The round() function is C98/C++11 only, which is why there is a special
KiROUND() function that is used almost everywhere for rounding.
This collects the remaining stragglers.
---
common/gal/graphics_abstraction_layer.cpp | 16
pcbnew/dialogs/dialog_move_exact.cpp | 2 +-
Hi,
Am 05.11.2015 um 03:45 schrieb Mark Roszko:
>> I have repeatedly asked that devs do incremental changes so I can
>> easily review them.
> It's hard to do refactoring when your one commit has to throw out the
> kitchen sink.
However, it is often possible to submit multiple commits of the
Hi,
this should get us over the next three months.
Simon
---
copyright.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/copyright.h b/copyright.h
index 79a50ec..4110151 100644
--- a/copyright.h
+++ b/copyright.h
@@ -12,8 +12,8 @@ may choose to document this
Hi,
I'm trying to figure out who is owning the *_SCREEN object pointed to by
EDA_DRAW_FRAME::m_currentScreen.
In EDA_DRAW_FRAME::~EDA_DRAW_FRAME(), the object pointed to is deleted,
which suggests to me that the draw frame has some sort of ownership,
however this will only delete the last object
Hi,
CMake provides a readymade test for the GCC/CLang -fvisibility= flag. This
removes our own test and uses the result from CMake.
I've left the test for "NOT APPLE" in place -- is there a particular reason
why we do that?
Simon
---
CMakeLists.txt | 16
SCREEN objects are typically owned by the DRAW_FRAME they are attached to.
SCH_SCREEN objects may additionally be shared between one DRAW_FRAME, and
multiple SCH_SHEET objects, so they are reference counted.
To make this distinction more obvious, and avoid potential errors, move the
pointer
Hi,
small cleanup: there are a few unnecessary casts because the covariant
return already does the right thing. Where we get the screen object only in
order to get the active layer, use the accessor function directly.
Simon
---
include/wxBasePcbFrame.h | 4 ++--
Hi,
On 06.09.2015 13:11, Mário Luzeiro wrote:
> what is the c++ standard? that kicad is building? Please make a X in one of
> the following options:
Currently we appear to be using C++98, with a number of nonstandard GNU
extensions.
C++11 would be nice. C++14 is too new, compiler support is
Hi,
this updates our copy of FindwxWidgets.cmake to a state that is fairly
close to what is in the CMake repository. The remaining improvements over
their version are
- version check
- forward compatibility with 3.1 (looking for wx-config-3.1)
- handling for the "webview" component
These
Hi,
CMake's test for OpenMP is generic enough to work anywhere, so there is no
good reason to make it conditional on gcc/clang.
This moves the OpenMP handling to the CMakeLists.txt toplevel.
Simon
---
CMakeLists.txt | 33 +++--
1 file changed, 19 insertions(+),
Hi,
As boost::polygon is gone, we can drop the workaround for older gcc.
Simon
---
CMakeLists.txt | 15 ---
1 file changed, 15 deletions(-)
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 6e79fc9..b76c48d 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -171,21 +171,6 @@
Hi,
probably an oversight after removing debug code. This is not needed, and in
the middle of a source file, between functions.
Simon
---
pcbnew/autorouter/rect_placement/rect_placement.cpp | 1 -
1 file changed, 1 deletion(-)
diff --git
Hi,
CMake provides a simple declarative statement to enable PIC, so no compiler
dependent handling is required. There is also no need to tell Boost to
build with -fPIC, their build system is smart enough.
Simon
---
CMakeLists.txt| 13 -
Hi,
this is an improved version of the earlier patch, using the visibility
handling in CMake 3.3, and falling back to changing the global CXXFLAGS on
previous versions.
As CMake 3.3 attempts to be compatible with earlier versions, the new
behaviour needs to be explicitly enabled until 3.3
Hi,
this allows to undo the grouping in the pin table, by right clicking on the
column header of the currently active grouping column -- this allows you to
go back to a fully flat table.
Simon
---
eeschema/dialogs/dialog_lib_edit_pin_table.cpp | 6 +++---
1 file changed, 3 insertions(+), 3
netlist_reader.h is also used by pcbcommon, so a dependency is necessary.
