Index | Thread | Search

From:
Stuart Henderson <stu@spacehopper.org>
Subject:
Re: [NEW PORT] devel/fast-float - Fast and exact implementation of the C++ from_chars functions
To:
Thomas Frohwein <tfrohwein@fastmail.com>, Jose Maldonado <josemald89@gmail.com>, <ports@openbsd.org>
Date:
Tue, 13 Feb 2024 19:14:18 +0000

Download raw body.

Thread
I'd prefer to skip NO_BUILD unless it fails without - if a newer version 
starts doing something in that stage and you have NO_BUILD set and don't 
notice it'll end up running build stages in "make fake"

-- 
  Sent from a phone, apologies for poor formatting.

On 13 February 2024 18:51:29 Thomas Frohwein <tfrohwein@fastmail.com> wrote:

> On Tue, Feb 13, 2024 at 05:00:08PM +0000, Stuart Henderson wrote:
>> On 2024/02/12 13:37, Jose Maldonado wrote:
>> >
>> > Hello everyone! New port here.
>> >
>> > In this case I bring you devel/fast-float, a series of C/C++ headers
>> > that seeks to offer us numerical operations 4 to 10x times faster than
>> > standard libraries (GCC and Clang).
>> >
>> > These libraries in question are on GitHub [1] and are part of the new
>> > dependencies for libplacebo (>=v6.338.0).
>> >
>> > These libraries are necessary to compile the new version of libplacebo
>> > and bump new version of mpv (v0.37.0)
>> >
>> > [1] https://github.com/fastfloat/fast_float
>> >
>> >
>> > --
>> > *********************************************************
>> > Dios en su cielo, todo bien en la Tierra
>>
>> updated tgz attached is OK sthen@, here's the diff:
>>
>>
>>
>> diff --git a/devel/fast-float/Makefile b/devel/fast-float/Makefile
>> old mode 100755
>> new mode 100644
>> index 3cfef11..0366819
>> --- a/devel/fast-float/Makefile
>> +++ b/devel/fast-float/Makefile
>> @@ -1,4 +1,4 @@
>> -COMMENT =	Fast and exact implementation of the C++ from_chars
>> +COMMENT =	fast and exact implementation of from_chars for float/double
>>
>>  V =		6.1.0
>>  PKGNAME =	fast-float-${V}
>> @@ -8,17 +8,9 @@ GH_PROJECT =	fast_float
>>
>>  CATEGORIES =	devel
>>
>> -HOMEPAGE =	https://github.com/fastfloat/fast_float
>> -
>> -# Apache 2.0 - BOOST 1.0 - MIT
>> +# Apache 2.0, Boost 1.0, or MIT
>>  PERMIT_PACKAGE =	Yes
>>
>> -BULID_DEPENDS =	devel/cmake \
>> -		devel/ninja
>> -
>>  MODULES =	devel/cmake
>>
>> -NO_BUILD =	Yes
>
> This one actually seems to be a NO_BUILD=Yes as it only installs header
> files.
>
> $ make build
> ...
> Run Build Command(s): /usr/local/bin/ninja -v -j 1
> ninja: no work to do.
>
>> -NO_TEST =	Yes
>> -
>>  .include <bsd.port.mk>
>> diff --git a/devel/fast-float/pkg/DESCR b/devel/fast-float/pkg/DESCR
>> old mode 100755
>> new mode 100644
>> index d03526b..94eef24
>> --- a/devel/fast-float/pkg/DESCR
>> +++ b/devel/fast-float/pkg/DESCR
>> @@ -1 +1,3 @@
>> -Fast and exact implementation of the C++ from_chars functions.
>> +The fast_float library provides fast header-only implementations for
>> +the C++ from_chars() functions for 'float' and 'double' as well as
>> +integer types.