profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/ryandesign/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.
Ryan Schmidt ryandesign Austin, TX he/him

ryandesign/createinstalliso 2

createinstalliso creates an iso disk image from a macOS install application for use with virtualization software running on Mac hardware.

ryandesign/cliclick 1

a fork of Carsten Blüm's cliclick with support for Mac OS X 10.5 and later (see the "legacy" branch). cliclick is a macOS CLI tool for emulating mouse and keyboard events

ryandesign/aircrack-ng 0

WiFi security auditing tools suite

ryandesign/angband 0

A free, single-player roguelike dungeon exploration game

ryandesign/ast 0

This is the AT&T Software Technology ast software download site from AT&T Research. The AT&T AST OpenSource Software Collection provides an overview and Practical Reusable UNIX Software provides a historical perspective. You can join the ast and uwin mailgroups, or monitor the mail archives. There is also an AT&T internal talk. The man page index contains most of the ast commands and library functions; command man pages are also available at runtime using the --man or --html option of any ast command.

ryandesign/autotrace 0

bitmap to vector graphics converter

ryandesign/buildbot 0

Python-based continuous integration testing framework; your pull requests are more than welcome!

ryandesign/cartman 0

Command line Trac tools

push eventmacports/macports-ports

Ryan Schmidt

commit sha 52d58e6b06eead7d8d512333afaa8165d8f9f7e7

mongodb: Update to 5.0.1

view details

push time in a day

push eventmacports/macports-ports

Ryan Schmidt

commit sha 3e9627748bfcdd3695ecb1f020fb43f8132cade8

pianod: Update to 343 Closes: https://trac.macports.org/ticket/61658

view details

push time in a day

push eventmacports/macports-ports

Kevin Horton

commit sha d9f563b91480e18c779a62c7594e0824f2be0ab9

convertall: New port, version 0.8.0

view details

push time in 3 days

PR merged macports/macports-ports

ConvertAll: New portfile submission. type: enhancement type: submission

ConvertAll: New portfile submission. First submission by Kevin Horton

I have never created a Portfile before, and am a git newbie, so make no assumptions

Description

Type(s)
  • [x] enhancement
Tested on

macOS 11.4 20F71 x86_64 Xcode 12.5.1 12E507

Verification <!-- (delete not applicable items) -->

Have you

  • [x] followed our Commit Message Guidelines?
  • [x] squashed and minimized your commits?
  • [x] checked that there aren't other open pull requests for the same change?
  • [ ] referenced existing tickets on Trac with full URL? <!-- Please don't open a new Trac ticket if you are submitting a pull request. -->
  • [x] checked your Portfile with port lint?
  • [x] tried existing tests with sudo port test?
  • [x] tried a full install with sudo port -vst install?
  • [x] tested basic functionality of all binary files?
+46 -0

5 comments

1 changed file

khorton

pr closed time in 3 days

push eventkhorton/macports-ports

Ryan Schmidt

commit sha d3bd004f0f17f268e0ffec8486caab914e04f0ca

convertall: Fixes as described in PR

view details

push time in 3 days

pull request commentmacports/macports-ports

p5.34-*: add subports for perl 5.34

Filed https://trac.macports.org/ticket/63265 to see if that can be changed.

This is my proposed modification to perl5: https://github.com/ryandesign/macports-ports/tree/perl5

Pushed a revised version based on 63265 having been resolved. Still untested though. Does it seem reasonable to you?

mojca

comment created time in 3 days

push eventryandesign/macports-ports

Ryan Schmidt

commit sha 8e7efad9e94f0c517e0209a2b9be5bbb96438e70

perl5: Use newer compiler for perl_tsa_mutex_lock Use a newer compiler (Apple clang 6.1 or later or open source clang 3.6 or later) to ensure that perl_tsa_mutex_lock gets compiled in, which is needed if any module would ever get compiled with such a newer compiler. See: https://github.com/Perl/perl5/issues/18991 Perl is only in the clang dependency chain if clang's +analyzer variant is selected, which it no longer is by default on 10.6 or earlier. For clang 3.3 this happened in 9a0f6dd; for clang 3.4 in 640b7e0; for clang 3.7 in 9809d4b; for clang 5.0, 6.0, 7.0, 8.0, 9.0, and 10 in 6b338b8; and for clang 11, 12, and devel in 2648d97. Closes: https://trac.macports.org/ticket/57648

view details

push time in 3 days

pull request commentmacports/macports-ports

ConvertAll: New portfile submission.