---
common/CMakeLists.txt | 1 +
1 file changed, 1 insertion(+)
diff --git a/common/CMakeLists.txt b/common/CMakeLists.txt
index 2b558c4..2d843cf 100644
--- a/common/CMakeLists.txt
+++ b/common/CMakeLists.txt
@@ -388,6
Hi,
this is an actual problem with VS 2015, which uses different name mangling
rules for "struct" and "class". This makes all instances use "struct".
Simon
---
pcbnew/dialogs/dialog_track_via_properties.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
Hi,
this adds a number of missing includes. On gcc, other system includes use
these internally, but that is not guaranteed.
Simon
---
common/gal/graphics_abstraction_layer.cpp | 2 ++
common/grid_tricks.cpp | 2 ++
common/selcolor.cpp
Hi,
On 14.09.2015 04:28, Mark Roszko wrote:
> Now anyone using the newer git-remote-bzr is in for fun as well
> because there is no such thing as "origin/HEAD".
origin/HEAD is automatically created with git-remote-bzr, but not
restored if lost.
You can recreate it via
echo "ref:
Hi,
> Looks ok to me, except that the naming of the _git_LONG_HASH_ORIGIN
> variable does not match what it is, it should have been
> _git_LONG_HASH_LOCAL instead.
This completely fails to compile for me:
| [ 11%] Generating version string header
| -- Using Git to determine build version
Hi,
On 13.09.2015 17:55, Lorenzo Marcantonio wrote:
> I suppose that is talking about the implementation of va_start: address
> of last+sizeof of last gives the starting point of the va_list (in the
> 'correct' stack direction, obviously); am I correct?
Indeed. The standard is written so va_*
Hi,
On 13.09.2015 20:29, Simon Richter wrote:
> This completely fails to compile for me:
Okay, when I wipe my local repository and repeat the checkout, it works.
Having a fallback to origin/master and/or more graceful handling of a
missing origin/HEAD reference would be great though.
Si
Hi,
while investigating an internal compiler error[1] in MSVC, I ran across
this function.
My feeling is that it is broken:
For int:
top_left = std::numeric_limits::min() / 2 +
std::numeric_limits::epsilon();
size = std::numeric_limits::max() -
Hi,
On 14.09.2015 19:10, Mark Roszko wrote:
> Weird, its supposed to be 0 for ints.
> http://en.cppreference.com/w/cpp/types/numeric_limits/epsilon
I was referring to the bit where it says "It is only meaningful if
std::numeric_limits::is_integer == false". I think it's not a big
problem for
Hi,
Am 14.09.2015 um 23:58 schrieb Adam Wolf:
[.deb packages]
> Isn't that what's in the reynauld PPA?
No, that's for Ubuntu -- they happen to use the same package format, but
packages are generally not expected to be compatible.
Simon
signature.asc
Description: OpenPGP digital
Hi,
On 19.09.2015 23:05, Wayne Stambaugh wrote:
> No such document but basically gcc on linux and windows and clang on
> osx. The official policy on MS VS is that we will not support it.
VS 2015 is broken to the extent that it fails to compile some of the
template code in VECTOR2<> and BOX<>,
Hi,
Am 04.12.2015 um 06:46 schrieb Ike Shields:
> In the print screen on R/H side;
>
> Under Copper layers Tick is on “B.cu”
In the main window, on the right hand side, there are two tabs, "Layers"
and "Render". The former is overridden by the layer selection in the
print dialog, but the
Hi,
I've made a backport of RC1 to Debian jessie -- release backports will
happen once the release has hit unstable. The same packages have been
uploaded to Debian, so if you have jessie-backports configured in your
sources.list, your system should silently switch over to the release
package
Hi,
On 08.12.2015 22:52, Wayne Stambaugh wrote:
> I asked Simon if he could figure out how to make it less noisy. I
> haven't heard back from him yet.
