Seth Hillbrand a écrit :
> You will need to upgrade ngspice as well. The variable is renamed to
> NG_BOOL.
Thanks. Build is done.
JB
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchp
As an aside, it would be very helpful when typedefing a new value to
include a define (like #define HAS_NG_BOOL 1) with it. That would allow
us to test for the new signature when compiling.
Done
Holger
___
Mailing list: https://launchpad.net
You will need to upgrade ngspice as well. The variable is renamed to
NG_BOOL.
-Seth
On Wed, May 5, 2021 at 12:21 PM BERTRAND Joël
wrote:
> Thanks for your patch.
>
> Nevertheless, after make clean && make, build process now aborts
> with :
>
> [ 47%] Building CXX object
> eesch
Thanks for your patch.
Nevertheless, after make clean && make, build process now aborts with :
[ 47%] Building CXX object
eeschema/CMakeFiles/eeschema_kiface_objects.dir/sim/ngspice.cpp.o
In file included from /home/bertrand/git/kicad/eeschema/sim/ngspice.cpp:31:
/home/bertrand/gi
I have conditionally adjusted the signature based on a secondary tag that
also seems new since NGSPICE34. We can't adjust the API without packaging
sharedspice.h ourselves, which I don't think we want.
As an aside, it would be very helpful when typedefing a new value to
include a define (like #de
Am 04.05.2021 um 23:32 schrieb Seth Hillbrand:
We can adjust our interface to cope with the change to int as an
underlying type.
That is great. Thus we may avoid any potential problems.
However, one concern is that the typedef exists unscoped inside a header
file. The term "BOOL" is rela
Jon Evans a écrit :
> What version of ngspice are you using?
I don't know, maybe a version pulled last week. But yesterday, I have
rebuild ngspice from scratch (I have deleted my build directory).
In ngspice header, I can read :
/*
...
All boolean signals (BOOL) are of type int.
We can adjust our interface to cope with the change to int as an underlying
type.
However, one concern is that the typedef exists unscoped inside a header
file. The term "BOOL" is relatively generic and so there are likely to be
some conflicts with other libraries that use this term differently.
The last commit to ngspice master branch is
https://sourceforge.net/p/ngspice/ngspice/ci/ded0f79c01079eaa6f4cdbd49292897c73273e23/
ded0f79c0 ("Avoid type 'bool', as ist is a built-in type in C++, not
equaling 'int' which ngspice is using.", 2021-03-28)
This patch is a reaction to bug report
Looks like Holger just added this:
https://github.com/imr/ngspice/commit/ded0f79c01079eaa6f4cdbd49292897c73273e23
I'm not sure why. I think Holger is on this list, so hopefully he can help
explain the reasoning.
-Seth
On Tue, May 4, 2021 at 2:04 PM Jon Evans wrote:
> What version of ngspice
What version of ngspice are you using?
Sourceforge is down right now so I can't check the latest tree easily, but
in my cached version, the ControlledExit function signature in
sharedspice.h is using bool (not BOOL), and so is KiCad.
Is there some version of ngspice where those function signature
Thanks, Orson. 5672 builds and seems to work fine for me on a non-retina
system against wxWidgets without the retina patch. Perhaps worth a mention of
that new patch in Documentation/compiling/mac-osx.txt for retina users not
building with the script.
Garth
> On May 20, 2015, at 3:46 PM, Ma
I also have a non-retina Mac.
Bob G
On 05/20/2015 06:19 PM, Adam Wolf wrote:
Thanks. I think Andy said he had a non-retina Mac. I can also hook up
to an external non-retina display.
Adam Wolf
On May 20, 2015 5:47 PM, "Maciej Sumiński" mailto:maciej.sumin...@cern.ch>> wrote:
Hi,
Th
If the nightlies go tonight, you can test those tomorrow.
Adam Wolf
On Wed, May 20, 2015 at 5:28 PM, Andy Peters wrote:
> >>
> >> On May 20, 2015 5:47 PM, "Maciej Sumiński"
> wrote:
> >> Hi,
> >>
> >> There were already 2 people today who had the same problem. I have just
> >> changed the patc
>>
>> On May 20, 2015 5:47 PM, "Maciej Sumiński" wrote:
>> Hi,
>>
>> There were already 2 people today who had the same problem. I have just
>> changed the patch, so the modified wx headers contain a #define to
>> indicate that you are building using patched version. I apologize for
>> disturban
Thanks. I think Andy said he had a non-retina Mac. I can also hook up to
an external non-retina display.
Adam Wolf
On May 20, 2015 5:47 PM, "Maciej Sumiński" wrote:
> Hi,
>
> There were already 2 people today who had the same problem. I have just
> changed the patch, so the modified wx headers
Hi,
There were already 2 people today who had the same problem. I have just
changed the patch, so the modified wx headers contain a #define to
indicate that you are building using patched version. I apologize for
disturbances.
I only had a Mac with Retina screen at hand, so I could not test other
Thanks.
Adam Wolf
On Wed, May 20, 2015 at 1:07 PM, Johannes Maibaum
wrote:
> Hi Adam,
>
> have you rebuilt wxWidgets with the retina display patch Orson commited in
> r5666? The missing symbols are introduced there.
>
>
> Best,
> Johannes
>
>
>
> Am 20.05.2015 um 21:51 schrieb Adam Wolf :
>
> I
Hi Adam,
have you rebuilt wxWidgets with the retina display patch Orson commited in
r5666? The missing symbols are introduced there.
Best,
Johannes
> Am 20.05.2015 um 21:51 schrieb Adam Wolf :
>
> Is anyone else having build issues on r5670 like:
>
> /Users/jenkins/remoteroot/workspace/KiC
19 matches
Mail list logo
