Hi Wayne,
Thanks for the heads up. I've pushed a fix to both branches. I'm not
sure why that ever worked silently on Linux GCC/Msys2 GCC/MSVC, but
using std:: is certainly correct. Perhaps a header has a sneaky type
alias somewhere.
Cheers,
John
On Fri, Mar 29, 2019 at 12:30 PM Wayne Stambaugh
e the headers separately to
> catch issues like this. There is helpful instructions in the coding
> policy on how to do this. Given the number of headers included in the
> KiCad source files, it's easy to overlook this.
>
> Cheers,
>
> Wayne
>
> On 3/29/19 8:53 AM, Jo
Hi all,
I have pushed some changes to the format check git hook, which means
that when you commit with the hooks on:
* You won't be nagged about changes in generated files
* You won't be nagged about unstaged files
Also provide a handy fixer/checker/amender script that uses the same
logic and da
Hi Wayne,
On 29 March 2019 22:30:27 GMT, Wayne Stambaugh wrote:
>
> I don't know if I want to force this even though I think it's a good
> idea. So few of the source files in KiCad adhere to this that it would
> confuse new developers. However, devs should be aware of the issue.
Sure, it's mor
Hi,
I'm trying to get the eeschema tests building and working, but I have
an issue with a segfault on init (the existing test itself will not
work either as it has hardcoded data paths, but it doesn't get that
far, so baby steps!).
It looks like the init of wxCursor in simulate.cpp line #144 is
e
Hi,
Just a quick plug for two fairly useful ways to "modularise" debug to
prevent insane amounts of logging overhead:
* trace masking, so you can turn on only what is needed to log. If you
want to conditionally trace more complex data, you can use
wxLog::IsAllowedTraceMask to prevent doing useles
committed to both master and 5.1 so fixes that come with their
own test cases can be easily applied to both branches.
Cheers,
John
On Sun, Mar 31, 2019 at 11:50 PM John Beard wrote:
>
> Hi,
>
> I'm trying to get the eeschema tests building and working, but I have
> an issue wit
Hi,
I think auto is useful when it avoids repetition of useless things,
and doubly so for things where the type is unwieldy, like iterators,
and, particularly, lambdas. For example:
auto up = std::make_unique( ... ); // or make_shared
auto it = aVec.begin(); // or cbegin or returns from std::find
Hi Jeff,
I think the current default is somewhat reasonable. If editing a
multi-unit part with substantially similar units, you are probably as
likely to want to keep your current zoom as reset it. And zoom reset
is a single keypress for when it isn't.
As this is likely to be one of those things
Hi Tom,
Can the hyperlink export test not exist in the existing qa_pcbnew
test? The code is all in pcbnew's library, so why not just mirror the
source layout and put your test as
"qa/pcbnew/exporters/test_export_hyperlynx.cpp"?
Otherwise we will eventually have dozens and dozens of tiny unit test
to the tools executable, not a whole new
statically-linked executable.
Cheers,
John
On Fri, Apr 5, 2019 at 3:55 PM Tomasz Wlostowski
wrote:
>
> On 05/04/2019 16:48, John Beard wrote:
> > Hi Tom,
> >
> > Can the hyperlink export test not exist in the existing qa_pcbnew
> >
Hi,
I'm trying to get the 3d viewer hooked up to the common settings, so I
can change the DPI scaling.
The 3d viewer frame is a KIWAY_PLAYER, but doesn't seem to participate
in the Kiway ecosystem, as it's never created by a call to Player().
This means it's never included in common settings upda
d.
>
>
> > On 8 Apr 2019, at 09:56, John Beard wrote:
> >
> > Hi,
> >
> > I'm trying to get the 3d viewer hooked up to the common settings, so I
> > can change the DPI scaling.
> >
> > The 3d viewer frame is a KIWAY_PLAYER, but does
Hi Wayne,
What is the policy for minor string changes going into 5.1.2?