>>> CMake Warning (dev) at utils/idftools/CMakeLists.txt:17 (add_executable):
>>> Policy CMP0063 is not set: Honor visibility properties for
Hi,
Am 03.12.2015 um 20:11 schrieb Jon Neal:
> With this patch when you add new layers in pcbnew they will be selected
> by default for plotting.
that is LP#1507822. -- https://bugs.launchpad.net/kicad/+bug/1507822
Simon
signature.asc
Description: OpenPGP digital signature
Hi,
> - set( BOOST_CFLAGS "cflags=${PIC_FLAG}" )
> - set( BOOST_CXXFLAGS "cxxflags=${PIC_FLAG}" )
> because cmake will only set -fpic in CFLAGS but the entire cmake file
> uses BOOST_CFLAGS and never passes the plain CFLAGS
The Boost build knows on its own that it needs -fPIC -- except if you
Hi,
I'm thinking about adding some support for pin and unit swapping. To
make it maximally useful, I'd like to generate a list of alternative
mappings inside the netlist, and add the actual swap functionality
inside pcbnew -- so if it turns out that the layout becomes easier if
two pins are
Hi,
Am 14.12.2015 um 16:01 schrieb Mark Roszko:
> Stupid question, doesn't this change remove -O2 entirely for all gcc
> builds? Previously it defaulted to -O2 if the bug wasn't present. Now
> it never optimizes as -O0 is the default.
The default depends on the build type -- if your source tree
Hi Wayne,
On 11.12.2015 20:23, Wayne Stambaugh wrote:
> Not until after the new schematic and symbol library file formats are
> implemented. Support for pin and unit swapping will be built in to the
> new symbol library file format.
That would be fairly orthogonal to what I'm suggesting -- I'd
Hi Wayne,
On 04.01.2016 00:50, Wayne Stambaugh wrote:
> I forgot to delete download_avhttp.cmake and the avhttp source zip file
> so this patch wont be necessary.
There is still a paragraph on KICAD_SKIP_BOOST in the documentation.
Simon
signature.asc
Description: OpenPGP digital
Hi,
this removes two leftover references to KICAD_SKIP_BOOST.
Simon
---
CMakeModules/download_avhttp.cmake | 9 +
Documentation/development/compiling.md | 8
2 files changed, 1 insertion(+), 16 deletions(-)
diff --git a/CMakeModules/download_avhttp.cmake
Hi,
On 04.01.2016 04:50, Jon Neal wrote:
> Is there any chance that the CMAKE_BUILD_TYPE flag could be changed from
> Product to Debug by default? It is the only flag I have to include when
> I run cmake now and I often forget to set it.
The goal is to make it easy for the people who want to
Hi,
On 21.12.2015 08:35, Lorenzo Marcantonio wrote:
> After this I'm *not* advocating a binary format, there would be no
> advantage for it in pcbnew. There are no performance restraints in
> either speed or space for kicad files so just use whatever is easier for
> you to manage (i.e.
Hi,
> Just a quick heads up. I just made my travel arrangements for FOSDEM
> 2016. That always makes it feel official to me. Please try to attend
> if you can manage it. I would like to meet as many of the KiCad
> developers as possible. I can tell you that EDA dev room dinner
> Saturday
Hi,
this patch set implements a summary line for the pin table, listing the
currently defined pins, to quickly check whether any pins have been
forgotten.
If pins 1,2,3,4,5,6,8,9,10,11,12 are defined, the status bar then reads
"1-6,8-12".
Simon
Simon Richter (5):
Split off
/eeschema/pin_number.cpp b/eeschema/pin_number.cpp
new file mode 100644
index 000..9f5f9ea
--- /dev/null
+++ b/eeschema/pin_number.cpp
@@ -0,0 +1,134 @@
+/*
+ * This program source code file is part of KiCad, a free EDA CAD application.
+ *
+ * Copyright (C) 2015 Simon Richter
+ * Copyright (C) 20
Below the pin table, display which pins are currently defined, in order to
find gaps.
---
eeschema/dialogs/dialog_lib_edit_pin_table.cpp | 27 +++
eeschema/dialogs/dialog_lib_edit_pin_table.h | 2 +
.../dialogs/dialog_lib_edit_pin_table_base.cpp | 3 +
This class wraps the comparison function in a way that is compatible with
std::map and std::set. This, too is generally useful, so it should be moved
to the generic header.
