Then it downloaded the whole program. Second time this happened. Is this normal ? KenW
![]() |
0 |
![]() |
On Tue, 10 Jul 2018 15:29:44 -0600, KenW wrote: > Then it downloaded the whole program. Second time this happened. > Is this normal ? > > > KenW Quite common for me. Just happened updating 52.8.? to 52.9.1 on Win 10. Happens with Firefox too. -- (Remove any numerics from my email address.)
![]() |
0 |
![]() |
On Wed, 11 Jul 2018 03:56:17 -0500, Dave Royal <dave@dave123royal.com> wrote: >On Tue, 10 Jul 2018 15:29:44 -0600, KenW wrote: > >> Then it downloaded the whole program. Second time this happened. >> Is this normal ? >> >> >> KenW > >Quite common for me. Just happened updating 52.8.? to 52.9.1 on Win 10. >Happens with Firefox too. At least I am not alone !! KenW
![]() |
0 |
![]() |
On 11-Jul-2018 07:53, KenW wrote: > On Wed, 11 Jul 2018 03:56:17 -0500, Dave Royal <dave@dave123royal.com> > wrote: > >> On Tue, 10 Jul 2018 15:29:44 -0600, KenW wrote: >> >>> Then it downloaded the whole program. Second time this happened. >>> Is this normal ? >>> >>> >>> KenW >> >> Quite common for me. Just happened updating 52.8.? to 52.9.1 on Win 10. >> Happens with Firefox too. > > At least I am not alone !! > > > KenW > Same thing has been happening to me lately, I thought it was because I was using the portable version but I guess wrong.
![]() |
0 |
![]() |
EnDeeGee <invalid@invalid.invalid> Wrote in message: > On 11-Jul-2018 07:53, KenW wrote: >> On Wed, 11 Jul 2018 03:56:17 -0500, Dave Royal <dave@dave123royal.com> >> wrote: >> >>> On Tue, 10 Jul 2018 15:29:44 -0600, KenW wrote: >>> >>>> Then it downloaded the whole program. Second time this happened. >>>> Is this normal ? >>>> >>>> >>>> KenW >>> >>> Quite common for me. Just happened updating 52.8.? to 52.9.1 on Win 10. >>> Happens with Firefox too. >> >> At least I am not alone !! >> >> >> KenW >> > > Same thing has been happening to me lately, I thought it was because I > was using the portable version but I guess wrong. > I've not been keeping score but I suspect on this Win10 laptop it happens more often than not. I assume it's when the 'delta' update fails for some reason but I don't know why it would: there's nothing special about the setup. I've always suspected it's because there is another (limited) user with Fx/TB open - but I've never tested the theory: the 'full' update always works. Only irritating thing is the 'it failed' dialogue is always behind the newly-reopened-but-not-updated-program window - so I sometimes don't notice. -- (Remove numerics from my email address)
![]() |
0 |
![]() |
On 7/10/2018 5:29 PM, KenW wrote: > Then it downloaded the whole program. Second time this happened. > Is this normal ? > > > KenW > I just experienced this doing some testing of version 52.8.0, then updating to version 52.9.1. The only explanation I can think of, is that Thunderbird just moved to taskcluster for building the application. When doing pre-release testing we always get a complete update because there are no partial updates. I'm not sure why it is looking for a partial update.
