kicad-developers team mailing list archive
-
kicad-developers team
-
Mailing list archive
-
Message #32617
Re: Simulator towards 5.0
All of your arguments are valid. For me personally, I'm fine with
enabling everything that's ready for release since I build from source
and it's not an issue on any of the platforms I use for development. If
that is the consensus, then I will enable them. I'm just trying to be
fair to our package devs.
On 12/27/2017 05:42 PM, Nick Østergaard wrote:
> 2017-12-27 23:04 GMT+01:00 Wayne Stambaugh <stambaughw@xxxxxxxxx
> <mailto:stambaughw@xxxxxxxxx>>:
>
> Here are my thoughts on the current options that are disable be default.
> Nothing is experimental except USE_WX_GRAPHICS_CONTEXT which really
> shouldn't be used and the WX_OVERLAY which is macos specific.
>
> KICAD_SCRIPTING, I'm fine with setting this to ON now that we have a
> sane solution for Python scripting on windows. All other platforms
> should have python 2 support for the foreseeable future.
>
> KICAD_SCRIPTING_MODULES, this can be enabled as well sense it goes hand
> in hand with KICAD_SCRIPTING.
>
> KICAD_SCRIPTING_WXPYTHON, this is not as clear cut as it would seem.
> Sometimes the wxPython build gets out of sync with the wxWidgets builds
> on certain platforms which is known to cause issues. Enabling this
> could cause issues for package devs.
>
>
> It will also if they enable it explicitly... On multiple platforms this
> ABI check between wx and wxpython is not marked as fatal and only
> servers as a hint that you should poke the maintainers to rebuild
> wxpython. For some reason they seems to remember to rebuild wx when the
> compiler gets updated but not wxpython. But they should really do it,
> there is no reason no to.
>
>
>
> KICAD_SCRIPTING_ACTION_MENU AFAIK depends on KICAD_SCRIPTING_WXPYTHON so
> enabling this will dependent upon enabling KICAD_SCRIPTING_WXPYTHON.
>
>
> And then what? The packager can still disable those options if he needs
> to. IMHO it is better that the packager does not need to explicitly
> enable any options to get a kicad build configuration that we advertise
> as working features. All those features do work on all the three major
> platforms. Linux, windows and macos.
>
>
>
> KICAD_USE_OCE, I'm not so comfortable enabling this by default due to
> issues between version of OCE. It might be best if this is left up to
> the package devs.
>
>
> What issues are we talking about here? If we never enable it we will not
> discover the issue as quickly as we could. We are not removing control
> from the packager here. He can still disable the option if it causes him
> any isses.
>
>
>
> KICAD_USE_SPICE, I'm not comfortable enabling this due to the fact that
> some linux packages of ngspice do not build with --libngspice enabled
> which will cause build config issues. I would to default to the package
> devs to enable this as necessary.
>
>
> This is the issue of the packager on that system. If he is lazy he can
> start by disabling this option untill he gets time to fix his platforms
> ngspice package. It should be good now that a release of ngspice have
> been made that has our required fixes. If the ngspice people did not
> make that release I would be okay with having it off by default.
>
>
>
> On 12/09/2017 11:46 AM, Kristoffer Ödmark wrote:
> > I think that this message is important, and I feel a decision on this
> > matter features has to be done by the project manager, basically
> give a
> > pointer to what should be used to as far extent as possible.
> >
> > Wayne, this is a package dev that wants to know this and while I
> do not
> > know who are package devs, the only package dev has expressed a
> need for
> > this decision. Please dont just ignore it.
> >
> > And yes, the coming patch submissions is dependent on what
> features are
> > deemed "standard"
> >
> > -Kristoffer
> >
> > On 12/07/2017 08:58 PM, Simon Richter wrote:
> >> Hi,
> >>
> >> On 07.12.2017 19:08, Wayne Stambaugh wrote:
> >>
> >>> Yes, that is exactly what I'm saying. it is possible kicad
> could have
> >>> different feature sets depending on the availability of
> dependencies on
> >>> the target platform. The kicad project has no control over
> this. If a
> >>> platform doesn't have dependency support for spice, they can still
> >>> provide a kicad package without spice support. That's better
> than no
> >>> kicad.
