Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.project > #13094 > unrolled thread

SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?]

Back to article view | Back to linux.debian.project

Contents

  SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] "Andrew M.A. Cater" <amacater@einval.com> - 2022-11-23 19:20 +0100
    Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] The Wanderer <wanderer@fastmail.fm> - 2022-11-23 20:20 +0100
      Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] "Andrew M.A. Cater" <amacater@einval.com> - 2022-11-23 20:50 +0100
      ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?]) "G. Branden Robinson" <g.branden.robinson@gmail.com> - 2022-11-23 22:20 +0100
        Re: ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?]) "Andrew M.A. Cater" <amacater@einval.com> - 2022-11-23 23:20 +0100
          Re: ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?]) Sam Hartman <hartmans@debian.org> - 2022-11-23 23:40 +0100
        Re: ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?]) Sam Hartman <hartmans@debian.org> - 2022-11-23 23:40 +0100
    Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] Sam Hartman <hartmans@debian.org> - 2022-11-23 20:50 +0100
      Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] Ansgar <ansgar@43-1.org> - 2022-11-24 00:00 +0100
        Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] The Wanderer <wanderer@fastmail.fm> - 2022-11-24 00:30 +0100
        Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] Sam Hartman <hartmans@debian.org> - 2022-11-24 01:20 +0100
    Re: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?] Diederik de Haas <didi.debian@cknow.org> - 2022-11-23 22:00 +0100

#13094 — SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?]

All,

Thank you for your considered opinions thus far. We have various developers
who have written defending free speech: we've had others who have expressed
various reservations with one aspect or other of the status quo.

There's been a grudging consensus that this is *hard*. It's very hard indeed
to draw good conclusions as to what to do when everyone agrees that something
could be done and disagrees with what that should be.

Notably, Sam Hartman and Branden Robinson have pointed up flaws with the
existing categorisations and with a blanket removal based on preference.
It's also noticeable that this largely comes down to consideration of 
fortunes in English - almost nothing has been said about other fortunes files
or other languages, though Sam talked about cultural perceptions.

A serious suggestion: it is not necessary for Debian to package fortune files
at all.

 The single collection we have is largely a random collection from BSD of
1995 vintage, itself representative of one Unix site in 1995 or earlier.
The upstream Github repository is potentially only one of many disparate sites
on the 'Net and the English language collection doesn't reflect the languages
of Debian users worldwide.

PyPi has a fortunes-mod equivalent to read fortunes files: it doesn't
necessarily include strfile but it will handle pre-existing fortune files.
It should be open to anybody to make their own fortunes files - just as
anyone can make a mix of their own music on their favourite music player.

If Debian doesn't distribute fortune files but instead provides the means
for users to make/download their own choice, nothing is lost. Debian is
not responsible for maintaining any file content, whether questionable
or unobjectionable depending on viewpoint, and we lose the burden of 
translation, maintenance and policing of content. 

This also means that anyone who wishes can add the *missing* content requested
in the bugs over years into their own files at their own risk.

Your thoughts, again, please.

Andy Cater. 

[toc] | [next] | [standalone]

#13095

[Multipart message — attachments visible in raw view] — view raw

On 2022-11-23 at 13:06, Andrew M.A. Cater wrote:

> All,
> 
> Thank you for your considered opinions thus far. We have various
> developers who have written defending free speech: we've had others
> who have expressed various reservations with one aspect or other of
> the status quo.
> 
> There's been a grudging consensus that this is *hard*. It's very hard
> indeed to draw good conclusions as to what to do when everyone agrees
> that something could be done and disagrees with what that should be.
> 
> Notably, Sam Hartman and Branden Robinson have pointed up flaws with
> the existing categorisations and with a blanket removal based on
> preference. It's also noticeable that this largely comes down to
> consideration of fortunes in English - almost nothing has been said
> about other fortunes files or other languages, though Sam talked
> about cultural perceptions.
> 
> A serious suggestion: it is not necessary for Debian to package
> fortune files at all.

I find this suggestion demotivating and discouraging.

> The single collection we have is largely a random collection from BSD
> of 1995 vintage, itself representative of one Unix site in 1995 or
> earlier.

I believe that this statement is inaccurate. There are parts of the
collection which are Debian-specific (the earliest of which, per the
changelog, were added in 1999), and others which have been added far
more recently than 1995 (there have been what seem like substantive
additions at least as recently as 2006).

The way the Debian packaging splits the collection into various files,
which I understand is not necessarily done upstream, can also be valuable.

> The upstream Github repository is potentially only one of many
> disparate sites on the 'Net and the English language collection
> doesn't reflect the languages of Debian users worldwide.

Can you point to the repository you're referencing?

I wanted to check that repository to see whether it had the
Debian-specific parts of the collection, but since I can't find it, I
can't verify that before sending this.

The only repository URLs I find in the package metadata are the
Vcs-Browser and Vcs-Git URLs from 'apt-cache showsrc fortunes-off'
(which shows the information for the fortune-mod source package), and
those are under anonscm.debian.org.

/usr/share/doc/fortune-mod/README.gz lists several URLs, at least one of
which appears to be a repository, but it isn't on GitHub and leads to a
404.

The files under /usr/share/doc/fortunes*/ don't seem to list any URLs at
all - except for one in changelog.Debian.gz, which dates from 1998 and
is about the addition of the 'perl' fortunes file.

The only upstream I can find referenced is the references in
changelog.debian.gz to "Pascal Hakim", but no apparent place to find
whatever upstream that person might host seems to be mentioned.

> PyPi has a fortunes-mod equivalent to read fortunes files: it
> doesn't necessarily include strfile but it will handle pre-existing
> fortune files. It should be open to anybody to make their own
> fortunes files - just as anyone can make a mix of their own music on
> their favourite music player.
> 
> If Debian doesn't distribute fortune files but instead provides the
> means for users to make/download their own choice, nothing is lost.
> Debian is not responsible for maintaining any file content, whether
> questionable or unobjectionable depending on viewpoint, and we lose
> the burden of translation, maintenance and policing of content.

