• Re: Bye-Bye Dialup USA

    From Lawrence =?iso-8859-13?q?D=FFOliveiro?=@[email protected] to comp.misc on Thu Aug 21 21:48:14 2025
    From Newsgroup: comp.misc

    On Wed, 20 Aug 2025 18:44:40 -0400, InterLinked wrote:

    On 8/20/2025 6:08 PM, Lawrence D’Oliveiro wrote:

    On Wed, 20 Aug 2025 09:37:24 -0400, InterLinked wrote:

    There is a lot of stuff in the field that uses 300 baud modems for
    telemetry.

    I have a customer who does a lot of that, up and down the country.
    They use wireless connections (formerly 3G, now 4G) for that.

    Everything I deal with is on POTS lines.

    Seems like an expensive and unwieldy way to do it. The remote sensors
    might need to go months between inspections. They need to, not only
    withstand the elements, but have an adequate power supply. A wireless
    data connection means they can make a connection, exchange data, and
    disconnect again, all within a fraction of the time (and power
    consumption) it takes to do a modem handshake.

    Low-speed modem protocols without error correction also tend to be
    a decent quality test for voice connections.

    A more accurate test would surely involve actual voices.

    It can be hard to test things like latency and compression purely
    from just voice.

    Sure it is. All you need is the right instrumentation and testing
    standards to measure that voice. You *do* have standards, don’t you?

    Bad or just-okay connections tend to be forgiving for voice but less
    so for data.

    Another reason not to use them.

    Here <https://www.sierrawireless.com/> is the sort of comms modules
    that customer is using.
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From InterLinked@[email protected] to comp.misc on Thu Aug 21 18:34:28 2025
    From Newsgroup: comp.misc

    On 8/21/2025 5:48 PM, Lawrence D’Oliveiro wrote:
    On Wed, 20 Aug 2025 18:44:40 -0400, InterLinked wrote:

    On 8/20/2025 6:08 PM, Lawrence D’Oliveiro wrote:

    On Wed, 20 Aug 2025 09:37:24 -0400, InterLinked wrote:

    There is a lot of stuff in the field that uses 300 baud modems for
    telemetry.

    I have a customer who does a lot of that, up and down the country.
    They use wireless connections (formerly 3G, now 4G) for that.

    Everything I deal with is on POTS lines.

    Seems like an expensive and unwieldy way to do it. The remote sensors
    might need to go months between inspections. They need to, not only
    withstand the elements, but have an adequate power supply. A wireless
    data connection means they can make a connection, exchange data, and disconnect again, all within a fraction of the time (and power
    consumption) it takes to do a modem handshake.

    I can't share more details about this, but telephone lines make
    infinitely more sense for this sort of thing, because they already have
    phone line connections so adding a modem to it is the natural thing to
    do and adds no cost overhead.

    And aside from that, that's how it's been done for the last thirty years
    at least, so I doubt it will change now.

    Low-speed modem protocols without error correction also tend to be
    a decent quality test for voice connections.

    A more accurate test would surely involve actual voices.

    It can be hard to test things like latency and compression purely
    from just voice.

    Sure it is. All you need is the right instrumentation and testing
    standards to measure that voice. You *do* have standards, don’t you?

    Bad or just-okay connections tend to be forgiving for voice but less
    so for data.

    Another reason not to use them.

    Here <https://www.sierrawireless.com/> is the sort of comms modules
    that customer is using.

    I'm not sure what you're trying to say - if you have a crummy phone connection, don't use it?

    I was referring to different tests that could be done to evaluate the suitability of a long-distance provider for voice usage. Some people
    care how their phone calls sound and want them to sound good, but I get
    that you might not be one of them.
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From Lawrence =?iso-8859-13?q?D=FFOliveiro?=@[email protected] to comp.misc on Fri Aug 22 04:12:14 2025
    From Newsgroup: comp.misc

    On Thu, 21 Aug 2025 18:34:28 -0400, InterLinked wrote:

    I was referring to different tests that could be done to evaluate the suitability of a long-distance provider for voice usage.

    I thought you were talking about data usage, not voice usage, which you
    said tended to be more “forgiving”.
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From InterLinked@[email protected] to comp.misc on Fri Aug 22 09:23:57 2025
    From Newsgroup: comp.misc

    On 8/22/2025 12:12 AM, Lawrence D’Oliveiro wrote:
    On Thu, 21 Aug 2025 18:34:28 -0400, InterLinked wrote:

    I was referring to different tests that could be done to evaluate the
    suitability of a long-distance provider for voice usage.

    I thought you were talking about data usage, not voice usage, which you
    said tended to be more “forgiving”.

