Opened 4 weeks ago

Last modified 3 weeks ago

#70611 new defect

alps @2.3.0_16: error: integer value -1 is outside the valid range of values [0, 3] for the enumeration type 'udt_builtin_mixture_enum' [-Wenum-constexpr-conversion]

Reported by: jcampo-ruiz Owned by:
Priority: Normal Milestone:
Component: ports Version: 2.10.1
Keywords: Cc: galexv@…
Port: alps

Description (last modified by ryandesign (Ryan Carsten Schmidt))

Working with Sonoma 14.6.1 in a MacBook M3 max. Xcode 15.4 The "alps" port failed to be built and finally it is not install.

--->  Fetching archive for alps
--->  Attempting to fetch alps-2.3.0_16+applications+python27.darwin_23.arm64.tbz2 from https://packages.macports.org/alps
--->  Attempting to fetch alps-2.3.0_16+applications+python27.darwin_23.arm64.tbz2 from https://mse.uk.packages.macports.org/alps
--->  Attempting to fetch alps-2.3.0_16+applications+python27.darwin_23.arm64.tbz2 from http://fco.it.packages.macports.org/alps
--->  Building alps
Error: Failed to build alps: command execution failed    
Error: See /opt/local/var/macports/logs/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_science_alps/alps/main.log for details.
Error: Follow https://guide.macports.org/#project.tickets if you believe there is a bug.
Error: Processing of port alps failed

Attachments (1)

main.log.bz2 (173.0 KB) - added by jmroot (Joshua Root) 4 weeks ago.

Download all attachments as: .zip

Change History (11)

comment:1 Changed 4 weeks ago by reneeotten (Renee Otten)

Priority: HighNormal
Type: requestdefect

comment:2 Changed 4 weeks ago by jmroot (Joshua Root)

Summary: Error: Failed to build alps: command execution failedalps @2.3.0_16 build failure
:info:build /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_science_alps/alps/work/alps-2.3.0-src-with-boost/boost/boost/mpl/aux_/integral_wrapper.hpp:73:31: error: integer value -1 is outside the valid range of values [0, 3] for the enumeration type 'udt_builtin_mixture_enum' [-Wenum-constexpr-conversion]

Changed 4 weeks ago by jmroot (Joshua Root)

Attachment: main.log.bz2 added

comment:3 Changed 4 weeks ago by ryandesign (Ryan Carsten Schmidt)

Description: modified (diff)
Summary: alps @2.3.0_16 build failurealps @2.3.0_16: error: integer value -1 is outside the valid range of values [0, 3] for the enumeration type 'udt_builtin_mixture_enum' [-Wenum-constexpr-conversion]

This is a well-known problem with older versions of boost that we have already fixed in the boost ports but the alps port uses a bundled copy of boost. We'll either need to patch the copy of boost used here or switch the alps port to use one of the MacPorts versions of boost.

The alps port's homepage has disappeared. Maybe a new version which might already fix this problem is available if the project has moved its homepage elsewhere.

comment:5 Changed 4 weeks ago by ryandesign (Ryan Carsten Schmidt)

Yes, that is an archived version of the homepage that no longer exists which describes version 2.3.0 which we already have in MacPorts.

It looks like the new homepage is https://alpscore.org on which a version 2.3.1 is available; we should try updating to that to see if it solves this problem.

comment:6 Changed 3 weeks ago by jcampo-ruiz

I add a new piece of information on this subject.

The port "alps 2.3.0_13" works very well in both arm64 and x86_64 architectures with Sonoma (darwin_23). The problem is within the upgraded version 2.3.0_16.

If you enter the command "port outdated" then you obtain

The following installed ports are outdated:
alps                    2.3.0_13 < 2.3.0_16

If you try to upgrade to the version 2.3.0_16 with the command sudo port upgrade outdated then it appears the error message

--->  Building alps
Error: Failed to build alps: command execution failed

However the version alps 2.3.0_13 still works. My question is that I want to instal alps 2.3.0_13, or something that works, in my new M3 chip computer

Thanks for your time.

Last edited 3 weeks ago by ryandesign (Ryan Carsten Schmidt) (previous) (diff)

comment:7 Changed 3 weeks ago by ryandesign (Ryan Carsten Schmidt)

The problem isn't caused by a new revision of the alps port. It's caused by using a newer compiler. The old version of the boost library contained within alps contains a programming mistake that old compilers allowed with just a warning. New compilers like the version of clang in Xcode 15.3 and later consider this mistake to be an error. The problem can be worked around by using an older compiler or by downgrading the warning back to an error. Instead of either of those, we should fix the port by correcting the programming mistake e.g. possibly by upgrading to a newer version of alps which may contain a newer version of boost that no longer has this mistake.

comment:8 in reply to:  3 Changed 3 weeks ago by ryandesign (Ryan Carsten Schmidt)

Replying to ryandesign:

or switch the alps port to use one of the MacPorts versions of boost.

The alps port used to use the bundled boost, then switched to a MacPorts boost, and was then switched back to using the bundled boost to fix #60825; there I explained why I thought at the time that using the bundled boost was preferable. However back then MacPorts only offered the single boost port and updating it to a new version had a tendency to break ports using it. Since then MacPorts has begun to offer versioned boost ports so that a port can depend on whatever version of boost it needs and fixes can be centralized in those boost ports, so maybe using using a MacPorts boost is now preferable again.

Replying to ryandesign:

It looks like the new homepage is https://alpscore.org on which a version 2.3.1 is available; we should try updating to that to see if it solves this problem.

Perhaps I am mistaken. We already have a separate ALPSCore port for version 2.2.0 of the software that is available from alpscore.org. Is this different from the software that is in the alps port? The alps port has existed since 2011; see #29430. The ALPSCore port has existed since 2015; see #47350.

comment:9 Changed 3 weeks ago by ryandesign (Ryan Carsten Schmidt)

Cc: galexv@… added

comment:10 in reply to:  9 Changed 3 weeks ago by jcampo-ruiz

Replying to ryandesign:

Thank you very much for your interest to solve this issue.

Unfortunately the alps and alpsCORE libraries are not doing the same. The alps library is the most powerfull (it includes much more functionalities) and therefore it is used extensively among the condensed matter physics community.

Note: See TracTickets for help on using tickets.