« first day (2371 days earlier)      last day (2945 days later) » 

08:00
Hm. This seems interesting
SO you have one boot loader that loads another one on disk that list that particular disks list of bootables?
Thats my old thnkpad :)
no i dont use 3rd party bootloading, or other bootloaders. just boot menu from the motherboard. that IS why, i want to avoid that.
when that stuff craps out, people tend to re-install, the whole mess all over again. I cannot afford to reinstall.
Oh you use the EFI menu?
Thanks for that link. I have 5 IBM and Lenovos sitting around here and now I am glad I can upgrade them.
@RichManson used to be a Thinkpad fan. Not a fan of recent moves tho.
I don't understand why that article says that some didn't like the move to 16:9 ratio displays from 4:3
08:06
whatever the motherboard provide, before it was bios boot menu, today it has some cute software :-) but it is still a basic hardware "bios" to me, just more cute software stuff that created more vulnerabilities , that does the same basic thing a different way .
I didn't :p
Point to a disk, point to a partition, fire up the OS from there.
Vertical space is kinda Nice.
They are saying preference.
Yup
Same reason we have 3:2 screens
08:10
@JourneymanGeek amusingly, Apple's Retina laptops of late are 16:10, which is slightly better vertically than 16:9
I still boot a Phenom machine on Dual 4:3 17" monitors :) Here I thought I was behind the times. Didn't know there was a folloing for that aspec ratio :P
Bob
Bob
@allquixotic I wonder why they, of all people, don't do 3:2. iPad, anyone?
oh wait those are 4:3 o.O
the reason I like my ultrawide on my main desktop is because, even though I have tons of horizontal, even the vertical is generous enough that pages don't feel "cut off" or "constrained" at the default DPI (no custom scaling), and the res is low enough that I can read it from a natural viewing distance
on a laptop I actually like 16:10 or 4:3 more than 16:9, and an ultrawide laptop would be dumb
@RichManson if you remember back when this "change" that MS did , linux people were afraid that microsoft was trying to lock them out completly. chainloading one os from another oses boot structure , again where do you go, when either one breaks?
@Bob and then to keep you guessing, the iPhone is 16:9 :P (actual native 1080p on the Plus models)
08:17
.. . you go back to the drawing board. me i just boot up the other one :-)
Well, vertical space is absolute. I actually like my old r60 better than an r61 (at 14 inches) but at smaller sizes widescreen is fine. At 27 inches.....
the only time I really notice vertical space is (1) the RA star wall -- on vertical constrained devices I can only see like 5 things on the star wall; and (2) Eclipse and Visual Studio, which have so much widget decoration that you barely get to see a viewport of your code on low-res devices
I might try your method on my Phenom machine. I don't think that is UEFI. The MB is 2010
but the MBP and the main desktop are enough that I'm happy
Wel I purchased it in 2010. Geez its old. I still don't want to retire it yet
08:19
@RichManson i think i explain my method, more to excuse the fact that i can't help you for yours :-)
@JourneymanGeek I have T60 and T61's
I have an X60 and X61T around somewhere that are completely outdated and disused
@Psycogeek Your method sounds interesting. It works for you and may have its merits. I'd like to see the drawings just because I like to study different multi boot systems. Kinda a geek that way.
@JourneymanGeek so do we call that a Hackinpad? :P
LOL I didn't know Superuser had Clippie
I missed that guy since Office XP or Office 2003
@allquixotic Hackithink
Padahack
Thinkiehackie
HackieThnkie?
Padawackahackie
Thinkiewinkiehackiepackie LOl
08:31
@allquixotic they will play HD movies , and browse the web? I am going mobile eventually , could be you should sell them to me :-) Gonna get me a weak touch tablet of 10+inch size someday.
oh wait , touching the 10 inch, thats not safe for work.