    I'm talking about using data calls as a general test, whether using for
    data or voice. Because voice is more forgiving, it's lousy as a test
    mechanism - it's hard to objectively compare phone calls by ear (except
    maybe testing latency by DTMF response time) but data calls give you
    hard, well, *data* that can be used for comparison. A phone call that
    works better for data is most likely going to be better for voice, too.
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From Lawrence =?iso-8859-13?q?D=FFOliveiro?=@[email protected] to comp.misc on Fri Aug 22 22:22:20 2025
    From Newsgroup: comp.misc

    On Fri, 22 Aug 2025 09:23:57 -0400, InterLinked wrote:

    On 8/22/2025 12:12 AM, Lawrence D’Oliveiro wrote:

    On Thu, 21 Aug 2025 18:34:28 -0400, InterLinked wrote:

    I was referring to different tests that could be done to evaluate the
    suitability of a long-distance provider for voice usage.

    I thought you were talking about data usage, not voice usage, which you
    said tended to be more “forgiving”.

    I'm talking about using data calls as a general test, whether using for
    data or voice.

    Not sure why, given that data calls are really only a test of data transmission, and voice calls really only a test of voice.

    Because voice is more forgiving, it's lousy as a test
    mechanism ...

    Testing voice calls is done by ... actually making voice calls. Or rather, transmitting audio signals of various types and measuring the results. Ask
    any telephone engineer.
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From jerk-o@[email protected] to comp.misc on Sun Aug 31 18:36:23 2025
    From Newsgroup: comp.misc

    On Sun, 10 Aug 2025 08:11:54 -0000 (UTC), Lawrence D'Oliveiro
    <[email protected]d> wrote
    US-based ISP America On-Line (AOL) will finally turn off its dialup
    Internet service at the end of September ending 34 years of operation.

    Does this mean the Eternal September will finally come to an end?
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From noreply@[email protected] to comp.misc on Sun Aug 31 23:51:20 2025
    From Newsgroup: comp.misc

    On Sun, 31 Aug 2025 18:36:23 -0700, jerk-o <[email protected]> wrote:
    Does this mean the Eternal September will finally come to an end?

    not likely, since the phrase "the eternal september" (popularized after its namesake usenet article was posted 1994-02-09 <2j9vhj$[email protected]>) reflected the consensus that the flood of mischief-makers was already never ending thirty-one years ago . . . operation eternal september, infiltration
    and occupation of usenet newsgroups (collectively known as the "troll farm") has thus continued with relentless attacks on active newsgroups to this day

    aol's new (in 1994) "usenet gateway" encouraged everyone to join in the fun
    and along with other isps usenet access via the internet became commonplace, but aol's historic contribution is indelible . . . freeware newsreaders and public (nntp) news servers also made usenet increasingly popular by the y2k, after which the signal-to-noise ratio tended to skyrocket, and googlegroups (after absorbing dejanews 2001-02-12) made it that much easier for everyone with a web browser to participate (but after around 2004-11-29, gg's "beta" made it progressively less reliable and more difficult to use for searching usenet archives), but by then usenet was mostly saturated with "troll farm" activity . . . even so, usenet was initially designed to withstand calamity
    so it has remained popular, mostly for its plain text and binary newsgroups that are unmoderated, so if you like the untamed wild west, usenet is ideal

    (using Tor Browser 14.5.6) https://duckduckgo.com/?q=america+online+shut&ia=web&assist=true
    AOL is shutting down its dial-up internet service after 30 years, with the >discontinuation set for the end of September 2025. This marks the end of an era
    for many users who experienced the early days of the internet through AOL's >services. Yahoo The Guardian
    AOL Shutting Down Dial-Up Internet Service
    Overview
    AOL, originally known as America Online, is discontinuing its dial-up internet >service after 30 years. This decision marks the end of an era for many users who
    experienced the early days of the internet through AOL's distinctive dial-up >connections.
    Key Details
    Shutdown Date: The dial-up service will officially cease operations on
    September 30, 2025.
    Historical Significance: AOL was a pioneer in providing internet access to
    millions of Americans, especially during the 1990s and early 2000s. At its
    peak, it had over 30 million subscribers.
    Decline of Dial-Up: The rise of broadband and wireless internet has led to a
    significant decline in dial-up users. As of 2023, only about 163,401
    households in the U.S. relied solely on dial-up, representing just over 0.13%
    of all internet subscriptions.
    Company Evolution
    Founding: AOL was founded in 1985 as Quantum Computer Services and rebranded
    in 1991. It became known for its iconic "You've got mail" notification.
    Ownership Changes: AOL has undergone several ownership changes, including a
    merger with Time Warner in 2000 and later acquisitions by Verizon and Apollo
    Global Management.
    Service Changes: In addition to shutting down dial-up, AOL has previously
    discontinued its Instant Messenger service in 2017 and continues to offer
    email and other online services.
    This shutdown reflects the broader trend of moving away from older internet >technologies as faster and more reliable options become the norm.
    Yahoo Wikipedia
    [end quoted "search assist"]

    (neodome's sporadic [s***] tag, sometimes it doesn't, but other times "spam") --- Synchronet 3.21a-Linux NewsLink 1.2
  • From Marco Moock@[email protected] to comp.misc on Mon Sep 1 17:53:28 2025
    From Newsgroup: comp.misc

    On 31.08.2025 18:36 Uhr jerk-o wrote:

    On Sun, 10 Aug 2025 08:11:54 -0000 (UTC), Lawrence D'Oliveiro <[email protected]d> wrote
    US-based ISP America On-Line (AOL) will finally turn off its dialup >Internet service at the end of September ending 34 years of
    operation.