![]() |
0 |
![]() |
WaltS48 <schw01@REMOVEverizon.net> Wrote in message: > On 7/10/2018 5:29 PM, KenW wrote: >> Then it downloaded the whole program. Second time this happened. >> Is this normal ? >> > > I just experienced this doing some testing of version 52.8.0, then > updating to version 52.9.1. > > The only explanation I can think of, is that Thunderbird just moved to > taskcluster for building the application. When doing pre-release testing > we always get a complete update because there are no partial updates. > > I'm not sure why it is looking for a partial update. > I thought it offered delta/partial updates for normal update paths to recent versions. So 52.9 to 52.9.1 would be a delta. But if 52.9 rollout was stopped, as Wayne said, then 52.8.0 to 52.9.1 would be a normal update path - and I think that's the one that failed for me too. You wouldn't normally know it's a delta - unless it fails. I've only seen this on Windows because I don't do this sort of update in Linux or SFOS. IME delta updates of OSes fail when the file to be updated fails an integrity check - because I patched it, or it's corrupt - or compromised. I wonder if Moz gets metrics for these failed updates. Probably. -- (Remove numerics from my email address)
![]() |
0 |
![]() |
On 12/07/18 22:20, Dave Royal wrote: > WaltS48 <schw01@REMOVEverizon.net> Wrote in message: >> On 7/10/2018 5:29 PM, KenW wrote: >>> Then it downloaded the whole program. Second time this happened. >>> Is this normal ? >>> >> >> I just experienced this doing some testing of version 52.8.0, then >> updating to version 52.9.1. >> >> The only explanation I can think of, is that Thunderbird just moved to >> taskcluster for building the application. When doing pre-release testing >> we always get a complete update because there are no partial updates. >> >> I'm not sure why it is looking for a partial update. >> > I thought it offered delta/partial updates for normal update paths > to recent versions. So 52.9 to 52.9.1 would be a delta. But if > 52.9 rollout was stopped, as Wayne said, then 52.8.0 to 52.9.1 > would be a normal update path - and I think that's the one that > failed for me too. > > You wouldn't normally know it's a delta - unless it fails. I've > only seen this on Windows because I don't do this sort of update > in Linux or SFOS. > > IME delta updates of OSes fail when the file to be updated fails > an integrity check - because I patched it, or it's corrupt - or > compromised. > > I wonder if Moz gets metrics for these failed updates. Probably. My Mint 17.3 x64 update manager showed TB 52.9.1 available yesterday. It updated from 52.8.0 without problem. -- Jeff
![]() |
0 |
![]() |
On 07/13/2018 02:52 AM, Jeff Layman wrote: > On 12/07/18 22:20, Dave Royal wrote: >> WaltS48 <schw01@REMOVEverizon.net> Wrote in message: >>> On 7/10/2018 5:29 PM, KenW wrote: >>>> Then it downloaded the whole program. Second time this happened. >>>> Is this normal ? >>>> >>> >>> I just experienced this doing some testing of version 52.8.0, then >>> updating to version 52.9.1. >>> >>> The only explanation I can think of, is that Thunderbird just moved to >>> taskcluster for building the application. When doing pre-release testing >>> we always get a complete update because there are no partial updates. >>> >>> I'm not sure why it is looking for a partial update. >>> >> I thought it offered delta/partial updates for normal update paths >> to recent versions. So 52.9 to 52.9.1 would be a delta. But if >> 52.9 rollout was stopped, as Wayne said, then 52.8.0 to 52.9.1 >> would be a normal update path - and I think that's the one that >> failed for me too. >> >> You wouldn't normally know it's a delta - unless it fails. I've >> only seen this on Windows because I don't do this sort of update >> in Linux or SFOS. >> >> IME delta updates of OSes fail when the file to be updated fails >> an integrity check - because I patched it, or it's corrupt - or >> compromised. >> >> I wonder if Moz gets metrics for these failed updates. Probably. > > My Mint 17.3 x64 update manager showed TB 52.9.1 available yesterday. It > updated from 52.8.0 without problem. > Available from Ubuntu this morning and updated without a problem. My test above was done on Windows 10. Testing updating from TB 52.8.0 to 52.9.1 in my test user account on Linux worked without a problem.
![]() |
0 |
![]() |
WaltS48 <schw01@REMOVEverizon.net> Wrote in message: > > Available from Ubuntu this morning and updated without a problem. > > My test above was done on Windows 10. > > Testing updating from TB 52.8.0 to 52.9.1 in my test user account on > Linux worked without a problem. > SuSE package updates - rpms - are often deltas and I've never known one fail and try again like Mozilla updates do. I've never noticed whether my Debian apt updates are deltas. It's as if Mozilla's update system expects delta updates to fail. I think this is a Windows-only feature. I've seen it a few times over the years but never as frequently as recently under Win 10 - which is otherwise pretty good IMO. Doesn't bother me enough to investigate further. But if it happened on my Linux boxes I would! -- (Remove numerics from my email address)
![]() |
0 |
![]() |