At the same time, the criterium for less-than is changed from "equal to -1"
to "smaller than 0", because technically the
The "set of pin numbers" functionality is also useful outside the pin table
dialog, so create a wrapper class that exposes the relevant interfaces.
---
eeschema/dialogs/dialog_lib_edit_pin_table.cpp | 2 +-
eeschema/pin_number.h | 21 +
2 files
For pin numbers ending in digits, consecutive numbers are collapsed to
ranges for a better overview.
---
eeschema/pin_number.cpp | 49 +++--
1 file changed, 39 insertions(+), 10 deletions(-)
diff --git a/eeschema/pin_number.cpp
As boost::polygon is gone, we can drop the workaround for older gcc.
---
CMakeLists.txt | 15 ---
1 file changed, 15 deletions(-)
diff --git a/CMakeLists.txt b/CMakeLists.txt
index ec89978..cd3e1bc 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -171,21 +171,6 @@ if(
Hi,
On 01.12.2015 07:52, Ike Shields wrote:
> I did an update and it back that I was up to date?
Also, I just noticed. The bzr 4027 version is ancient. The 4.0.0 version
is not contained in Debian stable, and will not be included there
either, as Debian never updates stable releases.
I believe
This simply moves the OpenMP handling code out of the gcc specific block
with no code changes. This is safe because OpenMP detection is portable.
---
CMakeLists.txt | 33 +++--
1 file changed, 19 insertions(+), 14 deletions(-)
diff --git a/CMakeLists.txt
CMake 3.0 defines two new variables,
* CMAKE_CXX_VISIBILITY_PRESET and
* CMAKE_VISIBILITY_INLINES_HIDDEN
to control whether symbols not explicitly tagged for export are implicitly
exported. Because only version 3.3 and following also applies that to
static libraries when in 3.3 mode,
CMake provides a simple declarative statement to enable PIC, so no compiler
dependent handling is required. There is also no need to tell Boost to
build with -fPIC, their build system is smart enough.
---
CMakeLists.txt| 13 -
CMakeModules/download_boost.cmake |
Hi,
On 01.12.2015 07:52, Ike Shields wrote:
> I am having trouble printing the last part (print to copper. it is the
> printed to photo paper then laminated onto Copper) when I print all I
> get is the small drill holes and solder pads no tracks.
Could it be that some of the "Layer" or
Hi,
On 01.12.2015 09:07, jp charras wrote:
>> As boost::polygon is gone, we can drop the workaround for older gcc.
> boost::polygon is gone, not the gcc bug.
> This is likely a gcc bug, not a boost::polygon bug.
Indeed, but the code that triggered the bug is gone, so it shouldn't
affect us any
Hi Chris,
Am 13.01.2016 um 02:23 schrieb Chris Pavlina:
> This patch doesn't apply to the latest revision for me, at very least
> the line numbers are off. Might have to rebase.
Ah yes, that is on top of the other stack of patches I sent, where I
split out the comparison function into a new
Hi Nick,
Am 13.01.2016 um 09:03 schrieb Nick Østergaard:
> There is no attached patch.
Yes, if there are multiple patches I occasionally write a cover letter
marked [0/x] that contains the blurb that shouldn't go into the commit
log -- so "git am" can import my commit messages directly. This is
Hi,
this brings our copy of FindwxWidgets.cmake in sync with the version in the
CMake repository. Since last September, CMake includes our changes (version
check, wx 3.1, support for webview component), and these have been released
with CMake 3.4.
The change to drop the -isystem special handling
---
eeschema/dialogs/dialog_lib_edit_pin_table.cpp | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/eeschema/dialogs/dialog_lib_edit_pin_table.cpp b/eeschema/dialogs/dialog_lib_edit_pin_table.cpp
index 63d8226..52c399f 100644
---
This is in anticipation of the introduction of icons -- retrieving all the
icons and throwing them away during sorting takes ages.
---
eeschema/dialogs/dialog_lib_edit_pin_table.cpp | 88 ++
1 file changed, 47 insertions(+), 41 deletions(-)
diff --git
Hi,
these are two small pin table improvements: a direct string access avoiding
the wxVariant wrapping/unwrapping for internal functions (major speedup in
sorting), and a way to undo the grouping in the pin table in order to get a
flat table.
