Nikita Nemkin added the comment:
> Unfortunately there is no official maintained list if those symbols, so is
> taking a while
On Windows there is PC/python3dll.c.
--
___
Python tracker
<https://bugs.python.org/i
Nikita Nemkin added the comment:
Thanks for reviewing my patches so quickly, Steve.
There's one last Windows-specific patch left
(https://bugs.python.org/issue41038). Single-line change.
--
___
Python tracker
<https://bugs.python.org/issue41
Change by Nikita Nemkin :
--
keywords: +patch
pull_requests: +20239
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/21070
___
Python tracker
<https://bugs.python.org/issu
New submission from Nikita Nemkin :
Visual Studio projects need a bit of grooming.
* File filters don't reflect recent file movements and additions.
* Solution file is missing _zoneinfo project entries, resulting in dirty repo
on every save.
* bdist_wininst project hasn't been updated
Nikita Nemkin added the comment:
Please take your time.
FWIW, I tested this patch by 1) comparing dumpbin /exports output; 2) comparing
bunaries themselves (.dll and .lib) and 3) linking xxlmited module.
It helps that 7-zip can open .dll and .lib files files as if they were archives
Change by Nikita Nemkin :
--
keywords: +patch
pull_requests: +20209
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/21037
___
Python tracker
<https://bugs.python.org/issu
New submission from Nikita Nemkin :
pyshellext uses MIDL to generate a header, whose only purpose is to define a
class GUID. MIDL generation step can be replaced with a simple #define. This
doesn't really matter for VS, but other build systems (CMake, probably Meson
too) will benefit
Change by Nikita Nemkin :
--
keywords: +patch
pull_requests: +20178
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/21004
___
Python tracker
<https://bugs.python.org/issu
New submission from Nikita Nemkin :
Every Windows project has a custom target (included from pyproject.props) that
generates a header with definitions for resource files.
Those definitions (PYTHON_DLL_NAME and FIELD3) can be passed directly to
resource compiler.
Another definition
Nikita Nemkin added the comment:
Shouldn't the following files be deleted too?
Include/bitset.h
Include/grammar.h
Include/graminit.h
Include/parsetok.h
Include/node.h
Python/graminit.c
Parser/node.c
Also declarations:
PyNode_Compile in Include/compile.h
PyParser_SimpleParse* in Include
Change by Nikita Nemkin :
--
keywords: +patch
pull_requests: +20163
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/20989
___
Python tracker
<https://bugs.python.org/issu
New submission from Nikita Nemkin :
python3.dll build process can be simplified if we use linker comment #pragma
instead of .def files. Custom build targets become unnecessary and hardcoded
Python DLL name can be replaced with a macro.
Also, python3.dll doesn't need DllMain and can be built
Change by Nikita Nemkin :
--
keywords: +patch
pull_requests: +20159
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/20985
___
Python tracker
<https://bugs.python.org/issu
New submission from Nikita Nemkin :
In absence of explicit declaration, resource compiler uses system
codepage. When this codepage is DBCS or UTF-8, Python's copyright
string is corrupted, because it contains copyright sign encoded
as \xA9.
The fix is to explicitly declare codepage 1252
New submission from Nikita Nemkin :
Windows build system always builds liblzma with optimizations disabled (/Od),
even in Release configuration.
Compared to optimized build (/O2), compression speed is 2-2.5x slower and
module size is 30% larger.
--
components: Build
messages: 370702
Nikita Nemkin added the comment:
COM should be initialized on demand by C modules that need it. They might need
STA or MTA, it's their choice to make.
Python core (ceval and builtins) doesn't need COM and shouldn't impose COM
threading models on any threads it creates. Things like -X:STA
New submission from Nikita Nemkin:
Currently array module only provides platform-dependent type specifiers.
It would be very nice to have platform-independent specifiers in addition to
that, matching the struct module.
For example array('H') - an array of little-endian 2-byte integers
17 matches
Mail list logo
