code logs -> 2018 -> Wed, 15 Aug 2018< code.20180814.log - code.20180816.log >
--- Log opened Wed Aug 15 00:00:18 2018
00:51 celticminstrel [celticminst@Nightstar-v1qb0r.dsl.bell.ca] has joined #code
00:51 mode/#code [+o celticminstrel] by ChanServ
01:28 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds]
01:50
<&[R]>
<Mahal> if you're on power it shouldn't happen. <-- it did
02:01 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code
02:01 mode/#code [+qo Vornicus Vornicus] by ChanServ
02:07 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds]
02:39
<&[R]>
And now it's doing an update, which means it will forget a random subselection of settings
02:40
<&[R]>
(I have no clue how that was ever considered acceptable)
03:56 Degi [Degi@Nightstar-hltc4r.dyn.telefonica.de] has quit [Connection closed]
04:32 celticminstrel [celticminst@Nightstar-v1qb0r.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
06:46 Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code
06:46 mode/#code [+qo Vornicus Vornicus] by ChanServ
07:10 himi [sjjf@Nightstar-1drtbs.anu.edu.au] has quit [Ping timeout: 121 seconds]
07:11 Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Connection closed]
07:12 Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has joined #code
07:12 mode/#code [+qo Vorntastic Vorntastic] by ChanServ
07:18 Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code
07:18 mode/#code [+qo Vornicus Vornicus] by ChanServ
07:23 Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds]
08:04 mac is now known as macdjord|slep
09:04 himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code
09:04 mode/#code [+o himi] by ChanServ
09:32 Kindamoody[zZz] is now known as Kindamoody
10:31 himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds]
11:20 Degi [Degi@Nightstar-m4ietf.dyn.telefonica.de] has joined #code
12:35 Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code
12:36 mode/#code [+qo Vornicus Vornicus] by ChanServ
13:08 Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds]
13:59 Kindamoody is now known as Kindamoody|afk
14:44 Degi [Degi@Nightstar-m4ietf.dyn.telefonica.de] has quit [Connection closed]
16:00 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code
16:00 mode/#code [+qo Vornicus Vornicus] by ChanServ
17:32 Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has quit [[NS] Quit: Connection closed for inactivity]
17:55
<&[R]>
Dear MS: On the edition of Windows that you make all computers that an SMB would purchase have, could you kindly add an option entiled "this is a fucking work computer, don't reboot it without the user's permission because it takes a fucking hour+ to install updates and pisses of my clients." I belive the option is fairly self-documenting in it's purpose. Thank you.
17:56
<&[R]>
PS: when logining in after a reboot, an option titled "don't try restart half my applications because you pick all of the ones that do heavy disk I/O on start-up and it makes logining in take another 40 minutes" I believe it's purpose is equally clear
17:57
<&[R]>
Thanks. Until then I'll be spending time trying to move to Linux ASAP.
17:59 mode/#code [-M] by abudhabi
17:59
<@abudhabi>
Test.
17:59
<&[R]>
<abudhabi> Test. <-- Test.
18:00
<@abudhabi>
[R]: I would just suggest never rebooting without the user's permission at all. Ever.
18:00
<&[R]>
So would I
18:00
<&[R]>
But it's MS. They do everything half-asses
18:00
<&[R]>
assed*
18:06
<&[R]>
I honestly couldn't expect them to do something properly. Especially one in the UX area.
18:08
<@TheWatcher>
[R]: I'm using win10 pro, and while that option isn't obvious, you can do it with group policies
18:08
<@TheWatcher>
I believe it doesn't work on anything but pro, though
18:09
<&[R]>
Right
18:11
<&[R]>
Hence my specification about the Windows version on consumer laptops
18:11
<&[R]>
Which is Home
18:12
<@TheWatcher>
Yeah. For Home, Microsoft assumes the user is either a drooling moron or a technophobe, and has decided what's best for them
18:13
<&[R]>
Conveniently forgetting all the SMBs that won't upgrade to Pro or whatever because there's no reason, the comnputer alread has Windows
19:30
<&McMartin>
To be fair, I'm pretty good at this stuff
19:30
<&McMartin>
And group policies tend to make me run screaming
19:30
<&McMartin>
That's a different skill they didn't teach us developers. >_>
19:31
<&McMartin>
(And oh god our entry point is terrifying. This is clearly a technology that *only8 the IT guys were supposed to use, through MS interfaces, once you were outside of MS itself)
20:11 Degi [Degi@Nightstar-m4ietf.dyn.telefonica.de] has joined #code
20:45
< Mahal>
So, [R], would you prefer that your SMB laptops never got patched, circa every OS before WIn10
20:45
< Mahal>
or that Microsoft at least made a token attempt to keep their users more secure in the biggest consumer market in the world?
