On 07/30/2018 08:40 PM, Jean Lucas via aur-requests wrote:

I suggest you link to a release announcement or tarball next time you flag a package out-of-date.

I suggest you become a competent maintainer. There is positively no requirement for out-of-date messages to contain either one, though it certainly does not hurt.

Unflagging it and refusing to update out of spite, when it took me all of one minute to find the release announcement linked right on the hugely visible download page (for their actual website, not the out of date url in the PKGBUILD), is... kind of odd.

And if you're simply the type of maintainer that doesn't really care enough about this package to, when informed of a new release, check up on it yourself, then I'd have to ask why you even bother.

...

Now admittedly this package uses a different url scheme than the one provided by the current download page, and at the same time the hdf5 people are exceedingly weird (for many reasons), but OTOH it looks like the download url in question is not very stable in the first place. Also maybe they stopped using it if they're not updating that directory anymore.

But I don't really see this as a reason to silently do nothing without bothering to check.

On 07/30/2018 09:35 PM, Jean Lucas via aur-requests wrote:

The unflagging was not done out of spite. I overlooked the fact that the previous URL was out-of-date. Will try to be more competent next time.

Okay. So. How about saying that next time?

"My apologies for unflagging it before, I looked in the download directory and on the homepage and didn't see a new release. I didn't realize they'd switched locations. Next time, if an update needs more than a simple pkgver bump, could you please link the new location to aid the maintainer in verifying and applying the update."

Or even:

"You keep flagging this out of date, but I cannot find any such update on the upstream url. *Where* exactly is this update of yours supposed to be?"

But instead, you challenged the reporter by trying to apply some sort of fundamental rule to the AUR, that people who click the "Flag package out-of-date" button must be doing it horribly wrong if they don't fill out a complex report and back up their assertion with links.

...

People are human, they make mistakes and overlook things. That's understandable. I'm a bit more astonished by a response that makes me think you consider a release announcement link to be obligatory and won't act without one, and doesn't carry any implication at all that you actually looked to see if there was an update but didn't see one.

When I read that message, it made me think "oh, the maintainer just looked at the flag message, didn't see a link which could be checked, and said: 'whatever, too much bother' then unflagged it".

From the sound of it, @eleftg received the same impression.