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


Groups > alt.comp.software.firefox > #16805 > unrolled thread

Unwanted update to 150 just happened

Started byThe Real Bev <bashley101@gmail.com>
First post2026-04-25 08:45 -0700
Last post2026-04-27 04:07 +0000
Articles 20 on this page of 91 — 15 participants

Back to article view | Back to alt.comp.software.firefox


Contents

  Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-25 08:45 -0700
    Re: Unwanted update to 150 just happened Retirednoguilt <HapilyRetired@fakeaddress.invalid> - 2026-04-25 13:12 -0400
      Re: Unwanted update to 150 just happened Andy Burns <usenet@andyburns.uk> - 2026-04-25 18:18 +0100
        Re: Unwanted update to 150 just happened Nobody <jock@soccer.com> - 2026-04-25 10:26 -0700
          Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-25 11:01 -0700
            Re: Unwanted update to 150 just happened Dave Royal <dave@dave123royal.com> - 2026-04-28 10:14 +0100
              Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-28 12:40 -0700
                Re: Unwanted update to 150 just happened Dave Royal <dave@dave123royal.com> - 2026-04-28 22:21 +0100
                  Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-28 19:36 -0700
                    Re: Unwanted update to 150 just happened dillinger <dillinger@invalid.not> - 2026-04-29 15:45 +0200
                      Re: Unwanted update to 150 just happened Richmond <dnomhcir@gmx.com> - 2026-04-29 15:04 +0100
                        Re: Unwanted update to 150 just happened Andy Burns <usenet@andyburns.uk> - 2026-04-29 15:10 +0100
                          Re: Unwanted update to 150 just happened Richmond <dnomhcir@gmx.com> - 2026-04-29 15:21 +0100
                            Re: Unwanted update to 150 just happened dillinger <dillinger@invalid.not> - 2026-04-29 17:26 +0200
                            Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-29 19:41 -0700
                      Re: Unwanted update to 150 just happened Richmond <dnomhcir@gmx.com> - 2026-04-29 16:47 +0100
                        Re: Unwanted update to 150 just happened Andy Burns <usenet@andyburns.uk> - 2026-04-29 16:52 +0100
                          Re: Unwanted update to 150 just happened Richmond <dnomhcir@gmx.com> - 2026-04-29 16:54 +0100
                            Re: Unwanted update to 150 just happened Andy Burns <usenet@andyburns.uk> - 2026-04-29 17:00 +0100
                              Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-05-02 05:41 +0000
                        Re: Unwanted update to 150 just happened dillinger <dillinger@invalid.not> - 2026-04-29 18:11 +0200
                          Re: Unwanted update to 150 just happened Dave Royal <dave@dave123royal.com> - 2026-04-29 20:20 +0100
                            Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-29 20:30 -0700
                              Re: Unwanted update to 150 just happened Richmond <dnomhcir@gmx.com> - 2026-04-30 09:10 +0100
                                Re: Unwanted update to 150 just happened Andy Burns <usenet@andyburns.uk> - 2026-04-30 10:03 +0100
                              Re: Unwanted update to 150 just happened dillinger <dillinger@invalid.not> - 2026-04-30 05:48 +0200
                                Re: Unwanted update to 150 just happened Richmond <dnomhcir@gmx.com> - 2026-04-30 14:23 +0100
                                Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-30 20:16 -0700
                                  Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-05-03 21:47 +0000
                                    Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-03 19:33 -0700
                              Re: Unwanted update to 150 just happened Dave Royal <dave@dave123royal.com> - 2026-05-01 07:04 +0000
                                Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-01 09:33 -0700
                                  Re: Unwanted update to 150 just happened Dave Royal <dave@dave123royal.com> - 2026-05-01 18:00 +0100
                                  Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-05-03 21:51 +0000
                                Re: Unwanted update to 150 just happened Dave Royal <dave@dave123royal.com> - 2026-05-03 11:52 +0000
                    Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-05-01 08:09 +0000
                      Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-01 09:17 -0700
                        Re: Unwanted update to 150 just happened Frank Miller <miller@posteo.ee> - 2026-05-01 19:10 +0200
                          Re: Unwanted update to 150 just happened R Daneel Olivaw <Danni@hyperspace.vogon.gov.invalid> - 2026-05-01 20:16 +0200
                            Re: Unwanted update to 150 just happened Nobody <jock@soccer.com> - 2026-05-01 12:57 -0700
                              Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-02 11:58 -0700
                                Re: Unwanted update to 150 just happened Frank Miller <miller@posteo.ee> - 2026-05-02 21:15 +0200
                                  Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-02 12:53 -0700
                                    Re: Unwanted update to 150 just happened Frank Miller <miller@posteo.ee> - 2026-05-02 22:25 +0200
                                      Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-02 22:10 -0700
                                Re: Unwanted update to 150 just happened Nobody <jock@soccer.com> - 2026-05-02 16:40 -0700
                                  Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-05-02 22:01 -0700
                                    Re: Unwanted update to 150 just happened Nobody <jock@soccer.com> - 2026-05-03 18:49 -0700
                            Re: Unwanted update to 150 just happened Frank Miller <miller@posteo.ee> - 2026-05-01 22:09 +0200
                        Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-05-02 12:08 +1200
                          Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-05-02 00:11 +0000
    Re: Unwanted update to 150 just happened R Daneel Olivaw <Danni@hyperspace.vogon.gov.invalid> - 2026-04-25 20:13 +0200
      Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-25 11:25 -0700
        Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-25 11:35 -0700
          Re: Unwanted update to 150 just happened "Alan K." <alan@invalid.com> - 2026-04-25 14:56 -0400
            Re: Unwanted update to 150 just happened -- SOLVED! The Real Bev <bashley101@gmail.com> - 2026-04-25 12:41 -0700
              Re: Unwanted update to 150 just happened -- SOLVED! "Alan K." <alan@invalid.com> - 2026-04-25 15:56 -0400
              Re: Unwanted update to 150 just happened -- SOLVED! Frank Miller <miller@posteo.ee> - 2026-04-25 23:03 +0200
                Re: Unwanted update to 150 just happened -- SOLVED! The Real Bev <bashley101@gmail.com> - 2026-04-25 14:51 -0700
                  Re: Unwanted update to 150 just happened -- SOLVED! "Alan K." <alan@invalid.com> - 2026-04-25 18:12 -0400
                    Re: Unwanted update to 150 just happened -- SOLVED! The Real Bev <bashley101@gmail.com> - 2026-04-25 22:13 -0700
            Re: Unwanted update to 150 just happened R Daneel Olivaw <Danni@hyperspace.vogon.gov.invalid> - 2026-04-26 10:12 +0200
        Re: Unwanted update to 150 just happened Frank Miller <miller@posteo.ee> - 2026-04-25 21:40 +0200
    Re: Unwanted update to 150 just happened VanguardLH <V@nguard.LH> - 2026-04-25 14:21 -0500
    Re: Unwanted update to 150 just happened -- SOLVED! The Real Bev <bashley101@gmail.com> - 2026-04-25 12:45 -0700
      Re: Unwanted update to 150 just happened -- SOLVED! The Real Bev <bashley101@gmail.com> - 2026-04-25 13:01 -0700
    Re: Unwanted update to 150 just happened micky <NONONOmisc07@fmguy.com> - 2026-04-25 19:58 -0400
      Re: Unwanted update to 150 just happened Retirednoguilt <HapilyRetired@fakeaddress.invalid> - 2026-04-26 07:42 -0400
        Re: Unwanted update to 150 just happened R Daneel Olivaw <Danni@hyperspace.vogon.gov.invalid> - 2026-04-26 18:46 +0200
          Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-26 12:11 -0700
            Re: Unwanted update to 150 just happened Nobody <jock@soccer.com> - 2026-04-26 15:47 -0700
              Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-26 17:05 -0700
                Re: Unwanted update to 150 just happened Nobody <jock@soccer.com> - 2026-04-26 18:49 -0700
                  Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-26 21:48 -0700
                    Re: Unwanted update to 150 just happened Retirednoguilt <HapilyRetired@fakeaddress.invalid> - 2026-04-27 13:42 -0400
                      Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-27 11:28 -0700
                        Re: Unwanted update to 150 just happened Retirednoguilt <HapilyRetired@fakeaddress.invalid> - 2026-04-27 16:01 -0400
                          Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-28 12:45 -0700
                        Re: Unwanted update to 150 just happened "Adam H. Kerman" <ahk@chinet.com> - 2026-04-28 00:23 +0000
                      Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-04-28 08:36 +0000
                        Re: Unwanted update to 150 just happened "Adam H. Kerman" <ahk@chinet.com> - 2026-04-28 12:35 +0000
                        Re: Unwanted update to 150 just happened Retirednoguilt <HapilyRetired@fakeaddress.invalid> - 2026-04-28 15:42 -0400
                    Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-04-28 01:14 +0000
                      Re: Unwanted update to 150 just happened "Adam H. Kerman" <ahk@chinet.com> - 2026-04-28 02:51 +0000
                        Re: Unwanted update to 150 just happened The Real Bev <bashley101@gmail.com> - 2026-04-28 12:51 -0700
                          Re: Unwanted update to 150 just happened Char Jackson <none@none.invalid> - 2026-05-08 15:06 -0500
                Re: Unwanted update to 150 just happened micky <NONONOmisc07@fmguy.com> - 2026-04-26 22:27 -0400
                Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-04-28 01:12 +0000
            Re: Unwanted update to 150 just happened Retirednoguilt <HapilyRetired@fakeaddress.invalid> - 2026-04-27 13:24 -0400
            Re: Unwanted update to 150 just happened R Daneel Olivaw <Danni@hyperspace.vogon.gov.invalid> - 2026-04-28 09:15 +0200
        Re: Unwanted update to 150 just happened Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-04-27 04:07 +0000

