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

Groups > comp.misc > #18038 > unrolled thread

python to deprecate NNTP

Back to article view | Back to comp.misc

Contents

  python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-23 17:44 -0400
    Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-23 21:49 +0000
      Re: python to deprecate NNTP not@telling.you.invalid (Computer Nerd Kev) - 2019-05-23 22:59 +0000
        Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-24 01:51 +0000
        Re: python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-23 22:24 -0400
          Re: python to deprecate NNTP Bob Eager <news0073@eager.cx> - 2019-05-24 07:52 +0000
            Re: python to deprecate NNTP mm0fmf <none@invalid.com> - 2019-05-29 18:06 +0100
          Re: python to deprecate NNTP Richard Kettlewell <invalid@invalid.invalid> - 2019-05-24 13:25 +0100
          Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-24 14:54 +0000
            Re: python to deprecate NNTP Sylvia Else <sylvia@email.invalid> - 2019-05-25 23:41 +1000
              Re: python to deprecate NNTP Huge <Huge@nowhere.much.invalid> - 2019-05-25 16:35 +0000
                Re: python to deprecate NNTP kludge@panix.com (Scott Dorsey) - 2019-05-26 13:38 -0400
                  Re: python to deprecate NNTP not@telling.you.invalid (Computer Nerd Kev) - 2019-05-26 23:21 +0000
                  Re: python to deprecate NNTP Huge <Huge@nowhere.much.invalid> - 2019-05-27 11:06 +0000
              Re: python to deprecate NNTP Mike Spencer <mds@bogus.nodomain.nowhere> - 2019-05-25 14:43 -0300
      Re: python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-23 20:37 -0400
        Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-24 15:00 +0000
        Re: python to deprecate NNTP Mike Spencer <mds@bogus.nodomain.nowhere> - 2019-05-24 17:31 -0300
          Re: python to deprecate NNTP Huge <Huge@nowhere.much.invalid> - 2019-05-25 07:38 +0000
          Re: python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-26 01:36 +0000
      Re: python to deprecate NNTP mm0fmf <none@invalid.com> - 2019-05-29 18:05 +0100
    Re: python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-23 17:51 -0400
    Re: python to deprecate NNTP Sylvia Else <sylvia@email.invalid> - 2019-05-24 14:29 +1000
      Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-24 15:01 +0000
        Re: python to deprecate NNTP Jerry Peters <jerry@example.invalid> - 2019-05-24 20:25 +0000
          Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-24 21:32 +0000
            Re: python to deprecate NNTP not@telling.you.invalid (Computer Nerd Kev) - 2019-05-25 03:23 +0000
            Re: python to deprecate NNTP Jerry Peters <jerry@example.invalid> - 2019-05-25 20:32 +0000
              Re: python to deprecate NNTP Dan Espen <dan1espen@gmail.com> - 2019-05-25 18:08 -0400
                Re: python to deprecate NNTP Mike Spencer <mds@bogus.nodomain.nowhere> - 2019-05-25 21:27 -0300
              Re: python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-27 21:12 -0400
                Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-28 03:06 +0000
                  Re: python to deprecate NNTP Richard Kettlewell <invalid@invalid.invalid> - 2019-05-28 08:16 +0100
                    Re: python to deprecate NNTP Rich <rich@example.invalid> - 2019-05-28 15:19 +0000
                      Re: python to deprecate NNTP Richard Kettlewell <invalid@invalid.invalid> - 2019-05-28 18:26 +0100
                Re: python to deprecate NNTP kludge@panix.com (Scott Dorsey) - 2019-05-28 14:46 -0400
    Re: python to deprecate NNTP "polk" <polka@dot.dot> - 2019-05-25 15:14 +0000
      Re: python to deprecate NNTP RS Wood <rsw@therandymon.com> - 2019-05-26 01:37 +0000
        Re: python to deprecate NNTP Huge <Huge@nowhere.much.invalid> - 2019-05-26 05:45 +0000
    Re: python to deprecate NNTP Robert Munyer <robertm@not-for-mail.invalid> - 2019-06-07 23:03 +0000

Page 2 of 2 — ← Prev page 1 [2]

#18077

On 23/05/2019 22:49, Rich wrote:
> RS Wood <rsw@therandymon.com> wrote:
>> https://www.python.org/dev/peps/pep-0594/#nntplib
>>
>> This affects me, at least.  The little script I use to post articles
>> from my RSS reader relies on python.  Boo!  Why can't developers just
>> leave working shit alone?
>
> Or... switch to another language for your 'little script':
>
> https://www.tcl.tk/community/tcl2004/Tcl2003papers/kupries-doctools/tcllib/modules/nntp/nntp.man
>
> Handles both reading and posting.
>
>> I'll wind up maintaining a VM with an older distro or something, just
>> to do what I like to do.  I hate that!
>
> An alternative, but if you do *hate* it enough, maybe you might
> consider the switch above?
>

tcl is the work of Satan.

