> >
>> >
>> >
>> > --
>> > *Anatole Tresch*
>> > Java Engineer & Architect, JSR Spec Lead
>> > Glärnischweg 10
>> > CH - 8620 Wetzikon
>> >
>> > *Switzerland, Europe Zurich, GMT+1*
>> > *Twitter: @atsticks*
>> > *Blogs: **http://javaremarkables.blogspot.ch/
>> > <http://javaremarkables.blogspot.ch/>*
>> >
>> > *Google: atsticksMobile +41-76 344 62 79*
>> >
>>
>
Re: Complexity Discussion
Posted by Anatole Tresch <at...@gmail.com>.
I think that would be better than move all to dormant. Nevertheless still
not satisfied with the way it was done by Mark, but at least have no good
discussions, so I will not create elephants from flies... ;)
John D. Ament <jo...@apache.org> schrieb am Sun Dec 28 2014 at
14:10:21:
> So maybe two branches of the code base?
>
> The original sitting on a 0.1-prototype and maybe the new stuff on a
> 0.2-prototype branch (since I saw that Mark had set his code as
> 0.2-snapshot)? This way no one uses master for a little while until all
> this gets settled and if one branch wants stuff from the other they can
> rebase?
>
> We do have to remember that the original intention of this podling, like
> pretty much all other podlings, is to start from a donated code base w/ a
> set of initial committers on the project. The project's going to evolve
> over time based on who's working on it.
>
> John
>
> On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch <at...@gmail.com>
> wrote:
>
> > Hi, on the other side, we could also just let it be where it is and I
> will
> > adapt things along our discussions
> > All we have discussed so far are no complete overhauls of everything. The
> > API part is quite near already what we discuss, and the
> > core gets simplifies as well, when things are getting clearer.
> > This has the advantage that we have more or less rather quickly a running
> > API/implementation, which does, what we have discussed. Similarly we
> have a
> > POC that we really have understood everything correctly. When I see that
> > something is getting completely differently, I would remove the
> > implementation parts and e.g. just add the interfaces as needed, so we
> see
> > that we have work to be done in the corresponding areas.
> >
> > I think that way, we get both the advantage to have some kind of POC and
> on
> > the other side the open discussions about how things should be...
> >
> >
> > 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rm...@gmail.com>:
> >
> > > We rediscuss basis and startiong from an empty branch while we dont
> agree
> > > is sane IMO....also means we dont import anything while it is not the
> > case.
> > > Le 27 déc. 2014 15:15, "John D. Ament" <jo...@apache.org> a
> écrit :
> > >
> > > > No it's not a technical requirement. There's no reason to move stuff
> > for
> > > > the same of moving stuff.
> > > >
> > > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg <st...@yahoo.de>
> > > wrote:
> > > >
> > > > > The reason for moving it into a separate directory is purely
> > technical.
> > > > > That way it is handled with GIT is that you can move files
> _without_
> > > > > loosing history of the content.
> > > > >
> > > > > So while we are moving bits back we still do have all the original
> > > > > contributions in place without loosing anything. It's just easier
> to
> > > move
> > > > > around files inside the same branch than to switch between multiple
> > of
> > > > > those.
> > > > >
> > > > >
> > > > > LieGrue,
> > > > > strub
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
> > > > werner.keil@gmail.com>
> > > > > wrote:
> > > > > > > E.g. DeviceMap emerged over time, even with multiple languages
> > > other
> > > > > than
> > > > > > Java, though you could call some of it "dormant".
> > > > > >
> > > > > > What we did there is a "contrib" section where new parts
> gradually
> > > got
> > > > > > refactored from, but as Tamaya already got a new structure,
> > > licensing,
> > > > > etc.
> > > > > > calling its root "contrib", "dormant" or
> > > > > > "stuffSomeDontLike" seems rather
> > > > > > useless.
> > > > > >
> > > > > > Better tag or even branch, then refactor what's worth changing
> and
> > > > merge
> > > > > > the results back if ALL who care about it agree it's good.