I sharply disagree that nothing is lost, but I don't seem to have the
emotional energy to try to explain why without becoming argumentative
and probably just making things worse overall. (I have held back a draft
which makes the attempt, but which I suspect distinctly fails.)

> This also means that anyone who wishes can add the *missing* content
> requested in the bugs over years into their own files at their own
> risk.
> 
> Your thoughts, again, please.

I am reasonably certain that this would just lead to far fewer people
bothering to make use of the fortunes database(s) at all, thereby
creating a self-fulfilling prophecy about how irrelevant this is in the
modern world.

From my perspective, this whole discussion looks like someone whom I've
respected coming in and proposing to take away one of the small things
I somewhat like having around, and that taking-away happening almost
immediately despite the existence of pushback over it, and then that
person reacting to the pushback by proposing to take away a *bigger*
thing that I even *more* like having around. I imagine it's not hard to
see how that could be upsetting or demotivating.

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man.         -- George Bernard Shaw

[toc] | [prev] | [next] | [standalone]

#13096

On Wed, Nov 23, 2022 at 02:14:38PM -0500, The Wanderer wrote:
> On 2022-11-23 at 13:06, Andrew M.A. Cater wrote:
> 
> > A serious suggestion: it is not necessary for Debian to package
> > fortune files at all.
> 
> I find this suggestion demotivating and discouraging.
> 
> > The single collection we have is largely a random collection from BSD
> > of 1995 vintage, itself representative of one Unix site in 1995 or
> > earlier.
> 

https://github.com/Distrotech/fortune-mod/tree/master/datfiles has the 
majority of what we have - and similarly for other Linux distros, I
believe.

> I believe that this statement is inaccurate. There are parts of the
> collection which are Debian-specific (the earliest of which, per the
> changelog, were added in 1999), and others which have been added far
> more recently than 1995 (there have been what seem like substantive
> additions at least as recently as 2006).
> 
> The way the Debian packaging splits the collection into various files,
> which I understand is not necessarily done upstream, can also be valuable.
> 

See above.

> > The upstream Github repository is potentially only one of many
> > disparate sites on the 'Net and the English language collection
> > doesn't reflect the languages of Debian users worldwide.
> 
> Can you point to the repository you're referencing?
> 
> I wanted to check that repository to see whether it had the
> Debian-specific parts of the collection, but since I can't find it, I
> can't verify that before sending this.
> 
> The only repository URLs I find in the package metadata are the
> Vcs-Browser and Vcs-Git URLs from 'apt-cache showsrc fortunes-off'
> (which shows the information for the fortune-mod source package), and
> those are under anonscm.debian.org.
> 
> /usr/share/doc/fortune-mod/README.gz lists several URLs, at least one of
> which appears to be a repository, but it isn't on GitHub and leads to a
> 404.
> 
> The files under /usr/share/doc/fortunes*/ don't seem to list any URLs at
> all - except for one in changelog.Debian.gz, which dates from 1998 and
> is about the addition of the 'perl' fortunes file.
> 
> The only upstream I can find referenced is the references in
> changelog.debian.gz to "Pascal Hakim", but no apparent place to find
> whatever upstream that person might host seems to be mentioned.
> 
> > PyPi has a fortunes-mod equivalent to read fortunes files: it
> > doesn't necessarily include strfile but it will handle pre-existing
> > fortune files. It should be open to anybody to make their own
> > fortunes files - just as anyone can make a mix of their own music on
> > their favourite music player.
> > 
> > If Debian doesn't distribute fortune files but instead provides the
> > means for users to make/download their own choice, nothing is lost.
> > Debian is not responsible for maintaining any file content, whether
> > questionable or unobjectionable depending on viewpoint, and we lose
> > the burden of translation, maintenance and policing of content.
> 
> I sharply disagree that nothing is lost, but I don't seem to have the
> emotional energy to try to explain why without becoming argumentative
> and probably just making things worse overall. (I have held back a draft
> which makes the attempt, but which I suspect distinctly fails.)
> 
> > This also means that anyone who wishes can add the *missing* content
> > requested in the bugs over years into their own files at their own
> > risk.
> > 
> > Your thoughts, again, please.
> 
> I am reasonably certain that this would just lead to far fewer people
> bothering to make use of the fortunes database(s) at all, thereby
> creating a self-fulfilling prophecy about how irrelevant this is in the
> modern world.
> 
> From my perspective, this whole discussion looks like someone whom I've
> respected coming in and proposing to take away one of the small things
> I somewhat like having around, and that taking-away happening almost
> immediately despite the existence of pushback over it, and then that
> person reacting to the pushback by proposing to take away a *bigger*
> thing that I even *more* like having around. I imagine it's not hard to
> see how that could be upsetting or demotivating.
> 

The way it was packaged 25 years ago doesn't necessarily mean that we
have to continue to do the same indefinitely - we now have a much wider
Internet to select from rather than storing program and files in the same
space.

I could see that being upsetting or demotivating - I'm not proposing
removing anything from the 'Net or files that already exist: I'm 
suggesting that finding consensus on actual maintenance of fortune 
files is harder than it originally appeared to me. 

I'm also not suggesting that we police anything - on the contrary, that we mightleave it to users to pick their own favourite quotes for their own fortune
files.

Again, thanks for your considered and considerate approach in your response 
which is appreciated.

With every good wish, as ever,

Andy Cater

> -- 
>    The Wanderer
> 
> The reasonable man adapts himself to the world; the unreasonable one
> persists in trying to adapt the world to himself. Therefore all
> progress depends on the unreasonable man.         -- George Bernard Shaw
> 

[toc] | [prev] | [next] | [standalone]

#13099 — ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?])

[Multipart message — attachments visible in raw view] — view raw

[I'm using the pseudonymous respondent's message to reply to Mr. Cater
as well.  Mind the angle brackets for quotation context.]

