[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[bug#56989] [PATCH v3] gnu: Add dbqn.

From: Maxime Devos
Subject: [bug#56989] [PATCH v3] gnu: Add dbqn.
Date: Mon, 8 Aug 2022 11:19:34 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Thunderbird/91.11.0

On 06-08-2022 04:20, Christopher Rodriguez wrote:

+(define-module (gnu packages bqn)

Copyright and license headers are missing. Also, usually we don't do per-package modules but rather thematic modules, though that's not a hard rule especially if there are technical problems with that.

+  #:use-module ((guix licenses) #:prefix license:)
+  #:use-module (guix gexp)
+  #:use-module (guix packages)
+  #:use-module (guix download)
+  #:use-module (guix git-download)
+  #:use-module (guix build-system copy)
+  #:use-module (guix build-system gnu)
+  #:use-module (guix utils)
+  #:use-module (guix deprecation)
I don't think you are using (guix deprecation) below?
+  #:use-module (gnu packages)
+  #:use-module (gnu packages libffi)
+  #:use-module (gnu packages base)
+  #:use-module (gnu packages pkg-config)
+  #:use-module (gnu packages llvm)
+  #:use-module (gnu packages java)
+  #:use-module (gnu packages compression))
+(define-public dbqn
+  (let ((commit "0bbe096fc07d278b679a8479318f1722d096a03e")
+        (revision "1"))
+    (package
+      (name "dbqn")
+      (version (git-version "0.2.1" revision commit))
+      (source (origin
+                (method git-fetch)
+                (uri (git-reference
+                      (url"";)
+                      (commit commit)))
+                (file-name (git-file-name name version))
+                (sha256
+                 (base32
+                  "1kxzxz2hrd1871281s4rsi569qk314aqfmng9pkqn8gv9nqhmph0"))))

I have looked at the 'v0.2.1' tag, and it points at the 0bbe096... commit, so you are actually packaging version 0.2.1, not some git commit after v0.2.1. As such, no need for (git-version ...), you can just write "0.2.1" there. Then 'revision' becomes unused and can be dropped, and 'commit' is only used in a single place anymore so it can be inlined.

+      (native-inputs (list `(,icedtea-8 "jdk") coreutils zip))
coreutils is an implicit (native-)input, likely no need to to mention it.

Also, the Makefile mentions that the executable to start things has #!/bin/bash -- to properly patch is when cross-compiling, you need 'bash-minimal' or 'bash' in inputs, otherwise it will be patched for the wrong architecture.

Also, that script runs 'java' -- make sure it is patched too such that java will actually be found -- and to patch it, you need to have icedtea:out or openjdk:out in 'inputs'.

+      (outputs '("out"))
That's the default, no need to set it.

+       (list #:imported-modules `(,@%gnu-build-system-modules (guix build
+                                                                    syscalls)
For formatting, (guix build syscalls) should be on a separate line.

+      (synopsis "BQN implementation based on dzaima/APL")
+      (description "BQN implementation based on dzaima/APL.")
The synopsis and description are identical, and this doesn't explain much to people who don't know what 'BQN' is. Can it be rewritten such that people not familiar with BQN can decide whether this ‘BQN’ is something that's useful for them? '(guix)Synopses and Descriptions' has more information.

+                            (lambda* (:#key tests?
+                                            #:allow-other-tags)
Why the : before #:key? Also, no need to break it in separate lines.

+                          (add-after 'install 'reorder-jar-content
+                            (lambda* (#:key outputs #:allow-other-keys)
+                              (apply (assoc-ref ant:%standard-phases
+                                                'reorder-jar-content)
+                                     #:outputs (list outputs))))
'output's is a list of strings, now you are giving reorder-jar-content a list of lists of strings However, looking at (guix build ant-build-system), it looks like it just wants a list of strings.

As such, maybe it should be:

(add-after 'install 'reorder-jar-content
  (assoc-ref ant:%standard-phases 'reorder-jar-content))

? (untested)  Possibly likewise for the other phases.

(the implementation is even under single-license GPL!) [...]
You are writing license:expat in the 'license' field, not license:gplN/license:gplN+. Is it Expat or is it GPL?

+      (synopsis "Official BQN sources in BQN")
If they are just sources, you can do (define bqn-bytecode-sources (origin ...))

+      (home-page"";)
+      (license license:gpl3))))
The 'LICENSE' file says something different. I don't think that's the home page, maybe <> instead?

+                              (chmod "BQN" 493)
Hexadecimal would be clearer.
+       "The expected implementation for the BQN language,
+according to the official documentation of that specification.")
expected -> standard (what's 'expected' depends on the user, they might want a different implementation), unless it's not the standard implementation.

'documentation of that specification' -> 'the specification (pleonasm)

And maybe remove 'official', given the absence of 'official' in the descriptions of, say, guile, gcc, openjdk and java, this sounds marketing and unfair to me.


+  (let* ((tag "0")
+         (revision "1")
+         (commit "fd17b144483549dbd2bcf23e3a37a09219171a99")
+         (hash "1rr4l7ijzcg25n2igi1mzya6qllh5wsrf3m5i429rlgwv1fwvfji")
+         (version (git-version tag revision commit)))
I'm not seeing a '0' tag anywhere in the repository -- I dont see any tags at all tere.

+    (inputs (list bqn-bytecode-sources libffi singeli-bootstrap))))
For cross-compilation, I would have expected sengili-bootstrap in native-inputs, not inputs, assuming that it is used as a compiler. Does cross-compilation (with --target) work?


Attachment: OpenPGP_0x49E3EE22191725EE.asc
Description: OpenPGP public key

Attachment: OpenPGP_signature
Description: OpenPGP digital signature

reply via email to

[Prev in Thread] Current Thread [Next in Thread]