> > > > > > Not sure, if Apache uses e.g. Gerrit, but it feels some of us
> could
> > > use
> > > > > >
> > > > > > that;-)
> > > > > >
> > > > > > Werner
> > > > > >
> > > > >
> > > >
> > >
> >
> >
> >
> > --
> > *Anatole Tresch*
> > Java Engineer & Architect, JSR Spec Lead
> > Glärnischweg 10
> > CH - 8620 Wetzikon
> >
> > *Switzerland, Europe Zurich, GMT+1*
> > *Twitter: @atsticks*
> > *Blogs: **http://javaremarkables.blogspot.ch/
> > <http://javaremarkables.blogspot.ch/>*
> >
> > *Google: atsticksMobile +41-76 344 62 79*
> >
>
Re: Complexity Discussion
Posted by "John D. Ament" <jo...@apache.org>.
So maybe two branches of the code base?
The original sitting on a 0.1-prototype and maybe the new stuff on a
0.2-prototype branch (since I saw that Mark had set his code as
0.2-snapshot)? This way no one uses master for a little while until all
this gets settled and if one branch wants stuff from the other they can
rebase?
We do have to remember that the original intention of this podling, like
pretty much all other podlings, is to start from a donated code base w/ a
set of initial committers on the project. The project's going to evolve
over time based on who's working on it.
John
On Sun Dec 28 2014 at 8:04:51 AM Anatole Tresch <at...@gmail.com> wrote:
> Hi, on the other side, we could also just let it be where it is and I will
> adapt things along our discussions
> All we have discussed so far are no complete overhauls of everything. The
> API part is quite near already what we discuss, and the
> core gets simplifies as well, when things are getting clearer.
> This has the advantage that we have more or less rather quickly a running
> API/implementation, which does, what we have discussed. Similarly we have a
> POC that we really have understood everything correctly. When I see that
> something is getting completely differently, I would remove the
> implementation parts and e.g. just add the interfaces as needed, so we see
> that we have work to be done in the corresponding areas.
>
> I think that way, we get both the advantage to have some kind of POC and on
> the other side the open discussions about how things should be...
>
>
> 2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rm...@gmail.com>:
>
> > We rediscuss basis and startiong from an empty branch while we dont agree
> > is sane IMO....also means we dont import anything while it is not the
> case.
> > Le 27 déc. 2014 15:15, "John D. Ament" <jo...@apache.org> a écrit :
> >
> > > No it's not a technical requirement. There's no reason to move stuff
> for
> > > the same of moving stuff.
> > >
> > > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg <st...@yahoo.de>
> > wrote:
> > >
> > > > The reason for moving it into a separate directory is purely
> technical.
> > > > That way it is handled with GIT is that you can move files _without_
> > > > loosing history of the content.
> > > >
> > > > So while we are moving bits back we still do have all the original
> > > > contributions in place without loosing anything. It's just easier to
> > move
> > > > around files inside the same branch than to switch between multiple
> of
> > > > those.
> > > >
> > > >
> > > > LieGrue,
> > > > strub
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
> > > werner.keil@gmail.com>
> > > > wrote:
> > > > > > E.g. DeviceMap emerged over time, even with multiple languages
> > other
> > > > than
> > > > > Java, though you could call some of it "dormant".
> > > > >
> > > > > What we did there is a "contrib" section where new parts gradually
> > got
> > > > > refactored from, but as Tamaya already got a new structure,
> > licensing,
> > > > etc.
> > > > > calling its root "contrib", "dormant" or
> > > > > "stuffSomeDontLike" seems rather
> > > > > useless.
> > > > >
> > > > > Better tag or even branch, then refactor what's worth changing and
> > > merge
> > > > > the results back if ALL who care about it agree it's good.