Page 2 of 5 — ← Prev page 1 [2] 3 4 5  Next page →


#16950

Fromdillinger <dillinger@invalid.not>
Date2026-04-29 18:11 +0200
Message-ID<a7n9cm-arov3.ln1@spock.lan>
In reply to#16946
On 29/04/2026 17:47, Richmond wrote:
> dillinger <dillinger@invalid.not> writes:
> 
>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>> On 4/28/26 14:21, Dave Royal wrote:
>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>
>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>
>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns <usenet@andyburns.uk>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>
>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>
>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>> install them"
>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>
>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install' then that
>>>>>>>> radio button changes to 'automatically install/recommended'.
>>>>>>>>
>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates but
>>>>>>>> lemme
>>>>>>>> do the work' setting.
>>>>>>>
>>>>>>> I have had that setting for decades -- and still do.  Some years
>>>>>>> ago it stopped working and some kind person posted the
>>>>>>> policies.json fix -- which worked until today.  Posted herewith:
>>>>>>>
>>>>>>> Make a new firefox/distribution/ subdirectory which will contain
>>>>>>> the file policies.json which contains
>>>>>>> {
>>>>>>>     "policies": {
>>>>>>>       "DisableAppUpdate": true
>>>>>>>     }
>>>>>>> }
>>>>>>
>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>    download, or did you create one - on Linux?
>>>>>
>>>>> Someone posted about that file long ago, and the one I created with
>>>>> pico worked for years.
>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>    problem - got there.
>>>>> They appeared when Firefox chose to update itself against my express
>>>>> wishes.  I suspect that Firefox did it of its own volition -- or
>>>>> simple clumsiness.  Core dump complete.
>>>>
>>>> You /found/ them after the unwanted update. Were they responsible
>>>>    for it? I doubt it: line endings - of either type - are ignored
>>>>    in a json file.
>>>
>>> The windows endings damaged the file such that it no longer served the
>>> purpose.  Fixing it worked.
>>>
>>
>> Firefox doesn't care about line endings.
>>
>> Just for the sake of it I tested it for you, I created a policies.json
>> which blocks about:config and tested it with both CRLF (windows) and LF
>> (unix) line endings.
>>
>> The result are the same for both, it works:
>> Blocked Page, Firefox can’t connect to the server at about:config
>>
>> ^M is not a corruption, this is how cat -A shows CR, your AI needs an
>> update.
>>
>> CRLF (windows):
>> cat -A policies.json
>> {^M$
>>    "policies": {^M$
>>      "BlockAboutConfig": true^M$
>>    }^M$
>> }^M$
>>
>> LF (unix):
>> cat -A policies.json
>> {$
>>    "policies": {$
>>      "BlockAboutConfig": true$
>>    }$
>> }$
> 
> You tested whether it blocked about:config, but you didn't test whether
> it blocked an update, or if the update altered the json file.

I tested policies.json, which was reported corrupted, with both CRLF and 
LF, any policy will do, we don't need to wait for the next update.
Policies.json is not included by default, updates will either not touch 
it at all or remove it completely, depending on how you update.
Since we can't see the original file there is no way to tell how it was 
changed.

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


#16951

FromDave Royal <dave@dave123royal.com>
Date2026-04-29 20:20 +0100
Message-ID<10stlmh$51ep$1@dont-email.me>
In reply to#16950
dillinger <dillinger@invalid.not> Wrote in message:

> On 29/04/2026 17:47, Richmond wrote:
>> dillinger <dillinger@invalid.not> writes:
>> 
>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>
>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>
>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns <usenet@andyburns.uk>
>>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>
>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>
>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>> install them"
>>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>>
>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install' then that
>>>>>>>>> radio button changes to 'automatically install/recommended'.
>>>>>>>>>
>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates but
>>>>>>>>> lemme
>>>>>>>>> do the work' setting.
>>>>>>>>
>>>>>>>> I have had that setting for decades -- and still do.  Some years
>>>>>>>> ago it stopped working and some kind person posted the
>>>>>>>> policies.json fix -- which worked until today.  Posted herewith:
>>>>>>>>
>>>>>>>> Make a new firefox/distribution/ subdirectory which will contain
>>>>>>>> the file policies.json which contains
>>>>>>>> {
>>>>>>>>     "policies": {
>>>>>>>>       "DisableAppUpdate": true
>>>>>>>>     }
>>>>>>>> }
>>>>>>>
>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>    download, or did you create one - on Linux?
>>>>>>
>>>>>> Someone posted about that file long ago, and the one I created with
>>>>>> pico worked for years.
>>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>>    problem - got there.
>>>>>> They appeared when Firefox chose to update itself against my express
>>>>>> wishes.  I suspect that Firefox did it of its own volition -- or
>>>>>> simple clumsiness.  Core dump complete.
>>>>>
>>>>> You /found/ them after the unwanted update. Were they responsible
>>>>>    for it? I doubt it: line endings - of either type - are ignored
>>>>>    in a json file.
>>>>
>>>> The windows endings damaged the file such that it no longer served the
>>>> purpose.  Fixing it worked.
>>>>
>>>
>>> Firefox doesn't care about line endings.
>>>
>>> Just for the sake of it I tested it for you, I created a policies.json
>>> which blocks about:config and tested it with both CRLF (windows) and LF
>>> (unix) line endings.
>>>
>>> The result are the same for both, it works:
>>> Blocked Page, Firefox can’t connect to the server at about:config
>>>
>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs an
>>> update.
>>>
>>> CRLF (windows):
>>> cat -A policies.json
>>> {^M$
>>>    "policies": {^M$
>>>      "BlockAboutConfig": true^M$
>>>    }^M$
>>> }^M$
>>>
>>> LF (unix):
>>> cat -A policies.json
>>> {$
>>>    "policies": {$
>>>      "BlockAboutConfig": true$
>>>    }$
>>> }$
>> 
>> You tested whether it blocked about:config, but you didn't test whether
>> it blocked an update, or if the update altered the json file.
> 
> I tested policies.json, which was reported corrupted, with both CRLF and 
> LF, any policy will do, we don't need to wait for the next update.
> Policies.json is not included by default, updates will either not touch 
> it at all or remove it completely, depending on how you update.
> Since we can't see the original file there is no way to tell how it was 
> changed.

Why might an update remove policies.json? I assume that the
 firefox updater will read it if it's there but never write to it.
 The point of it is that it's user, or organisation,
 controlled.

I did wonder if updates might /update/ policies.json, for changes
 in format say, in the way they update a profile. But I doubt
 it.

On Debian I have tarballs of Fx nightly and TB beta, in addtion to
 packaged versions for regular use. They do not include a
 distribution/ directory or a template policies.json.
-- 
Remove numerics from my email address.