At 2022-11-23T14:14:38-0500, The Wanderer wrote:
> On 2022-11-23 at 13:06, Andrew M.A. Cater wrote:
> > Thank you for your considered opinions thus far. We have various
> > developers who have written defending free speech: we've had others
> > who have expressed various reservations with one aspect or other of
> > the status quo.
> > 
> > There's been a grudging consensus that this is *hard*.

I gather that you don't join in that consensus, because your
prescriptions are quick and easy.  Mr. Dowland's assessment of everyone
who wants his action reversed as being desirous of association with
racism, sexism, and pro-Nazi sentiment[1] is facile, hasty, and
fallacious.

Neither you nor he, therefore, is well placed to present a
(presumptively neutral) summary of the discussion.  (Neither am I.)

> > Notably, Sam Hartman and Branden Robinson have pointed up flaws with
> > the existing categorisations and with a blanket removal based on
> > preference. It's also noticeable that this largely comes down to
> > consideration of fortunes in English - almost nothing has been said
> > about other fortunes files or other languages, though Sam talked
> > about cultural perceptions.
> > 
> > A serious suggestion: it is not necessary for Debian to package
> > fortune files at all.

I'm going to have to add "a serious suggestion" to "honestly" and "trust
me" as linguistic tags that flag a declaration as deceptive.

Have you worked on embedded systems, ever?  It's not _necessary_ for
Debian to package much of anything.  We could arguably serve just as
well as "universal OS" by providing only a nucleus, say, a high-quality
microkernel.[2]  Minimalism has never been an objective of the package
archive.  This fact has been so transparently obvious for so long that
it is difficult for me to maintain the presumption that you are arguing
in good faith.

> I find this suggestion demotivating and discouraging.

Sorry to hear that.  My own reaction is better termed "pissed off".

> I believe that this statement is inaccurate. There are parts of the
> collection which are Debian-specific (the earliest of which, per the
> changelog, were added in 1999), and others which have been added far
> more recently than 1995 (there have been what seem like substantive
> additions at least as recently as 2006).

Yes, some of them were collected by Joseph Carter ("knghtbrd"), a former
Debian developer, ca. 2000 and for some time afterward.

> The way the Debian packaging splits the collection into various files,
> which I understand is not necessarily done upstream, can also be
> valuable.

I agree with this.  If I were maintainer I think I'd thus segregate the
sort of mathy stuff I'd like to see, in acknowledgement that some people
just aren't edified or provoked to thought by such things.  Similarly,
some people just aren't up to being randomly confronted by exemplars of
human folly, which characterizes the 1.5% or so of the fortunes-off
package at issue, even if they _type_ "fortune" at a shell prompt.

> > The upstream Github repository is potentially only one of many
> > disparate sites on the 'Net and the English language collection
> > doesn't reflect the languages of Debian users worldwide.

Is anyone being prevented from submitting fortune collections localized
to other languages?  Several are already present.[3]  Would you be able
to recognize a Hitler quote rendered in Cyrillic?  Might such a quote
not carry context for a person with family who participated in the Great
Patriotic War that you don't share?  Perhaps some derogatory quip about
the inferiority of the Slavs, when Russians culturally well remember
hoisting the hammer and sickle over the Reichstag on 2 May 1945 (as my
countryfolk do a similar event at Iwo Jima).

Your response, to exclude it all and _foreclose as a matter of policy_
the curation and maintenance of such resources amounts to declaring the
Debian Project impotent to address the challenges here.

It's my turn to cite §2.1.1 of the Debian Constitution as Steve Langasek
did to me.  You are not _required_ to deal with fortune cookies.  You
don't have to install them or use them.  There are many packages that
any given user will never employ, or even be aware of.

The Project does not require your paternalistic supervision.

Remolding the archive to more closely resemble your personal corpus of
installed packages is a curious way to serve the principles of plurality
and diversity.

> > If Debian doesn't distribute fortune files but instead provides the
> > means for users to make/download their own choice, nothing is lost.
> > Debian is not responsible for maintaining any file content, whether
> > questionable or unobjectionable depending on viewpoint, and we lose
> > the burden of translation, maintenance and policing of content.

If you don't want these responsibilities, don't adopt them.  Same goes
for Mr. Dowland.

> I am reasonably certain that this would just lead to far fewer people
> bothering to make use of the fortunes database(s) at all, thereby
> creating a self-fulfilling prophecy about how irrelevant this is in
> the modern world.

The people who try to ban books in the United States employ a similar
multi-level strategy.  If censorship of the local public library's
collection is unsuccessful, they propose withdrawal of funds for the
library, removing a much greater volume of common resources from public
facility.[4]  No doubt to the satisfaction of the oligopolistic
publishing industry, which views public libraries as frustrating its
extraction of monopoly rents.[5]

> From my perspective, this whole discussion looks like someone whom
> I've respected coming in and proposing to take away one of the small
> things I somewhat like having around, and that taking-away happening
> almost immediately despite the existence of pushback over it, and then
> that person reacting to the pushback by proposing to take away a
> *bigger* thing that I even *more* like having around. I imagine it's
> not hard to see how that could be upsetting or demotivating.

I have to disagree with Sam a bit here, while "be bold" is _often_ good
advice, especially in volunteer communities, it is not always.  Apart
from the present contretemps, I can think of another (again from the
United States).[6]

Apparently the package's maintainer has not done an upload since
2013.[7]  Even groff has managed to release twice since then![8]

I therefore solicit a volunteer who is fully up-to-date with packaging
processes to privately advise me so that I can avoid making tedious
mistakes.  E.g., I don't know if filing bugs against "wnpp" is still
necessary or recommended for adopting a package with nine years of
inattention.  Please email me privately if you feel yourself qualified
and are supportive of my intended action.  You don't need to be
enthusiastic about the package's _content_; I am not myself--there is a
lot of crap in there IMO, even apart from the stuff at issue here.