> > > > > Not sure, if Apache uses e.g. Gerrit, but it feels some of us could
> > use
> > > > >
> > > > > that;-)
> > > > >
> > > > > Werner
> > > > >
> > > >
> > >
> >
>
>
>
> --
> *Anatole Tresch*
> Java Engineer & Architect, JSR Spec Lead
> Glärnischweg 10
> CH - 8620 Wetzikon
>
> *Switzerland, Europe Zurich, GMT+1*
> *Twitter: @atsticks*
> *Blogs: **http://javaremarkables.blogspot.ch/
> <http://javaremarkables.blogspot.ch/>*
>
> *Google: atsticksMobile +41-76 344 62 79*
>
Re: Complexity Discussion
Posted by Anatole Tresch <at...@gmail.com>.
Hi, on the other side, we could also just let it be where it is and I will
adapt things along our discussions
All we have discussed so far are no complete overhauls of everything. The
API part is quite near already what we discuss, and the
core gets simplifies as well, when things are getting clearer.
This has the advantage that we have more or less rather quickly a running
API/implementation, which does, what we have discussed. Similarly we have a
POC that we really have understood everything correctly. When I see that
something is getting completely differently, I would remove the
implementation parts and e.g. just add the interfaces as needed, so we see
that we have work to be done in the corresponding areas.
I think that way, we get both the advantage to have some kind of POC and on
the other side the open discussions about how things should be...
2014-12-28 9:56 GMT+01:00 Romain Manni-Bucau <rm...@gmail.com>:
> We rediscuss basis and startiong from an empty branch while we dont agree
> is sane IMO....also means we dont import anything while it is not the case.
> Le 27 déc. 2014 15:15, "John D. Ament" <jo...@apache.org> a écrit :
>
> > No it's not a technical requirement. There's no reason to move stuff for
> > the same of moving stuff.
> >
> > On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg <st...@yahoo.de>
> wrote:
> >
> > > The reason for moving it into a separate directory is purely technical.
> > > That way it is handled with GIT is that you can move files _without_
> > > loosing history of the content.
> > >
> > > So while we are moving bits back we still do have all the original
> > > contributions in place without loosing anything. It's just easier to
> move
> > > around files inside the same branch than to switch between multiple of
> > > those.
> > >
> > >
> > > LieGrue,
> > > strub
> > >
> > >
> > >
> > >
> > >
> > > > On Saturday, 27 December 2014, 14:28, Werner Keil <
> > werner.keil@gmail.com>
> > > wrote:
> > > > > E.g. DeviceMap emerged over time, even with multiple languages
> other
> > > than
> > > > Java, though you could call some of it "dormant".
> > > >
> > > > What we did there is a "contrib" section where new parts gradually
> got
> > > > refactored from, but as Tamaya already got a new structure,
> licensing,
> > > etc.
> > > > calling its root "contrib", "dormant" or
> > > > "stuffSomeDontLike" seems rather
> > > > useless.
> > > >
> > > > Better tag or even branch, then refactor what's worth changing and
> > merge
> > > > the results back if ALL who care about it agree it's good.
> > > > Not sure, if Apache uses e.g. Gerrit, but it feels some of us could
> use
> > > >
> > > > that;-)
> > > >
> > > > Werner
> > > >
> > >
> >
>
--
*Anatole Tresch*
Java Engineer & Architect, JSR Spec Lead
Glärnischweg 10
CH - 8620 Wetzikon
*Switzerland, Europe Zurich, GMT+1*
*Twitter: @atsticks*
*Blogs: **http://javaremarkables.blogspot.ch/
<http://javaremarkables.blogspot.ch/>*
*Google: atsticksMobile +41-76 344 62 79*
Re: Complexity Discussion
Posted by Romain Manni-Bucau <rm...@gmail.com>.
We rediscuss basis and startiong from an empty branch while we dont agree
is sane IMO....also means we dont import anything while it is not the case.
Le 27 déc. 2014 15:15, "John D. Ament" <jo...@apache.org> a écrit :
> No it's not a technical requirement. There's no reason to move stuff for
> the same of moving stuff.