Simon
Simon Richter (2):
pin table: Add
Hi,
On 02.06.2016 02:32, Wayne Stambaugh wrote:
> Ahh, the old double negative. Using !NDEBUG would be ugly but I suppose
> it would be effective. I'm not sure why cmake isn't defining DEBUG for
> debug builds. I thought at one time it did but I could be wrong.
We define it manually in the
There is little to no reason why this variable should be static, given that
it is cleared after each iteration.
---
eeschema/sch_line.cpp | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/eeschema/sch_line.cpp b/eeschema/sch_line.cpp
index 72efb5c..050b786 100644
---
This is unnecessary -- explicitly specifying the regular copy ctor will
inhibit any generated versions anyway.
---
include/plugins/3dapi/ifsg_node.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/include/plugins/3dapi/ifsg_node.h b/include/plugins/3dapi/ifsg_node.h
index 6b3641e..f949c80
SCREEN objects are typically owned by the DRAW_FRAME they are attached to.
SCH_SCREEN objects may additionally be shared between one DRAW_FRAME, and
multiple SCH_SHEET objects, so they are reference counted.
To make this distinction more obvious, and avoid potential errors, move the
pointer
Simon Richter (16):
Fix spelling "propage" -> "propagate"
Add missing dependency github_plugin -> pcbcommon
Clarify ERC: we're iterating netlist items, not nets
Explain how ERC works.
Cache netlist item during ERC
Check sorting of netlist during ERC
Kill unuse
These are never shared, so std::unique_ptr works fine, and std::unique_ptr
has a specialisation for array types that uses "operator delete[]", so this
is still correct.
---
common/gal/opengl/opengl_gal.cpp | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git
This function does not need to modify object state.
---
eeschema/class_netlist_object.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/eeschema/class_netlist_object.h b/eeschema/class_netlist_object.h
index 9608f90..403bade 100644
--- a/eeschema/class_netlist_object.h
+++
This is just a speed optimization -- preallocate four items to avoid
reallocations.
---
eeschema/sch_line.cpp | 1 +
1 file changed, 1 insertion(+)
diff --git a/eeschema/sch_line.cpp b/eeschema/sch_line.cpp
index 050b786..6d86b5d 100644
--- a/eeschema/sch_line.cpp
+++ b/eeschema/sch_line.cpp
This function is unused, and looks similar to the tree builder above, so it
might likely be left over from a refactoring effort.
---
.../netlist_exporters/netlist_exporter_generic.cpp | 73 --
.../netlist_exporters/netlist_exporter_generic.h | 7 ---
2 files changed, 80
This is a simple 1:1 replacement for all cases where shared_ptr is not used
to point at an array (array handling is not in C++11).
---
3d-viewer/3d_mesh_model.h | 4 ++--
common/gal/opengl/opengl_gal.cpp | 2 +-
common/geometry/hetriang.cpp | 34
The exception objects caught are either not referenced at all, or only in
debug builds. This avoids the warnings for the unused variables.
---
eeschema/class_libentry.cpp| 2 +-
eeschema/class_library.cpp | 4 ++--
eeschema/cross-probing.cpp | 2 +-
These switch statements only have a single "default" case, so they are unneeded.
---
bitmap2component/bitmap2cmp_gui.cpp | 10 +-
pcb_calculator/pcb_calculator.cpp | 12 +---
2 files changed, 2 insertions(+), 20 deletions(-)
diff --git a/bitmap2component/bitmap2cmp_gui.cpp
Hi,
On 12.06.2016 22:22, Wayne Stambaugh wrote:
> Plotting is definitely more accurate than printing although I wont
> completely rule it out. I make a lot of prototype boards and I've found
> through trial and error the plotting is the way to go.
I think it is wrong in both dialogs.
Whether
This shouldn't affect code because these are compile time constants anyway,
but gets rid of some annoying warnings about casting integers to pointers
of different size.