Cheers,
John
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More
Hi Simon,
Thanks for catching. Pushed to both branches.
Cheers,
John
On Mon, Apr 15, 2019 at 11:35 PM Simon Richter wrote:
>
>
> This is not technically correct (the Boost dependency is introduced through
> libcommon), but less invasive as it doesn't pull in libcommon on the linker
> command l
Hi,
Occasionally, there are references made to performance during
processing of very large >8 layer PCBs (and the schematics thereof).
However, I don't have any such PCBs to uses as a reference, and there
are (reasonably, due to the size) none in the demos folder.
I can make faked-up PCBs program
Hi,
The following members are not initialised in ZONE_CONTAINER's
constructor, but they are accessed in the Setter functions:
* m_ThermalReliefCopperBridge
* m_ZoneMinThickness
This was picked up with valgrind memcheck:
valgrind --tool=memcheck qa/pcbnew_tools/qa_pcbnew_tools pcb_parser
/tmp/ci
==by 0xD4A3CD: PCB_PARSER::parseBOARD() (pcb_parser.cpp:486)
==12798==by 0xD4A0E9: PCB_PARSER::Parse() (pcb_parser.cpp:465)
Cheers,
John
On Tue, Apr 16, 2019 at 5:38 PM jp charras wrote:
>
> Le 16/04/2019 à 17:59, John Beard a écrit :
> > Hi,
> >
> > The followin
Hi Jeff,
On Thu, Apr 18, 2019 at 12:02 AM Jeff Young wrote:
> I wonder if this is because I have OCE turned off so the code-under-test
> isn’t being built?
Yes, that's exactly the reason. I've fixed it by aborting the test
CMake file if the required lib isn't in place.
Cheers,
John
_
Hi Simon,
I committed this patch. Thank you for the contribution!
John
On Mon, Apr 15, 2019 at 8:53 PM Simon Richter wrote:
>
>
> Some translations may need this.
> ---
> eeschema/netlist_object.cpp | 2 +-
> eeschema/tools/sch_editor_control.cpp | 4 ++--
> 2 files changed, 3 insert
e string
> freeze. I would prefer that we keep the changes to a minimum. The two
> patches that Simon submitted are fine.
>
> Cheers,
>
> Wayne
>
> On 4/15/19 6:13 PM, John Beard wrote:
> > Hi Wayne,
> >
> > What is the policy for minor st
+1
Reminds me of the scoring we used to do code review on Gerrit, with a different
threshold. That worked well.
Is there a minimum time to wait for a -1? If a reviewer didn't see the mail
before three +1s, their veto is too late. But if they were checking their mail
earlier, the veto would co
understands.
Cheers,
John
[1]:
https://jenkins.simonrichter.eu/job/linux-kicad-head/lastCompletedBuild/testReport/
From 8586df1dc314ed797debd943137eac04b3f7d519 Mon Sep 17 00:00:00 2001
From: John Beard
Date: Tue, 23 Apr 2019 08:58:17 +0100
Subject: [PATCH 1/4] Eeschema: build with object librar
Hi Adam,
The fix was only made on master. I have pushed it to 5.1 as commit
b6c9d4a88c279c4269cb22995765640de294a16e
Cheers,
John
On Wed, Apr 24, 2019 at 2:24 PM Adam Wolf wrote:
>
> I am seeing this again when I am trying to build 5.1.2 for macOSes
> older than 10.14. Does that make any sens
Hi,
Context: I have managed to break *another* build on a platform I don't
have (macos).