>
> On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg <st...@yahoo.de> wrote:
>
> > The reason for moving it into a separate directory is purely technical.
> > That way it is handled with GIT is that you can move files _without_
> > loosing history of the content.
> >
> > So while we are moving bits back we still do have all the original
> > contributions in place without loosing anything. It's just easier to move
> > around files inside the same branch than to switch between multiple of
> > those.
> >
> >
> > LieGrue,
> > strub
> >
> >
> >
> >
> >
> > > On Saturday, 27 December 2014, 14:28, Werner Keil <
> werner.keil@gmail.com>
> > wrote:
> > > > E.g. DeviceMap emerged over time, even with multiple languages other
> > than
> > > Java, though you could call some of it "dormant".
> > >
> > > What we did there is a "contrib" section where new parts gradually got
> > > refactored from, but as Tamaya already got a new structure, licensing,
> > etc.
> > > calling its root "contrib", "dormant" or
> > > "stuffSomeDontLike" seems rather
> > > useless.
> > >
> > > Better tag or even branch, then refactor what's worth changing and
> merge
> > > the results back if ALL who care about it agree it's good.
> > > Not sure, if Apache uses e.g. Gerrit, but it feels some of us could use
> > >
> > > that;-)
> > >
> > > Werner
> > >
> >
>
Re: Complexity Discussion
Posted by "John D. Ament" <jo...@apache.org>.
No it's not a technical requirement. There's no reason to move stuff for
the same of moving stuff.
On Sat Dec 27 2014 at 9:05:35 AM Mark Struberg <st...@yahoo.de> wrote:
> The reason for moving it into a separate directory is purely technical.
> That way it is handled with GIT is that you can move files _without_
> loosing history of the content.
>
> So while we are moving bits back we still do have all the original
> contributions in place without loosing anything. It's just easier to move
> around files inside the same branch than to switch between multiple of
> those.
>
>
> LieGrue,
> strub
>
>
>
>
>
> > On Saturday, 27 December 2014, 14:28, Werner Keil <we...@gmail.com>
> wrote:
> > > E.g. DeviceMap emerged over time, even with multiple languages other
> than
> > Java, though you could call some of it "dormant".
> >
> > What we did there is a "contrib" section where new parts gradually got
> > refactored from, but as Tamaya already got a new structure, licensing,
> etc.
> > calling its root "contrib", "dormant" or
> > "stuffSomeDontLike" seems rather
> > useless.
> >
> > Better tag or even branch, then refactor what's worth changing and merge
> > the results back if ALL who care about it agree it's good.
> > Not sure, if Apache uses e.g. Gerrit, but it feels some of us could use
> >
> > that;-)
> >
> > Werner
> >
>
Re: Complexity Discussion
Posted by Mark Struberg <st...@yahoo.de>.
The reason for moving it into a separate directory is purely technical.
That way it is handled with GIT is that you can move files _without_ loosing history of the content.
So while we are moving bits back we still do have all the original contributions in place without loosing anything. It's just easier to move around files inside the same branch than to switch between multiple of those.
LieGrue,
strub
> On Saturday, 27 December 2014, 14:28, Werner Keil <we...@gmail.com> wrote:
> > E.g. DeviceMap emerged over time, even with multiple languages other than
> Java, though you could call some of it "dormant".
>
> What we did there is a "contrib" section where new parts gradually got
> refactored from, but as Tamaya already got a new structure, licensing, etc.
> calling its root "contrib", "dormant" or
> "stuffSomeDontLike" seems rather
> useless.
>
> Better tag or even branch, then refactor what's worth changing and merge
> the results back if ALL who care about it agree it's good.
> Not sure, if Apache uses e.g. Gerrit, but it feels some of us could use
>
> that;-)
>
> Werner
>
Re: Complexity Discussion
Posted by Werner Keil <we...@gmail.com>.
E.g. DeviceMap emerged over time, even with multiple languages other than
Java, though you could call some of it "dormant".