Looks like the reason why the -p and -b flags you were sending to install.py were not having any effect is because the python portgroup by default sets build.target to build[python_get_defaults jobs_arg] which install.py wasn't expecting. I guess it stopped processing arguments after that. The solution is to clear build.target. Then the post-patch block can go away.

You refer to the repository in the github.setup line as "convertall" but the repository name is actually "ConvertAll". Although GitHub currently allows case-mismatched access, it's best to specify case correctly. That would change the port name capitalization to "ConvertAll" as well. We could either leave it that way or specify name convertall to keep the port name lowercase.

When we change the repo name, the distname won't be right anymore, since the distname should indeed be lowercase, so that'll need to be fixed.

The distfile extracts to a directory called just "ConvertAll" without a version number. Although MacPorts will handle that automatically now by creating a symlink for you when worksrcdir is wrong, I prefer to set worksrcdir correctly. It seems cleaner to me to have just the extracted directory in the work dir, rather than both an extracted directory and a symlink.

I mentioned in an earlier comment that, since install.py cannot separate build from install, I thought doing everything in the build phase was best. I've changed my mind. Because install.py doesn't really build anything—all it does is install things—it would be best to do it in the destroot phase.

The fact that nothing actually gets built, files just get copied, was a clue to the likelihood that this port does not install any architecture-specific files. That being the case, this needs to be indicated via the line supported_archs noarch.

We're also getting into the habit of always adding the revision line, even when it is 0, so I'll add that too.

Since you mentioned you'll be away, I hope you don't mind if I make the various corrections myself.

khorton

comment created time in 3 days

PR opened doug-101/ConvertAll

Fix typo: retuen -> return

Fix typos.

+2 -2

0 comment

2 changed files

pr created time in 3 days

create barnchryandesign/ConvertAll

branch : typos

created branch time in 3 days

Pull request review commentmacports/macports-ports

ConvertAll: New portfile submission.

+# -*- coding: utf-8; mode: tcl; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- vim:fenc=utf-8:ft=tcl:et:sw=4:ts=4:sts=4++PortSystem          1.0+PortGroup           github 1.0+PortGroup           python 1.0+PortGroup           qt5 1.0++github.setup        doug-101 ConvertAll 0.8.0 v+github.tarball_from releases++python.versions     34 35 36 37 38 39+python.default_version     39++qt5.min_version     5.4++name                convertall+version             0.8.0+license             GPL-2++categories          math science+platforms           darwin+maintainers         {@khorton kilohotel.com:kevin01} openmaintainer+description         Extremely flexible unit converter+long_description    ConvertAll has a large database of units, and allows conversions \+                    that use multiple units, e.g. convert from feet per decade to \+                    nautical miles per fortnight.+homepage            http://convertall.bellz.org++checksums           rmd160  89fcc2aa9bad7ecc1de7bee4edea68fdff39706a \+                    sha256  624c8a792b0bc7ff3776499c2c743b32273569efd0567615e570a7e739e8d521 \+                    size    281055++depends_lib-append  port:py${python.version}-pyqt5++use_configure       no+build.cmd           ${python.bin} install.py+build.args          -p ${prefix} \+                    -b ${destroot}++destroot {}++post-patch {+    reinplace "s|/usr/local|${prefix}|g" ${worksrcpath}/install.py+    reinplace "s|buildRoot = '/'|buildRoot = '${workpath}/destroot'|g" ${worksrcpath}/install.py

If I understand correctly, I shouldn't hard code /opt/local into a patch. The patch would need to replace the existing /usr/local with some placeholder, then I would need a post-patch reinplace to replace that placeholder with ${prefix}.

Yes, that is a pattern you will see used by many ports.