Just a matter of shoving a huge battery pack on it , so it runs for 10 hours non-stop
I also have a r51 (which is PATA so kinda useless) and a x220 (which is actually not too bad, though I wouldn't mind higher res screen)
@allquixotic my dad runs a upgraded r61. Not quite useless once you get rid of the spinning rust
@Nzall yeah, that came up a few hours ago
short version: one of the DBAs accidentally nuked their production database
ah, missed that
@Nzall ouchhhhh
what a nightmare
LVM, pg_brokenshit commands... yuck
08:39
actually its a little worse than that
Note: I did check back to around 11 PM last night in the history of this chatroom
I don't envy that sysadmin
their backups were borken too
@Nzall meh, stuff gets repeated a lot
@allquixotic heh, seems more like a dev sysadminning than a proper sysadmin
@Nzall meh, stuff gets repeated a lot
@allquixotic >_>
@Nzall with people who have similar interests, occationally stuff gets repeated decades hours apart
08:40
@JourneymanGeek Was that in this chatroom?
@Nzall I coulda sworn it was.
because I got the transcript loaded to 3 PM yesterday, and a search on Gitlab finds nothing
apart from what I posted
I also clearly am suffering from confusion a glitch in the matrix.
08:43
user image
2
that throws him off everytime :-)
aww <3
@JourneymanGeek Spinning rust? You mean went from HDD to SSD?
@RichManson almost all my systems use SSD for boot
I have one spare box that runs on HDDs, and I still use it for bulk storage and backups
I bought an SSD and was going to use it on this machine I'm building but can't find it :(
can't help you with that ;p
08:56
OMG its 4am here. I'm outa here
0
Q: Always display vote split

Radial AppsSince I made it to 1k reputation (yay!), I can now see the vote split by clicking on the current vote status of a question or answer. Is there any way so that I can always see these when I am logged in (so that I need not click it every time)

@JourneymanGeek Problem with that IBM link is the site they reference that is upgrading those is in all Chinese with no English translation
Yeah, that is likely a PITA, but not uncommon
@JourneymanGeek have you dabbled in watch dogs II yet?
09:11
@Psycogeek not yet. Waiting for it to go on sale
Hi!
@StackExchange Even though I was crazy about earning it, I've never actually used this privilege since I earned it.
@Burgi lol
@Burgi hahahehe
Is this question acceptable here?
0
Q: update pgp subkey signature

kinnlaMy public pgp public key has a subkey with a signature (self signed) that contains an outdated reference to a policy. How can I update the signature and remove the reference? I had the same problem with my main key and its UIDs. I could solve it by deleting the signatures and self-signing them a...

@CodesInChaos probably
09:52
@CodesInChaos no mention of the environment tho
its very low quality, and would likely be thrown back without edits
11:00
@Nzall Somehow disallow rm -rf for the PostgreSQL data directory? Unsure if this is feasible, or necessary once we have proper backups
xsigh
/Hi, were gitlab, and we dont have proper bqackups
@bob :( crash! lol, thanks for the tool, looking now...
did a firmware update for board that said "increases stability"... one week and no crash, so, thought all was good... then, out of the blue, BSOD :(
Bob
Bob
11:33
@WilliamHilsum ow
@RichManson ok, so the MSR does exist, but has absolutely nothing to do with the boot process
@Rahul2001 Win7 does not create a separate recovery partition
@djsmiley2k untested backups arn't backups
2
Bob
Bob
(the MSR doesn't store anything significant either - it substitutes for unallocated space...)
I feel so dirty clicking "use express settings" on a Win10 setup
11:55
@Bob huh?
@djsmiley2k apparently, you can add the -a flag to the directory so it can be changed (amended), but not deleted
o_O?
as root?
Bob
Bob
@Rahul2001 Sorry, meant to ping @RichManson with that one
@JourneymanGeek My backups are "tested"!
Trial by fire, but that counts right? Right?
Lol
@Bob I periodically do test restores :p
i dont care how big ur data is, it isnt hard to 'restore' to some cloufd instances and verify it works
12:03
@djsmiley2k It may not be that simple, depending on how interoperability-heavy your system is. If your service interacts with tons of other services, you may not be able to fully verify that it works
Bob
Bob
@djsmiley2k Yes it is. Impossibly so.
the data, not the service
Bob
Bob
Restoring my 200 GB image to a "cloud" instance would take a literal year to upload.
Wait, no. A couple months if I max it out. The year fig was for the full 1+ TB backup.
you have somewhere else you can fire it up @bob?
im talking about checking the data iss valid, not literal firing up of a duplicate instance
Bob
Bob
@djsmiley2k In a VM, if I really wanted to.
