John Larkin <jlarkin@highland_atwork_technology.com> wrote:

> On Mon, 30 Aug 2021 22:46:56 -0000 (UTC), Steve Wilson
> <spamme@not.com> wrote:
>
>>John Larkin <jlarkin@highland_atwork_technology.com> wrote:
>>
>>[...]
>>
>>> Now FF is popping up a giant ugly block in the middle of my screen, no
>>> matter what I'm doing, announcing yet another update.
>>>
>>> And it looks like they have sold out to Google.
>>
>>What version are you running?
>
> 91.0.2 64 bit. It says I'm up to date, so why the popups?

Did you modify the about:config update parameters?

--
The best designs occur in the theta state. - sw

On 31/8/21 6:03 am, Robert Baer wrote:
> Clifford Heath wrote:
>> On 30/8/21 4:56 am, Jeroen Belleman wrote:
>>> The web was supposed to be [...] reliable
>>
>> But it's made of software. We had no idea at the time that software
>> which seemed to be "mostly working" was in fact so easy to break in a
>> myriad of ways... almost all of them unavoidable now, and certainly at
>> the time.
>>
>> CH
>   Excuse me, but there are ways to write bug-free programs, ways of
> proving them, in fact there are now languages built for that.
>
>   Even 15 years ago (roughly) it was easy to make a program that had no
> "memory leaks" and no "buffer overflow"; ways to make such
> funny-business impossible.

That's hilarious, coming from a guy who still pines for Windows 95 and
wonders why he can never get his software to work.

Joe gave you the correct answer.

On 8/30/2021 3:32 PM, Joe Gwinn wrote:
> On Mon, 30 Aug 2021 13:03:23 -0700, Robert Baer
> <robertbaer@localnet.com> wrote:
>
>> Clifford Heath wrote:
>>> On 30/8/21 4:56 am, Jeroen Belleman wrote:
>>>> The web was supposed to be [...] reliable
>>>
>>> But it's made of software. We had no idea at the time that software
>>> which seemed to be "mostly working" was in fact so easy to break in a
>>> myriad of ways... almost all of them unavoidable now, and certainly at
>>> the time.
>>>
>>> CH
>> Excuse me, but there are ways to write bug-free programs, ways of
>> proving them, in fact there are now languages built for that.
>
> Yes, it's possible to approach that ideal, but it is too expensive for
> anything like a browser. And too slow by decades.

What this glosses over is that it inherently limits the rate
of evolution. If you can't reify changes in a timely manner,
then you'll not encounter many real changes!

We'd still be using *gopher* clients!

Likewise, it limits the complexity of any potential solution
to only those things that can be completely codified and tested
in some timeframe.

>> Even 15 years ago (roughly) it was easy to make a program that had no
>> "memory leaks" and no "buffer overflow"; ways to make such
>> funny-business impossible.

Why can't we design power supplies that never fail? Surely,
that's a more mature technology, right? (Oh, you can? It just
costs a lot more than you are willing to spend and ends up
physically larger than the device it seeks to power??)

And, what about things like *cars*? Surely the volume is high enough
to justify "getting it right" and then repeating the process over
and over...

> Yes, but it's quite the process.
>
> .<https://en.wikipedia.org/wiki/DO-178C>
>
> It only cost ~100 times as much as full DoD mission code.

On 8/30/21 6:57 PM, Steve Wilson wrote:
> John Larkin <jlarkin@highland_atwork_technology.com> wrote:
>
>> On Mon, 30 Aug 2021 22:46:56 -0000 (UTC), Steve Wilson
>> <spamme@not.com> wrote:
>>
>>> John Larkin <jlarkin@highland_atwork_technology.com> wrote:
>>>
>>> [...]
>>>
>>>> Now FF is popping up a giant ugly block in the middle of my
>>>> screen, no matter what I'm doing, announcing yet another
>>>> update.
>>>>
>>>> And it looks like they have sold out to Google.
>>>
>>> What version are you running?
>>
>> 91.0.2 64 bit. It says I'm up to date, so why the popups?
>
> Did you modify the about:config update parameters?

Why would you modify those individually? Twenty-two of them?

Edit > Settings (linux) , the General group, has a Firefox Updates section.

This page
https://support.mozilla.org/en-US/kb/firefox-options-preferences-and-settings?as=u&utm_source=inproduct#w_general-panel
says:
"Firefox Updates
This is where you can check your update history or change update
settings for Firefox."

There's nothing there that I can change, but that's probably because my
FF and its updates are coming thru Ubuntu repos, not from Mozilla.

John Larkin <jlarkin@highland_atwork_technology.com> wrote:

> On Mon, 30 Aug 2021 22:46:56 -0000 (UTC), Steve Wilson
> <spamme@not.com> wrote:
>
>>John Larkin <jlarkin@highland_atwork_technology.com> wrote:
>>
>>[...]
>>
>>> Now FF is popping up a giant ugly block in the middle of my screen, no
>>> matter what I'm doing, announcing yet another update.
>>>
>>> And it looks like they have sold out to Google.
>>
>>What version are you running?
>
> 91.0.2 64 bit. It says I'm up to date, so why the popups?

Unfortunately, I cannot duplicate your results as I am running 32-bit
88.0.1.

However, if 91.0.2 is similar, you should have a box in the upper right
corner marked "Show only modified preferences". Click on it and any
parameters that have changed will appear.

Most parameters have a little counterclockwise arrow in the far right. If
you click on it, the parameter will go back to the original value. This
should restore the system to the original configuration and eliminate the
problem.