P.S: Use ${destroot} here, not ${workpath}/destroot.

    reinplace "s|/usr/local|${prefix}|g" ${worksrcpath}/install.py
    reinplace "s|buildRoot = '/'|buildRoot = '${destroot}'|g" ${worksrcpath}/install.py
khorton

comment created time in 3 days

PullRequestReviewEvent

Pull request review commentmacports/macports-ports

ConvertAll: New portfile submission.

+# -*- coding: utf-8; mode: tcl; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- vim:fenc=utf-8:ft=tcl:et:sw=4:ts=4:sts=4++PortSystem          1.0+PortGroup           github 1.0+PortGroup           python 1.0+PortGroup           qt5 1.0++github.setup        doug-101 ConvertAll 0.8.0 v+github.tarball_from releases++python.versions     34 35 36 37 38 39+python.default_version     38++qt5.min_version     5.4++name                convertall+version             0.8.0+license             GPL-2++categories          math science+platforms           darwin+maintainers         {@khorton kilohotel.com:kevin01}+description         Extremely flexible unit converter+long_description    ConvertAll has a large database of units, and allows conversions \+                    that use multiple units, e.g. convert from feet per decade to \+                    nautical miles per fortnight.+homepage            http://convertall.bellz.org++checksums           rmd160  89fcc2aa9bad7ecc1de7bee4edea68fdff39706a \+                    sha256  624c8a792b0bc7ff3776499c2c743b32273569efd0567615e570a7e739e8d521 \+                    size    281055++depends_lib-append  port:py${python.version}-pyqt5++use_configure       no+build.cmd           ${python.bin} install.py+build.args          -p ${prefix} \+                    -b ${destroot}++destroot {}

As long as the files to be installed end up in the destroot directory by the time that the destroot phase is over, that's fine. Typically a build would happen in the build phase and the built files would be copied into the destroot directory in the destroot phase. But if this build system does not easily accommodate splitting those tasks it is acceptable for both tasks to occur in a single phase (either build or destroot) and for the other one to be made a no-op. The mongodb Portfile is an example of a port that already does that. My feeling today is that in this case having both build and destroot tasks take place in the build phase is preferable, since users usually expect the build phase to take longer than the destroot phase. (MacPorts also adds a -j flag to enable parallel building in the build phase but not in the destroot phase, though that is not relevant to this port since its build command is not make.) In the mongodb port, the build installs some things to the wrong places within the destroot directory, so I use a very brief destroot phase to clean that up.

khorton

comment created time in 3 days

PullRequestReviewEvent

Pull request review commentmacports/macports-ports

ConvertAll: New portfile submission.

+# -*- coding: utf-8; mode: tcl; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- vim:fenc=utf-8:ft=tcl:et:sw=4:ts=4:sts=4++PortSystem          1.0+PortGroup           github 1.0+PortGroup           python 1.0+PortGroup           qt5 1.0++github.setup        doug-101 convertall 0.8.0 v+github.tarball_from releases++python.default_version 39++qt5.min_version     5.4++version             0.8.0

You can remove this version line; the github.setup line sets the version for you.

khorton

comment created time in 3 days

PullRequestReviewEvent

Pull request review commentmacports/macports-ports

libreoffice: update to 7.2.0.1

 if {${os.platform} eq "darwin" && ${os.major} < 19} {     } } -patchfiles \-    patch-configure.diff \-    patch-ucb_ucp_webdav_webdavprovider_cxx.diff \-    patch-unpack-sources-fix-for-bsd-find.diff- set product_name LibreOffice  pre-configure {     system "cd ${worksrcpath} && env NOCONFIGURE=1 ./autogen.sh"+    reinplace "/# if the systemd-detect-virt command/d" Makefile.in

I assume this reinplace is meant to be equivalent to this upstream commit?

https://github.com/LibreOffice/core/commit/cfab695d0e3bd5889cfa4b6e81300dd11d4eef72

Patchfiles are preferred over reinplaces. Can a patch be used instead? If Makefile.in is a generated file, can you patch the file it is generated from?