Is there an list of CI builders which we expect to be passing at any
given time? So far, I know of:
* Simon's Jenkins[1], which has a Msys2, MSVC and Debian builds that
fire on SCM changes. I often do build
On Wed, Apr 24, 2019 at 4:51 PM Steven A. Falco wrote:
>
> It is not quite a CI builder, but there is also the Fedora nightly that
> happens on Copr:
> https://copr.fedorainfracloud.org/coprs/g/kicad/kicad/
That's useful to know, thanks
> I see some failures there on kicad2step. Here is a repr
Hi Antonio,
On 25/04/2019 10:39, Antonio Vázquez Blanco wrote:
I've been playing around a little bit with KiCad source code lately and
in the forums [1] I was encouraged to write to the dev mailing list in
order to get feedback on how to improve the current library error checking.
This looks
On 24/04/2019 18:07, Adam Wolf wrote:
I will set up a new view for the macbuilder that should be all green,
and I can also set it up to push build failures to a list.
Good idea. I did the same for Simon's Jenkins:
https://jenkins.simonrichter.eu/view/KiCad%20Status
Once you have such a view,
On 29/04/2019 04:25, Jon Evans wrote:> Is there any particular
philosophical or technical reason why
> ADVANCED_CONFIG was created as a read-only system (i.e. to modify any
> values, you are supposed to edit the text config file and restart the
> program)?
More of a philosophical reason.
Mostly
On 29/04/2019 12:41, John Beard wrote:
Then you might start running into consistency issues with multiple
users of A_C getting different answers.
Notably, there is a specific case that illustrates where global state
can cause Bad Things (TM) right now: unit tests.
Say you had a mutable
On 30/04/2019 16:01, Jeff Young wrote:
Primarily for performance reasons.
WRT performance, I did a few benchmarks for reference (on Linux)
Loading this large CIAA PCB[1] allocates, out of a peak usage of 467MB
of heap with a 0.01% threshold:
* 9.6MB of std::basic_string::_M_assign
* 9.4M
On 30/04/2019 18:19, Jeff Young wrote:
I was referring to UCS-2 or UCS-4. I’m evidently behind the times, though,
because I now see that UTF-32 and UCS-4 are equivalent.
(Which means that both some of John’s original premises and my quote in teal
below were wrong: UTF32 is indeed a one:one ma
On 30/04/2019 21:55, Jeff Young wrote:
I think we’re a long way from handling Hiragana, Katakana or Kanji. Probably
the same for Tamil or Telegu, although I know nothing about them.
We can already accept Unicode from many string user inputs. For example
a net name can be "🇺🇸" (this is 2 code
On 01/05/2019 10:53, Mário Luzeiro wrote:
Hi Henner,
I didn't get a chance to test the patch yet but
I was thinking adding another options to implement some kind of spiral
rendering from the center :) (as usually the board is centered)
Assuming the reason for the Morton code in the existing
Hi Wayne,
I don't see any issues on the Jenkins Msys2 and MSVC builds. In fact,
it's green across the board:
https://jenkins.simonrichter.eu/view/KiCad%20Status/
What is the error? And when did it go wrong?
Cheers,
John
On 01/05/2019 12:42, Wayne Stambaugh wrote:
Anyone else having build
On 01/05/2019 12:53, Wayne Stambaugh wrote:
I'll try a clean build to see if that fixes it. I rarely have to do
that. I haven't done a windows build in a while so I was a bit
surprised. I've attached the build error.
Hmm, have you upgraded Boost or something? IIRC, Boost 1.70 was just
rele
On 01/05/2019 12:53, Jeff Young wrote:
Someone on the forums had to turn off the qa tests on Linux to get it to build:
https://forum.kicad.info/t/call-for-testers-eemodern/16663/8
This bit, right?
.../qa/utils/kicad2step/pcb/test_base.cpp: In member function ‘void
PcbBase::SexprTo2DPosAndRot:
On 01/05/2019 13:36, Wayne Stambaugh wrote:
On 5/1/2019 8:27 AM, John Beard wrote:
On 01/05/2019 12:53, Jeff Young wrote:
Someone on the forums had to turn off the qa tests on Linux to get it
to build:
It fails during linking not compiling. I also tried clang but I get the
same error. Here
On 01/05/2019 13:57, Mário Luzeiro wrote:
Hi John,
yeah the Morton code is to improve cache hits.