---
include/gal/opengl/vertex_common.h | 24
1 file changed, 12 insertions(+), 12 deletions(-)
diff
Hi,
On 05.06.2016 20:12, Chris Pavlina wrote:
> Uh, has anyone besides Nick and Simon seen this?
Windows builds against 1.60 still AFAIK, mainly because FindBoost.cmake
has an explicit list of versions to look for, so it isn't a simple drop
in replacement.
Simon
signature.asc
Hi,
On 09.06.2016 20:28, Mário Luzeiro wrote:
[Using iterator instead of const_iterator as place hint for std::list]
> This workaround works!
I think this gcc version we can support with only little effort, since
we need to have a non-const reference to the container anyway, so
getting a
---
common/gal/opengl/cached_container.cpp | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/common/gal/opengl/cached_container.cpp b/common/gal/opengl/cached_container.cpp
index 40d4969..2029929 100644
--- a/common/gal/opengl/cached_container.cpp
+++
---
common/gal/opengl/gpu_manager.cpp | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/common/gal/opengl/gpu_manager.cpp b/common/gal/opengl/gpu_manager.cpp
index 2cba5cc..f14a994 100644
--- a/common/gal/opengl/gpu_manager.cpp
+++ b/common/gal/opengl/gpu_manager.cpp
@@
Hi,
On 15.06.2016 08:53, jp charras wrote:
> I do not understand why net ties are related to schematic cleanup.
> They are 2 separate features.
For the "net ties as special junctions" implementation, the cleanup code
needs to be adjusted -- that's where they come together.
> Although no
This is just a comment explaining the use of the local variables.
---
eeschema/dialogs/dialog_erc.cpp | 9 +
1 file changed, 9 insertions(+)
diff --git a/eeschema/dialogs/dialog_erc.cpp b/eeschema/dialogs/dialog_erc.cpp
index 2f9a1f2..7323a64 100644
--- a/eeschema/dialogs/dialog_erc.cpp
This renames the "net", "lastNet" and "nextNet" variables to "item",
"lastItem" and "nextItem", respectively, because these refer to netlist
items, not nets.
---
eeschema/dialogs/dialog_erc.cpp | 26 +-
1 file changed, 13 insertions(+), 13 deletions(-)
diff --git
The ERC code depends on netlist items to be sorted by net code, so verify
that in debug builds. While this condition is stricter than necessary, it
should still hold with the current code, and provide a good canary if a
change to the sorting code might break ERC.
---
The GitHub plugin needs the generated PCB parser code, so introduce an
artificial dependency.
---
pcbnew/CMakeLists.txt | 2 ++
1 file changed, 2 insertions(+)
diff --git a/pcbnew/CMakeLists.txt b/pcbnew/CMakeLists.txt
index e77cbd9..aea5469 100644
--- a/pcbnew/CMakeLists.txt
+++
This adds a suffix "_idx" to the indexes into the list, and introduces local
copies of the pointers to the objects we are looking at, in order to have a
shorter way of addressing them.
---
eeschema/dialogs/dialog_erc.cpp | 30 --
1 file changed, 16 insertions(+), 14
---
utils/idftools/idf2vrml.cpp | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/utils/idftools/idf2vrml.cpp b/utils/idftools/idf2vrml.cpp
index 05e375e..a62db0f 100644
--- a/utils/idftools/idf2vrml.cpp
+++ b/utils/idftools/idf2vrml.cpp
@@ -100,7 +100,7 @@ static const
From: Simon Richter <ge...@kiwi.hogyros.de>
This function has two conflicting definitions in the "kicad" executable and
the other wrapper programs. As the kifaces can be loaded from either, this
silently assumes compatible data layout for the PGM_KICAD and PGM_BASE
types when pas
This makes sure that the declaration is always the same within a single
module, so all users have the same idea of whether the symbol is imported.
In all dynamically linked objects, all users should be aware that the
symbol is imported, while in the statically linked bitmap2component
program, no
While defining functions in another namespace is technically allowed as
long as the definition can be matched to a declaration, this can lead to
ambiguous resolutions, such as here (GAL vs KIGFX).
---
pcbnew/ratsnest_viewitem.cpp | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff
CURL on Windows uses OpenSSL on MSYS, and SChannel on MSVC.