In the short run, my plan would be to ensure that the package is
policy-compliant (and therefore can't be sniped again before release on
a technicality by some "deeply principled" person), evaluate the handful
of categories that are suggestive of group prejudice, and decide on some
kind of updated disclaimer (if necessary) so that the package can warn
about itself.  In the years since fortunes-mod was first packaged for
Debian, the abbreviation "NSFW" has become commonly known for tagging of
materials like this.  That may be more communicative than an "-off"
suffix in the 2020s.  But if a rename will frustrate the package's
progress through the incoming queue, I won't do it before the bookworm
release.  The NSFW annotation can easily go in the package description.

In the long run, I expect to curate as suits me personally; I'll
document my actions (of course) and if someone wants to package the
items I throw away, that's up to them.  One can expect the addition of
items reflecting my personal interests, like math, retrocomputing, music
theory, and Catalan revolutionaries.

I don't have the desire or--especially--the time to undertake major
changes in the package before the bookworm release.  I have something
else I'm working that brings me joy and I'd like to get that solid for
GNU release and thence into Debian.  Approximately 400 upstream bugs
fixed--it does an old XFree86 package maintainer's heart good.[9][10]

Anyone who helps has my thanks in advance.

Regards,
Branden

[1] https://lists.debian.org/debian-project/2022/11/msg00062.html

[2] And do I ever have one in mind I'd like to tell people about!  It
    would please me much more than fighting over elementary principles
    of open society.

[3] https://packages.debian.org/search?keywords=fortunes

[4] https://www.washingtonpost.com/nation/2022/08/24/michigan-library-defunded-gender-queer/
[5] https://www.publishersweekly.com/pw/by-topic/industry-news/publisher-news/article/89038-over-the-past-25-years-the-big-publishers-got-bigger-and-fewer.html

[6] As usual, it was our conservo-glibertarian friends who were behind
    this (q.v. "Free State Project").

    "...At issue was the local school budget, which funds a school for
    children through fourth grade in town and covers tuition for older
    students to attend private and public schools in neighboring towns.
    At a poorly attended annual meeting in March, voters approved a
    measure to cut the school budget by more than half, from $1.7
    million to $800,000.

    The cut would have transformed the education system in Croydon,
    replacing the public school system with one run by cheaper, private
    companies that offer individualized programs, largely online.
    Croydon would have become the first community in New Hampshire where
    two companies, Prenda and Kaipod, would have become the default
    education providers.

    Both companies have been championed by New Hampshire Education
    Commissioner Frank Edelblut, who helped them secure contracts with
    the state of New Hampshire. ..."

    https://www.nhpr.org/nh-news/2022-05-08/croydon-voters-overturn-school-budget-cut

[7] https://tracker.debian.org/pkg/fortune-mod
[8] https://ftp.gnu.org/gnu/groff/
[9] https://git.savannah.gnu.org/cgit/groff.git/tree/ANNOUNCE
[10] https://git.savannah.gnu.org/cgit/groff.git/tree/NEWS

[toc] | [prev] | [next] | [standalone]

#13101 — Re: ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?])

On Wed, Nov 23, 2022 at 03:05:29PM -0600, G. Branden Robinson wrote:
> [I'm using the pseudonymous respondent's message to reply to Mr. Cater
> as well.  Mind the angle brackets for quotation context.]
> 
> At 2022-11-23T14:14:38-0500, The Wanderer wrote:
> > On 2022-11-23 at 13:06, Andrew M.A. Cater wrote:
> > > Thank you for your considered opinions thus far. We have various
> > > developers who have written defending free speech: we've had others
> > > who have expressed various reservations with one aspect or other of
> > > the status quo.
> > > 
> > > There's been a grudging consensus that this is *hard*.
> 

No - there is a consensus that splitting things based on cultural preference
is hard - you and Sam both agree on that point - various other people in
the discussion have had other viewpoints. Notably, I'm not actually 
suggesting that it's straightforward for all the reasons put forward by
everyone in the discussion.

> I gather that you don't join in that consensus, because your
> prescriptions are quick and easy.  Mr. Dowland's assessment of everyone
> who wants his action reversed as being desirous of association with
> racism, sexism, and pro-Nazi sentiment[1] is facile, hasty, and
> fallacious.
> 

The reason I brought this to debian-project is absolutely that CT had a query
from someone living in Germany as to the desirability of Nazi quotations in
fortunes-off and the suggestion from that person that this *might* be illegal
to host in Germany, Austria [Czech Republic and France at least also have 
similar laws I believe]. As I wrote, I wasn't sure that Nazi quotes were still
there - they are - and you were helpful in identifying where several of them
are.

Notably, I haven't assessed anybody's motives in this.

> Neither you nor he, therefore, is well placed to present a
> (presumptively neutral) summary of the discussion.  (Neither am I.)
> 

I don't have a categorical view one way or another on this hence bringing it
to this list. It did seem to me that some of the quotations wouldn't fit well
with the Code of Conduct. If you take Sam's view, that's OK, because this is
a game and we shouldn't apply the spirit of the Code of Conduct to software

Some of the people replying have one view, some another.

> > > Notably, Sam Hartman and Branden Robinson have pointed up flaws with
> > > the existing categorisations and with a blanket removal based on
> > > preference. It's also noticeable that this largely comes down to
> > > consideration of fortunes in English - almost nothing has been said
> > > about other fortunes files or other languages, though Sam talked
> > > about cultural perceptions.
> > > 
> > > A serious suggestion: it is not necessary for Debian to package
> > > fortune files at all.
> 
> I'm going to have to add "a serious suggestion" to "honestly" and "trust
> me" as linguistic tags that flag a declaration as deceptive.
> 

It should be obvious from Debian list archives that I try to think through
what I write and consider who is reading it. It's not a frivolous, spur of 
the moment sentence: it's not axiomatic that we should still package
fortunes and translate them. Nor is it necessary for us to police what others
would choose to read or select for themselves.