What we did there is a "contrib" section where new parts gradually got
refactored from, but as Tamaya already got a new structure, licensing, etc.
calling its root "contrib", "dormant" or "stuffSomeDontLike" seems rather
useless.
Better tag or even branch, then refactor what's worth changing and merge
the results back if ALL who care about it agree it's good.
Not sure, if Apache uses e.g. Gerrit, but it feels some of us could use
that;-)
Werner
Re: Complexity Discussion
Posted by Werner Keil <we...@gmail.com>.
You should conduct a vote first before doing crap all by yourself;-)
Am 27.12.2014 11:09 schrieb "Mark Struberg" <st...@yahoo.de>:
> Hi!
>
> It would eventually help to stop adding fresh sources and functionality
> and start DISCUSSING :)
> This is not a discussion right now. More like some people arguing and
> complaining and others don't listening much but just continuing with the
> original direction. This is not how open source works! Actually this is not
> how teamwork is done!
>
>
>
> To get rid of this I suggest to start again from scratch.
>
> I gonna move all the current code to a 'dormant' directory and create a
> fresh structure.
>
>
> Then we will discuss one feature after the other. And if there is a good
> fit in the original source then we gonna move it back out of 'dormant' and
> tinker on it until we finally are all satisfied with it. That way we will
> just have the bits we really need and we also don't loose any history for
> the parts which already exist.
>
>
>
> Oh and one very important message: please don't get this as me being a
> mean jerk. If you look at CODI and then DeltaSpike you will see that we
> re-thought our own ideas MANY times and tried to become better each time.
> There was btw also a config system which I wrote long before CODI and also
> didn't met our quality criteria. And we also re-wrote the DS config a few
> times before we merged it into master.
>
>
>
> LieGrue,
> strub
>
>
>
>
>
>
> > On Wednesday, 24 December 2014, 0:21, Anatole Tresch <an...@apache.org>
> wrote:
> > > Dear all
> >
> > to make further discussions hopefully less harmful and easier, I will
> > remove the clutter that is still in there, especially in core, but also
> in
> > API. The corresponding cleanup ticket is still in progress. Basically the
> > *PropertySource* interface will expect 1-2 methods be exact the same as
> in
> > Deltaspike, so we hopefully get a better starting point for further
> > evaluation ;)
> >
> > Best and good night and merry X-Mas
> > Anatole
> >
>
Re: Complexity Discussion
Posted by Mark Struberg <st...@yahoo.de>.
Hi!
It would eventually help to stop adding fresh sources and functionality and start DISCUSSING :)
This is not a discussion right now. More like some people arguing and complaining and others don't listening much but just continuing with the original direction. This is not how open source works! Actually this is not how teamwork is done!
To get rid of this I suggest to start again from scratch.
I gonna move all the current code to a 'dormant' directory and create a fresh structure.
Then we will discuss one feature after the other. And if there is a good fit in the original source then we gonna move it back out of 'dormant' and tinker on it until we finally are all satisfied with it. That way we will just have the bits we really need and we also don't loose any history for the parts which already exist.
Oh and one very important message: please don't get this as me being a mean jerk. If you look at CODI and then DeltaSpike you will see that we re-thought our own ideas MANY times and tried to become better each time. There was btw also a config system which I wrote long before CODI and also didn't met our quality criteria. And we also re-wrote the DS config a few times before we merged it into master.
LieGrue,
strub
> On Wednesday, 24 December 2014, 0:21, Anatole Tresch <an...@apache.org> wrote:
> > Dear all
>
> to make further discussions hopefully less harmful and easier, I will
> remove the clutter that is still in there, especially in core, but also in
> API. The corresponding cleanup ticket is still in progress. Basically the
> *PropertySource* interface will expect 1-2 methods be exact the same as in
> Deltaspike, so we hopefully get a better starting point for further
> evaluation ;)
>
> Best and good night and merry X-Mas
> Anatole
>