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

#18040

On Thu, 23 May 2019 17:44:27 -0400
RS Wood <rsw@therandymon.com> wrote:

> https://www.python.org/dev/peps/pep-0594/#nntplib
> 
> This affects me, at least.  The little script I use to post articles
> from my RSS reader relies on python.  Boo!  Why can't developers just
> leave working shit alone?
> 
> I'll wind up maintaining a VM with an older distro or something, just
> to do what I like to do.  I hate that!

Just remembered another example: I had an old x86 box and a Hayes
serial modem and thought it would be fine to try to build a dial up
server.  I got nowhere fast, although let's admit my skills could use
some sharpening. Most how-tos depend on older Linux kernels and
pre-systemd systems with working getty, expect, and inittab configs.  I
got annoyed with it and gave up.  I can totally imagine a situation
where it would be useful to install a little dial up server somewhere
and, eg. run a BBS or something.

The web is obviously moving so fast browsers from 2 years ago
struggle.  Yeah yeah, in the name of progress you have to slough off
the dead wood, etc.  Blah blah blah.  

Off to pour myself a whiskey ... see ya!

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

#18046

On 24/05/2019 7:44 am, RS Wood wrote:
> https://www.python.org/dev/peps/pep-0594/#nntplib
> 
> This affects me, at least.  The little script I use to post articles
> from my RSS reader relies on python.  Boo!  Why can't developers just
> leave working shit alone?

Can't you just keep a copy of it?

Sounds like it's not being maintained anyway.

Sylvia.

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

#18051

Sylvia Else <sylvia@email.invalid> wrote:
> On 24/05/2019 7:44 am, RS Wood wrote:
>> https://www.python.org/dev/peps/pep-0594/#nntplib
>> 
>> This affects me, at least.  The little script I use to post articles 
>> from my RSS reader relies on python.  Boo!  Why can't developers 
>> just leave working shit alone?
> 
> Can't you just keep a copy of it?
> 
> Sounds like it's not being maintained anyway.

What is there to 'maintain' when the protocol it speaks (NNTP) itself 
is not changing?

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

#18052

Rich <rich@example.invalid> wrote:
> Sylvia Else <sylvia@email.invalid> wrote:
>> On 24/05/2019 7:44 am, RS Wood wrote:
>>> https://www.python.org/dev/peps/pep-0594/#nntplib
>>> 
>>> This affects me, at least.  The little script I use to post articles 
>>> from my RSS reader relies on python.  Boo!  Why can't developers 
>>> just leave working shit alone?
>> 
>> Can't you just keep a copy of it?
>> 
>> Sounds like it's not being maintained anyway.
> 
> What is there to 'maintain' when the protocol it speaks (NNTP) itself 
> is not changing?
> 

Because the people in charge of Python keep changing it, addding new
features, but also deprecating old ones. Or changes like 'print'
becoming a function rather than a statement. Then there are the
strings changing from characters to unicode characters.

I *liked* Python 2.6, it was a nice, simple language to learn with
lots of stuff supported by included and optional libraries. Then TPTB
got on a functional kick an started complicating a nice simpe
langauge.

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

#18054

Jerry Peters <jerry@example.invalid> wrote:
> Rich <rich@example.invalid> wrote:
>> Sylvia Else <sylvia@email.invalid> wrote:
>>> On 24/05/2019 7:44 am, RS Wood wrote:
>>>> https://www.python.org/dev/peps/pep-0594/#nntplib
>>>> 
>>>> This affects me, at least.  The little script I use to post articles 
>>>> from my RSS reader relies on python.  Boo!  Why can't developers 
>>>> just leave working shit alone?
>>> 
>>> Can't you just keep a copy of it?
>>> 
>>> Sounds like it's not being maintained anyway.
>> 
>> What is there to 'maintain' when the protocol it speaks (NNTP) itself 
>> is not changing?
>> 
> 
> Because the people in charge of Python keep changing it, addding new
> features, but also deprecating old ones. Or changes like 'print'
> becoming a function rather than a statement. Then there are the
> strings changing from characters to unicode characters.

So, the 'maintence' is not because of the NNTP module, it is self 
imposed because they (those in charge) can't leave well enough alone 
(or at least provide proper backwards compatibility so old stuff just 
keeps working without needing an update).

