Hi all,
As people may have seen there's a pre-release here:
http://downloads.xiph.org/releases/flac/beta/
Specifically:
flac-1.3.1pre1.tar.xz : The source code
flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries
Please test.
I'm particularly interested in hearing about the windows binaries
which were cross compiled from Linux to Windows. Unfortunately
there is a bug in Wine which prevents me from running either of
the testsuites. Maybe someone on windows could grab the source
and the binaries, build from source, copy the binaries into the
built flac tree and then run the tests.
Cheers,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/
On 25 Nov 2014, at 08:43, Erik de Castro Lopo <mle+la at mega-nerd.com> wrote:> Hi all, > > As people may have seen there's a pre-release here: > > http://downloads.xiph.org/releases/flac/beta/ > > Specifically: > > flac-1.3.1pre1.tar.xz : The source code > flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries > > Please test. > > I'm particularly interested in hearing about the windows binaries > which were cross compiled from Linux to Windows. Unfortunately > there is a bug in Wine which prevents me from running either of > the testsuites. Maybe someone on windows could grab the source > and the binaries, build from source, copy the binaries into the > built flac tree and then run the tests. > > Cheers, > ErikI seem to recall from 1.3.0 that there was an issue with the encoder string (for instance "reference LibFLAC 1.3.1 20141125"). I believe this is set at various locations which are also platform dependent and in 1.3.0 some places initially still contained the old version. Unless this was simplified in the mean time and I missed that, it would be useful to test if the resulting flac files on the various platforms report the correct version.
MauritsVB wrote:> I seem to recall from 1.3.0 that there was an issue with the encoder > string (for instance "reference LibFLAC 1.3.1 20141125"). I believe > this is set at various locations which are also platform dependent and > in 1.3.0 some places initially still contained the old version. > > Unless this was simplified in the mean time and I missed that, it would > be useful to test if the resulting flac files on the various platforms > report the correct version.Good catch. I correctly fixed all the version strings, but the date string (eg 20130526) was not fixed, but will be. Cheers, Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/
Op 25-11-14 om 10:22 schreef MauritsVB:> Unless this was simplified in the mean time and I missed thatIt was in this commit: http://git.xiph.org/?p=flac.git;a=commit;h=f931d134112e327c5ccb88a0a8f2eb2abfa281e3
On 25 Nov 2014, at 08:43, Erik de Castro Lopo <mle+la at mega-nerd.com> wrote:> Hi all, > > As people may have seen there's a pre-release here: > > http://downloads.xiph.org/releases/flac/beta/ > > Specifically: > > flac-1.3.1pre1.tar.xz : The source code > flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries > > Please test. > > I'm particularly interested in hearing about the windows binaries > which were cross compiled from Linux to Windows. Unfortunately > there is a bug in Wine which prevents me from running either of > the testsuites. Maybe someone on windows could grab the source > and the binaries, build from source, copy the binaries into the > built flac tree and then run the tests. > > Cheers, > ErikIs there any objection to posting the news about the pre-release in this topic on HydrogenAudio (http://www.hydrogenaud.io/forums/index.php?showtopic=101082)? People have been using that topic to test various Git versions over the past few months and it could get this pre-version some extra eyeballs over the next 24 hours.
