protege: init at 5.5.0 #97365
protege: init at 5.5.0 #97365
Conversation
| pname = "protege"; | ||
| version = "5.5.0"; | ||
|
|
||
| src = fetchzip { |
There was a problem hiding this comment.
I'm not absolutely against using this zip file, but there's a new part of the manual explaining how to build maven projects in nixpkgs from source:
https://nixos.org/manual/nixpkgs/unstable/#maven
Could you please give it a try?
There was a problem hiding this comment.
Thank you for the review!
I didn't have time to try the maven packaging yet, however @JonathanReeve would like to start using the package using the current solution proposed by @mgttlinger .
Could we maybe merge this version if it's ok and refine upon it on the future?
There was a problem hiding this comment.
@doronbehar the "blessed" technique suggested in the manual apparently fails to detect the dependency for org.apache.felix:maven-bundle-plugin. As the translation from pom.xml to project-info.json is not fully clear to me, I also went for a direct JAR package.
|
There's merge conflicts. |
|
I'd love to see this package included. What can be done to fix it up? One thing is, I notice it doesn't build unless I also include But then it seems to give lots of errors while running it, so I'm not sure what's going on here. |
|
@mucaho, could you take another look at this, when you get a chance? I'd love to be able to use this package, but I don't know enough about Java packaging to get this to work. |
mucaho
force-pushed
the
feature/protege_init
branch
from
b56cfa1
to
ffcea2f
Compare
|
|
||
| meta = { | ||
| description = "Ontology editor and framework for building intelligent systems"; | ||
| license = lib.licenses.bsd2; |
There was a problem hiding this comment.
Protege itself is released under a 2-clause BSD, but this is the distribution version, which includes third-party plugins. These come with a variety of licenses that can be extracted from the individual JARs.
From what I could see, there are pieces that come under Apache 2.0, Eclipse Public License 1.0 and LGPL 3.0 (don't know if in the "only" or "or later" form, though). It may be good to include these additional licenses here.
There was a problem hiding this comment.
The official github protege distribution license is BSD-2 https://github.com/protegeproject/protege-distribution .
Putting other licenses in the meta field makes it sound like the user may choose which license terms they accept, which is not the case here.
There was a problem hiding this comment.
According to the relevant manual section, a list is used to indicate that different parts of the package are licensed differently. I agree that this can be confusing, but the solution would be to improve the meta-attribute themselves.
Also, as far as I understand the license field is used by binary caches to determine if a package can be cached and redistributed. If one of the licenses is "unfree", then I suppose the whole package would be marked "uncacheable" as to comply to licensing regulations.
| @@ -25161,6 +25161,8 @@ in | |||
|
|
|||
| protonmail-bridge = callPackage ../applications/networking/protonmail-bridge { }; | |||
|
|
|||
| protege = callPackage ../applications/science/logic/protege { }; | |||
There was a problem hiding this comment.
I am not sure if Protege should be lumped together with things like theorem provers, and it's more of a development/content creation tool than a scientific instrument. The base application doesn't even include a reasoner, for instance.
But this is just a matter of opinions.
There was a problem hiding this comment.
Agreed, putting it here was unintentional.
|
Closing in favor of #127396 |
Motivation for this change
Adds Protege, a free, open-source ontology editor and framework for building intelligent systems.
Things done
sandboxinnix.confon non-NixOS linux)Not applicable here, not a nixos module
nix-shell -p nixpkgs-review --run "nixpkgs-review wip"It's a new application, no packages do or will depend upon it
./result/bin/)Tested also using runtime-sandboxing (
nix-shell -I nixpkgs=. --pure -p protege)nix path-info -Sbefore and after)Couldn't run the command due to error, but it's around 60MB large
error: Please be informed that this pseudo-package is not the only part of Nixpkgs that fails to evaluate. You should not evaluate entire Nixpkgs without some special measures to handle failing packages, like those taken by Hydra.No documentation needed here
Yes, except for the maintainers field. Happy to take a look and update it from time to time, but can't fully commit as maintainer.