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

#18055

Rich <rich@example.invalid> wrote:
> Jerry Peters <jerry@example.invalid> wrote:
>> Rich <rich@example.invalid> wrote:
>>> 
>>> What is there to 'maintain' when the protocol it speaks (NNTP) itself 
>>> is not changing?
>> 
>> Because the people in charge of Python keep changing it, addding new
>> features, but also deprecating old ones. Or changes like 'print'
>> becoming a function rather than a statement. Then there are the
>> strings changing from characters to unicode characters.
> 
> So, the 'maintence' is not because of the NNTP module, it is self 
> imposed because they (those in charge) can't leave well enough alone 
> (or at least provide proper backwards compatibility so old stuff just 
> keeps working without needing an update).

Yep, hence when looking for Python scripts you find the old "works
with Python version x, and version y, maybe version z". Or abandoned
scripts written for "Python" which just mysteriously refuse to work.

So while it may make it easier for the Python developers to implement
new ideas, the burden of the changes is either passed onto script
writers (who now have to work out ways to maintain backwards
compatibility in their own code), or onto script users who are going
to end up keeping multiple versions of Python installed if they want
to use scripts from multiple outside sources (this burden may then be
passed on to distibution package maintainers).

But that's hardly uncommon - as a look at the tangle of Debian
packages available for such software "dependencies" shows.

-- 
__          __
#_ < |\| |< _#

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

#18061

Rich <rich@example.invalid> wrote:
> Jerry Peters <jerry@example.invalid> wrote:
>> Rich <rich@example.invalid> wrote:
>>> Sylvia Else <sylvia@email.invalid> wrote:
>>>> On 24/05/2019 7:44 am, RS Wood wrote:
>>>>> https://www.python.org/dev/peps/pep-0594/#nntplib
>>>>> 
>>>>> This affects me, at least.  The little script I use to post articles 
>>>>> from my RSS reader relies on python.  Boo!  Why can't developers 
>>>>> just leave working shit alone?
>>>> 
>>>> Can't you just keep a copy of it?
>>>> 
>>>> Sounds like it's not being maintained anyway.
>>> 
>>> What is there to 'maintain' when the protocol it speaks (NNTP) itself 
>>> is not changing?
>>> 
>> 
>> Because the people in charge of Python keep changing it, addding new
>> features, but also deprecating old ones. Or changes like 'print'
>> becoming a function rather than a statement. Then there are the
>> strings changing from characters to unicode characters.
> 
> So, the 'maintence' is not because of the NNTP module, it is self 
> imposed because they (those in charge) can't leave well enough alone 
> (or at least provide proper backwards compatibility so old stuff just 
> keeps working without needing an update).

Yep! 
I suspect the unicode changes are the deal breaker. A string now has
to be in the correct encoding, if it's not you get an exception. The
problem becomes *knowing* the correct encoding for arbitrary text
you're receiving over the net. A byte string would work, but they
aren't supported as completely as a unicode string. This BTW is
claimed to be an improvement by the Python gurus.

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

#18062

Jerry Peters <jerry@example.invalid> writes:

> Rich <rich@example.invalid> wrote:
>> Jerry Peters <jerry@example.invalid> wrote:
>>> Rich <rich@example.invalid> wrote:
>>>> Sylvia Else <sylvia@email.invalid> wrote:
>>>>> On 24/05/2019 7:44 am, RS Wood wrote:
>>>>>> https://www.python.org/dev/peps/pep-0594/#nntplib
>>>>>> 
>>>>>> This affects me, at least.  The little script I use to post
>>>>>> articles from my RSS reader relies on python.  Boo!  Why can't
>>>>>> developers just leave working shit alone?
>>>>> 
>>>>> Can't you just keep a copy of it?
>>>>> 
>>>>> Sounds like it's not being maintained anyway.
>>>> 
>>>> What is there to 'maintain' when the protocol it speaks (NNTP)
>>>> itself is not changing?
>>>> 
>>> 
>>> Because the people in charge of Python keep changing it, addding new
>>> features, but also deprecating old ones. Or changes like 'print'
>>> becoming a function rather than a statement. Then there are the
>>> strings changing from characters to unicode characters.
>> 
>> So, the 'maintence' is not because of the NNTP module, it is self
>> imposed because they (those in charge) can't leave well enough alone
>> (or at least provide proper backwards compatibility so old stuff just
>> keeps working without needing an update).
>
> Yep!  I suspect the unicode changes are the deal breaker. A string now
> has to be in the correct encoding, if it's not you get an
> exception. The problem becomes *knowing* the correct encoding for
> arbitrary text you're receiving over the net. A byte string would
> work, but they aren't supported as completely as a unicode
> string. This BTW is claimed to be an improvement by the Python gurus.