If you are really brave, you can downgrade Firefox to a previous version
and see if that eliminates the problem. Just copy your firefox folder to a
backup folder in case something goes wrong. You can find previous versions
at

https://ftp.mozilla.org/pub/firefox/releases/

Firefox Setup 88.0.1.exe is at

https://ftp.mozilla.org/pub/firefox/releases/88.0.1/win64/en-US/Firefox%
20Setup%2088.0.1.exe

You may have to reinstall your profile after downgrading, so save your home
url and any extensions you have installed.

I still have not had any upgrade requests since the first one after
modifying the about:config parameters shown here:

app.update.BITS.enabled true
app.update.auto.migrated true
app.update.backgroundMaxErrors 10
app.update.badgeWaitTime 345600
app.update.channel release
app.update.checkInstallTime true
app.update.checkInstallTime.days 100000
app.update.download.attempts 0
app.update.download.promptMaxAttempts 0
app.update.elevate.attempts 0
app.update.elevation.promptMaxAttempts 0
app.update.interval 10000000
app.update.langpack.enabled true
app.update.log false
app.update.log.file false
app.update.notifyDuringDownload false
app.update.promptWaitTime 691200
app.update.service.enabled false
app.update.staging.enabled false
app.update.timerFirstInterval 10000000
app.update.timerMinimumDelay 1000000
browser.region.update.enabled false
browser.search.update.interval 1000000
browser.search.update.log false

It should be possible to tame Firefox with the right parameter. Since there
is no documentation from mozilla, we have to find it ourself. But when we
do find it, the update problem should be fixed forever.

My installation seems to be stable. No funny bugs, and no update requests.

--
The best designs occur in the theta state. - sw

On 2021-08-31 00:32, Joe Gwinn wrote:
> On Mon, 30 Aug 2021 13:03:23 -0700, Robert Baer
> <robertbaer@localnet.com> wrote:
>
>> Clifford Heath wrote:
>>> On 30/8/21 4:56 am, Jeroen Belleman wrote:
>>>> The web was supposed to be [...] reliable
>>>
>>> But it's made of software. We had no idea at the time that software
>>> which seemed to be "mostly working" was in fact so easy to break in a
>>> myriad of ways... almost all of them unavoidable now, and certainly at
>>> the time.
>>>
>>> CH
>> Excuse me, but there are ways to write bug-free programs, ways of
>> proving them, in fact there are now languages built for that.
>
> Yes, it's possible to approach that ideal, but it is too expensive for
> anything like a browser. And too slow by decades.
>
>
>> Even 15 years ago (roughly) it was easy to make a program that had no
>> "memory leaks" and no "buffer overflow"; ways to make such
>> funny-business impossible.
>
> Yes, but it's quite the process.
>
> .<https://en.wikipedia.org/wiki/DO-178C>
>
> It only cost ~100 times as much as full DoD mission code.
>
> Joe Gwinn
>

Software adheres very much to "There is never time to do it
right, but always time to do it over."

It's the same mindset that seems to think it's normal to
reboot a machine when you install new software. People
have been conditioned to accept the unacceptable.

Jeroen Belleman

On 8/31/2021 1:59 AM, Jeroen Belleman wrote:
> On 2021-08-31 00:32, Joe Gwinn wrote:
>> On Mon, 30 Aug 2021 13:03:23 -0700, Robert Baer
>> <robertbaer@localnet.com> wrote:
>>
>>> Clifford Heath wrote:
>>>> On 30/8/21 4:56 am, Jeroen Belleman wrote:
>>>>> The web was supposed to be [...] reliable
>>>>
>>>> But it's made of software. We had no idea at the time that software
>>>> which seemed to be "mostly working" was in fact so easy to break in a
>>>> myriad of ways... almost all of them unavoidable now, and certainly at
>>>> the time.
>>>>
>>>> CH
>>> Excuse me, but there are ways to write bug-free programs, ways of
>>> proving them, in fact there are now languages built for that.
>>
>> Yes, it's possible to approach that ideal, but it is too expensive for
>> anything like a browser. And too slow by decades.
>>
>>
>>> Even 15 years ago (roughly) it was easy to make a program that had no
>>> "memory leaks" and no "buffer overflow"; ways to make such
>>> funny-business impossible.
>>
>> Yes, but it's quite the process.
>>
>> .<https://en.wikipedia.org/wiki/DO-178C>
>>
>> It only cost ~100 times as much as full DoD mission code.
>
> Software adheres very much to "There is never time to do it
> right, but always time to do it over."

But is that the developer's fault? Or, the company's?

Before you answer, look at FOSS projects. In theory, they have
as much time as they choose (there are no "market pressures"
driving the financial aspects of the transaction). So, they
should be the most rigorously planned and implemented, right?

[Planning, testing, etc. are boring activities. "Let the user
tell us what's broke..."]

> It's the same mindset that seems to think it's normal to
> reboot a machine when you install new software. People
> have been conditioned to accept the unacceptable.

To some extent, yes. But, how many users are actually *confident*
in their knowledge of how a particular tool/device is *supposed*
to work?

How many have ever read a manual about a particular application?

Users boil software interactions down to a series of questions that
they pose of OTHERS, never of a manual: "How do I..." They aren't
interested in how it's *supposed* to work, just how to get it to
solve their immediate problem.

So, if something doesn't work "as it should", the user is never sure
whether they did something wrong *or* the program farted. And, if
someone else comes along and manages to get the program to do what
they'd intended, that's effectively proof that they obviously did
something wrong, previously.

I can recall arguments I've had with compiler vendors where they
try to dismiss a claim I've made against their tools. Not
easily intimidated, I refine my test case to something short and
sweet (so they can't blame it on some other aspect of the code
or wave it off as TL;DR).

"Here. Run this through your compiler and see what you get.
Then, try to explain how the generated code can *claim* to
be true to my source! Or, some excuse for my code not
being compliant with current standards. I'll just wait here,
for your reply..."

"Ooops!"