Miller 6 #851

johnkerl · 2022-01-10T00:52:57Z

johnkerl
Jan 10, 2022
Maintainer

Please see What's new in Miller 6 for complete information.

This discussion was created from the release Miller 6.

johnkerl · 2022-01-10T01:01:05Z

johnkerl
Jan 10, 2022
Maintainer Author

@jauderho re #821 and #822 -- https://github.com/johnkerl/miller/runs/4756000917?check_suite_focus=true looks great!! :)

6 replies

johnkerl Jan 10, 2022
Maintainer Author

@jauderho this is resolved.

For posterity, what I did to trigger a re-run was:

Goreleaser binary names #852
Made a temp release https://github.com/johnkerl/miller/releases/tag/v6.0.0-goreleaserfix (which I afterward deleted)
Download the .tar.gz files to local disk
tar ztf / tar zxf the one for darwin amd64 (which I'm on) & verify mlr existed & was executable
Rename all the .tar.gz files to have the filenames at https://github.com/johnkerl/miller/releases/tag/v6.0.0
Update the checksums txt file with the correct names, and verified shasum -a 256 was unchanged
Went back to https://github.com/johnkerl/miller/releases/tag/v6.0.0
Unattached the previous .tar.gz and checksums txt files
Uploaded the new ones from local disk

Thanks again for setting this up BTW!! I regret only that I didn't catch the miller / mlr issue sooner on 6.0.0.rc1; I didn't look carefully. Besides that though -- this go-releaser is a fantastic feature! :D

jauderho Jan 10, 2022

@johnkerl Congrats on the release. I'll have to go update my miller Docker builds to track the release instead of just git now.

Your fix for re-releasing is a lot more goreleaser than I have previously worked with before. Definitely good to know in case I have to trigger a re-release for something else later.

I'm not 100% sure but setting draft: true in the release section of the yaml might help with the re-release. See https://goreleaser.com/customization/release/

jauderho Jan 11, 2022

@johnkerl I've figured out how to use goreleaser to create Windows zip files so that can be part of the goreleaser process. Similarly, I've also figured out how to include deb/rpm builds. I am doing more testing but will submit a new pull request once I'm satisfied.

Preview is here: https://github.com/jauderho/bl3auto/blob/main/.goreleaser.yml

johnkerl Jan 11, 2022
Maintainer Author

@jauderho awesome!! :)

jauderho Jan 12, 2022

See PR #866 and #867

BEFH · 2022-01-10T01:23:24Z

BEFH
Jan 10, 2022

@johnkerl, should conda compile from https://github.com/johnkerl/miller/archive/refs/tags/v6.0.0.tar.gz or https://github.com/johnkerl/miller/releases/download/v6.0.0/miller-6.0.0.tar.gz? Traditionally I've used the tag (as did the original maintainers), but it looks like the files are different.

6 replies

BEFH Jan 10, 2022

It seems to have worked on MacOS, and it passed all of the unit tests on linux, but then when I ran mlr --version, it now segfaults. Also does it just running mlr. I'm trying to rebuild now.

Reading symbols from mlr...
[New LWP 79998]
Core was generated by `./mlr --version'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x00007fce97fff0a0 in ?? ()

BEFH Jan 10, 2022

False alarm; it worked the second time.

BEFH Jan 10, 2022

I've initiated the pull request and am waiting for CI checks to run on Azure. I will merve when and if they complete.

BEFH Jan 10, 2022

OK, it segfaulted on the linux build check, just like the first time I ran it locally. Thoughts?

https://dev.azure.com/conda-forge/84710dde-1620-425b-80d0-4cf5baca359d/_apis/build/builds/438468/logs/10

johnkerl Jan 10, 2022
Maintainer Author

@BEFH from
https://dev.azure.com/conda-forge/84710dde-1620-425b-80d0-4cf5baca359d/_apis/build/builds/438468/logs/10
I'm seeing a lot of Unknown format for lib{varies}.a -- maybe an unclean cross-compile with remnant artifacts from building for another architecture -- ?

