We are pleased to announce that conda 4.14.0 & conda-build 3.22.0 releases are now available, featuring
⇒ rename conda environments
⇒ channel notifications
⇒ better error handling
⇒ and more!
Many thanks to all our contributors, especially the 14 new contributors who helped make this happen
⇒ rename conda environments
⇒ channel notifications
⇒ better error handling
⇒ and more!
Many thanks to all our contributors, especially the 14 new contributors who helped make this happen
So I ran into this trying to update figure out the right workflow for my system
conda/conda#5821
So its clear what I want to do is not the "conda approved" workflow so I am trying to figure out the right method.
I have software repo A that provides a conda environment to run it. Then I have software repo B that uses repo A as a dependency but also adds a number of additional dependencies that are not needed for repo A.
It seems like what I would want is to specify a new environment in repo B that is something like "these dependencies plus whatever is in the repo A environment file" and for the package manager to solve for if that's possible (and it will be) and build me that environment. If its not possible it would throw and error just like it does if I make a single environment file that isn't possible, like asking for specific versions of packages that don't go together.
What is the suggested workflow for this? Would env update second_file.yaml work to add second set of dependencies?
I installed a tool with micromamba on a CentOS 7 machine. The install worked fine, but when I try to run it, I get an error:
ImportError: libffi.so.7: cannot open shared object file: No such file or directory I see libffi.so.6 in the system lib64, and libffi.so.8 in my ~/micromamba/lib (it's a symbolic link to somewhere else in the hierarchy). Is this most likely a problem with the conda package?
1 reply
I am having issues with installing an old package on windows that used to install well and I can no longer install it - I wonder what is the problem - it used to work and no longer works. This is anaconda 2 based and uses python 2 yet it does no longer install even with an older anaconda - here is the command I am using
conda install mist -c jacob-barhak - here is the error I am getting CondaMultiError: CondaHTTPError: HTTP 404 NOT FOUND for url <https://conda.anaconda.org/jacob-barhak/win-64/win-64\inspyred-1.0-py27_0.tar.bz2> - can anyone with a windows machine can try to verify this is what they get...
I looked the anaconda cloud and the file is there. see: https://anaconda.org/jacob-barhak/inspyred I changed it access to public yesterday since I though it will be a source for the error, yet it seems the problem is elsewhere now...
I located someone else who seemed to have a problem like mine and was no answered on stackoverflow: https://stackoverflow.com/questions/59335034/condahttperror-http-404-not-found-for-dask-core-2-7-0-py-0-tar-bz2
does anyone have any clue on why old packages cannot be located anymore?
I also tried conda install inspyred -c jacob-barhak separately - it seems the problem is with this specific package and it is either a conda error or a cloud flare error since there is CF message attached to the http error. Here is how it loos like:
Downloading and Extracting Packages
inspyred-1.0 | 78 KB | | 0%
CondaMultiError: CondaHTTPError: HTTP 404 NOT FOUND for url <https://conda.anaconda.org/jacob-barhak/win-64/win-64\inspyred-1.0-py27_0.tar.bz2>
Elapsed: 00:00.423906
CF-RAY: 74c04c37b9b174a1-LHR
An HTTP error occurred when trying to retrieve this URL.
HTTP errors are often intermittent, and a simple retry will get you on your way.Is there a way to tell if this is a conda issue or a cloud flare issue?
1 reply
Hi All, a draft conda Discourse Forum is now up. See GitHub for the announcement, and for the call for help. My hope is to make this live by the end of September. Please provide feedback in GitHub or here. Thanks.
🚀 We are elated to announce the release of #conda 22.9.0, the first version following CEP 8 (Conda Enhancement Proposal) in our new bi-monthly release cycle.
We are (also!) pleased to announce that the conda Community Forum is now open. Questions, answers, discussions, and news about the conda ecosystem. It is (or will be!) all here.
@thath posted in Conda.org is comming and we want your help! - https://conda.discourse.group/t/conda-org-is-comming-and-we-want-your-help/99/1
@SandeepAllampalli posted in CondaSSLError: OpenSSL appears to be unavailable on this machine. OpenSSL is required to download and install packages - https://conda.discourse.group/t/condasslerror-openssl-appears-to-be-unavailable-on-this-machine-openssl-is-required-to-download-and-install-packages/103/1
Conda users: we want to hear from you! Complete a survey about your experience using conda, and enter a raffle for a chance to win a $150 Amazon gift-card.
The results of the survey will be summarized and published, and then used to help guide future directions for conda.
Interested in providing your feedback? Click here to fill out the survey and tell us your experience using conda.
Complete the survey before 11:59 pm EST on 11/16/2022 for a chance to win a $150 Amazon gift card. The winner of the raffle will be notified via email.