---
common/kicad_curl/kicad_curl.cpp | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/common/kicad_curl/kicad_curl.cpp b/common/kicad_curl/kicad_curl.cpp
index a5670a7..fb4413c 100644
---
Support for C99 in C++ programs is a gcc extension.
---
eeschema/autoplace_fields.cpp | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/eeschema/autoplace_fields.cpp b/eeschema/autoplace_fields.cpp
index 9af6247..96ee6c0 100644
--- a/eeschema/autoplace_fields.cpp
+++
The combination of
- template resolution
- optimization
- structured exception handling
- chained operator=
appears to trigger a bug in the compiler.
As this code is nonsensical anyway (numeric_limits<>::min() returns the
smallest positive value for floating point types, but the largest
to use wxApp etc for option processing
Simon Richter (11):
Use std::atomic for portable locale init counting
Add missing C++ stdlib headers
Open namespace around definitions
Avoid C99 style compound statement
Avoid conflicting declaration for Pgm()
Make sure KIFACE_GETTER linkage is consistent
The GCC standard library headers often include other headers, which makes
some code compile that forgets to include several headers.
---
common/gal/graphics_abstraction_layer.cpp | 2 ++
common/grid_tricks.cpp | 2 ++
common/selcolor.cpp|
---
common/common.cpp | 10 +++---
include/common.h | 4 +++-
2 files changed, 6 insertions(+), 8 deletions(-)
diff --git a/common/common.cpp b/common/common.cpp
index 388a803..976e419 100644
--- a/common/common.cpp
+++ b/common/common.cpp
@@ -64,14 +64,12 @@ EDA_COLOR_Tg_GhostColor;
---
utils/idftools/idf2vrml.cpp | 2 --
1 file changed, 2 deletions(-)
diff --git a/utils/idftools/idf2vrml.cpp b/utils/idftools/idf2vrml.cpp
index a62db0f..47821d0 100644
--- a/utils/idftools/idf2vrml.cpp
+++ b/utils/idftools/idf2vrml.cpp
@@ -52,8 +52,6 @@
#include
#include
#include
From: Cirilo Bernardo
---
utils/idftools/idf2vrml.cpp | 267 +---
1 file changed, 126 insertions(+), 141 deletions(-)
diff --git a/utils/idftools/idf2vrml.cpp b/utils/idftools/idf2vrml.cpp
index df66e4b..05e375e 100644
---
On Windows, performance measurement is best performed using
QueryPerformanceCounter, which reports ticks with an arbitrary but fixed
frequency.
---
include/profile.h | 32
1 file changed, 32 insertions(+)
diff --git a/include/profile.h b/include/profile.h
index
The same class is defined below the 3d_rendering/ directory as well, with
more functionality.
---
3d-viewer/3d_draw.cpp| 2 +-
3d-viewer/3d_draw_board_body.cpp | 2 +-
3d-viewer/CImage.cpp | 440 ---
3d-viewer/CImage.h
This is unnecessary -- explicitly specifying the regular copy ctor will
inhibit any generated versions anyway.
---
include/plugins/3dapi/ifsg_node.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/include/plugins/3dapi/ifsg_node.h b/include/plugins/3dapi/ifsg_node.h
index 6b3641e..f949c80
Hi,
there was a small hiccup with a missing symbol in Release builds, which
I also saw in the MSVC builds, and expected to be affecting those only
because I don't have explicit logic to set DEBUG and NDEBUG in
CMakeLists.txt, assuming that CMake does that for me.
CMake does not set DEBUG in
Hi,
for net ties, I have to touch the schematic cleanup, because that will
merge adjacent wires going in the same direction, which is not useful
for me, as I need all wires going into a tie as separate nets (except
for ERC, where they are connected).
I'd like to generalize this to "wires are
These are a tiny bit more readable, and do not depend on all branches to
correctly advance the loop variable.
---
eeschema/sch_screen.cpp | 27 +++
1 file changed, 7 insertions(+), 20 deletions(-)
diff --git a/eeschema/sch_screen.cpp b/eeschema/sch_screen.cpp
index
101 - 200 of 898 matches
Mail list logo