Op 25-11-14 om 09:43 schreef Erik de Castro Lopo:> I'm particularly interested in hearing about the windows binaries > which were cross compiled from Linux to Windows.I've ran test_flac.sh, test_metaflac.sh and test_streams.sh. If the other tests (like ./test_libFLAC.sh) have to be run, the corresponding binaries have to be supplied :) Anyway, test_flac.sh runs fine except for the metadata tests and test_metaflac.sh has the same metadata problem. This is explained on line 1157 of test_flac.sh> # minor danger, changing vendor strings might change the > length of the > # VORBIS_COMMENT block, but if we add "^ length: " to the > patterns, > # we lose info about PADDING size that we need > grep -Ev '^ vendor string: |^ m..imum .....size: ' | sed > -e 's/, stream_offset.*//'test_streams.sh runs without problems too
On 25.11.2014 11:43, Erik de Castro Lopo wrote:> > flac-1.3.1pre1.tar.xz : The source code > flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries > > Please test. > > I'm particularly interested in hearing about the windows binaries > which were cross compiled from Linux to Windows. Unfortunately > there is a bug in Wine which prevents me from running either of > the testsuites. Maybe someone on windows could grab the source > and the binaries, build from source, copy the binaries into the > built flac tree and then run the tests. >MSYS2 + i686-w64-mingw32-gcc-4.9.1 -> builds ok, passes all tests, although there were some strange messages about compression ration being > 1.0 in two seek tests. Dropped x86 binaries into the same tree, ran the tests - passes (same ratio warnings as above). Dropped x86-64 binaries into the same tree, ran the tests - also passes (same ratio warnings as above). Logs of the output of all three "make check"s are available on demand (uncompressed 1MB per log, wouldn't want to send this to the ML). I did have to apply some compatibility patches, as usual, but nothing major. Patches are attached. -- O< ascii ribbon - stop html email! - www.asciiribbon.org -------------- next part --------------
MauritsVB wrote:> Is there any objection to posting the news about the pre-release in > this topic on HydrogenAudio (http://www.hydrogenaud.io/forums/index.php?showtopic=101082)?All good! Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/
On Nov 25 00:43:22, mle+la at mega-nerd.com wrote:> Hi all, > > As people may have seen there's a pre-release here: > > http://downloads.xiph.org/releases/flac/beta/ > > Specifically: > > flac-1.3.1pre1.tar.xz : The source code > flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries > > Please test.The flac binary builds and runs fine on current OpenBSD/i386 and OpenBSD/armv7. I am at a loss though running the test suite $ cd test $ make check $ make check Making check in cuesheets Making check in flac-to-flac-metadata-test-files Making check in metaflac-test-files Making check in pictures make test_libFLAC.sh test_libFLAC++.sh test_grabbag.sh test_flac.sh test_metaflac.sh test_replaygain.sh test_seeking.sh test_streams.sh test_compression.sh `test_libFLAC.sh' is up to date. `test_libFLAC++.sh' is up to date. `test_grabbag.sh' is up to date. `test_flac.sh' is up to date. `test_metaflac.sh' is up to date. `test_replaygain.sh' is up to date. `test_seeking.sh' is up to date. `test_streams.sh' is up to date. `test_compression.sh' is up to date. FLAC__TEST_LEVEL=1 FLAC__TEST_WITH_VALGRIND=no ./test_libFLAC.sh /bin/sh: ./test_libFLAC.sh: No such file or directory *** Error 1 in /home/hans/src/flac-1.3.1pre1/test (Makefile:689 'check') Am I missing something obvious? Is running the test suite described somewhere in the documentation? I coudn't find it. Jan
On 25.11.2014 16:03, Jan Stary wrote:> On Nov 25 00:43:22, mle+la at mega-nerd.com wrote: >> Hi all, >> >> As people may have seen there's a pre-release here: >> >> http://downloads.xiph.org/releases/flac/beta/ >> >> Specifically: >> >> flac-1.3.1pre1.tar.xz : The source code >> flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries >> >> Please test. > > > I am at a loss though running the test suite > > $ cd test > $ make check > /bin/sh: ./test_libFLAC.sh: No such file or directory > *** Error 1 in /home/hans/src/flac-1.3.1pre1/test (Makefile:689 'check') > > Am I missing something obvious? Is running the test suite > described somewhere in the documentation? I coudn't find it. >I ran `make check' from the toplevel, not from the tests subdir, although that shouldn't make a difference. I do hope you're running the testsuite in-tree. It doesn't work if you have a separate build directory, although the error you'd get in that case is different from the one you are getting right now, so i don't know. -- O< ascii ribbon - stop html email! - www.asciiribbon.org -------------- next part -------------- A non-text attachment was scrubbed... Name: 0x922360B0.asc Type: application/pgp-keys Size: 1717 bytes Desc: not available Url : http://lists.xiph.org/pipermail/flac-dev/attachments/20141125/1868dba0/attachment.key -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature Url : http://lists.xiph.org/pipermail/flac-dev/attachments/20141125/1868dba0/attachment.pgp
Erik de Castro Lopo <mle+la at mega-nerd.com> ?????(?) ? ????? ?????? Tue, 25 Nov 2014 11:43:22 +0300:> > flac-1.3.1pre1.tar.xz : The source code > flac-1.3.1pre1-win.zip : Windows 32 and 64 bit binaries > > Please test.Binaries contain debug info, so it's possible to reduce their size. For example, 32-bit: for flac.exe - from 1080 to 670 kB, for metaflac.exe - from 802 to 421 kB.