    Does this mean the Eternal September will finally come to an end?

    IIRC AOL already closed Usetnet access in the early 2000s.
    --
    kind regards
    Marco

    Send spam to [email protected]

    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From candycanearter07@[email protected] to comp.misc on Tue Sep 2 21:50:04 2025
    From Newsgroup: comp.misc

    jerk-o <[email protected]> wrote at 01:36 this Monday (GMT):
    On Sun, 10 Aug 2025 08:11:54 -0000 (UTC), Lawrence D'Oliveiro
    <[email protected]d> wrote
    US-based ISP America On-Line (AOL) will finally turn off its dialup >>Internet service at the end of September ending 34 years of operation.

    Does this mean the Eternal September will finally come to an end?


    Hopefully not, I'm using their server to post! :D
    --
    user <candycane> is generated from /dev/urandom
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From Anthk NM@[email protected] to comp.misc on Sun Nov 2 22:35:47 2025
    From Newsgroup: comp.misc

    On 2025-08-14, Retrograde <[email protected]d> wrote:
    On Sun, 10 Aug 2025 08:11:54 -0000 (UTC)
    Lawrence D'Oliveiro <[email protected]d> wrote:

    US-based ISP America On-Line (AOL) will finally turn off its dialup
    Internet service at the end of September

    There are other dial-up providers beside AOL. That kind of connection
    is only good for email however and maybe Usenet. Impossible to surf
    the modern web. I remember when webpages strove to keep an individual
    page size below 30KB. Long ago.

    I used Gopher, Usenet and IRC with 2.7 KBPS. Some OPUS audio
    stations (16 KBPS) were usable with mplayer and streaming
    caching options.
    Mosh for SSH worked like a dream, too; and I could
    properly read some RSS feeds and web sites with lynx/links
    and some of them, with edbrowse. Nowadays with
    gemini (check comp.infosystems.gemini) and a proxy
    to the web I almost can read news at great speeds
    thru News Waffle at gemini://gemi.dev

    You could be surprised what could you do with
    constrained connections.
    --- Synchronet 3.21a-Linux NewsLink 1.2
  • From Anthk NM@[email protected] to comp.misc on Sun Nov 2 22:35:49 2025
    From Newsgroup: comp.misc

    On 2025-08-16, Mike Spencer <[email protected]e> wrote:

    Retrograde <[email protected]d> writes:

    On Sun, 10 Aug 2025 08:11:54 -0000 (UTC)
    Lawrence D'Oliveiro <[email protected]d> wrote:

    US-based ISP America On-Line (AOL) will finally turn off its dialup
    Internet service at the end of September

    There are other dial-up providers beside AOL. That kind of connection
    is only good for email however and maybe Usenet. Impossible to surf
    the modern web.

    There's a work-around that can help a little for site you visit often, assuming that useful info will render w/o js, assuming you have a
    resident web server on localhost and can write some perl code.

    Put a link on your home page on localhost to a cgi-bin script. (You
    *do* keep a home page on localhost, don't you? ;-) Cause that link to
    send the real URL as data.

    Create a cgi-bin perl script that reads the request from your bowser,
    then uses wget or similar to fetch the target page.

    The script reads in whatever is sent into a perl variable, then use
    regexps to elide all IMG and SCRIPT tags/blocks, elides STYLE and SVG
    blocks, elides and LINK tags the fetch or prefetch other data.

    Re-writing and anchor tags that point back to the remote host so that
    they point to the script instead (handing the script the real URL as
    data) is also good but a little more trouble.

    Script then sends the result of the editing process back to your
    browser.

    I've only been off dial-up for five years. This hack sped up several
    sites. I still use some of the scripts to get rid of unwanted STYLE
    and js.

    Useless, of course, for all-js social media sites but I don't do those anyway.

    I remember when webpages strove to keep an individual page size
    below 30KB. Long ago.

    Now some email has more than 30KB in headers, not to mention
    unwarranted HTML with huge STYLE blocks.


    Just get a Gemini client. Bombadillo/Amfora, Kristall, Lagrange (the easiest. Head to gemini://gemi.dev, enter to the News Waffle service.
    Input any news or blog URL, the full one, with a preceeding https://,
    such as https://arstechnica.com

    Say hello to bandwidth saves down to a 5% of the original.
    --- Synchronet 3.21a-Linux NewsLink 1.2