It is a serious suggestion because it's thought through: you may note
from what I write that I'm endeavouring to be even-handed and transparent.
I also try to write clear prose and not weasel words.

Just because it was done that way in 1995 doesn't mean we have to do this now.

> Have you worked on embedded systems, ever?  It's not _necessary_ for
> Debian to package much of anything.  We could arguably serve just as
> well as "universal OS" by providing only a nucleus, say, a high-quality
> microkernel.[2]  Minimalism has never been an objective of the package
> archive.  This fact has been so transparently obvious for so long that
> it is difficult for me to maintain the presumption that you are arguing
> in good faith.
> 

I'm not arguing in *bad* faith; I'm thinking that we don't have to package
everything that we always have just because we've always done it that way

> > I find this suggestion demotivating and discouraging.
> 
> Sorry to hear that.  My own reaction is better termed "pissed off".
> 
> > I believe that this statement is inaccurate. There are parts of the
> > collection which are Debian-specific (the earliest of which, per the
> > changelog, were added in 1999), and others which have been added far
> > more recently than 1995 (there have been what seem like substantive
> > additions at least as recently as 2006).
> 
> Yes, some of them were collected by Joseph Carter ("knghtbrd"), a former
> Debian developer, ca. 2000 and for some time afterward.
> 
> > The way the Debian packaging splits the collection into various files,
> > which I understand is not necessarily done upstream, can also be
> > valuable.
> 
> I agree with this.  If I were maintainer I think I'd thus segregate the
> sort of mathy stuff I'd like to see, in acknowledgement that some people
> just aren't edified or provoked to thought by such things.  Similarly,
> some people just aren't up to being randomly confronted by exemplars of
> human folly, which characterizes the 1.5% or so of the fortunes-off
> package at issue, even if they _type_ "fortune" at a shell prompt.
> 

With the possible exception of the knightbrd files, the files are
segregated fairly well following the Github repository which itself looks
to have been derived from ibiblio.org

> > > The upstream Github repository is potentially only one of many
> > > disparate sites on the 'Net and the English language collection
> > > doesn't reflect the languages of Debian users worldwide.
> 
> Is anyone being prevented from submitting fortune collections localized
> to other languages?  Several are already present.[3]  Would you be able
> to recognize a Hitler quote rendered in Cyrillic?  Might such a quote
> not carry context for a person with family who participated in the Great
> Patriotic War that you don't share?  Perhaps some derogatory quip about
> the inferiority of the Slavs, when Russians culturally well remember
> hoisting the hammer and sickle over the Reichstag on 2 May 1945 (as my
> countryfolk do a similar event at Iwo Jima).
> 

I might well be able to recognise a Hitler quote in Cyrillic, yes - that's me.
I'm not guessing as to the context other people may have: I do note that
there are relatively few languages represented in the translations of fortune 
files in Debian. I could quite easily be missing out on something interesting
in French, Spanish or Italian, for example - I don't have context for
non-European languages apart from Welsh.

> Your response, to exclude it all and _foreclose as a matter of policy_
> the curation and maintenance of such resources amounts to declaring the
> Debian Project impotent to address the challenges here.
> 

It's a relatively straightforward approach that simplifies things. It 
would also propagate to Ubuntu and other Debian-derived distributions if
they took the change forward.

> It's my turn to cite §2.1.1 of the Debian Constitution as Steve Langasek
> did to me.  You are not _required_ to deal with fortune cookies.  You
> don't have to install them or use them.  There are many packages that
> any given user will never employ, or even be aware of.
> 
> The Project does not require your paternalistic supervision.
> 

I'm not supervising the Project - we're all equals here, no?

> Remolding the archive to more closely resemble your personal corpus of
> installed packages is a curious way to serve the principles of plurality
> and diversity.
> 

Everyone has their own Debian ...

> > > If Debian doesn't distribute fortune files but instead provides the
> > > means for users to make/download their own choice, nothing is lost.
> > > Debian is not responsible for maintaining any file content, whether
> > > questionable or unobjectionable depending on viewpoint, and we lose
> > > the burden of translation, maintenance and policing of content.
> 
> If you don't want these responsibilities, don't adopt them.  Same goes
> for Mr. Dowland.
> 

I'm not adopting them: I'm suggesting that the project disclaim the
principle of selection, translation and maintenance in favour of letting
a user decide on what they want in 2022. 

> > I am reasonably certain that this would just lead to far fewer people
> > bothering to make use of the fortunes database(s) at all, thereby
> > creating a self-fulfilling prophecy about how irrelevant this is in
> > the modern world.
> 
> The people who try to ban books in the United States employ a similar
> multi-level strategy.  If censorship of the local public library's
> collection is unsuccessful, they propose withdrawal of funds for the
> library, removing a much greater volume of common resources from public
> facility.[4]  No doubt to the satisfaction of the oligopolistic
> publishing industry, which views public libraries as frustrating its
> extraction of monopoly rents.[5]
> 
> > From my perspective, this whole discussion looks like someone whom
> > I've respected coming in and proposing to take away one of the small
> > things I somewhat like having around, and that taking-away happening
> > almost immediately despite the existence of pushback over it, and then
> > that person reacting to the pushback by proposing to take away a
> > *bigger* thing that I even *more* like having around. I imagine it's
> > not hard to see how that could be upsetting or demotivating.
> 
> I have to disagree with Sam a bit here, while "be bold" is _often_ good
> advice, especially in volunteer communities, it is not always.  Apart
> from the present contretemps, I can think of another (again from the
> United States).[6]
> 
> Apparently the package's maintainer has not done an upload since
> 2013.[7]  Even groff has managed to release twice since then![8]
> 

That's part of the point - there have been a bunch of NMUs and various hands
involved. The current maintainer is a Debian Maintainer who would probably
welcome help.