> >>
> >> Right, but we still need to give some guidance on the status of
> >> features. As it is now, new features are introduced as default-off
> >> "experimental" stuff that is only to be used by a select few, then at
> >> some point we enable it for nightlies, mostly driven by
> availability of
> >> dependencies and a vague feeling that features should be tested,
> and at
> >> some point the feature has become something that should have been
> >> enabled by default a long time ago (but nobody can tell the exact
> date
> >> when).
> >>
> >> We need a bit of a process here to promote feature status, e.g.
> >>
> >> experimental new stuff, not for general use
> >> optional => add to nightlies
> >> standard => enable by default
> >> required => remove option
> >>
> >> I think all the new features for v5 have reached "optional"
> status, and
> >> they have been enabled in nightlies as far as possible. The next
> >> question is whether they are "standard" and should be part of the
> stable
> >> release, including questions by users and a commitment to file
> >> compatibility. Probably yes, but this is a project management
> decision
> >> that will affect requirements for patch submissions in the next
> release
> >> cycle, so it needs to be an explicit decision.
> >>
> >> Simon
> >>
> >>
> >>
> >> _______________________________________________
> >> Mailing list: https://launchpad.net/~kicad-developers
> <https://launchpad.net/~kicad-developers>
> >> Post to : kicad-developers@xxxxxxxxxxxxxxxxxxx
> <mailto:kicad-developers@xxxxxxxxxxxxxxxxxxx>
> >> Unsubscribe : https://launchpad.net/~kicad-developers
> <https://launchpad.net/~kicad-developers>
> >> More help : https://help.launchpad.net/ListHelp
> <https://help.launchpad.net/ListHelp>
> >>
> >
> > _______________________________________________
> > Mailing list: https://launchpad.net/~kicad-developers
> <https://launchpad.net/~kicad-developers>
> > Post to : kicad-developers@xxxxxxxxxxxxxxxxxxx
> <mailto:kicad-developers@xxxxxxxxxxxxxxxxxxx>
> > Unsubscribe : https://launchpad.net/~kicad-developers
> <https://launchpad.net/~kicad-developers>
> > More help : https://help.launchpad.net/ListHelp
> <https://help.launchpad.net/ListHelp>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~kicad-developers
> <https://launchpad.net/~kicad-developers>
> Post to : kicad-developers@xxxxxxxxxxxxxxxxxxx
> <mailto:kicad-developers@xxxxxxxxxxxxxxxxxxx>
> Unsubscribe : https://launchpad.net/~kicad-developers
> <https://launchpad.net/~kicad-developers>
> More help : https://help.launchpad.net/ListHelp
> <https://help.launchpad.net/ListHelp>
>
>
Follow ups
References
-
Simulator towards 5.0
From: Kristoffer Ödmark, 2017-12-04
-
Re: Simulator towards 5.0
From: Nick Østergaard, 2017-12-05
-
Re: Simulator towards 5.0
From: Wayne Stambaugh, 2017-12-05
-
Re: Simulator towards 5.0
From: Nick Østergaard, 2017-12-05
-
Re: Simulator towards 5.0
From: Maciej Sumiński, 2017-12-05
-
Re: Simulator towards 5.0
From: Nick Østergaard, 2017-12-05
-
Re: Simulator towards 5.0
From: Maciej Sumiński, 2017-12-05
-
Re: Simulator towards 5.0
From: Simon Richter, 2017-12-05
-
Re: Simulator towards 5.0
From: Kristoffer Ödmark, 2017-12-06
-
Re: Simulator towards 5.0
From: Simon Richter, 2017-12-06
-
Re: Simulator towards 5.0
From: Kristoffer Ödmark, 2017-12-06
-
Re: Simulator towards 5.0
From: Wayne Stambaugh, 2017-12-07
-
Re: Simulator towards 5.0
From: kristoffer Ödmark, 2017-12-07
-
Re: Simulator towards 5.0
From: Wayne Stambaugh, 2017-12-07
-
Re: Simulator towards 5.0
From: Simon Richter, 2017-12-07
-
Re: Simulator towards 5.0
From: Kristoffer Ödmark, 2017-12-09
-
Re: Simulator towards 5.0
From: Wayne Stambaugh, 2017-12-27
-
Re: Simulator towards 5.0
From: Nick Østergaard, 2017-12-27