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


#16958

FromThe Real Bev <bashley101@gmail.com>
Date2026-04-29 20:30 -0700
Message-ID<10suick$c4n8$2@dont-email.me>
In reply to#16951
On 4/29/26 12:20, Dave Royal wrote:
> dillinger <dillinger@invalid.not> Wrote in message:
> 
>> On 29/04/2026 17:47, Richmond wrote:
>>> dillinger <dillinger@invalid.not> writes:
>>> 
>>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>
>>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>
>>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns <usenet@andyburns.uk>
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>>
>>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>>
>>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>>> install them"
>>>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>>>
>>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install' then that
>>>>>>>>>> radio button changes to 'automatically install/recommended'.
>>>>>>>>>>
>>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates but
>>>>>>>>>> lemme
>>>>>>>>>> do the work' setting.
>>>>>>>>>
>>>>>>>>> I have had that setting for decades -- and still do.  Some years
>>>>>>>>> ago it stopped working and some kind person posted the
>>>>>>>>> policies.json fix -- which worked until today.  Posted herewith:
>>>>>>>>>
>>>>>>>>> Make a new firefox/distribution/ subdirectory which will contain
>>>>>>>>> the file policies.json which contains
>>>>>>>>> {
>>>>>>>>>     "policies": {
>>>>>>>>>       "DisableAppUpdate": true
>>>>>>>>>     }
>>>>>>>>> }
>>>>>>>>
>>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>>    download, or did you create one - on Linux?
>>>>>>>
>>>>>>> Someone posted about that file long ago, and the one I created with
>>>>>>> pico worked for years.
>>>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>>>    problem - got there.
>>>>>>> They appeared when Firefox chose to update itself against my express
>>>>>>> wishes.  I suspect that Firefox did it of its own volition -- or
>>>>>>> simple clumsiness.  Core dump complete.
>>>>>>
>>>>>> You /found/ them after the unwanted update. Were they responsible
>>>>>>    for it? I doubt it: line endings - of either type - are ignored
>>>>>>    in a json file.
>>>>>
>>>>> The windows endings damaged the file such that it no longer served the
>>>>> purpose.  Fixing it worked.
>>>>>
>>>>
>>>> Firefox doesn't care about line endings.
>>>>
>>>> Just for the sake of it I tested it for you, I created a policies.json
>>>> which blocks about:config and tested it with both CRLF (windows) and LF
>>>> (unix) line endings.
>>>>
>>>> The result are the same for both, it works:
>>>> Blocked Page, Firefox can’t connect to the server at about:config
>>>>
>>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs an
>>>> update.
>>>>
>>>> CRLF (windows):
>>>> cat -A policies.json
>>>> {^M$
>>>>    "policies": {^M$
>>>>      "BlockAboutConfig": true^M$
>>>>    }^M$
>>>> }^M$
>>>>
>>>> LF (unix):
>>>> cat -A policies.json
>>>> {$
>>>>    "policies": {$
>>>>      "BlockAboutConfig": true$
>>>>    }$
>>>> }$
>>> 
>>> You tested whether it blocked about:config, but you didn't test whether
>>> it blocked an update, or if the update altered the json file.
>> 
>> I tested policies.json, which was reported corrupted, with both CRLF and 
>> LF, any policy will do, we don't need to wait for the next update.
>> Policies.json is not included by default, updates will either not touch 
>> it at all or remove it completely, depending on how you update.
>> Since we can't see the original file there is no way to tell how it was 
>> changed.
> 
> Why might an update remove policies.json? I assume that the
>   firefox updater will read it if it's there but never write to it.
>   The point of it is that it's user, or organisation,
>   controlled.
> 
> I did wonder if updates might /update/ policies.json, for changes
>   in format say, in the way they update a profile. But I doubt
>   it.
> 
> On Debian I have tarballs of Fx nightly and TB beta, in addtion to
>   packaged versions for regular use. They do not include a
>   distribution/ directory or a template policies.json.

No, they're not included with the tarball.  "Some kind person" posted 
the fix at some point in the past (earliest I have is Firefox 72, July 
7,2020) posted how to do it.  If I want to use a newer version I don't 
update.  I download the tarball, untar it into its own subdirectory and 
run it from within that subdirectory (/blabla/firefox/firefox -P &). 
After running it once I exit and copy over (cp -rf *) my previous 
profile into the newly created profile.  Accordingly, I have at least a 
dozen fully functional firefoxes.  TMI, but there it is.

Worked fine until somebody said that FF148 had some sort of security 
problem, at which point I did an update to 149 rather than building a 
whole new installation.  I think I renamed the policies.json file, which 
would have been the simplest thing. Sorry I can't be more certain.  Then 
I would have renamed it back.  And then I got the unwanted surprise of 
the 150 update.  I turned the problem over to Perplexity, which provided 
the solution.  If anyone is interested I can have Perplexity generate a 
nice pdf file of the entire interchange.

I assume that firefox does some variant of cp -rf * when it updates and 
then does some profile-creation stuff and god knows what else.  It must 
also use some touch-like command, becuase the date of the policies.json 
file was the same as the date on all the other FF 150 subdirectories and 
files. I don't remember looking at the FF149/firefox files because I 
would have had no reason to. And that must be when the mischief occurred.

AH-HAH!  At some point when I first heard about the policies.json thing 
I was unsure about where to put it, so I put one in the /firefox 
subdirectory.  It's still there, and contains the ^M characters. They 
seem to be M- instead of ^M, for what it's worth. Its date is the same 
as all the others -- April 24 19:37.

cat -A policies.json
{$
M-  "policies": {$
M-  M-  "DisableAppUpdate": true$
M-  }$
}$
$

-- 
Cheers, Bev
   "Nothing in the universe can withstand the relentless application
    of brute force and ignorance."         -- Frd, via Dennis (evil)

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


#16964

FromRichmond <dnomhcir@gmx.com>
Date2026-04-30 09:10 +0100
Message-ID<82jyto27il.fsf@example.com>
In reply to#16958
The Real Bev <bashley101@gmail.com> writes:

> On 4/29/26 12:20, Dave Royal wrote:
>> dillinger <dillinger@invalid.not> Wrote in message:
>> 
>>> On 29/04/2026 17:47, Richmond wrote:
>>>> dillinger <dillinger@invalid.not> writes:
>>>> 
>>>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>
>>>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>>
>>>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns
>>>>>>>>>>> <usenet@andyburns.uk>
>>>>>>>>>>> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>>>
>>>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>>>> install them" and as best as I can remember, its never
>>>>>>>>>>>> disobeyed ...
>>>>>>>>>>>
>>>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install'
>>>>>>>>>>> then that radio button changes to 'automatically
>>>>>>>>>>> install/recommended'.
>>>>>>>>>>>
>>>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates
>>>>>>>>>>> but lemme do the work' setting.
>>>>>>>>>>
>>>>>>>>>> I have had that setting for decades -- and still do.  Some
>>>>>>>>>> years ago it stopped working and some kind person posted the
>>>>>>>>>> policies.json fix -- which worked until today.  Posted
>>>>>>>>>> herewith:
>>>>>>>>>>
>>>>>>>>>> Make a new firefox/distribution/ subdirectory which will
>>>>>>>>>> contain the file policies.json which contains { "policies": {
>>>>>>>>>> "DisableAppUpdate": true } }
>>>>>>>>>
>>>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>>>    download, or did you create one - on Linux?
>>>>>>>>
>>>>>>>> Someone posted about that file long ago, and the one I created
>>>>>>>> with pico worked for years.  > Just wondering how these Windows
>>>>>>>> line endings - if that was the > problem - got there.  They
>>>>>>>> appeared when Firefox chose to update itself against my express
>>>>>>>> wishes.  I suspect that Firefox did it of its own volition --
>>>>>>>> or simple clumsiness.  Core dump complete.
>>>>>>>
>>>>>>> You /found/ them after the unwanted update. Were they
>>>>>>>    responsible for it? I doubt it: line endings - of either type
>>>>>>>    - are ignored in a json file.
>>>>>>
>>>>>> The windows endings damaged the file such that it no longer
>>>>>> served the purpose.  Fixing it worked.
>>>>>>
>>>>>
>>>>> Firefox doesn't care about line endings.
>>>>>
>>>>> Just for the sake of it I tested it for you, I created a
>>>>> policies.json which blocks about:config and tested it with both
>>>>> CRLF (windows) and LF (unix) line endings.
>>>>>
>>>>> The result are the same for both, it works: Blocked Page, Firefox
>>>>> can’t connect to the server at about:config
>>>>>
>>>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs
>>>>> an update.
>>>>>
>>>>> CRLF (windows): cat -A policies.json {^M$ "policies": {^M$
>>>>> "BlockAboutConfig": true^M$ }^M$ }^M$
>>>>>
>>>>> LF (unix): cat -A policies.json {$ "policies": {$
>>>>> "BlockAboutConfig": true$ }$ }$ >>>> You tested whether it blocked
>>>>> about:config, but you didn't test >>>> whether >>>> it blocked an
>>>>> update, or if the update altered the json file.  >>> I tested
>>>>> policies.json, which was reported corrupted, with both >>> CRLF
>>>>> and LF, any policy will do, we don't need to wait for the next >>>
>>>>> update.  >>> Policies.json is not included by default, updates
>>>>> will either not >>> touch it at all or remove it completely,
>>>>> depending on how you >>> update.  >>> Since we can't see the
>>>>> original file there is no way to tell how it >>> was changed.  >>
>>>>> Why might an update remove policies.json? I assume that the >>
>>>>> firefox updater will read it if it's there but never write to it.
>>>>> >> The point of it is that it's user, or organisation, >>
>>>>> controlled.  >> I did wonder if updates might /update/
>>>>> policies.json, for changes >> in format say, in the way they
>>>>> update a profile. But I doubt >> it.  >> On Debian I have tarballs
>>>>> of Fx nightly and TB beta, in addtion to >> packaged versions for
>>>>> regular use. They do not include a >> distribution/ directory or a
>>>>> template policies.json.
>
> No, they're not included with the tarball.  "Some kind person" posted
> the fix at some point in the past (earliest I have is Firefox 72, July
> 7,2020) posted how to do it.  If I want to use a newer version I don't
> update.  I download the tarball, untar it into its own subdirectory
> and run it from within that subdirectory (/blabla/firefox/firefox -P
> &). After running it once I exit and copy over (cp -rf *) my previous
> profile into the newly created profile.  Accordingly, I have at least
> a dozen fully functional firefoxes.  TMI, but there it is.
>
> Worked fine until somebody said that FF148 had some sort of security
> problem, at which point I did an update to 149 rather than building a
> whole new installation.  I think I renamed the policies.json file,
> which would have been the simplest thing. Sorry I can't be more
> certain.  Then I would have renamed it back.  And then I got the
> unwanted surprise of the 150 update.  I turned the problem over to
> Perplexity, which provided the solution.  If anyone is interested I
> can have Perplexity generate a nice pdf file of the entire
> interchange.
>
> I assume that firefox does some variant of cp -rf * when it updates
> and then does some profile-creation stuff and god knows what else.  It
> must also use some touch-like command, becuase the date of the
> policies.json file was the same as the date on all the other FF 150
> subdirectories and files. I don't remember looking at the
> FF149/firefox files because I would have had no reason to. And that
> must be when the mischief occurred.
>
> AH-HAH!  At some point when I first heard about the policies.json
> thing I was unsure about where to put it, so I put one in the /firefox
> subdirectory.  It's still there, and contains the ^M characters. They
> seem to be M- instead of ^M, for what it's worth. Its date is the same
> as all the others -- April 24 19:37.
>
> cat -A policies.json {$ M- "policies": {$ M- M- "DisableAppUpdate":
> true$ M- }$ }$ $