@djsmiley2k I can just mount the images.
But doing anything "cloud" is a no-go.
12:07
is this for a business @bob?
Bob
Bob
There's also the massive issue of data security.
@djsmiley2k Nup. Not this case anyway.
But I've been an admin at a place with ADSL. No backup uploading happening there either.
k, im talking about businesses
Bob
Bob
Also the backups had financial data, so no cloud bs going on either.
@bob offsite stored backups?
Bob
Bob
@djsmiley2k Yea, they were sneakernetted out.
12:09
that works too
still no reason not to test the data
Gitlab is livestreaming their recovery
Bob
Bob
@djsmiley2k They were auto-tested by the backup tool (random files verified) and occasionally mounted to manually double-check.
That was just for data integrity though.
We didn't have a full restore test nor a hardware failure recovery plan.
@RichManson Ok. So Windows always creates a MSR only if the drive is GPT. Win7 chucks recovery (winre.wim) onto C:\Recovery\. Win10 creates the extra partition for recovery. All boot data is still stored in the ESP directly.
brb purging VMs
Hm... xkcd.com/1037 isn't loading in chrome for me... Anyone else getting this issue?
Bob
Bob
@Rahul2001 blank in ff
blank in chrome here
secure connection issue
try loading it over HTTP
check the browser console
@Nzall Chrome redirects to https :/
Bob
Bob
@Rahul2001 remove xkcd from hsts chrome://net-internals/#hsts
@Bob Too much work for a comic :/
Bob
Bob
oh wait xkcd 301's it to https
12:22
:(
Bob
Bob
lol the umwelt server 503's anyway
i remmeber that one
I didn't understand it then I don't understand it now
@Burgi Basically, it switches image based on tons of variables, most notably your browser and your network IP
hm
I finally found a situation where licecap does not work well
12:54
@JourneymanGeek Capturing your Minecraft Let's Play?
no
Trying to capture my answer here
0
A: Graduated brightness adjustment in Windows for TN monitor

Journeyman GeekI'm not massively happy with the results (though I suspect its more my image editing skills than anything else), and it needs some tweaking for your specific monitor but in theory I could get this to work with a few small tools and some sneakiness. I've not tried this for gaming (and it probably...

wut wut wut
the more i read about this gitlab thing, the more I feel like they don't have a clue how to monitor infra.
backups that had 'succeeded' but were only a few kb's in size
wut
@djsmiley2k they trusted the code too much
trust, but verify
D:
I just can't comprehend how that happens
right, now I'm happy with that answer <3
13:08
I mean at jobs I've had, I'll randomly go and check out the backups, how long do they run, etc
just because I do stuff like that, becaue thats' what I do
:/
maybe I am good at my job :o
I find it hard to believe.
Probably so do they!
a customers website (hosted and maintained by a 3rd party) has been down for 8 hours
it took a email to the customer's CTO for the 3rd party to acknowledge the failure
@Burgi you sure you don't wanna talk to us? XD
the 3rd party is a rival marketing agency
lol fail
13:23
this customer splits their business across two marketing agencies for "reasons"
it just surprises me that neither the hosting company or the other agency had proactive monitoring in place
minidump shows nvidia again :(
so annoying...
@WilliamHilsum what nvidia card is it?
ah
Bob
Bob
13:49
@WilliamHilsum what's the bugcheck code?
@Bob BugCheck 133, {0, 501, 500, 0} - is that it?
Bob
Bob
> Bug Check 0x133 DPC_WATCHDOG_VIOLATION
sounds nasty :(
@WilliamHilsum What are the args when you !analyze -v it?
oh wait 501 looks right
@WilliamHilsum are you able to test with the GPU out? onboard graphics?
@WilliamHilsum oh yea, did the memtest check out?
I had to get on with work / wasn't able to complete the memory test, but, did about 12 hours which was a few passes without issue
I never had a single bluescreen in the 2/3 years I've had this pc before I put the GPU in
heat - cpu averages 50 degrees - I thought it was only 35 when I first got it - so, I may try a new fan... I also have a good PSU, but, can't rule that out at all...
    DPC_WATCHDOG_VIOLATION (133)
The DPC watchdog detected a prolonged run time at an IRQL of DISPATCH_LEVEL
or above.
