Arch Linux User Repository

bertptrs, thank you for your actions and position. Considering the time I updated this package last was almost a year ago and there have been at least a couple of updates to pacman in the meantime; quite probably my results will be the same as yours today.

I would suggest the both of you turn over a new leaf. The previous interactions were not good, I recognize that, but please, turn over a new leaf. I want to be done with this.

Toropisco, you should be more open to suggestions from people in the package comments. That is specifically what the package comments are for. People needn't be an authority figure to be right. Also refrain from name-calling going forward.

Xiota, try and be more tactful when suggesting changes. Don't just say "do this." You can stand to write a couple words with an explanation and point people the right way. Demands from volunteers are rarely appreciated.

Consider this a warning to you both, and consider the Arch Linux Code of Conduct going forward. https://terms.archlinux.org/docs/code-of-conduct/

As a show of good faith, you could start by removing the more heated comments from before.

@bertptrs Difficult to assume "good faith" when toropisco has written to users, "you really are an idiot", "IDOT", "entitled petty child", "Go take a hike" (across multiple packages). He has a habit of leaving packages in broken, unusable state for years, blaming users and Arch developers for problems, and refusing to fix them until a PM tells him to. This package is a case in point. Packaging issues and harmful behavior since at least 2022. Blames pacman for bugs that have never been described. Refused to fix until PM comment a few days ago.

Thank you for updating the package. I will deny the pending request. Some more notes:

Using conflicts and provides proved insuficient in my tests before uploading previously.

If you can provide a reproducer for this issue, please report it so we can take care of it: https://gitlab.archlinux.org/pacman/pacman/-/issues

I don't know what you tested so I can't comment on that.

I insist on providing a copy of the OFL-1.0 license because it is not included in the core licenses package; perhaps you can nudge that oversight in the right direction.

The OFL cannot unfortunately cannot be provided by the licenses package as it is ever so slightly different for every package; the copyright notice in the first line mentions the author. As such, every package must provide their own copy, because it's (usually) not the same. You are doing the correct thing here. For completeness, you could update the license field to its SPDX identifier, OFL-1.0, without the custom: prefix. It is in fact a valid SPDX license.

The MIT and BSD licenses have a similar issue, see: https://wiki.archlinux.org/title/PKGBUILD#license

@bertptrs

Also, I would like to ask everyone in this discussion to take a more  constructive and less hostile tone, and assume good faith from the others. 
Everyone here stands to benefit from well-maintained and correct PKGBUILDs. We can have a productive discussion about improvements.
I'm sure everyone would appreciate that outcome.

I agree; yet, I must emphasize I have dislike entitled drive-by comments that do not start nor encourage a healthy and constructive discussion.

On your observations, I've uploaded a modified PKBUILD. BUT, do take note; that I used replaces because pacman IS broken with AUR packages; core repositories are managed just fine. There is no point in having both the variable and the opticals installed as they conflict each other. Fontconfig cannot distinguish them and I have observed that both freetype and harfbuzz seem rather confused with the situation. Using conflicts and provides proved insuficient in my tests before uploading previously. I insist on providing a copy of the OFL-1.0 license because it is not included in the core licenses package; perhaps you can nudge that oversight in the right direction.

Also, FYI, I've worked with fonts for several decades including contributing patches to the originam AMS TeX Type 1 font releases available in CTAN (I've been mad enough to write custom optimized PostSript hinting subroutines by hand). Check the date in these files for another reference of the TeX engine I used to write my thesis and dissertation.

Toropisco, please either fix the package to not use replaces incorrectly, or orphan it so other people might fix it. These package do not replace each other. Replaces is for "this is the new package, these old packages should never be used again." This is not how you use this directive, and also not a bug in pacman. For what it's worth, I am a Package Maintainer, you can check the website.

Also, I would like to ask everyone in this discussion to take a more constructive and less hostile tone, and assume good faith from the others. Everyone here stands to benefit from well-maintained and correct PKGBUILDs. We can have a productive discussion about improvements. I'm sure everyone would appreciate that outcome.

The AUR rules of submission are not predicated on whether the submitter thinks there is something wrong with pacman.

Before I make a decision I need to clarify two things: First, are you a TU (provide unquestinable proof) or a drive-by enforcer? Second, will you take over the package to make sure the enforcing is done? It would make my life easier to just drop the corpse on you, but only if you are a TU (again, provide unquestionable proof here, in public).

Thank you in advance.

The AUR rules of submission are not predicated on whether the submitter thinks there is something wrong with pacman.

This package and otf-literata violate AUR submission guidelines, 1.1 subpoint 5: https://wiki.archlinux.org/title/AUR_submission_guidelines#Rules_of_submission

Please fix both.

If you fix pacman, I'll do.