The plot thickens. Here is what my favourite AI said:

"In your case, the indentation uses UTF-8 encoded non-breaking spaces
(or similar Unicode whitespace) instead of regular ASCII spaces
(0x20). For example, M- is likely a non-breaking space (U+00A0, encoded
as 0xC2 0xA0 in UTF-8).

Why this matters for a JSON file: Most JSON parsers only treat ASCII
whitespace (regular space, tab, newline) as insignificant
whitespace. Non-breaking spaces can cause parse errors depending on the
parser.  "

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


#16966

FromAndy Burns <usenet@andyburns.uk>
Date2026-04-30 10:03 +0100
Message-ID<n5gk7aFqjvdU2@mid.individual.net>
In reply to#16964
Richmond wrote:

> Most JSON parsers only treat ASCII
> whitespace (regular space, tab, newline) as insignificant
> whitespace. Non-breaking spaces can cause parse errors depending on the
> parser.

Yeah, that'd b0rk it ...

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


#16970

Fromdillinger <dillinger@invalid.not>
Date2026-04-30 05:48 +0200
Message-ID<p10bcm-jq3.ln1@spock.lan>
In reply to#16958
On 30/04/2026 05:30, The Real Bev wrote:
> On 4/29/26 12:20, Dave Royal wrote:
>> dillinger <dillinger@invalid.not> Wrote in message:
>>
>>> On 29/04/2026 17:47, Richmond wrote:
>>>> dillinger <dillinger@invalid.not> writes:
>>>>
>>>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>
>>>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>>
>>>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns 
>>>>>>>>>>> <usenet@andyburns.uk>
>>>>>>>>>>> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>>>
>>>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>>>> install them"
>>>>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>>>>
>>>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install' 
>>>>>>>>>>> then that
>>>>>>>>>>> radio button changes to 'automatically install/recommended'.
>>>>>>>>>>>
>>>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates but
>>>>>>>>>>> lemme
>>>>>>>>>>> do the work' setting.
>>>>>>>>>>
>>>>>>>>>> I have had that setting for decades -- and still do.  Some years
>>>>>>>>>> ago it stopped working and some kind person posted the
>>>>>>>>>> policies.json fix -- which worked until today.  Posted herewith:
>>>>>>>>>>
>>>>>>>>>> Make a new firefox/distribution/ subdirectory which will contain
>>>>>>>>>> the file policies.json which contains
>>>>>>>>>> {
>>>>>>>>>>     "policies": {
>>>>>>>>>>       "DisableAppUpdate": true
>>>>>>>>>>     }
>>>>>>>>>> }
>>>>>>>>>
>>>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>>>    download, or did you create one - on Linux?
>>>>>>>>
>>>>>>>> Someone posted about that file long ago, and the one I created with
>>>>>>>> pico worked for years.
>>>>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>>>>    problem - got there.
>>>>>>>> They appeared when Firefox chose to update itself against my 
>>>>>>>> express
>>>>>>>> wishes.  I suspect that Firefox did it of its own volition -- or
>>>>>>>> simple clumsiness.  Core dump complete.
>>>>>>>
>>>>>>> You /found/ them after the unwanted update. Were they responsible
>>>>>>>    for it? I doubt it: line endings - of either type - are ignored
>>>>>>>    in a json file.
>>>>>>
>>>>>> The windows endings damaged the file such that it no longer served 
>>>>>> the
>>>>>> purpose.  Fixing it worked.
>>>>>>
>>>>>
>>>>> Firefox doesn't care about line endings.
>>>>>
>>>>> Just for the sake of it I tested it for you, I created a policies.json
>>>>> which blocks about:config and tested it with both CRLF (windows) 
>>>>> and LF
>>>>> (unix) line endings.
>>>>>
>>>>> The result are the same for both, it works:
>>>>> Blocked Page, Firefox can’t connect to the server at about:config
>>>>>
>>>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs an
>>>>> update.
>>>>>
>>>>> CRLF (windows):
>>>>> cat -A policies.json
>>>>> {^M$
>>>>>    "policies": {^M$
>>>>>      "BlockAboutConfig": true^M$
>>>>>    }^M$
>>>>> }^M$
>>>>>
>>>>> LF (unix):
>>>>> cat -A policies.json
>>>>> {$
>>>>>    "policies": {$
>>>>>      "BlockAboutConfig": true$
>>>>>    }$
>>>>> }$
>>>>
>>>> You tested whether it blocked about:config, but you didn't test whether
>>>> it blocked an update, or if the update altered the json file.
>>>
>>> I tested policies.json, which was reported corrupted, with both CRLF 
>>> and LF, any policy will do, we don't need to wait for the next update.
>>> Policies.json is not included by default, updates will either not 
>>> touch it at all or remove it completely, depending on how you update.
>>> Since we can't see the original file there is no way to tell how it 
>>> was changed.
>>
>> Why might an update remove policies.json? I assume that the
>>   firefox updater will read it if it's there but never write to it.
>>   The point of it is that it's user, or organisation,
>>   controlled.
>>
>> I did wonder if updates might /update/ policies.json, for changes
>>   in format say, in the way they update a profile. But I doubt
>>   it.
>>
>> On Debian I have tarballs of Fx nightly and TB beta, in addtion to
>>   packaged versions for regular use. They do not include a
>>   distribution/ directory or a template policies.json.
> 
> No, they're not included with the tarball.  "Some kind person" posted 
> the fix at some point in the past (earliest I have is Firefox 72, July 
> 7,2020) posted how to do it.  If I want to use a newer version I don't 
> update.  I download the tarball, untar it into its own subdirectory and 
> run it from within that subdirectory (/blabla/firefox/firefox -P &). 
> After running it once I exit and copy over (cp -rf *) my previous 
> profile into the newly created profile.  Accordingly, I have at least a 
> dozen fully functional firefoxes.  TMI, but there it is.
> 
> Worked fine until somebody said that FF148 had some sort of security 
> problem, at which point I did an update to 149 rather than building a 
> whole new installation.  I think I renamed the policies.json file, which 
> would have been the simplest thing. Sorry I can't be more certain.  Then 
> I would have renamed it back.  And then I got the unwanted surprise of 
> the 150 update.  I turned the problem over to Perplexity, which provided 
> the solution.  If anyone is interested I can have Perplexity generate a 
> nice pdf file of the entire interchange.
> 
> I assume that firefox does some variant of cp -rf * when it updates and 
> then does some profile-creation stuff and god knows what else.  It must 
> also use some touch-like command, becuase the date of the policies.json 
> file was the same as the date on all the other FF 150 subdirectories and 
> files. I don't remember looking at the FF149/firefox files because I 
> would have had no reason to. And that must be when the mischief occurred.
> 
> AH-HAH!  At some point when I first heard about the policies.json thing 
> I was unsure about where to put it, so I put one in the /firefox 
> subdirectory.  It's still there, and contains the ^M characters. They 
> seem to be M- instead of ^M, for what it's worth. Its date is the same 
> as all the others -- April 24 19:37.
> 
> cat -A policies.json
> {$
> M-  "policies": {$
> M-  M-  "DisableAppUpdate": true$
> M-  }$
> }$
> $
> 

Why cat -A? Check man cat.
In the output of cat -A "M-" is a tab, "^M" is a CR and "$" is a LF
If you copy those literally into a script it will very likely produce an 
error.