BEFH · 2022-01-10T03:08:25Z

BEFH
Jan 10, 2022

I don't know how that would happen; Azure is creating a brand-new clean environment for each build using Docker. It might be packages, even though the requirements are the same: This worked on my test setup: ``` _go_select: 2.2.0-nocgo conda-forge _libgcc_mutex: 0.1-conda_forge conda-forge _openmp_mutex: 4.5-1_gnu conda-forge binutils_impl_linux-64: 2.36.1-h193b22a_2 conda-forge binutils_linux-64: 2.36-hf3e587d_4 conda-forge gcc_impl_linux-64: 11.2.0-h82a94d6_11 conda-forge gcc_linux-64: 11.2.0-h39a9532_4 conda-forge go: 1.16.10-he1b62bb_0 conda-forge kernel-headers_linux-64: 2.6.32-he073ed8_15 conda-forge ld_impl_linux-64: 2.36.1-hea4e1c9_2 conda-forge libgcc-devel_linux-64: 11.2.0-h0952999_11 conda-forge libgcc-ng: 11.2.0-h1d223b6_11 conda-forge libgomp: 11.2.0-h1d223b6_11 conda-forge libsanitizer: 11.2.0-he4da1e4_11 conda-forge libstdcxx-ng: 11.2.0-he4da1e4_11 conda-forge make: 4.3-hd18ef5c_1 conda-forge sysroot_linux-64: 2.12-he073ed8_15 conda-forge ``` And this failed on azure: ``` _go_select: 2.3.0-cgo conda-forge _libgcc_mutex: 0.1-conda_forge conda-forge _openmp_mutex: 4.5-1_gnu conda-forge binutils_impl_linux-64: 2.36.1-h193b22a_2 conda-forge binutils_linux-64: 2.36-hf3e587d_4 conda-forge gcc_impl_linux-64: 9.4.0-h03d3576_11 conda-forge gcc_linux-64: 9.4.0-h391b98a_4 conda-forge go: 1.16.10-h17d2266_0 conda-forge kernel-headers_linux-64: 2.6.32-he073ed8_15 conda-forge ld_impl_linux-64: 2.36.1-hea4e1c9_2 conda-forge libgcc-devel_linux-64: 9.4.0-hd854feb_11 conda-forge libgcc-ng: 11.2.0-h1d223b6_11 conda-forge libgfortran-ng: 11.2.0-h69a702a_11 conda-forge libgfortran5: 11.2.0-h5c6108e_11 conda-forge libgomp: 11.2.0-h1d223b6_11 conda-forge libsanitizer: 9.4.0-h79bfe98_11 conda-forge libstdcxx-ng: 11.2.0-he4da1e4_11 conda-forge make: 4.3-hd18ef5c_1 conda-forge sysroot_linux-64: 2.12-he073ed8_15 conda-forge ``` I just don't understand how it compiles and passes all the tests, but when it's copied and tested a second time, it segfaults.

…

On Sun, Jan 9, 2022 at 9:55 PM John Kerl ***@***.***> wrote: From https://dev.azure.com/conda-forge/84710dde-1620-425b-80d0-4cf5baca359d/_apis/build/builds/438468/logs/10 I'm seeing a lot of Unknown format for lib{varies}.a -- maybe an unclean cross-compile with remnant artifacts from building for another architecture -- ? — Reply to this email directly, view it on GitHub <#851 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAZ2Z2CUQWRXQ2NY6KKOTY3UVJDB3ANCNFSM5LSP4T5Q> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>

7 replies

johnkerl Jan 10, 2022
Maintainer Author

@BEFH worth a try -- ?

johnkerl Jan 10, 2022
Maintainer Author

@BEFH I will need to do a local Conda build like we did before to help debug this

BEFH Jan 10, 2022

Give me a little while longer on this; I'm trying to work it out now. The version that works on my computer is failing on conda-forge, and the version that works on conda-forge is failing on my computer, so I am trying to figure it out.