Elm version 18, temporary upgrade? - elm

I'm currently using Elm version 17.1 and want to try an application that uses version 18. I'm on Ubuntu.
Is there a way to temporarily upgrade to version 18 and then revert afterwards?
Otherwise what is the best approach?

You can use the npm package elm-version-manager to install multiple versions of Elm alongside each other (with the command evm install 0.18.0 or evm install 0.17.1) and switch which one is 'active' (with evm use 0.18.0 or evm use 0.17.1) systemwide.

you can use npm. As without global flag, npm packages can be installed for local project folder.

Related

How to install specific version of Cap'n Proto (v0.8.0)?

sudo apt install capnproto
This installs v0.7.0 in my ubuntu 20.04. However, i require v0.8.0 to be installed.
Also, i need v0.6.1 to be installed additionally for backward compatibility.
Any solution for these two cases ?
You will probably need to install from source rather than use a distro package. Unfortunately, it sounds like the Ubuntu distro package hasn't been updated in a while.
Instructions for building and installing from source can be found here: https://capnproto.org/install.html
Old versions are available by changing the version number in the download URL to whichever version you need.
Note that all versions of Cap'n Proto are backwards-compatible, so there should be no need to install older versions, unless you need to run a specific complied binary that was linked against a specific old version.

What is the safest way to get a newer version of a repo from yum on Amazon Linux?

I need a newer version of binutils on Amazon Linux to compile a piece of needed software. This is due to a bug in version 2.29 which is the latest available in their repo. What would be the best way to add another repo, and will this mess up future uses of the yum command as I add and remove non-approved repos? Can I just add in a repo from a similar flavor like CentOS? Does Amazon have a less "Safe" version that has up to date software in it?
For me the safest way to get newer version on your VM is to compile it and install it in different place like /opt/binutils or /usr/local/binutils and use absolute paths for utils.

Upgrading "icu" pacman package

Today I tried to upgrade the "icu" pacman package, because I need it for NodeJS to work.
After I installed the newest version my system completely broke, I can't even start it up anymore.
Does anybody know how to fix the problem with Node or how to correctly upgrade the "icu" package?
You shouldn't upgrade individual packages. Instead, you should upgrade the entire system. The reason for this is that the dependencies that exist between packages will break sooner or later if you perform partial upgrades.
If you can't boot your system any more, try booting on the installation media, mount the installed disk, arch-chroot into it and run pacman -Syu.
The proper way to upgrade Arch linux is described here:
https://wiki.archlinux.org/title/System_maintenance#Upgrading_the_system

Why are the bob databases (e.g. bob.db.arface) not available for Python 3.8

I have lately installed the latest version of Bob (particularly, the package bob.bio.face) (9.0) using conda and following the installation instructions. This automatically installed python version 3.8. When I try to run
bob bio pipelines vanilla-biometrics arface gabor_graph
I get the error: ModuleNotFoundError: No module named 'bob.db.arface'
So, I try to install bob.db.arface: conda install bob.db.arface. For some reason, this wants to install a whole new python version (3.7), and reinstall all bob packages.
Looking at the pipelines of bob.db.arface (https://gitlab.idiap.ch/bob/bob.db.arface/pipelines/45435), I can see that only python 3.6 and 3.7 is supported. It seems to be that this is true for many of the bob.db packages. Will there be versions for python 3.8 available, or do I need to downgrade to python 3.7?
In the new release of bob we deprecated all database packages and porting them one by one (once we have the time for it) to the new database interface (https://gitlab.idiap.ch/bob/bob.bio.face/-/issues/54).
If you want to use the old bob.dbs and the legacy database support (https://www.idiap.ch/software/bob/docs/bob/bob.bio.base/stable/legacy.html), please, rely on bob for py37.
Cheers
You may pip install the bob.db.arface package or use buildout but know that most bob.db. packages are deprecated in favor of the csv format. https://groups.google.com/g/bob-devel/c/6-4SxluHHrA

Redhawk IDE will not display Components

When I launch the REDHAWK IDE via Eclipse, I cannot see the installed components (SigGen, fastFilter, etc). I can see the components just fine if I use the command line to create a project. I'm convinced it has to be a path or variable issues, but I just don't know what to reconfigure.
I'm using REDHAWK Version 2.1.0 and on CentOS7
OSSIEHOME is set to the /usr/local/redhawk/core
SDRROOT is set to /var/redhawk/sdr
Any suggestions?
I found a solution by reinstalling and including every Redhawk package that I could find using "yum list redhawk*"
I have the same problem and it turns out I was using a Java version higher than 8 which is not supported by RedHawk. Make sure you use Java 8 to launch the IDE otherwise it does not work. Submitted a DR but because JDK8 still supported they do not feel the need to fix it.
I followed the official installation instructions for RedHawk 2.3.0 on Centos 7.9 and had this same issue, I fixed it by opening a terminal in the Redhawk installation directory and running
sudo yum install redhawk-basic-components-2.3.0-5.el7.x86_64.rpm
After which the components appeared in the IDE without a restart