On 2014-11-25 7:36 AM, lvqcl wrote:> Binaries contain debug info, so it's possible to reduce their size. For example, > 32-bit: for flac.exe - from 1080 to 670 kB, for metaflac.exe - from 802 to 421 kB.Yes, please strip the windows binaries. That's preferred by the Windows community, as I was reminded every time I built opus binary packages... -r
On Nov 25 00:43:22, mle+la at mega-nerd.com wrote:> http://downloads.xiph.org/releases/flac/beta/ > Please test.Running 'make fullcheck' fails for me like this: $ pwd /home/hans/src/flac-1.3.1pre1/test $ make fullcheck FLAC__TEST_LEVEL=1 FLAC__TEST_WITH_VALGRIND=no ./test_libFLAC.sh /bin/sh: ./test_libFLAC.sh: No such file or directory *** Error 1 in /home/hans/src/flac-1.3.1pre1/test (Makefile:702 'fullcheck') $ ls -l ./test_libFLAC.sh -rwxr-xr-- 1 hans hans 1275 Nov 25 04:42 ./test_libFLAC.sh $ ./test_libFLAC.sh ksh: ./test_libFLAC.sh: No such file or directory I suppose this is due to the #!/bin/bash - indeed, bash does not exist here. Is there a reason the test scripts are calling bash? It doesn't seem there is soemthing bash-specific that plain old sh could not do. Replacing '#!/bin/bash -e' with '#!/bin/sh -e' and removing a few pieces of bash-specific syntax apparently runs the tests just fine. (That's not to say they pass.) I believe it also makes the test scripts more portable; diff below. The following tests go fine: test_libFLAC.sh test_libFLAC++.sh test_grabbag.sh test_flac.sh The metaflac test fails like this: Generating stream... flac 1.3.0, Copyright (C) 2000-2009, 2011-2013 Josh Coalson & Xiph.Org Foundation flac comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under certain conditions. Type `flac' for details. 80000+0 records in 156+1 records out 80000 bytes transferred in 0.182 secs (437977 bytes/sec) -: Verify OK, wrote 9004 bytes, ratio=0.112 test case00: --list... OK test case01: STREAMINFO --show-* shortcuts... OK test case02: --add-padding... OK test case03: --set-tag=ARTIST... OK test case04: --set-tag=ARTIST... OK test case05: --set-tag=ARTIST... OK test case06: --set-tag=TITLE... OK test case07: --show-vendor-tag --show-tag=ARTIST... ERROR: metadata does not match expected metaflac-test-files/case07-expect.meta The replaygain test fails; reading the test script, I don't think the problem is in trying to call mawk (which does not exist here). Generating stream... flac 1.3.0, Copyright (C) 2000-2009, 2011-2013 Josh Coalson & Xiph.Org Foundation flac comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under certain conditions. Type `flac' for details. 80000+0 records in 156+1 records out 80000 bytes transferred in 0.194 secs (412063 bytes/sec) -: Verify OK, wrote 9004 bytes, ratio=0.112 ./test_replaygain.sh: line 85: mawk: command not found Testing FLAC replaygain 8000 (8000 x 1) ... -: ERROR: got partial sample The following tests pass fine: test_seeking.sh test_streams.sh test_compression.sh This is all on current OpenBSD/i386. Jan --- test_bins.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_bins.sh Tue Nov 25 23:27:47 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2001-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=../src/flac:$PATH PATH=../objs/$BUILD/bin:$PATH --- test_compression.sh Tue Nov 25 04:40:14 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_compression.sh Tue Nov 25 23:19:19 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2012 Xiph.Org Foundation @@ -17,7 +17,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=`pwd`/../src/flac:$PATH @@ -39,7 +39,7 @@ for k in 0 1 2 3 4 5 6 7 8 ; do echo "Error : Compression ${last_k} size $last_size >= compression $k size $size." exit 1 fi - let last_size=${size}+10 + last_size=$((${size}+10)) last_k=${k} rm -f ${fname} done --- test_flac.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_flac.sh Tue Nov 25 23:00:35 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2001-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh # we use '.' as decimal separator in --skip/--until tests export LANG=C LC_ALL=C --- test_grabbag.