Package Details: nautilus-checksums 0.1.0-1

Git Clone URL: https://aur.archlinux.org/nautilus-checksums.git (read-only, click to copy)
Package Base: nautilus-checksums
Description: Add checksums to Nautilus' properties window
Upstream URL: https://gitlab.gnome.org/madmurphy/nautilus-checksums
Licenses: GPL
Conflicts: nautilus-checksums-bin, nautilus-checksums-git
Submitter: grufo
Maintainer: grufo
Last Packager: grufo
Votes: 6
Popularity: 0.95
First Submitted: 2022-12-28 01:51 (UTC)
Last Updated: 2022-12-28 01:51 (UTC)

Pinned Comments

grufo commented on 2022-12-28 17:31 (UTC)

After installing this package don't forget to restart Nautilus, with

nautilus -q

Latest Comments

1 2 Next › Last »

Aftermath commented on 2023-05-27 15:21 (UTC)

@orlfman Thanks. I also am using it for more than almost a week and there is no problem or incompatibility.

orlfman commented on 2023-05-26 18:41 (UTC)

@arash12ir a bit late, but works fine on my end.

Aftermath commented on 2023-04-14 14:18 (UTC)

Does it work on Gnome/Nautilus 44 as well?

yochananmarqos commented on 2022-12-28 18:48 (UTC) (edited on 2022-12-28 18:48 (UTC) by yochananmarqos)

@grufo: It's not annoying, it's redundant and unnecessary. I think the example in the wiki explains it just fine.

All nautilus-checksums-git (and nautilus-checksums-bin if it's ever created) needs is the following:

provides=('nautilus-checksums')
conflicts=('nautilus-checksums')

Again, nautilus-checksums neither needs provides nor conflicts.

grufo commented on 2022-12-28 18:34 (UTC)

@yochananmarqos

Strangely enough you are not the first one that asks me to remove the $conflicts variable from one of the packages I maintain. I do not know why people find that array annoying, but I know that a part of the current text in the ArchWiki is the result of a discussion I had with one of the admins, complaining exactly about the fact that the ArchWiki is not clear enough and people kept asking to remove that array. Obviously the wiki is still not clear enough (in my opinion the example you mentioned should be removed).

Just out of curiosity: What is that makes you pay attention to the $conflicts array?

You can do this experiment:

  1. Take nautilus-checksums, remove only the $conflicts array and then install the package. Then take nautilus-checksums-git, remove only the $conflicts array and then install the package without uninstalling nautilus-checksums first.
  2. Do the same without removing the $conflicts array

You will realize that in the first case no information is provided to you about what is that causes the conflict, and that will make your life harder.

yochananmarqos commented on 2022-12-28 18:17 (UTC)

@grufo: Sorry, just realized I used the wrong link for Package relations, I mean to link PKGBUILD: Package relations

As per the example:

  • nautilus-checksums implicitly provides nautilus-checksums as the pkgname itself
  • nautilus-checksums-git provides nautilus-checksums and conflicts with nautilus-checksums
  • nautilus-checksums-bin provides nautilus-checksums and conflicts with nautilus-checksums, but does not need to explicitly conflict with nautilus-checksums-git since packages providing the same feature are implicitly in conflict.

grufo commented on 2022-12-28 17:55 (UTC) (edited on 2022-12-28 17:56 (UTC) by grufo)

@yochananmarqos

A package already provides and conflicts with itself.

With itself yes, but the -git version is not itself, but another package. The $conflicts array contains explicit package names, it is not a list of “provided” features that are in conflict with this package…

My internet connection is fine. As I said, I can download the with-configure tarball fine in my browser.

The with-configure tarball is much much faster and should be resilient against possible changes in Autotools (theoretically you should be able to compile it even twenty years from now). Let's see if your problem persists, I suspect that it had nothing to do with the PKGBUILD or the upstream link.

yochananmarqos commented on 2022-12-28 17:34 (UTC)

@grufo:

Good point about the license, I doesn't really matter then. I'm in the habit of matching the license with one of the common licenses in /usr/share/licenses/common/.

As far as the conflicts, nautilus-checksums-git should provide and conflict with nautilus-checksums. That's it. No provides or conflicts are needed at all for nautilus-checksums. A package already provides and conflicts with itself.

My internet connection is fine. As I said, I can download the with-configure tarball fine in my browser.

grufo commented on 2022-12-28 17:31 (UTC)

After installing this package don't forget to restart Nautilus, with

nautilus -q

grufo commented on 2022-12-28 17:17 (UTC) (edited on 2022-12-28 17:18 (UTC) by grufo)

@yochananmarqos

The license is GPL3

The conventions for PKGBUILDs are quite liberal concerning GPL versions. Quoting from the ArchWiki:

(L)GPL has many versions and permutations of those versions. For (L)GPL software, the convention is:

  • (L)GPL — (L)GPLv2 or any later version
  • (L)GPL2 — (L)GPL2 only
  • (L)GPL3 — (L)GPL3 or any later version

Why do you think it is important to write GPL3? To be exact (and pedantic) the extension is licensed under GPL-3-or-later.

There is no need for any conflicts, especially for packages that do not exist.

The package is in conflict with nautilus-checksums-git, and so it must be specified. The current version releases also pre-compiled binaries; if you make a PKGBUILD from that (nautilus-checksums-bin) it will be also in conflict with this package, and so it must be specified too.

See Package relations

Package relations does not mention the $conflicts variable. However it does mention it the PKGBUILD article, which correctly states that

When packages provide the same feature via the provides array, there is a difference between explicitly adding the alternative package to the conflicts array and not adding it. If the conflicts array is explicitly declared the two packages providing the same feature will be considered as alternative; if the conflicts array is missing the two packages providing the same feature will be considered as possibly cohabiting. Packagers should always ignore the content of the provides variable in deciding whether to declare a conflicts variable or not.

(emphasis mine)

There's something wrong with the source link. I can click and download it from here, however running updpkgsums downloads a broken tarball. Just use the release tag tarball and run ./bootstrap before ./configure.

Launching updpkgsums works well with me (it might be that your connection broke exactly in that moment?).