I just see it as......as long as I have the "option" to update using Terminal and "dnf"?.......then I will just ignore the big blue "Restart & Install" button!....
EGO II
On Mon, Dec 30, 2019, 1:39 PM Andy Paterson via users <users@xxxxxxxxxxxxxxxxxxxxxxx> wrote:
Please forgive my intrusion on this ott thread, but as a now retired “unix guy” using unix since xenix blah! And having written network device drivers before tcp/ip became the norm
In order to mmap a file it must be open, furthermore when a file is removed (ulink’ed) the kernel doesn't actually “delete” the file (& free its blocks etc) until it has an “open” count of zero - similarily with link count.
So if a .so or any other file is “replaced” the original file is still open & exists for any process that has it open!
The same thong applies for example with binary executable files.
So .... it seems to me that any update problems are limited to installing a new executable before installing a new version of a library (.so etc) and running the executable before its dependency is installed
Of course a new kernel & device drivers is a totally different issue!
This isn't Windows!
> On 30 Dec 2019, at 17:58, Richard Hughes <hughsient@xxxxxxxxx> wrote:
>
> On Mon, 30 Dec 2019 at 15:02, John Mellor <john.mellor@xxxxxxxxx> wrote:
>> I've complained about this issue before. Its a defective design
>> decision made by the Gnome people, some of whom I suspect to be
>> ex-Windows people trying to sabotage the desktop ;^0
>
> This is unacceptable.
>
>> but I have yet to either hear of or experience an actual
>> problem caused by not rebooting in 20 years of Linux use
>
> I was the person triaging these bugs for about the last decade. If you
> have a failure rate of 1/10000, and you have millions of users, you
> have tens of angry users EVERY DAY filing bugs that their root
> filesystem exploded or that their GUI application crashed while it was
> updated in the background, losing all their work. Offline updates has
> reduced this failure rate by about 3 orders of magnitude.
>
>> Unix is designed to prevent this problem
>
> That's nonsense, sorry.
>
>> as its actually a ruse by the Gnome developers to justify their broken design decision
>
> You're just being offensive now.
>
>> instead of just doing the simpler and easier code in the update app
>
> You're hilarious, and you you clearly don't actually understand how
> rpm deployment works, UXIX locking semantics, or modern Linux service
> or application design. Please self moderate your opinions in the
> future.
>
> Richard.
> _______________________________________________
> users mailing list -- users@xxxxxxxxxxxxxxxxxxxxxxx
> To unsubscribe send an email to users-leave@xxxxxxxxxxxxxxxxxxxxxxx
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/users@xxxxxxxxxxxxxxxxxxxxxxx
_______________________________________________
users mailing list -- users@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to users-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/users@xxxxxxxxxxxxxxxxxxxxxxx
_______________________________________________ users mailing list -- users@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to users-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@xxxxxxxxxxxxxxxxxxxxxxx