I find that ironic because all these encodings are the reason I dumped
Perl/Gtk for Python/QT.

I had a Perl/Gtk1 app that I wrote that became unbearably slow with Gtk2
and all it's character set stuff.

Back when I started, computers did upper case only and I was already
writing documentation in all upper case.  I was fine with this and I
guessed that in the future humanity would adapt and start using upper
case only for everything.

When lower case was introduced, I was skeptical that this was necessary
or a good idea.

Today, it's hard to believe computers support picture characters,
characters that go right to left, left to right, top to bottom, bottom
to top.  Some programs even let to mix directional text in the same
document.  Sort programs have to look at every character to figure out
how to compare characters.  It's completely crazy and a huge waste of
computer cycles.


-- 
Dan Espen

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

#18063

Dan Espen <dan1espen@gmail.com> writes:

> Today, it's hard to believe computers support picture characters,
> characters that go right to left, left to right, top to bottom, bottom
> to top.  Some programs even let to mix directional text in the same
> document.  Sort programs have to look at every character to figure out
> how to compare characters.  It's completely crazy and a huge waste of
> computer cycles.

What he said.  Computer character sets are wonderful if you need to
write Arabic, Russian, Japanese or even Hungarian.  Did you ever see a
movable type font for a chinese newspaper or a pre-computers Japanese
typewriter?

But my reading is 99% in English and I can live with the occasional
grammatically incorrect e-acute, c-cidilla or a-umlaut
substitution. Several of my correspondents seem to use mailer software
that insists on replacing the ASCII double- and single-quote chars with
the typographically correct UTF-8 left and right chars even though
nothing else in their messages needs anything but ASCII.

And I still cant figure out what the current strings(1) util is
doing.  Hunting for ASCII char strings in binaries now returns
unwanted garbage, presumably because of some LANG or charset setting.

Feh.
-- 
Mike Spencer                  Nova Scotia, Canada

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

#18070

On 2019-05-25, Jerry Peters <jerry@example.invalid> wrote:
> Rich <rich@example.invalid> wrote:

>> So, the 'maintence' is not because of the NNTP module, it is self 
>> imposed because they (those in charge) can't leave well enough alone 
>> (or at least provide proper backwards compatibility so old stuff just 
>> keeps working without needing an update).

So ... "We over-complicated things so badly that now we have to drop
basic stuff.  But it's not our fault, somehow."

I happen to like python, but this is not a great outcome, and it's not
flattering of python leadership.

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

#18071

RS Wood <rsw@therandymon.com> wrote:
> On 2019-05-25, Jerry Peters <jerry@example.invalid> wrote:
>> Rich <rich@example.invalid> wrote:
> 
>>> So, the 'maintence' is not because of the NNTP module, it is self 
>>> imposed because they (those in charge) can't leave well enough 
>>> alone (or at least provide proper backwards compatibility so old 
>>> stuff just keeps working without needing an update).
> 
> So ...  "We over-complicated things so badly that now we have to drop 
> basic stuff.  But it's not our fault, somehow."
> 
> I happen to like python, but this is not a great outcome, and it's 
> not flattering of python leadership.

No, it is not.  And it implies a "lost their way" aspect that just 
might result in python falling out of favor and fading away.

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

#18072

Rich <rich@example.invalid> writes:
> RS Wood <rsw@therandymon.com> wrote:
>> On 2019-05-25, Jerry Peters <jerry@example.invalid> wrote:
>>> Rich <rich@example.invalid> wrote:
>> 
>>>> So, the 'maintence' is not because of the NNTP module, it is self
>>>> imposed because they (those in charge) can't leave well enough
>>>> alone (or at least provide proper backwards compatibility so old
>>>> stuff just keeps working without needing an update).
>> 
>> So ...  "We over-complicated things so badly that now we have to drop
>> basic stuff.  But it's not our fault, somehow."

It seems like neither of you bothered to read the explanation of why
nntplib is causing practical trouble.

>> I happen to like python, but this is not a great outcome, and it's
>> not flattering of python leadership.
>
> No, it is not.  And it implies a "lost their way" aspect that just 
> might result in python falling out of favor and fading away.

This thread is getting pretty ridiculous. You can still use nntplib if
you want; at worst you’ll have to install it manually, and maybe fix the
occasional syntax issue if you want to use very new Python
implementations.  As for the implications for Python itself, there are
none, Usenet has fallen out of favour and is well advanced in the
process of fading away, so it shouldn’t be surprising that the rest of
the world doesn’t want to spend time and effort supporting it any more.