Arguments:
Arg1: 0000000000000000, A single DPC or ISR exceeded its time allotment. The offending
	component can usually be identified with a stack trace.
Arg2: 0000000000000501, The DPC time count (in ticks).
Arg3: 0000000000000500, The DPC time allotment (in ticks).
Arg4: 0000000000000000
Bob
Bob
@WilliamHilsum what was your old one?
Could just be a faulty GPU (or faulty PCIe)
Internal / HD graphics..
I was so happy with it! did everything I needed, but, my friends kept having a go and telling me to get a GPU!
Bob
Bob
14:03
@WilliamHilsum Can you borrow an old card?
>_>
@Bob I didn't know you could do per area ICC profiles
only one I have is a 8800GTX on my shelf somewhere... (customer cancelled order), never used it... I can try to get that tonight
but, not sure it will fit in this case!
Bob
Bob
@JourneymanGeek I'm not sure it's possible either!
...I got a half depth 1060, it made me think "have they left something out" to get it to fit, but, a friend has the same one without any issues
14:04
8800GTX?
Bob
Bob
@WilliamHilsum I mean, at this point it'd help to swap parts and see if that fixes it.
(oldschool yo)
Bob
Bob
@WilliamHilsum The 10xx gen is remarkably low power
My GTX1080 is happy with a single 8-pin PCIe power connector :P
a 1060 runs purely off PCIe slot power IIRC
:/ my 1060 has a PCIe power connector on top
I am jsut so annoyed... 1 week without issue, thought the motherboard firmware fixed it!
I think the 1060s need additional power
14:07
seriously wondering about just scrapping it off and getting a new i7 7 series! I only have a B series mobo/business one, so, just thinking of getting ... for a few hundred quid, thinking it isn't worth the trouble... but that feels lazy!
its the 50s that are passively bus powered
"Passively powered"
You mean slot powered
@Bob so is just about every non-extreme flagship in the last 10 years (barring the R9 290x)
The GTX 1080 actually has a higher TDP than the GTX 980
Bob
Bob
@noitsbecky my GTX 760 took one 6-pin and one 8-pin :\
pretty sure it didn't need anywhere near that capacity though
@Bob yeah a lot of manufacturers overspecced for compatibility/reliability/overclocking headroom reasons
The 760 is a 170W rated card. The 980 is rated at 165W, the 1080 is rated at 180W
Bearing in mind Maxwell introduced a very large efficiency gain using secret tiled rendering or some shit
@noitsbecky meh. My audience gets what I'm babbling ;p
14:18
Hmm GTX 780 was actually 250W, just over the one-8-pin limit, the 770 was 220W though
Bob
Bob
@noitsbecky oh, so my new card has a barely-higher TDP than my shitty old one? :P
@noitsbecky that would've been too close for comfort considering TDP doesn't necessary restrict peak power
@Bob Well, AMD did it with the RX 480... same goes for the GTX 1050
Bob
Bob
@noitsbecky yea, and fried some mobos in the process :\
Bear in mind exceeding the 75w limit on a PCIe slot is pretty bad, given most manufacturers don't engineer much additional headway into those things
But the AWG 18 conductors in an 8-pin 150w PCIe connector could happily carry 1000W without breaking a sweat
Bob
Bob
@noitsbecky kinda expensive to add a trace capable of carrying 6A in the first place, so I can see why they don't
@noitsbecky wasn't it the connectors that were most likely to overheat?
14:24
Hi!
Bob
Bob
@noitsbecky wait, can it actually?
not worth paying to upgrade imho :(
Bob
Bob
3x 12V wires, yea?
@Bob Probably.
@Bob Well AWG 20 can carry 10A if you're willing to tolerate some heat
@Bob 4x in an 8-pin...
@Bob It's not really one trace... it's more like 40
Bob
Bob
14:27
@noitsbecky not according to this :P
3x +12V, 5x ground
@noitsbecky hm, yea, I'm seeing 18AWG as rated at up to around 9A if single-core. That's without significant heat I guess. Some 300 W would be completely safe then, with 3 wires.