> I therefore solicit a volunteer who is fully up-to-date with packaging
> processes to privately advise me so that I can avoid making tedious
> mistakes.  E.g., I don't know if filing bugs against "wnpp" is still
> necessary or recommended for adopting a package with nine years of
> inattention.  Please email me privately if you feel yourself qualified
> and are supportive of my intended action.  You don't need to be
> enthusiastic about the package's _content_; I am not myself--there is a
> lot of crap in there IMO, even apart from the stuff at issue here.
> 

Great - thank you for stepping forward to help in getting the package into
shape.

> In the short run, my plan would be to ensure that the package is
> policy-compliant (and therefore can't be sniped again before release on
> a technicality by some "deeply principled" person), evaluate the handful
> of categories that are suggestive of group prejudice, and decide on some
> kind of updated disclaimer (if necessary) so that the package can warn
> about itself.  In the years since fortunes-mod was first packaged for
> Debian, the abbreviation "NSFW" has become commonly known for tagging of
> materials like this.  That may be more communicative than an "-off"
> suffix in the 2020s.  But if a rename will frustrate the package's
> progress through the incoming queue, I won't do it before the bookworm
> release.  The NSFW annotation can easily go in the package description.
> 

> In the long run, I expect to curate as suits me personally; I'll
> document my actions (of course) and if someone wants to package the
> items I throw away, that's up to them.  One can expect the addition of
> items reflecting my personal interests, like math, retrocomputing, music
> theory, and Catalan revolutionaries.
> 

There are a bunch of low priority bugs where folk have also suggested
either individual quotes or categories. One is for Terry Pratchett, for 
example.

> I don't have the desire or--especially--the time to undertake major
> changes in the package before the bookworm release.  I have something
> else I'm working that brings me joy and I'd like to get that solid for
> GNU release and thence into Debian.  Approximately 400 upstream bugs
> fixed--it does an old XFree86 package maintainer's heart good.[9][10]
> 
> Anyone who helps has my thanks in advance.
> 
> Regards,
> Branden
> 
> [1] https://lists.debian.org/debian-project/2022/11/msg00062.html
> 
> [2] And do I ever have one in mind I'd like to tell people about!  It
>     would please me much more than fighting over elementary principles
>     of open society.
> 
> [3] https://packages.debian.org/search?keywords=fortunes
> 
> [4] https://www.washingtonpost.com/nation/2022/08/24/michigan-library-defunded-gender-queer/
> [5] https://www.publishersweekly.com/pw/by-topic/industry-news/publisher-news/article/89038-over-the-past-25-years-the-big-publishers-got-bigger-and-fewer.html
> 
> [6] As usual, it was our conservo-glibertarian friends who were behind
>     this (q.v. "Free State Project").
> 
>     "...At issue was the local school budget, which funds a school for
>     children through fourth grade in town and covers tuition for older
>     students to attend private and public schools in neighboring towns.
>     At a poorly attended annual meeting in March, voters approved a
>     measure to cut the school budget by more than half, from $1.7
>     million to $800,000.
> 
>     The cut would have transformed the education system in Croydon,
>     replacing the public school system with one run by cheaper, private
>     companies that offer individualized programs, largely online.
>     Croydon would have become the first community in New Hampshire where
>     two companies, Prenda and Kaipod, would have become the default
>     education providers.
> 
>     Both companies have been championed by New Hampshire Education
>     Commissioner Frank Edelblut, who helped them secure contracts with
>     the state of New Hampshire. ..."
> 
>     https://www.nhpr.org/nh-news/2022-05-08/croydon-voters-overturn-school-budget-cut
> 
> [7] https://tracker.debian.org/pkg/fortune-mod
> [8] https://ftp.gnu.org/gnu/groff/
> [9] https://git.savannah.gnu.org/cgit/groff.git/tree/ANNOUNCE
> [10] https://git.savannah.gnu.org/cgit/groff.git/tree/NEWS

[toc] | [prev] | [next] | [standalone]

#13102 — Re: ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?])

>>>>> "Andrew" == Andrew M A Cater <amacater@einval.com> writes:

    Andrew> On Wed, Nov 23, 2022 at 03:05:29PM -0600, G. Branden Robinson wrote:
    >> [I'm using the pseudonymous respondent's message to reply to
    >> Mr. Cater as well.  Mind the angle brackets for quotation
    >> context.]
    >> 
    >> At 2022-11-23T14:14:38-0500, The Wanderer wrote:
    >> > On 2022-11-23 at 13:06, Andrew M.A. Cater wrote:
    >> > > Thank you for your considered opinions thus far. We have
    >> various > > developers who have written defending free speech:
    >> we've had others > > who have expressed various reservations with
    >> one aspect or other of > > the status quo.
    >> > > 
    >> > > There's been a grudging consensus that this is *hard*.
    >> 

    Andrew> No - there is a consensus that splitting things based on
    Andrew> cultural preference is hard - you and Sam both agree on that
    Andrew> point - various other people in the discussion have had
    Andrew> other viewpoints.

I don't think I have agreed that splitting things based on cultural
preference is hard.  I don't really think I think of things in a way
where I can imagine drawing conclusions about that.

I think that:

* There are hard issues involved

* There is cultural preference involved  to some degree.

[toc] | [prev] | [next] | [standalone]

#13103 — Re: ITA: fortunes-mod (was: SUMMARY [Was Re: Fortunes-off - do we need this as a package for Bookworm?])

[Multipart message — attachments visible in raw view] — view raw

n>>>>> "G" == G Branden Robinson <g.branden.robinson@gmail.com> writes:


    G> Neither you nor he, therefore, is well placed to present a
    G> (presumptively neutral) summary of the discussion.  (Neither am
    G> I.)

Branden, I'd like to push back on the idea that we want a summary from
someone neutral.
If we have that luxury--for example if we have a facilitator of
consensus building who happens not to have a strong opinion related to
the current consensus--that's great.

However, summaries are critical to consensus-building discussions.
We cannot have them without summaries.
The summaries are where someone tries to capture  where we are and see
if it "sticks."