-- 
https://www.greenend.org.uk/rjk/

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

#18073

Richard Kettlewell <invalid@invalid.invalid> wrote:
> Rich <rich@example.invalid> writes:
>> RS Wood <rsw@therandymon.com> wrote:
>>> On 2019-05-25, Jerry Peters <jerry@example.invalid> wrote:
>>>> Rich <rich@example.invalid> wrote:
>>> 
>>>>> So, the 'maintence' is not because of the NNTP module, it is self
>>>>> imposed because they (those in charge) can't leave well enough
>>>>> alone (or at least provide proper backwards compatibility so old
>>>>> stuff just keeps working without needing an update).
>>> 
>>> So ...  "We over-complicated things so badly that now we have to drop
>>> basic stuff.  But it's not our fault, somehow."
> 
> It seems like neither of you bothered to read the explanation of why
> nntplib is causing practical trouble.

Nor did you bother to enlighten us.  Care to explain/post a link (since 
you imply you have found such "explanation" by your comment above).

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

#18074

Rich <rich@example.invalid> writes:
> Richard Kettlewell <invalid@invalid.invalid> wrote:
>> Rich <rich@example.invalid> writes:
>>> RS Wood <rsw@therandymon.com> wrote:
>>>> On 2019-05-25, Jerry Peters <jerry@example.invalid> wrote:
>>>>> Rich <rich@example.invalid> wrote:
>>>> 
>>>>>> So, the 'maintence' is not because of the NNTP module, it is self
>>>>>> imposed because they (those in charge) can't leave well enough
>>>>>> alone (or at least provide proper backwards compatibility so old
>>>>>> stuff just keeps working without needing an update).
>>>> 
>>>> So ...  "We over-complicated things so badly that now we have to drop
>>>> basic stuff.  But it's not our fault, somehow."
>> 
>> It seems like neither of you bothered to read the explanation of why
>> nntplib is causing practical trouble.
>
> Nor did you bother to enlighten us. Care to explain/post a link (since
> you imply you have found such "explanation" by your comment above).

The URL is at the top of the thread. The explanation is neither long nor
complex.

-- 
https://www.greenend.org.uk/rjk/

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

#18075

RS Wood  <randy@therandymon.com> wrote:
>On 2019-05-25, Jerry Peters <jerry@example.invalid> wrote:
>> Rich <rich@example.invalid> wrote:
>
>>> So, the 'maintence' is not because of the NNTP module, it is self 
>>> imposed because they (those in charge) can't leave well enough alone 
>>> (or at least provide proper backwards compatibility so old stuff just 
>>> keeps working without needing an update).
>
>So ... "We over-complicated things so badly that now we have to drop
>basic stuff.  But it's not our fault, somehow."
>
>I happen to like python, but this is not a great outcome, and it's not
>flattering of python leadership.

Python 3 is like that.
There is great resistance in some quarters to move from Python 2 as a
consequence.
--scott
-- 
"C'est un Nagra. C'est suisse, et tres, tres precis."

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

#18058

> https://www.python.org/dev/peps/pep-0594/#nntplib

https://perldoc.perl.org/Net/NNTP.html

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

#18065

On 2019-05-25, polk <polka@dot.dot> wrote:
>> https://www.python.org/dev/peps/pep-0594/#nntplib
>
> https://perldoc.perl.org/Net/NNTP.html

But doesn't that send me down the Perl 5 vs Perl 6 rabbit hole? 

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

#18066

On 2019-05-26, RS Wood <rsw@therandymon.com> wrote:
> On 2019-05-25, polk <polka@dot.dot> wrote:
>>> https://www.python.org/dev/peps/pep-0594/#nntplib
>>
>> https://perldoc.perl.org/Net/NNTP.html
>
> But doesn't that send me down the Perl 5 vs Perl 6 rabbit hole? 

It looks like Python is going exactly the same way.


-- 
Today is Sweetmorn, the 73rd day of Discord in the YOLD 3185
       Comes in bells, your servant, don't forsake him 

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

#18134

RS Wood wrote:

> https://www.python.org/dev/peps/pep-0594/#nntplib

The text at that location seems to be saying that they never set up a
server for their automated test suite to communicate with, but instead
they just routinely abused someone else's real production server for
that purpose.

The obvious and predictable shortcoming of that arrangement is that
if the abused system happens to be down, or if it blocks or throttles
connections from the abusing system, the automated test suite fails.

Disclaimer:  I have no first-hand knowledge of any of this; I'm just
interpreting what I read at that URL.

-- 
e-mail: the TLD is COM, the subdomain is my surname,
and the local-part is the product of 86373 and 106472.

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | comp.misc

csiph-web