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


#16972

FromRichmond <dnomhcir@gmx.com>
Date2026-04-30 14:23 +0100
Message-ID<82ecjwintw.fsf@example.com>
In reply to#16970
dillinger <dillinger@invalid.not> writes:

> On 30/04/2026 05:30, The Real Bev wrote:
>> On 4/29/26 12:20, Dave Royal wrote:
>>> dillinger <dillinger@invalid.not> Wrote in message:
>>>
>>>> On 29/04/2026 17:47, Richmond wrote:
>>>>> dillinger <dillinger@invalid.not> writes:
>>>>>
>>>>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>
>>>>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>>>
>>>>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns
>>>>>>>>>>>> <usenet@andyburns.uk>
>>>>>>>>>>>> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>>>>
>>>>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>>>>> install them"
>>>>>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>>>>>
>>>>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install'
>>>>>>>>>>>> then that
>>>>>>>>>>>> radio button changes to 'automatically install/recommended'.
>>>>>>>>>>>>
>>>>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates but
>>>>>>>>>>>> lemme
>>>>>>>>>>>> do the work' setting.
>>>>>>>>>>>
>>>>>>>>>>> I have had that setting for decades -- and still do.  Some years
>>>>>>>>>>> ago it stopped working and some kind person posted the
>>>>>>>>>>> policies.json fix -- which worked until today.  Posted herewith:
>>>>>>>>>>>
>>>>>>>>>>> Make a new firefox/distribution/ subdirectory which will contain
>>>>>>>>>>> the file policies.json which contains
>>>>>>>>>>> {
>>>>>>>>>>>     "policies": {
>>>>>>>>>>>       "DisableAppUpdate": true
>>>>>>>>>>>     }
>>>>>>>>>>> }
>>>>>>>>>>
>>>>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>>>>    download, or did you create one - on Linux?
>>>>>>>>>
>>>>>>>>> Someone posted about that file long ago, and the one I created with
>>>>>>>>> pico worked for years.
>>>>>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>>>>>    problem - got there.
>>>>>>>>> They appeared when Firefox chose to update itself against my
>>>>>>>>> express
>>>>>>>>> wishes.  I suspect that Firefox did it of its own volition -- or
>>>>>>>>> simple clumsiness.  Core dump complete.
>>>>>>>>
>>>>>>>> You /found/ them after the unwanted update. Were they responsible
>>>>>>>>    for it? I doubt it: line endings - of either type - are ignored
>>>>>>>>    in a json file.
>>>>>>>
>>>>>>> The windows endings damaged the file such that it no longer
>>>>>>> served the
>>>>>>> purpose.  Fixing it worked.
>>>>>>>
>>>>>>
>>>>>> Firefox doesn't care about line endings.
>>>>>>
>>>>>> Just for the sake of it I tested it for you, I created a policies.json
>>>>>> which blocks about:config and tested it with both CRLF (windows)
>>>>>> and LF
>>>>>> (unix) line endings.
>>>>>>
>>>>>> The result are the same for both, it works:
>>>>>> Blocked Page, Firefox can’t connect to the server at about:config
>>>>>>
>>>>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs an
>>>>>> update.
>>>>>>
>>>>>> CRLF (windows):
>>>>>> cat -A policies.json
>>>>>> {^M$
>>>>>>    "policies": {^M$
>>>>>>      "BlockAboutConfig": true^M$
>>>>>>    }^M$
>>>>>> }^M$
>>>>>>
>>>>>> LF (unix):
>>>>>> cat -A policies.json
>>>>>> {$
>>>>>>    "policies": {$
>>>>>>      "BlockAboutConfig": true$
>>>>>>    }$
>>>>>> }$
>>>>>
>>>>> You tested whether it blocked about:config, but you didn't test whether
>>>>> it blocked an update, or if the update altered the json file.
>>>>
>>>> I tested policies.json, which was reported corrupted, with both
>>>> CRLF and LF, any policy will do, we don't need to wait for the
>>>> next update.
>>>> Policies.json is not included by default, updates will either not
>>>> touch it at all or remove it completely, depending on how you
>>>> update.
>>>> Since we can't see the original file there is no way to tell how
>>>> it was changed.
>>>
>>> Why might an update remove policies.json? I assume that the
>>>   firefox updater will read it if it's there but never write to it.
>>>   The point of it is that it's user, or organisation,
>>>   controlled.
>>>
>>> I did wonder if updates might /update/ policies.json, for changes
>>>   in format say, in the way they update a profile. But I doubt
>>>   it.
>>>
>>> On Debian I have tarballs of Fx nightly and TB beta, in addtion to
>>>   packaged versions for regular use. They do not include a
>>>   distribution/ directory or a template policies.json.
>> No, they're not included with the tarball.  "Some kind person"
>> posted the fix at some point in the past (earliest I have is Firefox
>> 72, July 7,2020) posted how to do it.  If I want to use a newer
>> version I don't update.  I download the tarball, untar it into its
>> own subdirectory and run it from within that subdirectory
>> (/blabla/firefox/firefox -P &). After running it once I exit and
>> copy over (cp -rf *) my previous profile into the newly created
>> profile.  Accordingly, I have at least a dozen fully functional
>> firefoxes.  TMI, but there it is.
>> Worked fine until somebody said that FF148 had some sort of security
>> problem, at which point I did an update to 149 rather than building
>> a whole new installation.  I think I renamed the policies.json file,
>> which would have been the simplest thing. Sorry I can't be more
>> certain.  Then I would have renamed it back.  And then I got the
>> unwanted surprise of the 150 update.  I turned the problem over to
>> Perplexity, which provided the solution.  If anyone is interested I
>> can have Perplexity generate a nice pdf file of the entire
>> interchange.
>> I assume that firefox does some variant of cp -rf * when it updates
>> and then does some profile-creation stuff and god knows what else. 
>> It must also use some touch-like command, becuase the date of the
>> policies.json file was the same as the date on all the other FF 150
>> subdirectories and files. I don't remember looking at the
>> FF149/firefox files because I would have had no reason to. And that
>> must be when the mischief occurred.
>> AH-HAH!  At some point when I first heard about the policies.json
>> thing I was unsure about where to put it, so I put one in the
>> /firefox subdirectory.  It's still there, and contains the ^M
>> characters. They seem to be M- instead of ^M, for what it's
>> worth. Its date is the same as all the others -- April 24 19:37.
>> cat -A policies.json
>> {$
>> M-  "policies": {$
>> M-  M-  "DisableAppUpdate": true$
>> M-  }$
>> }$
>> $
>> 
>
> Why cat -A? Check man cat.
> In the output of cat -A "M-" is a tab, "^M" is a CR and "$" is a LF
> If you copy those literally into a script it will very likely produce
> an error.

My manual didn't say that. I tested:

hexdump -C tab.txt 
00000000  09 0a                                             |..|
00000002
cat -A tab.txt 
^I$

So, tab is ^I, as you would expect if carriage return is ^M

I am starting to suspect that all this is a wild goose chase.

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


#16977

FromThe Real Bev <bashley101@gmail.com>
Date2026-04-30 20:16 -0700
Message-ID<10t15tn$14kib$1@dont-email.me>
In reply to#16970
On 4/29/26 20:48, dillinger wrote:
> On 30/04/2026 05:30, The Real Bev wrote:

>> AH-HAH!  At some point when I first heard about the policies.json thing 
>> I was unsure about where to put it, so I put one in the /firefox 
>> subdirectory.  It's still there, and contains the ^M characters. They 
>> seem to be M- instead of ^M, for what it's worth. Its date is the same 
>> as all the others -- April 24 19:37.
>> 
>> cat -A policies.json
>> {$
>> M-  "policies": {$
>> M-  M-  "DisableAppUpdate": true$
>> M-  }$
>> }$
>> $
>> 
> 
> Why cat -A? Check man cat.
> In the output of cat -A "M-" is a tab, "^M" is a CR and "$" is a LF
> If you copy those literally into a script it will very likely produce an
> error.

I don't use cat, so I assumed that the -A switch is supposed to show All 
characters.  That was just showing the corrupted/altered file because 
somebody said it was too bad it didn't exist any more..

-- 
Cheers, Bev
     Will give investment advice for food.

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


#17033

FromLawrence D’Oliveiro <ldo@nz.invalid>
Date2026-05-03 21:47 +0000
Message-ID<10t8fq0$37epi$1@dont-email.me>
In reply to#16977
On Thu, 30 Apr 2026 20:16:06 -0700, The Real Bev wrote:

> I don't use cat ...

WHAT?!!?? Whoever heard of a CLI veteran who never used cat??!!??

If anything, cat tends to be overused, at least by those at a certain
point on the Dunning-Kruger curve
<https://en.wiktionary.org/wiki/UUOC> ...