I think we're now all able to see what happens when people think the
summary didn't capture the discussion--it didn't stick.

This will advance the discussion far more than if each of us walked away
from the discussion making our own assumptions about where we reached
without sharing those assumptions.  As an example of why that's bad, the
decision to originally enable usrmerge in debootstrap cited a
debian-devel consensus that was never summarized.  Putting it mildly,
some people viewed the existence of that consensus differently than the
people proposing the debootstrap change.  If that consensus had been
summarized--even by someone who was not neutral--we would have
discovered a disconnect much sooner.  We might have been able to avoid
significant bad blood.

So, I absolutely think Andrew was in a position to summarize, and if he
didn't I hope you or I or someone else would have chosen to do so.
I'm very glad he did.

[toc] | [prev] | [next] | [standalone]

#13097

[Multipart message — attachments visible in raw view] — view raw

>>>>> "Andrew" == Andrew M A Cater <amacater@einval.com> writes:

    Andrew> All, Thank you for your considered opinions thus far. We
    Andrew> have various developers who have written defending free
    Andrew> speech: we've had others who have expressed various
    Andrew> reservations with one aspect or other of the status quo.

    Andrew> There's been a grudging consensus that this is *hard*. It's
    Andrew> very hard indeed to draw good conclusions as to what to do
    Andrew> when everyone agrees that something could be done and
    Andrew> disagrees with what that should be.

    Andrew> Notably, Sam Hartman and Branden Robinson have pointed up
    Andrew> flaws with the existing categorisations and with a blanket
    Andrew> removal based on preference.  It's also noticeable that this
    Andrew> largely comes down to consideration of fortunes in English -
    Andrew> almost nothing has been said about other fortunes files or
    Andrew> other languages, though Sam talked about cultural
    Andrew> perceptions.

I think Debian has a standard approach for dealing with these sorts of
hard issues: leave it up to the maintainer.
I think you'd need a much stronger consensus than you could show today
to do anything else.

So, I have an alternative:
This is not a project-level decision.
Whoever is maintaining fortune should take this discussion as input and
maintain the packages.
If we don't like how they are doing that, we use our normal mechanisms
(TC and GR) to override them.

If you want to maintain fortune-mod and remove all the fortunes, go do
that.
But don't be surprised if someone else wants to maintain a version with
fortunes.


    Andrew> A serious suggestion: it is not necessary for Debian to
    Andrew> package fortune files at all.

It is not necessary, but as I discussed in my long message about
removing software from Debian, removing creative content from Debian has
a chilling effect.
Even if you write a downloader,  so all I have to do is type
fortune-install really-disgusting-insensitive-stuff
then it's still less accessible  than  if I can type apt install
whatever:

*  It's not on our mirrors
* It's not on our media

And as I discussed in my  DebConf talk [1], that sort of access matters.
Not everyone has internet connectivity.

[1]:
https://debconf20.debconf.org/talks/32-when-we-virtualize-the-whole-internet/


What you propose is a very community team compromise: let us remove the
things that make us uncomfortable.
And when those things are not core to our mission, that's often a
reasonable approach.
Especially when I consider the precident this creates, I consider this
suggestion an escalation.
I do not support it.
I think it compromises the freedoms I said I care about.

[toc] | [prev] | [next] | [standalone]

#13104

On Wed, 2022-11-23 at 12:34 -0700, Sam Hartman wrote:
> It is not necessary, but as I discussed in my long message about
> removing software from Debian, removing creative content from Debian
> has a chilling effect.

I would very much prefer explicit sexual content over Nazi symbols. So
let me make a suggestion:

If we insist on keeping Nazi symbolics in Debian, let's at least also
add explicit quotes from descriptions of Hitler + Blondi sexual
adventures (which I'm sure must exist on the internet), various form of
sexual abuse (there is enough fictional content to quote) and more in
the fortune-off package or elsewhere.

Or are explicit sexual references so bad that Nazi symbols are okay for
freedom of speech, but sexual references not?

In a related question: should we accept packages that link to, say, the
Stormfront web site in user-facing material or source code? (That's not
really hypothetical: I believe we have at least one package referencing
a page including racism and Holocaust denial.)

What about software greeting the user with a friendly "Sieg Heil" ("not
illegal in all juristictions" / "in every jurisdiction in the world"
was used as an argument earlier in the thread...)?

And for the "chilling effect": Debian makes it fairly easy to add
additional software sources; it's not like ecosystems where one or few
providers have a near monopoly like platform App stores, credit card
companies, Twitter, Amazon, ... which use their power to remove content
they don't like. And that is despite providers like Twitter, Github and
so on even having legal protections for third-party content which
Debian would not fall under, i.e., they do more control while having
less legal risks (AFAIU).

Or if you like another comparison more: Debian is more a newspaper
(with selected articles) rather than a generic communication provider
like the post office that can be used to send any information. Not
being able to publish your article in the New York Times is not a
chilling effect or censorship, the post office refusing to send certain
mail on the other hand would be.

Ansgar

-- 
"He [Hitler] is an extreme masochist who derives pleasure from having a
woman squat over him while she urinates or defecates on his face."
  -- Walter C. Langer: A Psychological Analysis of Adolph Hitler:
     His Life and Legend; PDF page 102 in
     https://www.cia.gov/readingroom/docs/CIA-RDP78-02646R000600240001-5.pdf

[toc] | [prev] | [next] | [standalone]

#13105

[Multipart message — attachments visible in raw view] — view raw

On 2022-11-23 at 16:16, Ansgar wrote:

> On Wed, 2022-11-23 at 12:34 -0700, Sam Hartman wrote:
>> It is not necessary, but as I discussed in my long message about 
>> removing software from Debian, removing creative content from
>> Debian has a chilling effect.
> 
> I would very much prefer explicit sexual content over Nazi symbols.
> So let me make a suggestion:
> 
> If we insist on keeping Nazi symbolics in Debian, let's at least
> also add explicit quotes from descriptions of Hitler + Blondi sexual 
> adventures (which I'm sure must exist on the internet), various form
> of sexual abuse (there is enough fictional content to quote) and more
> in the fortune-off package or elsewhere.
> 
> Or are explicit sexual references so bad that Nazi symbols are okay
> for freedom of speech, but sexual references not?