Regarding the speed test, since OS are multi-tasking there could be some
interference on the results so 1s difference is not a very measurable
difference ( 4% ).
A possibility would be to run the
On 2 May 2019 17:47:59 BST, Seth Hillbrand wrote:
>Am 2019-05-02 12:37, schrieb Henner Zeller:
>> On Thu, 2 May 2019 at 09:31, Henner Zeller wrote:
>>>
>>> On Thu, 2 May 2019 at 09:25, Seth Hillbrand
>>> wrote:
>
>Any event, anyone see issues with the center spiral patch before it is
>merged
I had a quick look at this. I'm not sure expanding all search hits to show sub
units is the best fix here. It means for the 7400 series that you can fit only
3 or 4 search hits in the dialog at default size.
I think it would be reasonable to expand only as far as the LIBID nodes.
Cheers,
John
Hi Jeff,
I think it is the index access operator that performs this caching, to allow
you to access the n'th code point any number of times while only iterating the
string only once.
However, you can still use the iterator access safely. It is only index based
access that is cached and thread-
Hi Ian,
On 05/05/2019 15:58, Ian McInerney wrote:
I saw that on the todo list along with the
auto_ptr replacement and decided to take a stab at it.
Thank you for taking the initiative here, and thanks also for updating
the TODO list!
doing the auto_ptr replacement part since when I looked t
On 03/05/2019 19:28, Wayne Stambaugh wrote:
I'm guessing John used a later version of clang-format when he wrote the
commit hooks. John, any ideas how to fix this or do we force devs to
use clang-format > 3.8?
I'm on Arch, so I have quite recent clang-format (8.0.0). This
particular option h
On 06/05/2019 17:51, Reece R. Pollack wrote:
John,
I've already jumped to clang-format 6.0, which is one of the optional
installs for Mint 18. That works, once you get all the symlinks fixed,
Good to know, thanks for the update.
except it keeps wanting to reformat my switch statements like
On 06/05/2019 21:46, Wayne Stambaugh wrote:
John,
On 5/6/19 4:09 PM, John Beard wrote:
On 06/05/2019 17:51, Reece R. Pollack wrote:
John,
I've already jumped to clang-format 6.0, which is one of the optional
installs for Mint 18. That works, once you get all the symlinks fixed,
Go
On 06/05/2019 22:12, Wayne Stambaugh wrote:
On 5/6/2019 5:11 PM, John Beard wrote:
I still suggest to change it to false be default and allow developers to
manually align when they want (and then override the formatter). Then at
least the default behaviour is a valid formatting choice.
That
On 24/04/2019 13:38, John Beard wrote:
I have here a couple of patches that enable unit tests in eeschema's
library code.
this is a proposal to merge the changes attached, at the cost of
another heavy link. However, having a working eeschema test suite will
allow such things as eesch
On 10/05/2019 11:08, Jeff Young wrote:
Once we remove PCBNew’s legacy canvas I’d like to propose that we simplify the
hotkey architecture:
Get rid of the HotKey data-structures; have Actions specify a hotkey
key-combination.
+0 (strongly support stripping legacy hotkey infrastructure, but...
On 10/05/2019 11:53, Jeff Young wrote:
My concern with this is that the more spread out you store the info, the
more maps you need, and the more room for error you have (when maps are
missing keys, etc.).
I have a single big default list in mind, rather than many disparate
lists. This is co
Hi Steve,
Make test will run the tests. You need to run that after building. The
qa_all_tests can be used to build only tests, but it is redundant if you
already built everything.
Cheers,
John
On 10 May 2019 18:32:21 BST, "Steven A. Falco" wrote:
>I used the footprint wizard to create some f
On 15/05/2019 18:32, Jeff Young wrote:
This is probably a good time for a few words about the new code.
Looks nice!