20:45
< Mahal>
and, hey, if you want business class features, pay for the business class OS.
20:45
< Mahal>
I'm not a Microsoft fangirl or apologist
20:45
< Mahal>
but given the alternative I don't feel like this is a terrible one.
20:53
<&[R]>
I'd like more control over how the updates happen, the forced updates are BS and causing problems. I'd bitch a TON less if like every other OS they could do updates while the computer is usable.
20:53
< Mahal>
If you want control - pay for corporate control.
20:53
<&[R]>
Basically they chose every user-hostile option in the name of security.
20:53
< Mahal>
You can tune the reboots a little, and put them off till hours that suit you even on the home OS versions.
20:54
< Mahal>
I have one!
20:54
< Mahal>
Yeah, nah, sorry, that's not user hostility.
20:54
<&[R]>
When the machine is randomly unusable for long periods of billable time, that's hostiltiy
20:55
< Mahal>
then set the active hours on the device so it doesn't do updates then!
20:55
< Mahal>
and if you don't have the competency to change those settings, *don't bitch about it*
20:55
<&[R]>
Or you know, they could make it so you can just trivially upgrade to pro from a home-installed device
20:55
<&[R]>
But they don't
20:55
< Mahal>
Select the Start button, then select Settings > Update & security > Activation.
20:55
< Mahal>
Select Change product key, and then enter the 25-character Windows 10 Pro product key.
20:55
< Mahal>
Select Next to start the upgrade to Windows 10 Pro.
20:57
<&[R]>
Which costs the full cost of a new windows license when you're already getting a home one
20:57
< Mahal>
Then don't buy a Home one in the first place?
20:57
<&[R]>
Also lolwut at the compentency thing
20:57
< Mahal>
Look, I sympathise with your point, honestly.
20:57
< Mahal>
I do
20:57
<&[R]>
I'm complaining in the context of SMB
20:57
< Mahal>
but the actual probllem here is SMB's DON'T UNDERSTAND WHAT THEY'RE DOING
20:58
< Mahal>
and buy the wrong products, and don't want to pay for support
20:58
< Mahal>
all of which I get
20:58
< Mahal>
and as a result of that, you have the option of "unpatched unsecure shit that actually causes issues for everyone else"
20:58
<&[R]>
The people 1) most affected by this because they don't have the staff for it 2) the least likely to see any need to spring for pro
20:58
< Mahal>
or "well, it sucks to be you"
20:58
< Mahal>
"accept your updates, get on with it"
20:59
< Mahal>
I bought a Pro license for my personal laptop because I use it for work purposes as well as personal ones.
21:01
< Mahal>
You obviously *do* understand what you're doing, so ... buy pro licenses? explain to your clients what the difference between consumer and pro editions are?
21:01
< Mahal>
Or, sure, switch to Ubuntu, but that's not a feasible option for a *large chunk* of smb's
21:01
< Mahal>
(And potentially leaves them with unwarrantied devices depending on where they purchased them!)
21:08
<&[R]>
Whatever. An update should not cause significantly more downtime than a normal reboot
21:09
< Mahal>
and on a managed OS they don't -shrug- they'll install in the background and only prompt for reboots as required.
21:09
< Mahal>
(except for major version upgrades where a significant portion of the OS is changing, e.g. 1709 to 1803 was quite lengthy)
21:38 Kindamoody|afk is now known as Kindamoody
21:39 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds]
22:38
<&McMartin>
For the record Ubuntu 16.04 to 18.04 was eight hours
23:06
<&[R]>
But you could actually use your system for that time
23:07
<&[R]>
I don't care if an update takes a day or two to do, if the system is still usable during that time.
23:08
<&McMartin>
Oh. No
23:08
<&McMartin>
This was the "please have all applications closed during the upgrade"
23:08
<&McMartin>
part
23:08
<&[R]>
WTF
23:08
<&McMartin>
the download part was merely the amount of time it takes to download 2GB, which even on my shitty internet wasn't long
23:08
<&[R]>
Last I used Ubuntu you didn't need to worry about that shit
23:08
<&McMartin>
Yeah, they apparently noticed that people must have really missed the old console mode dialog boxes that stopped everything
23:08
<&McMartin>
... and worse, you couldn't see them
23:08
<&McMartin>
so now there are like five, but they're visible and GNOME-based.
23:09
<&McMartin>
and roughly evenly spaced. -_-
23:09
<&[R]>
... because that's the proper solution to that problem...