Tatsh

comment created time in 3 days

PullRequestReviewEvent

push eventmacports/macports-ports

Calvin Ardi

commit sha 7605e28b6d5cc29b69abc8c2c93af445f3eb5ed5

hugo: update to 0.85.0

view details

push time in 3 days

PR merged macports/macports-ports

hugo: update to 0.85.0 maintainer maintainer: open type: update

Description

hugo: update to 0.85.0

Tested on

macOS 10.15.7 19H1217 x86_64 Xcode 12.0 12A7209

Verification

Have you

  • [x] followed our Commit Message Guidelines?
  • [x] squashed and minimized your commits?
  • [x] checked that there aren't other open pull requests for the same change?
  • [x] checked your Portfile with port lint?
  • [x] tried a full install with sudo port -vst install?
  • [x] tested basic functionality of all binary files?
+4 -4

0 comment

1 changed file

cardi

pr closed time in 3 days

push eventmacports/macports-ports

Ralph Seichter

commit sha a266c3ebf4d87d89c51a81a6a15092df435d9e49

notmuch: Bump to version 0.32.2 Also upgrades from Python 3.8 to 3.9.

view details

push time in 3 days

PR merged macports/macports-ports

notmuch: Bump to version 0.32.2 maintainer maintainer: open

Also upgrades from Python 3.8 to 3.9.

+6 -9

0 comment

1 changed file

rseichter

pr closed time in 3 days

Pull request review commentmacports/macports-ports

Gildas for arm

 PortGroup           active_variants 1.1  name                gildas version             202107a-revision            1+revision            2 set my_version      [string tolower [clock format [clock scan 2000-[string range ${version} 4 5]-10] -format %b]][string range ${version} 2 3][string range ${version} 6 end] set my_machine      [if {[string match *arm* ${os.arch}]} {list arm64} {list x86_64}]

Based on this line, it appears that this port intends to be used only on arm64 or x86_64 systems. If that's so, could we add the line:

supported_archs arm64 x86_64

That way an appropriate error message will be emitted early on systems that are not arm64 nor x86_64 (such as i386, ppc, or ppc64).

bardeau

comment created time in 3 days

PullRequestReviewEvent
PullRequestReviewEvent

push eventmacports/macports-ports

Ryan Schmidt

commit sha 7d12bddfd9d9dd856ff66f22f6f51603e4bbfbad

archive_sites.tcl, mirror_sites.tcl: Remove nou.nc Remove nou.nc mirror. The administrators notified us they have removed the MacPorts files from the mirror because they were not accessed often and they ran out of disk space.

view details

push time in 3 days

pull request commentmacports/macports-ports

p5.34-*: add subports for perl 5.34

clang 11, 12, and devel still have +analyzer as an unconditional default intentionally

Filed https://trac.macports.org/ticket/63265 to see if that can be changed.

This is my proposed modification to perl5: https://github.com/ryandesign/macports-ports/tree/perl5

mojca

comment created time in 4 days

create barnchryandesign/macports-ports

branch : perl5

created branch time in 4 days

pull request commentmacports/macports-ports

p5.34-*: add subports for perl 5.34

maybe we can use clang-3.7 after all since it only requires port:perl5 with the +analyzer variant which is no longer the default on 10.6 and earlier thanks to 9809d4b

And for clang 5.0 thru 10 in 6b338b8b0a4fd729e64d35829b0fa3b6d6711b5e. clang 11, 12, and devel still have +analyzer as an unconditional default intentionally so we still need clang_dependency for those.

I admit it's probably worth checking whether the bootstrapping still works if we remove that again, it just that testing takes a while.

I can try a build in a clean prefix if we can agree on what changes to make.

In the upstream bug report you wrote:

There's probably no reason to worry about the scenario of building perl with New Clang and modules with Old Clang since ensuring that all modules build with New Clang should be straightforward

I was not planning on requiring all perl modules to build with New Clang as well. In my opinion that would introduce unnecessary compiler dependencies. If every perl module (on certain older systems) had to activate a clang compiler port to build, that would drastically slow down buildbot builds.

The perl_tsa_mutex stuff was added in https://github.com/Perl/perl5/commit/3baee7ccba467a3ddc1cb6bfb1a4cabd40cd7424 in 5.23.6 so we wouldn't need to revbump earlier perls.

mojca

comment created time in 4 days