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


#17036

FromThe Real Bev <bashley101@gmail.com>
Date2026-05-03 19:33 -0700
Message-ID<10t90hi$3au6n$1@dont-email.me>
In reply to#17033
On 5/3/26 14:47, Lawrence D’Oliveiro wrote:
> On Thu, 30 Apr 2026 20:16:06 -0700, The Real Bev wrote:
> 
>> I don't use cat ...
> 
> WHAT?!!?? Whoever heard of a CLI veteran who never used cat??!!??

Mostly I want to edit a file rather than just look at it.

And this is the most frequent ls alias (dird) I use and can't understand 
why anybody would want anything else as a default:: ls -altGF 
--color=yes |more
> If anything, cat tends to be overused, at least by those at a certain
> point on the Dunning-Kruger curve
> <https://en.wiktionary.org/wiki/UUOC> ...

Akin to Dunder-Mifflin, I understand...

-- 
Cheers, Bev
   "The Flat Earth Society has members all over the globe."
                                            -- Bob Henson

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


#16984

FromDave Royal <dave@dave123royal.com>
Date2026-05-01 07:04 +0000
Message-ID<10t1jag$17hp9$1@dont-email.me>
In reply to#16958
On Wed, 29 Apr 2026 20:30:28 -0700, The Real Bev wrote:

> On 4/29/26 12:20, Dave Royal wrote:
>> dillinger <dillinger@invalid.not> Wrote in message:
>> 
>>> On 29/04/2026 17:47, Richmond wrote:
>>>> dillinger <dillinger@invalid.not> writes:
>>>> 
>>>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>
>>>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>>
>>>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns
>>>>>>>>>>> <usenet@andyburns.uk>
>>>>>>>>>>> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>>>
>>>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>>>> install them"
>>>>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>>>>
>>>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install'
>>>>>>>>>>> then that radio button changes to 'automatically
>>>>>>>>>>> install/recommended'.
>>>>>>>>>>>
>>>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates
>>>>>>>>>>> but lemme do the work' setting.
>>>>>>>>>>
>>>>>>>>>> I have had that setting for decades -- and still do.  Some
>>>>>>>>>> years ago it stopped working and some kind person posted the
>>>>>>>>>> policies.json fix -- which worked until today.  Posted
>>>>>>>>>> herewith:
>>>>>>>>>>
>>>>>>>>>> Make a new firefox/distribution/ subdirectory which will
>>>>>>>>>> contain the file policies.json which contains {
>>>>>>>>>>     "policies": {
>>>>>>>>>>       "DisableAppUpdate": true
>>>>>>>>>>     }
>>>>>>>>>> }
>>>>>>>>>
>>>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>>>    download, or did you create one - on Linux?
>>>>>>>>
>>>>>>>> Someone posted about that file long ago, and the one I created
>>>>>>>> with pico worked for years.
>>>>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>>>>    problem - got there.
>>>>>>>> They appeared when Firefox chose to update itself against my
>>>>>>>> express wishes.  I suspect that Firefox did it of its own
>>>>>>>> volition -- or simple clumsiness.  Core dump complete.
>>>>>>>
>>>>>>> You /found/ them after the unwanted update. Were they responsible
>>>>>>>    for it? I doubt it: line endings - of either type - are ignored
>>>>>>>    in a json file.
>>>>>>
>>>>>> The windows endings damaged the file such that it no longer served
>>>>>> the purpose.  Fixing it worked.
>>>>>>
>>>>>>
>>>>> Firefox doesn't care about line endings.
>>>>>
>>>>> Just for the sake of it I tested it for you, I created a
>>>>> policies.json which blocks about:config and tested it with both CRLF
>>>>> (windows) and LF (unix) line endings.
>>>>>
>>>>> The result are the same for both, it works:
>>>>> Blocked Page, Firefox can’t connect to the server at about:config
>>>>>
>>>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs
>>>>> an update.
>>>>>
>>>>> CRLF (windows):
>>>>> cat -A policies.json {^M$
>>>>>    "policies": {^M$
>>>>>      "BlockAboutConfig": true^M$
>>>>>    }^M$
>>>>> }^M$
>>>>>
>>>>> LF (unix):
>>>>> cat -A policies.json {$
>>>>>    "policies": {$
>>>>>      "BlockAboutConfig": true$
>>>>>    }$
>>>>> }$
>>>> 
>>>> You tested whether it blocked about:config, but you didn't test
>>>> whether it blocked an update, or if the update altered the json file.
>>> 
>>> I tested policies.json, which was reported corrupted, with both CRLF
>>> and LF, any policy will do, we don't need to wait for the next update.
>>> Policies.json is not included by default, updates will either not
>>> touch it at all or remove it completely, depending on how you update.
>>> Since we can't see the original file there is no way to tell how it
>>> was changed.
>> 
>> Why might an update remove policies.json? I assume that the
>>   firefox updater will read it if it's there but never write to it.
>>   The point of it is that it's user, or organisation,
>>   controlled.
>> 
>> I did wonder if updates might /update/ policies.json, for changes
>>   in format say, in the way they update a profile. But I doubt it.
>> 
>> On Debian I have tarballs of Fx nightly and TB beta, in addtion to
>>   packaged versions for regular use. They do not include a
>>   distribution/ directory or a template policies.json.
> 
> No, they're not included with the tarball.  "Some kind person" posted
> the fix at some point in the past (earliest I have is Firefox 72, July
> 7,2020) posted how to do it.  If I want to use a newer version I don't
> update.  I download the tarball, untar it into its own subdirectory and
> run it from within that subdirectory (/blabla/firefox/firefox -P &).
> After running it once I exit and copy over (cp -rf *) my previous
> profile into the newly created profile.  Accordingly, I have at least a
> dozen fully functional firefoxes.  TMI, but there it is.
> 
> Worked fine until somebody said that FF148 had some sort of security
> problem, at which point I did an update to 149 rather than building a
> whole new installation.  I think I renamed the policies.json file, which
> would have been the simplest thing. Sorry I can't be more certain.  Then
> I would have renamed it back.  And then I got the unwanted surprise of
> the 150 update.  I turned the problem over to Perplexity, which provided
> the solution.  If anyone is interested I can have Perplexity generate a
> nice pdf file of the entire interchange.
> 
> I assume that firefox does some variant of cp -rf * when it updates and
> then does some profile-creation stuff and god knows what else.  It must
> also use some touch-like command, becuase the date of the policies.json
> file was the same as the date on all the other FF 150 subdirectories and
> files. I don't remember looking at the FF149/firefox files because I
> would have had no reason to. And that must be when the mischief
> occurred.

I assumed (having never noticed) that update would only update selected 
files.

Here on Debian 13 I have a tarball installation for Firefox nightly. I 
haven't used it recently: it was v145. (All the files were dated 9/25 but 
I /think/ that's because I copied it from another machines.) I just ran it 
and allowed it to update itself, which was to 152. All the files and 
directories are now dated today - which was not what I expected. But it 
may because of the jump from 145 to 152. Maybe single-version updates 
would do as I had expected.

AND it has created a firefox/distribution/distribution.ini file. I wrote 
earlier that it didn't have a distribution/ directory, so unless I missed 
it that is new since v145. (The Thunderbird beta tarball, which has been 
used more recently does not have a distribution directory.)

[Global]
id=mozilla-official
version=1.0
about=Mozilla Firefox Official Build

But still no policies.json. I've put one in there to see whether it gets 
touched at the next update.

Caveat: All this is with nightly, which may not update the same as 
release.
 
> AH-HAH!  [snip rest of post]
-- 
(Remove any numerics from my email address.)

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


#16990

FromThe Real Bev <bashley101@gmail.com>
Date2026-05-01 09:33 -0700
Message-ID<10t2kl4$1h9gu$3@dont-email.me>
In reply to#16984
On 5/1/26 00:04, Dave Royal wrote:
> On Wed, 29 Apr 2026 20:30:28 -0700, The Real Bev wrote:
<massive snippage>
>> I assume that firefox does some variant of cp -rf * when it updates and
>> then does some profile-creation stuff and god knows what else.  It must
>> also use some touch-like command, becuase the date of the policies.json
>> file was the same as the date on all the other FF 150 subdirectories and
>> files. I don't remember looking at the FF149/firefox files because I
>> would have had no reason to. And that must be when the mischief
>> occurred.
> 
> I assumed (having never noticed) that update would only update selected
> files.
> 
> Here on Debian 13 I have a tarball installation for Firefox nightly. I
> haven't used it recently: it was v145. (All the files were dated 9/25 but
> I /think/ that's because I copied it from another machines.) I just ran it
> and allowed it to update itself, which was to 152. All the files and
> directories are now dated today - which was not what I expected. But it
> may because of the jump from 145 to 152. Maybe single-version updates
> would do as I had expected.
> 
> AND it has created a firefox/distribution/distribution.ini file. 