Do you have anything specific to point to as constituting "Nazi
symbolics"?

I'm not sure I can think of *anything* in the fortunes-off package which
would actually be even Nazi propaganda, much less Nazi symbology or
symbolism (I'm not sure what "symbolics" here means). Even the "Mein
Kampf" reference that were cited by count as reason to remove it, when
examined directly, don't seem to qualify; like every other Hitler-et-al.
quote in there that I remember seeing on a recent skim-through with this
sort of question in mind, they seem to be there to present and remind us
of how horrible these things are *because it was Nazis saying them*.

It seems to be being taken as a given that any quote from a Nazi is
automatically objectionable enough to exclude just because of who said
it, *even if provoking thought about how objectionable those people are
is the point of including that quote*. (Or, as with most of the "Mein
Kampf" quotes, provoking thought about how objectionable X other thing
must be because this known-objectionable person said this about it.)

If there's going to be criticism of the package for its contents, we
should make sure that the contents actually match the characteristics
being criticized, and I'm not at all sure that that's been shown.

> In a related question: should we accept packages that link to, say,
> the Stormfront web site in user-facing material or source code?
> (That's not really hypothetical: I believe we have at least one
> package referencing a page including racism and Holocaust denial.)
> 
> What about software greeting the user with a friendly "Sieg Heil"
> ("not illegal in all juristictions" / "in every jurisdiction in the
> world" was used as an argument earlier in the thread...)?

I don't have the time or energy at the moment to properly engage with
those questions, but I do see sufficient reason to distinguish them from
the one(s) about fortunes-off, such that the answers to those (sets of)
questions could well be different.

> And for the "chilling effect": Debian makes it fairly easy to add
> additional software sources; it's not like ecosystems where one or few
> providers have a near monopoly like platform App stores, credit card
> companies, Twitter, Amazon, ... which use their power to remove content
> they don't like. And that is despite providers like Twitter, Github and
> so on even having legal protections for third-party content which
> Debian would not fall under, i.e., they do more control while having
> less legal risks (AFAIU).

That's true, and a good thing. It doesn't mean that removing such
content from the official repositories *doesn't* have a chilling effect,
however; at most it means that that effect is less than it otherwise
might be.

The term "chilling effect" is not about blocking access to, or ability
to get/use, a thing; it's about discouraging people from engaging in /
with that thing. (IIRC its modern usage dates back to Supreme Court
jurisprudence on the subject of freedom of speech, possibly in an
anti-war context.)

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man.         -- George Bernard Shaw

[toc] | [prev] | [next] | [standalone]

#13106

>>>>> "Ansgar" == Ansgar  <ansgar@43-1.org> writes:


    Ansgar> I would very much prefer explicit sexual content over Nazi
    Ansgar> symbols. So let me make a suggestion:

As would I.
If anyone wants to work on the challenges of adult content feel free to
reach out and I'd be happy to spend some of my time on that.

I am not particularly interested in  erotica in Debian, so much as
software to promote sexual freedom.

I think there are legal challenges to sexually explicit content in
Debian.
And yes, there absolutely are legal challenges to some things related to
nazis in in Debian.

I think one thing that Andrew and I both failed to include in our
summaries is that there is a strong consensus we should not distribute
things that are illegal to distribute.
I think there's a consensus that if parts of fortunes-off that are
illegal to distribute they should not be distributed.
I think there is disagreement about how credible any such claims are.

[toc] | [prev] | [next] | [standalone]

#13098

[Multipart message — attachments visible in raw view] — view raw

On Wednesday, 23 November 2022 19:06:33 CET Andrew M.A. Cater wrote:
> If Debian doesn't distribute fortune files but instead provides the means
> for users to make/download their own choice, nothing is lost. Debian is
> not responsible for maintaining any file content, whether questionable
> or unobjectionable depending on viewpoint, and we lose the burden of
> translation, maintenance and policing of content.

My problem with this all is that it was removed NOT because it violates the 
DFSG or that it was (properly) considered illegal.
Those are excellent reasons to remove it.

It was removed because someone found its contents objectionable.

Thereby you make the criteria subjective.
To be consistent you'd have to remove ALL packages/contents which someone 
could possibly find objectionable.

/me wonders whether he should grab popcorn or run

Here's an example: I find stoning, crucifixion and keeping people as slaves 
rather objectionable. So let's get rid of all Bible related packages?
Keep the package, but make the content downloadable from 'some' site?

Another: I find ML* the tool to single out (and normally 'punish' in some way) 
anyone who falls out of the statistical norm. So it reduces people to numbers 
and anyone who falls out of that statistical norm, is an anomaly and/or and 
outcast. If that's a value that Debian wants to promote, I'm out.

And to be consistent, you'd then have to scan ALL the content, including all 
the source code for phrases/expressions that someone may find objectionable.
How about projects who use 'master' as the name of their main development 
branch? Get rid of that too?

Good luck getting a new Stable release out within the next 10 years.


My main objections are thus:
- It's subjective. And (therefor) a slippery slope.
- If you do it, you have to do it across the board. Consistently.

You may find the fortunes-mod package distasteful, but if you like it or not, 
it is representative of that time. So from a historical perspective, that has 
value. Just as 'Mein Kampf' has value from that perspective. But it being 
illegal in some jurisdictions could prevent it from entering the archive, 
which is an excellent and objective reason.
Also, you/people have the option/freedom to NOT install the package.
Name it 'fortunes-offensive' so no one installs it 'by accident'.

My 0.02

*) It is irrelevant if my understanding of ML is incorrect. The criteria for 
inclusion in the archive has become subjective now.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.project

csiph-web