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_grabbag.sh Tue Nov 25 23:00:10 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2001-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=../src/test_grabbag/cuesheet:$PATH PATH=../src/test_grabbag/picture:$PATH --- test_libFLAC++.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_libFLAC++.sh Tue Nov 25 22:59:18 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2002-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=../src/test_libFLAC++:$PATH PATH=../objs/$BUILD/bin:$PATH --- test_libFLAC.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_libFLAC.sh Tue Nov 25 22:58:46 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2001-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=../src/test_libFLAC:$PATH PATH=../objs/$BUILD/bin:$PATH --- test_metaflac.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_metaflac.sh Tue Nov 25 23:02:07 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2002-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=`pwd`/../src/flac:$PATH PATH=`pwd`/../src/metaflac:$PATH --- test_replaygain.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_replaygain.sh Tue Nov 25 23:02:32 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2002-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=`pwd`/../src/flac:$PATH PATH=`pwd`/../src/metaflac:$PATH --- test_seeking.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_seeking.sh Tue Nov 25 23:02:54 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2004-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=../src/flac:$PATH PATH=../src/metaflac:$PATH --- test_streams.sh Tue Nov 25 04:42:25 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_streams.sh Tue Nov 25 23:03:11 2014 @@ -1,4 +1,4 @@ -#!/bin/bash -e +#!/bin/sh -e # FLAC - Free Lossless Audio Codec # Copyright (C) 2001-2009 Josh Coalson @@ -18,7 +18,7 @@ # restrictive of those mentioned above. See the file COPYING.Xiph in this # distribution. -source common.sh +. ./common.sh PATH=../src/flac:$PATH PATH=../src/test_streams:$PATH --- test_wrapper.sh Tue Nov 25 04:40:14 2014 +++ /home/hans/src/flac-1.3.1pre1/test/test_wrapper.sh Tue Nov 25 23:21:26 2014 @@ -1,12 +1,12 @@ -#!/bin/bash -e +#!/bin/sh -e # This test script should exit on the first failure. -./test_libFLAC.sh -./test_libFLAC++.sh -./test_grabbag.sh -./test_flac.sh -./test_metaflac.sh -./test_replaygain.sh -./test_seeking.sh -./test_streams.sh +sh -e ./test_libFLAC.sh +sh -e ./test_libFLAC++.sh +sh -e ./test_grabbag.sh +sh -e ./test_flac.sh +sh -e ./test_metaflac.sh +sh -e ./test_replaygain.sh +sh -e ./test_seeking.sh +sh -e ./test_streams.sh
Op 25-11-14 om 23:39 schreef Jan Stary:> Is there a reason the test scripts are calling bash?The change from sh to bash was made a little more than a year ago. The mailing list thread accompanying this change can be found here: http://lists.xiph.org/pipermail/flac-dev/2013-September/004374.html
LRN wrote:> MSYS2 + i686-w64-mingw32-gcc-4.9.1 -> builds ok, passes all tests, although > there were some strange messages about compression ration being > 1.0 in two > seek tests.Yes, that is benign. FLAC is good at compression music, but not so good at compressing random noise which is used in a couple of the tests.> I did have to apply some compatibility patches, as usual, but nothing major. > Patches are attached.Err, why are these patches needed? I thought FLAC compiled correctly on all major platforms with all major compilers. Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/
On Nov 25 00:43:22, mle+la at mega-nerd.com wrote:> http://downloads.xiph.org/releases/flac/beta/ > Please test.case00 of ./test_metaflac.sh fails, the difference in results being this: --- metaflac-test-files/case00-expect.meta Tue Apr 30 12:47:29 2013 +++ /tmp/m Wed Nov 26 22:07:11 2014 @@ -2,6 +2,10 @@ METADATA block #0 type: 0 (STREAMINFO) is last: false length: 34 + minimum blocksize: 1152 samples + maximum blocksize: 1152 samples + minimum framesize: 10 bytes + maximum framesize: 10 bytes sample_rate: 8000 Hz channels: 1 bits-per-sample: 8 @@ -12,11 +16,12 @@ METADATA block #1 is last: false length: 18 seek points: 1 - point 0: sample_number=0 + point 0: sample_number=0, stream_offset=0, frame_samples=1152 METADATA block #2 type: 4 (VORBIS_COMMENT) is last: false - length: 40 + length: 44 + vendor string: reference libFLAC 1.3.1pre1 20130526 comments: 0 METADATA block #3 type: 1 (PADDING) It's not just the vendor string, it seems the format of --list output has changed Jan