Thank you for helping to improve conda!
@tnabtaf (Dave Clements) posted in Tell us how you use conda! - https://conda.discourse.group/t/tell-us-how-you-use-conda/113/1
@baszalmstra posted in Https://prefix.dev - a fast package search website - https://conda.discourse.group/t/https-prefix-dev-a-fast-package-search-website/116/1
Hi, I have an env.yaml containing a pip: section. These packages are being built as wheels prior creating the env (residing in /root/.cache/pip, e.g. Docker env). However during the creation of the env, pip ignores the cache and tries to build these again. Is there any way to force pip to use the cache?
@tnabtaf (Dave Clements) posted in Conda-build 3.23.1 Released - https://conda.discourse.group/t/conda-build-3-23-1-released/123/1
@tnabtaf (Dave Clements) posted in 🎉 Conda 22.11.1 Release - https://conda.discourse.group/t/conda-22-11-1-release/139/1
Hi guys, not sure if this is the best channel, can we get some attention on this issue (regarding conda-forge package info not updated on anaconda.org) please? Thanks! https://github.com/conda/infra/discussions/649
1 reply
@tyler_wang (Hellcat) posted in how to build and install a self-made conda package - https://conda.discourse.group/t/how-to-build-and-install-a-self-made-conda-package/146/1
hello,
does anyone knows where CMAKE_ARGS is defined in conda-build ?
I do not seem to find anything in the sources.
I'm trying to figure out how to add flags for conda-forge/conda-forge.github.io#1859
does anyone knows where CMAKE_ARGS is defined in conda-build ?
I do not seem to find anything in the sources.
I'm trying to figure out how to add flags for conda-forge/conda-forge.github.io#1859
2 replies
So it's not conda-build defining it, but a package designed to export those variables upon environment activation, which conda-build triggers for the
build and host environment during the building phases.
@tnabtaf (Dave Clements) posted in Conda is now fiscally sponsored by NumFOCUS - https://conda.discourse.group/t/conda-is-now-fiscally-sponsored-by-numfocus/150/1
Hi all, I'm wondering if it's possible for root to use an unprivileged user's conda install without polluting that install with root-owned files. When another unprivileged user uses someone else's conda install, it seems to be clever enough to use
~/.conda for the pkgs and cache, but I don't see a way to force this for the case of root.
2 replies
Ah, I see, you just want to execute the existing
conda in some user's account, without adding root-owned stuff there... You can define temporary pkgs caches with CONDA_PKGS_DIRS. It might redownload things again, but it shouldn't touch the original pkgs location.
The "clever" approach to fallback to
~/.conda is out of necessity, I'd say. It doesn't change locations because it knows it's another user's property. It does it because it can't write there. By redefining CONDA_PKGS_DIRS (or adding a /root/.condarc with equivalent contents pkgs_dirs: [your_path]), you are mimicking this process, I think.
Is that somehow a compromise Nate Coraor ?
Yep, in fact that's what I'm doing: galaxyproject/ansible-miniconda#4 but I wasn't sure if there was another way to induce the "other unprivleged user" scenario directly.
2 replies
@Angel_Picos posted in Install miniconda3 for windows - https://conda.discourse.group/t/install-miniconda3-for-windows/169/1
In the
continuumio/anaconda3 Docker container I try to install the pyopenms package. But the package cannot be found in the bioconda channel. It is clearly there.(base) root@61346253da63:/# conda install -c bioconda pyopenms
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
PackagesNotFoundError: The following packages are not available from current channels:
- pyopenms
Current channels:
- https://conda.anaconda.org/bioconda/linux-aarch64
- https://conda.anaconda.org/bioconda/noarch
- https://repo.anaconda.com/pkgs/main/linux-aarch64
- https://repo.anaconda.com/pkgs/main/noarch
- https://repo.anaconda.com/pkgs/r/linux-aarch64
- https://repo.anaconda.com/pkgs/r/noarch
To search for alternate channels that may provide the conda package you're
looking for, navigate to
https://anaconda.org
and use the search bar at the top of the page.
Hi @lars20070 ! You are running Docker for
aarch64, but the package is only built for linux-64 (amd64 aka Intel/AMD). Is this Docker on Apple Silicon? You can use --platform linux/amd64 in your docker run command to use arch emulation for linux-64.
IIUC
bioconda uses conda-forge as base channel (not defaults)
@wolfv @jaimergp:matrix.org Thanks for the quick reply! Adding the
--platform linux/amd64 option and the conda-forge channel prevented the PackagesNotFoundError above. But # Solving environment: ...working... is not finished after 90 min. I am building the docker image on three machines: Apple Silicon, Windows WSL2 and a GitLab runner. On neither of them does solving the environment finish ...