With the same date as all the other new files.  I wonder when this 
started.  My previous hand-created firefox/distribution subdirectory 
(containing only policies.json) was 148 so it has to have been 149 or 150.

> I wrote
> earlier that it didn't have a distribution/ directory, so unless I missed
> it that is new since v145. (The Thunderbird beta tarball, which has been
> used more recently does not have a distribution directory.)
> 
> [Global]
> id=mozilla-official
> version=1.0
> about=Mozilla Firefox Official Build
> 
> But still no policies.json. I've put one in there to see whether it gets
> touched at the next update.
> 
> Caveat: All this is with nightly, which may not update the same as
> release.

I used nightlies for maybe a year before installing the release version. 
  Why would nightlies be any differet if the idea is to test before release?

I never want files to change creation/edit dates unless edits are 
actually made.  First thing I did was alias cp to cp -ia.  Why is -a NOT 
the default?

-- 
Cheers, Bev
  His men would follow him anywhere, but only out of morbid curiosity.

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


#16992

FromDave Royal <dave@dave123royal.com>
Date2026-05-01 18:00 +0100
Message-ID<10t2m7o$1i8ck$1@dont-email.me>
In reply to#16990
The Real Bev <bashley101@gmail.com> Wrote in message:

> On 5/1/26 00:04, Dave Royal wrote:
>> On Wed, 29 Apr 2026 20:30:28 -0700, The Real Bev wrote:
> <massive snippage>
>>> I assume that firefox does some variant of cp -rf * when it updates and
>>> then does some profile-creation stuff and god knows what else.  It must
>>> also use some touch-like command, becuase the date of the policies.json
>>> file was the same as the date on all the other FF 150 subdirectories and
>>> files. I don't remember looking at the FF149/firefox files because I
>>> would have had no reason to. And that must be when the mischief
>>> occurred.
>> 
>> I assumed (having never noticed) that update would only update selected
>> files.
>> 
>> Here on Debian 13 I have a tarball installation for Firefox nightly. I
>> haven't used it recently: it was v145. (All the files were dated 9/25 but
>> I /think/ that's because I copied it from another machines.) I just ran it
>> and allowed it to update itself, which was to 152. All the files and
>> directories are now dated today - which was not what I expected. But it
>> may because of the jump from 145 to 152. Maybe single-version updates
>> would do as I had expected.
>> 
>> AND it has created a firefox/distribution/distribution.ini file. 
> 
> With the same date as all the other new files.  I wonder when this 
> started.  My previous hand-created firefox/distribution subdirectory 
> (containing only policies.json) was 148 so it has to have been 149 or 150.
> 
>> I wrote
>> earlier that it didn't have a distribution/ directory, so unless I missed
>> it that is new since v145. (The Thunderbird beta tarball, which has been
>> used more recently does not have a distribution directory.)
>> 
>> [Global]
>> id=mozilla-official
>> version=1.0
>> about=Mozilla Firefox Official Build
>> 
>> But still no policies.json. I've put one in there to see whether it gets
>> touched at the next update.
>> 
>> Caveat: All this is with nightly, which may not update the same as
>> release.
> 
> I used nightlies for maybe a year before installing the release version. 
> Why would nightlies be any differet if the idea is to test before release?

To test Firefox itself, yes, but not necessarily to test the
 release process. 
Nightles (alphas) are inherently buggy, so I can imagine that it's
 easier and 
safer to release the complete program, and not add bugs as a
 result of the 
release process being buggy.

But I'm speculating. Maybe the Beta and Release versions are the same. 
My release version is an ESR Debian package updated by apt - the
 Debian 
package manager - so I don't know how a Release tarball updates itself.

> [snip]
-- 
Remove numerics from my email address.

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


#17034

FromLawrence D’Oliveiro <ldo@nz.invalid>
Date2026-05-03 21:51 +0000
Message-ID<10t8g1o$37epi$2@dont-email.me>
In reply to#16990
On Fri, 1 May 2026 09:33:40 -0700, The Real Bev wrote:

> I never want files to change creation/edit dates unless edits are
> actually made. First thing I did was alias cp to cp -ia. Why is -a
> NOT the default?

I tend to be a bit picky about timestamps, too. But I try to avoid
such aliasing of common commands. I use “cp -p” a lot when running
nonprivileged, or “cp --preserve=timestamps” when running privileged.

I suppose the thing with plain “cp” is that the updated modification
timestamp can be used to trigger build rules in software development
projects.

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


#17032

FromDave Royal <dave@dave123royal.com>
Date2026-05-03 11:52 +0000
Message-ID<10t7cuj$2rvuo$1@dont-email.me>
In reply to#16984
On Fri, 1 May 2026 07:04:48 -0000 (UTC), Dave Royal wrote:

> On Wed, 29 Apr 2026 20:30:28 -0700, The Real Bev wrote:
> 
>> On 4/29/26 12:20, Dave Royal wrote:
>>> dillinger <dillinger@invalid.not> Wrote in message:
>>> 
>>>> On 29/04/2026 17:47, Richmond wrote:
>>>>> dillinger <dillinger@invalid.not> writes:
>>>>> 
>>>>>> Op 29-04-2026 om 04:36 schreef The Real Bev:
>>>>>>> On 4/28/26 14:21, Dave Royal wrote:
>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>
>>>>>>>>> On 4/28/26 02:14, Dave Royal wrote:
>>>>>>>>>> The Real Bev <bashley101@gmail.com> Wrote in message:
>>>>>>>>>>
>>>>>>>>>>> On 4/25/26 10:26, Nobody wrote:
>>>>>>>>>>>> On Sat, 25 Apr 2026 18:18:07 +0100, Andy Burns
>>>>>>>>>>>> <usenet@andyburns.uk>
>>>>>>>>>>>> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> Retirednoguilt wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>> I've never had FF update me autonomously.
>>>>>>>>>>>>>
>>>>>>>>>>>>> I have it set for "Check for updates but let you choose to
>>>>>>>>>>>>> install them"
>>>>>>>>>>>>> and as best as I can remember, its never disobeyed ...
>>>>>>>>>>>>
>>>>>>>>>>>> Should a user (for whatever reason) 'uninstall/re-install'
>>>>>>>>>>>> then that radio button changes to 'automatically
>>>>>>>>>>>> install/recommended'.
>>>>>>>>>>>>
>>>>>>>>>>>> Semi-OT, Thunderbird conversely holds the 'check for updates
>>>>>>>>>>>> but lemme do the work' setting.
>>>>>>>>>>>
>>>>>>>>>>> I have had that setting for decades -- and still do.  Some
>>>>>>>>>>> years ago it stopped working and some kind person posted the
>>>>>>>>>>> policies.json fix -- which worked until today.  Posted
>>>>>>>>>>> herewith:
>>>>>>>>>>>
>>>>>>>>>>> Make a new firefox/distribution/ subdirectory which will
>>>>>>>>>>> contain the file policies.json which contains {
>>>>>>>>>>>     "policies": {
>>>>>>>>>>>       "DisableAppUpdate": true
>>>>>>>>>>>     }
>>>>>>>>>>> }
>>>>>>>>>>
>>>>>>>>>> Did that 'kind person' provide a policies.json file for you to
>>>>>>>>>>    download, or did you create one - on Linux?
>>>>>>>>>
>>>>>>>>> Someone posted about that file long ago, and the one I created
>>>>>>>>> with pico worked for years.
>>>>>>>>>> Just wondering how these Windows line endings - if that was the
>>>>>>>>>>    problem - got there.
>>>>>>>>> They appeared when Firefox chose to update itself against my
>>>>>>>>> express wishes.  I suspect that Firefox did it of its own
>>>>>>>>> volition -- or simple clumsiness.  Core dump complete.
>>>>>>>>
>>>>>>>> You /found/ them after the unwanted update. Were they responsible
>>>>>>>>    for it? I doubt it: line endings - of either type - are
>>>>>>>>    ignored in a json file.
>>>>>>>
>>>>>>> The windows endings damaged the file such that it no longer served
>>>>>>> the purpose.  Fixing it worked.
>>>>>>>
>>>>>>>
>>>>>> Firefox doesn't care about line endings.
>>>>>>
>>>>>> Just for the sake of it I tested it for you, I created a
>>>>>> policies.json which blocks about:config and tested it with both
>>>>>> CRLF (windows) and LF (unix) line endings.
>>>>>>
>>>>>> The result are the same for both, it works:
>>>>>> Blocked Page, Firefox can’t connect to the server at about:config
>>>>>>
>>>>>> ^M is not a corruption, this is how cat -A shows CR, your AI needs
>>>>>> an update.
>>>>>>
>>>>>> CRLF (windows):
>>>>>> cat -A policies.json {^M$
>>>>>>    "policies": {^M$
>>>>>>      "BlockAboutConfig": true^M$
>>>>>>    }^M$
>>>>>> }^M$
>>>>>>
>>>>>> LF (unix):
>>>>>> cat -A policies.json {$
>>>>>>    "policies": {$
>>>>>>      "BlockAboutConfig": true$
>>>>>>    }$
>>>>>> }$
>>>>> 
>>>>> You tested whether it blocked about:config, but you didn't test
>>>>> whether it blocked an update, or if the update altered the json
>>>>> file.
>>>> 
>>>> I tested policies.json, which was reported corrupted, with both CRLF
>>>> and LF, any policy will do, we don't need to wait for the next
>>>> update.
>>>> Policies.json is not included by default, updates will either not
>>>> touch it at all or remove it completely, depending on how you update.
>>>> Since we can't see the original file there is no way to tell how it
>>>> was changed.
>>> 
>>> Why might an update remove policies.json? I assume that the
>>>   firefox updater will read it if it's there but never write to it.
>>>   The point of it is that it's user, or organisation,
>>>   controlled.
>>> 
>>> I did wonder if updates might /update/ policies.json, for changes
>>>   in format say, in the way they update a profile. But I doubt it.
>>> 
>>> On Debian I have tarballs of Fx nightly and TB beta, in addtion to
>>>   packaged versions for regular use. They do not include a
>>>   distribution/ directory or a template policies.json.
>> 
>> No, they're not included with the tarball.  "Some kind person" posted
>> the fix at some point in the past (earliest I have is Firefox 72, July
>> 7,2020) posted how to do it.  If I want to use a newer version I don't
>> update.  I download the tarball, untar it into its own subdirectory and
>> run it from within that subdirectory (/blabla/firefox/firefox -P &).
>> After running it once I exit and copy over (cp -rf *) my previous
>> profile into the newly created profile.  Accordingly, I have at least a
>> dozen fully functional firefoxes.  TMI, but there it is.
>> 
>> Worked fine until somebody said that FF148 had some sort of security
>> problem, at which point I did an update to 149 rather than building a
>> whole new installation.  I think I renamed the policies.json file,
>> which would have been the simplest thing. Sorry I can't be more
>> certain.  Then I would have renamed it back.  And then I got the
>> unwanted surprise of the 150 update.  I turned the problem over to
>> Perplexity, which provided the solution.  If anyone is interested I can
>> have Perplexity generate a nice pdf file of the entire interchange.
>> 
>> I assume that firefox does some variant of cp -rf * when it updates and
>> then does some profile-creation stuff and god knows what else.  It must
>> also use some touch-like command, becuase the date of the policies.json
>> file was the same as the date on all the other FF 150 subdirectories
>> and files. I don't remember looking at the FF149/firefox files because
>> I would have had no reason to. And that must be when the mischief
>> occurred.
> 
> I assumed (having never noticed) that update would only update selected
> files.
> 
> Here on Debian 13 I have a tarball installation for Firefox nightly. I
> haven't used it recently: it was v145. (All the files were dated 9/25
> but I /think/ that's because I copied it from another machines.) I just
> ran it and allowed it to update itself, which was to 152. All the files
> and directories are now dated today - which was not what I expected. But
> it may because of the jump from 145 to 152. Maybe single-version updates
> would do as I had expected.
> 
> AND it has created a firefox/distribution/distribution.ini file. I wrote
> earlier that it didn't have a distribution/ directory, so unless I
> missed it that is new since v145. (The Thunderbird beta tarball, which
> has been used more recently does not have a distribution directory.)
> 
> [Global]
> id=mozilla-official version=1.0 about=Mozilla Firefox Official Build
> 
> But still no policies.json. I've put one in there to see whether it gets
> touched at the next update.