On Nov 25 00:43:22, mle+la at mega-nerd.com wrote:> http://downloads.xiph.org/releases/flac/beta/ > Please test../test_replaygain.sh fails for me in tonegenerator(), saying: ./test_replaygain.sh[91]: mawk: not found Testing FLAC replaygain 8000 (8000 x 1) ... -: ERROR: got partial sample Apparently, the tone-generating awk script does not work with my system's awk, which is "awk version 20110810" as distributed with current OpenBSD 5.6 GENERIC.MP#610 amd64. With mawk-1.3.4.20140914 it works fine (and ./test_replaygain.sh passes). Can anyone please shed some light on what's mawk-spacific about that awk script? It seems to be just a sine-generator. Here is the head of the hexdump of the two generated tones. Notice how the 2501's are preceded by the 0000's with mawk. awk: 0000000 0080 da00 7882 ffff daff 7882 0080 2500 0000010 877d 2501 877d ff7f daff 7882 ffff daff 0000020 7882 0080 2500 877d 2501 877d ff7f daff 0000030 7882 ffff daff 7882 0080 2500 877d 2501 0000040 877d ff7f daff 7882 ffff daff 7882 0080 0000050 2500 877d 2501 877d ff7f daff 7882 ffff 0000060 daff 7882 0080 2500 877d 2501 877d ff7f 0000070 daff 7882 ffff daff 7882 0080 2500 877d 0000080 2501 877d ff7f daff 7882 ffff daff 7882 0000090 0080 2500 877d 2501 877d ff7f daff 7882 mawk: 0000000 0080 da00 7882 ffff daff 7882 0080 2500 0000010 877d 0000 2501 877d ff7f daff 7882 ffff 0000020 daff 7882 0080 2500 877d 0000 2501 877d 0000030 ff7f daff 7882 ffff daff 7882 0080 2500 0000040 877d 0000 2501 877d ff7f daff 7882 ffff 0000050 daff 7882 0080 2500 877d 0000 2501 877d 0000060 ff7f daff 7882 ffff daff 7882 0080 2500 0000070 877d 0000 2501 877d ff7f daff 7882 ffff 0000080 daff 7882 0080 2500 877d 0000 2501 877d 0000090 ff7f daff 7882 ffff daff 7882 0080 2500 Jan
On Nov 26 22:39:27, hans at stare.cz wrote:> On Nov 25 00:43:22, mle+la at mega-nerd.com wrote: > > http://downloads.xiph.org/releases/flac/beta/ > > Please test. > > ./test_replaygain.sh fails for me in tonegenerator(), saying: > > ./test_replaygain.sh[91]: mawk: not found > Testing FLAC replaygain 8000 (8000 x 1) ... -: ERROR: got partial sample > > Apparently, the tone-generating awk script does not work with > my system's awk, which is "awk version 20110810" as distributed > with current OpenBSD 5.6 GENERIC.MP#610 amd64. > With mawk-1.3.4.20140914 it works fine > (and ./test_replaygain.sh passes). > > Can anyone please shed some light on what's mawk-spacific > about that awk script? It seems to be just a sine-generator. > > Here is the head of the hexdump of the two generated tones.Ha! If the final printf("%c", byte[bx]); is changed to printf("%d", byte[bx]); the results produced by mawk and awk are identical. So the diference must be just that mawk printf's "%c" differently. $ wc -c a ma 22000 a 24000 ma
On Nov 25 00:43:22, mle+la at mega-nerd.com wrote:> http://downloads.xiph.org/releases/flac/beta/ > Please test../test_bins.sh fails for me due to BINS_PATH=../../test_files/bins test -d ${BINS_PATH} || exit 77 This happens in flac-1.3.1pre1/test, so ../../test_files/bins jumps out of the flac tree; is this intended? Jan
Jan Stary wrote:> ./test_bins.sh fails for me due to > > BINS_PATH=../../test_files/bins > test -d ${BINS_PATH} || exit 77 > > This happens in flac-1.3.1pre1/test, > so ../../test_files/bins jumps out > of the flac tree; is this intended?Interesting. That file does not seem to ever have been part of the actual FLAC test suite and the last non-trival change to that file was: commit 83d914253edf252f8315d82a598ac93994e3ecbe Author: Josh Coalson <jcoalson at users.sourceforce.net> Date: Fri Nov 10 05:15:23 2006 +0000 add missing --lax I suspect that was a test script that Josh used on his own machine with some files that were kept outside the FLAC svn/git tree. I've made a note to myself to look at that more closely after this release. Cheers, Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/