23:09
<&McMartin>
But yeah, even if you ignore that
23:09
<&McMartin>
The "OK, we're removing these 500 packages now" was measured in hours.
23:09
<&[R]>
You know, instead of NOT having console programs start during package maintenance scripts
23:09
<&McMartin>
They fixed it1
23:10
<&McMartin>
They weren't console programs
23:10
<&[R]>
?
23:10
<&McMartin>
This was dist-upgrade requiring you to hit NEXT at each step, more or less.
23:10
<&[R]>
I've never seen something like that
23:10
<&McMartin>
Hours apart, because apt-get is really, really slow, at least when run on spinning rust
23:11
<&McMartin>
And yeah, this was also paired with explicit messaging to not be running any applications during the upgrade process
23:11
<&[R]>
All of my Linux update experience has been "I'm updating everything. ... Okay, I'm done now. You should reboot." And the system was usable during that process
23:12
<@TheWatcher>
Wait, reboot?
23:13
<&McMartin>
Yeah, dist-upgrades are... more dramatic.
23:13 * TheWatcher only ever does that when upgrading the kernel, otherwise runs lib_users to see which processes have old libs loaded and restarts them >.>
23:13
<&McMartin>
And yes, for some reason Fedora now reboots on every update ever if you go through the graphical installer
23:13
< Mahal>
TheWatcher: you are a power user, as opposed to a normal person :P
23:13
<&McMartin>
As opposed to sudo dnf upgrade which Just Upgrades Stuff And Is Done.
23:13
<@TheWatcher>
Mahal: fair
23:14
<&[R]>
dnf is which distro?
23:14
<&[R]>
Oh Fedora
23:15
<&McMartin>
So that might be GNOME 3 generally and not Fedora specifically.
23:15
< Mahal>
(OSX also does mandatory updates and reboots these days, with varying time factors required. I'd imagine that's the second heaviest consumer OS these days.)
23:16
< Mahal>
(As does iOS in the mobile space. Can't speak for ANdroid as I don't use it and I understand some of that relies on the carrier)
23:16
<&McMartin>
I mean, most of the time Windows updates do not require a reboot either, and the experience of that is basically that autoupdates that do not require a reboot didn't happen, so 100% of updates require reboots, froth, rage
23:16
< Mahal>
^
23:17
<&McMartin>
Also "I can't update Xcode while iTunes is running, press Continue to force-quit iTunes"
23:17 * Mahal nods
23:19
<&McMartin>
Which I haven't seen on Windows in years and years, despite Windows having non-advisory file locking and OSX not
23:21
< Mahal>
Office will usually ask you to close all Office products while updating itself, but that might be a different case.
23:22
<&McMartin>
Oh yeah, I'm not counting the case where the file *being replaced* must be closed.
23:22
<&[R]>
That's because Windows does manditory file-locking
23:22
<&McMartin>
Because yes, Windows will still want that.
23:22
<&[R]>
You can't update a bindary that's in use.
23:22
<&McMartin>
You can
23:23
<&[R]>
But no-one does.
23:23
<&McMartin>
the part you're objecting to is that when a file is deleted or replaced, the other processes that have that file open will notice
23:23
<&McMartin>
(And indeed that will only happen if everyone who opened it permitted concurrent modification by others)
23:24
<&McMartin>
the part I was objecting to was when an unrelated application must be closed to update you
23:24
<&McMartin>
This is also related to the most succinct version of Unix Brain Damage as experienced by everyone who isn't an old Unix Hand
23:25
<&McMartin>
which is that everything sucks because you can't arrange shared memory by having two processes open the same file, followed by deleting that file
23:25
<&[R]>
If you're updating office, you'd need all of office closed because they're all getting updated.
23:25
<&McMartin>
And obviously any system that doesn't permit this exact protocol to have that exact result must be literally insane and obviously of reprehensibly terrible design
23:26
<&McMartin>
But I think these days Linux and BSD both have actual APIs intended to provide shared memory control, so this comes up less
23:26
<@TheWatcher>
Yep, they do
23:26
<&[R]>
Linux has 7 or so now
23:27
<&[R]>
BBL, still working
23:40 himi [sjjf@Nightstar-1drtbs.anu.edu.au] has joined #code
23:40 mode/#code [+o himi] by ChanServ
23:57 * Mahal chortles
23:57
< Mahal>
https://www.reddit.com/r/talesfromtechsupport/comments/97hf4j/cthulhu_debugging/e489sct/
--- Log closed Thu Aug 16 00:00:19 2018
code logs -> 2018 -> Wed, 15 Aug 2018< code.20180814.log - code.20180816.log >

[ Latest log file ]