I'm reluctant to say that paywalls with no workaround should be banned outright, but obviously they're not covered by the "ok" policy.
Sometimes people post these and others respond with links to freely available versions, or articles about the work. In such cases we're happy to update the URLs.
We're not happy about announcement mills either (and those sites are penalized on HN), but that's arguably a separate problem.
My concern is with the interaction of the "original source" rule and the
"ok for paywalls with workarounds" rule preventing most articles on new
research. The "original source" papers are often locked behind paywalls
(or embargoes) when the insipid announcement mill advertisement articles
start appearing. Even supposedly reputable official university press
sites are thoroughly guilty these overtly promotional teaser articles.
The trouble is, lightweight advertisement articles are often the only
things we can (legally) access when the "news" first becomes public.
As much as I hate to admit it, the sad state of suckage for announcement
mills (including university press sites) actually does have some minor
advantages; which would you be more inclined to read and up-vote?
"Astronomers detect furthest galaxy yet with Keck telescope"
or
"Lyman-Alpha Emission From A Luminous Z=8.68 Galaxy: Implications For
Galaxies As Tracers Of Cosmic Reionization"
Non-Astronomers would be lucky if they understand the details presented
in just the abstract of the paper, and I say this as a non-astronomer
who does _NOT_ understand all of said details. Reading original source
papers takes far more effort than reading lightweight announcements, and
this gets to the fundamental question of, "What do we want HN to be?"
The status quo of interested HN users finding and comment-linking to the
original source papers (if available) on the puff-piece stories is a lot
of manual work and some stuff gets missed, but it really does tend to
work out reasonably well. If we forbid paywalls without workarounds and require original sources, then we will miss out on a lot of great new research. Besides infringement, there is no easy answer for this situation.
Sometimes people post these and others respond with links to freely available versions, or articles about the work. In such cases we're happy to update the URLs.
We're not happy about announcement mills either (and those sites are penalized on HN), but that's arguably a separate problem.