I'm seeing dozens (36 to be exact) of asserts on startup under GTK3:
./src/gtk/menu.cpp(729): assert "Assert failure" failed in SetBitmap():
only normal menu items can have bit
Hi Steve,
On 15/05/2019 22:42, Steven A. Falco wrote:
There are two things that would make this easier:
1) A way to assign a hot-key to a Symbol Editor grid setting; i.e. something similar to
the pcbnew "Fast Switching" feature in the Grid menu.
I have added the N/Shift+N next/previous grid
Hi Wayne,
On 21/05/2019 18:26, Wayne Stambaugh wrote:
John,
If this is not acceptable, would an option that allows the user to turn
this feature on or off be an acceptable solution. In some use cases, I
can see the value of the this patch.
I'm not saying it's unacceptable, I'm just not sure
iases for the 'fixes' alias, so let's do it for the
formatting too!
[1]:
http://docs.kicad-pcb.org/doxygen/md_Documentation_development_coding-style-policy.html
>From fb053d8cc27179542320bc3b297087aa28fd440a Mon Sep 17 00:00:00 2001
From: John Beard
Date: Wed, 22 May 2019 10:26
On 21/05/2019 22:27, Nick Østergaard wrote:
I think we should add notes in the kicad dev docs (some md file in the
doxygen docs) along the lines of
This email is not intended to be a discussion, just a +1 thing if you
like it and feedback on where we should document this.
+1. Consistent use of
Hi Wayne,
On 16/05/2019 20:12, Wayne Stambaugh wrote:
I absolutely would like to see more Eeschema unit testing. I'm going to
need it when I start on the new file format code. Of course the usual
caveats like don't break anything apply.
I have pushed the first couple of commits for this. It
On 23/05/2019 19:05, Wayne Stambaugh wrote:
/home/wayne/src/kicad-trunk/include/bitmap_base.h:34:1: error: declaration
conflicts with target of using declaration already in scope
class COLOR4D;
Odd, I don't see this at all, even on Jenkins. Must be a compiler thing?
COLOR4D is actually
Hi Wayne,
On 11/06/2019 14:15, Wayne Stambaugh wrote:
As of the commit b9e07f2a706b719a5f86b6372c5a7cda2f6a7c55 I am getting
the following build errors on linux using clang:
I just fixed these in 1e4abac57 as they were blocking me.
The first one, I bisected down to c13ef839c, though I do not
Hi Mark,
On 11/06/2019 19:20, Mark Roszko wrote:
Idea: quick tap of R just does 90. Long press R and it pops open a tiny
dialog with focus to enter a rotation number and then you can hit enter
and it just sets the rotation angle directly on the footprint
properties.
Ctrl+M will already do th
On 10/06/2019 23:35, Andrew Lutsenko wrote:
If adding 2 new default shortcuts is not ideal maybe there is a way to
have these as actions with no shortcuts and let the user choose if they
want to use them and assign any shortcut they want.Is that technically
supported with current hotkey framew
On 12/06/2019 15:33, Jeff Young wrote:
I’m putting the kicad gestures into the List HotKeys dialog.
Excellent!
Pan Up/Down shift-mouse-wheel
Pan Left/Right ctrl-mouse-wheel
Do we need also zoom?
Also, these are not the same when touchpad panning
("Inkscape mode") is on
On 12/06/2019 16:28, Reece R. Pollack wrote:
Most applications I've used provide these "gestures" with the mouse wheel:
Pan up/down: no modifier
Zoom in/out: Control
Pan left/right: Shift
This is exactly how it works in "touchpad pan" mode (an option in the
preferences panel).
If you have a
On 12/06/2019 16:45, Jon Evans wrote:
In my opinion, it would be more powerful to break out the individual
settings rather than keep around "touchpad pan mode", so that everyone
can be happy. I think this is the only solution when there are so many
different ways people could be used to coming
401 - 464 of 464 matches
Mail list logo