Yes, it did. After the update all files are dated identically, including 
policies.json.

Might it be changed - e.g. errors corrected, keys re-ordered, obsolete 
keys removed, file re-encoded, prettified? No idea.
 
> Caveat: All this is with nightly, which may not update the same as
> release.
>  
>> AH-HAH!  [snip rest of post]





-- 
(Remove any numerics from my email address.)

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


#16985

FromLawrence D’Oliveiro <ldo@nz.invalid>
Date2026-05-01 08:09 +0000
Message-ID<10t1n4h$196ee$1@dont-email.me>
In reply to#16919
On Tue, 28 Apr 2026 19:36:56 -0700, The Real Bev wrote:

> The windows endings damaged the file such that it no longer served
> the purpose. Fixing it worked.

In case it’s relevant, I keep seeing those “^M” sequences in postings
from you when I paste the text into Emacs.

I just find that odd, for a Linux user ...

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


#16989

FromThe Real Bev <bashley101@gmail.com>
Date2026-05-01 09:17 -0700
Message-ID<10t2jn7$1h9gu$2@dont-email.me>
In reply to#16985
On 5/1/26 01:09, Lawrence D’Oliveiro wrote:
> On Tue, 28 Apr 2026 19:36:56 -0700, The Real Bev wrote:
> 
>> The windows endings damaged the file such that it no longer served
>> the purpose. Fixing it worked.
> 
> In case it’s relevant, I keep seeing those “^M” sequences in postings
> from you when I paste the text into Emacs.
> 
> I just find that odd, for a Linux user ...

Blame Thunderbird.

-- 
Cheers, Bev
  His men would follow him anywhere, but only out of morbid curiosity.

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


#16993

FromFrank Miller <miller@posteo.ee>
Date2026-05-01 19:10 +0200
Message-ID<69F4DE79.7090604@backwurst.de>
In reply to#16989
The Real Bev wrote:
> On 5/1/26 01:09, Lawrence D’Oliveiro wrote:
>> On Tue, 28 Apr 2026 19:36:56 -0700, The Real Bev wrote:
>> 
>>> The windows endings damaged the file such that it no longer served
>>> the purpose. Fixing it worked.
>> 
>> In case it’s relevant, I keep seeing those “^M” sequences in postings
>> from you when I paste the text into Emacs.
>> 
>> I just find that odd, for a Linux user ...
> 
> Blame Thunderbird.

Yes. Just blame Thunderbird, blame Firefox, blame Mozilla as a whole.
Or better blame Microsoft or Linux, blame the developers of software.
But never take responsibility for all that weird mix up of an operating
system you created but don't understand yourself.

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


#16997

FromR Daneel Olivaw <Danni@hyperspace.vogon.gov.invalid>
Date2026-05-01 20:16 +0200
Message-ID<10t2qlo$2ktq4$1@paganini.bofh.team>
In reply to#16993
Frank Miller wrote:
> The Real Bev wrote:
>> On 5/1/26 01:09, Lawrence D’Oliveiro wrote:
>>> On Tue, 28 Apr 2026 19:36:56 -0700, The Real Bev wrote:
>>>
>>>> The windows endings damaged the file such that it no longer served
>>>> the purpose. Fixing it worked.
>>>
>>> In case it’s relevant, I keep seeing those “^M” sequences in postings
>>> from you when I paste the text into Emacs.
>>>
>>> I just find that odd, for a Linux user ...
>>
>> Blame Thunderbird.
> 
> Yes. Just blame Thunderbird, blame Firefox, blame Mozilla as a whole.
> Or better blame Microsoft or Linux, blame the developers of software.
> But never take responsibility for all that weird mix up of an operating
> system you created but don't understand yourself.
> 

You have no idea, https://www.youtube.com/watch?v=bOR38552MJA - Blame 
Canada!

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


#16998

FromNobody <jock@soccer.com>
Date2026-05-01 12:57 -0700
Message-ID<q11avk9bmm35eeqpcitnoa4k2n5ov75c1r@4ax.com>
In reply to#16997
On Fri, 1 May 2026 20:16:24 +0200, R Daneel Olivaw
<Danni@hyperspace.vogon.gov.invalid> wrote:

>Frank Miller wrote:
>> The Real Bev wrote:
>>> On 5/1/26 01:09, Lawrence D’Oliveiro wrote:
>>>> On Tue, 28 Apr 2026 19:36:56 -0700, The Real Bev wrote:
>>>>
>>>>> The windows endings damaged the file such that it no longer served
>>>>> the purpose. Fixing it worked.
>>>>
>>>> In case it’s relevant, I keep seeing those “^M” sequences in postings
>>>> from you when I paste the text into Emacs.
>>>>
>>>> I just find that odd, for a Linux user ...
>>>
>>> Blame Thunderbird.
>> 
>> Yes. Just blame Thunderbird, blame Firefox, blame Mozilla as a whole.
>> Or better blame Microsoft or Linux, blame the developers of software.
>> But never take responsibility for all that weird mix up of an operating
>> system you created but don't understand yourself.
>> 
>
>You have no idea, https://www.youtube.com/watch?v=bOR38552MJA - Blame 
>Canada!

OT: ha, ha funny... until that disgusting epithet hurled at a
well-known Nova Scotian singer.

Humour/satire doesn't carry the right to insult others by name.

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


Page 2 of 5 — ← Prev page 1 [2] 3 4 5  Next page →

Back to top | Article view | alt.comp.software.firefox


csiph-web