500 W should be fine, bit of heat
1000 W would be pushing it but it probably won't melt :P
Well no. Bear in mind a PCIe 1x slot carries like, 15w and a 16x slot provides 75w, the longer slot uses more pins for more power = not using one single 6A trace
@Bob Oh, yeah
Somehow I got it into my head EPS 12v and PCIe 8-pin were the same... probably because all my connectors and conductors are just entirely black.
Bob
Bob
@noitsbecky I almost plugged a PCIe one into my mobo :\
well, almost tried to. the whole keying thing is pretty neat.
Ah, yeah, it was my PSU that had them the same
(Modular PSU that lets you connect either PCIe 6/8-pin or EPS12v 8-pin to any connector on the back)
@Bob The tomshardware article talking about PSU AWGs said "even the thinnest AWG 20 cables still carry up to approximately 10A per wire"
A lot of manufacturers are putting additional connectors on the GTX 1050's anyway for overclocking headroom as well.
Bob
Bob
@noitsbecky I went off this chart
But as I say, 75w on the slot should really be adhered to, exceeding 150w on an 8-pin connector is a non-issue
Speaking of which I should mod my GPU BIOS again now I understand supply rails better
(When I bought it I modded it to allow 80w on the slot and 80w on each of the 6-pin connectors, turns out that's actually limiting my overclocking)
Bob
Bob
14:36
@noitsbecky except when my PSU shuts off that cable cause it went over the OCP limit... whoops? :P
@Bob lol
Your new PSU or old one? :-P
Bob
Bob
@noitsbecky new, ha
my old one would probably just explode
My HX850i allows you to configure either multiple 20A OCP rails or a single 50A rail or turn off OCP entirely
Bob
Bob
it's actually set to 40A (default) so 480W
There's no per-cable OCP though :-/
Bob
Bob
14:37
pretty sure my GPU would burst into flames first with that current
@noitsbecky yea, that's AXi only AFAIK
(Also those numbers I made up are just rough memory, but you get the general idea)
@Bob Oh you have the AXi?
That should be more configurable, no?
welp
i might be getting fired
Bob
Bob
@noitsbecky Yea, I can adjust OCP anywhere from 20A to 40A, and enable them selectively
don't think I can do single rail with configured OCP though
> Yes. With a default OCP of 40A on each 8-pin connector on the modular interface, there is MORE THAN ENOUGH power for any peripherals you plug into the PSU. The ability to toggle it to single +12V rail mode is only there for those people that refuse to believe anything other than "single +12V rail is better".
Huh maybe it is per connector
I can't remember, and not got time to read the manual atm. Still, close enough
14:39
the rival agency have kicked up a fuss that I have emailed the customers CTO
Bob
Bob
@noitsbecky would have to be, to make any sense. either per connector or groups of connectors
also apparently my GPU is currently not drawing any power at all from the cable
looks like it idles around 25% TDP
... going back, sorry! I have a gut feeling that there is something wrong with my motherboard... every crash I have ever had has been since putting the card in... It always seems to be 2/3 very close together... the stability patch (which goes back to 2015) seemed to make it last about a week...
Bob
Bob
pushing it up to 100% I see a 12A draw (PCIe #2)
12-16A
I'm just wondering if the B chipset board were designed for graphics being low cost/business ones... I just had an extra i7 from a project and bought a load of these in bulk :/
Bob
Bob
@noitsbecky ^ looks like it does peak at around 200W and hovers around 156W (13A) but tries to stay under 150W long-term
Maxing out the CPU pulls 7A from PCIe #1 (EPS)
14:43
@Burgi Explain to them it was necessary because they weren't doing their job monitoring the website and you didn't want the joint customer to suffer any more downtime than was necessary
oh i will
@Burgi cc customer ceo and cto
Bob
Bob
@WilliamHilsum a 1060 is hardly demanding and probably not uncommon in graphics design studios
@DavidPostill can't, they have requested i don't send any more emails on the subject
@Burgi huh? They are not your boss. Don't allow some other company to stop you doing the right thing.
14:46
my company has told me to stop emailing
@Burgi :/
the complaint went to the owner of our company
@Bob :( you are right... argh, just fancy screaming... maybe I should do a reinstall, probably have so much crap over the years as well which wouldn't help
Bob
Bob
@WilliamHilsum with random crashes that could just be luck
shrug
i haven't done anything wrong
Bob
Bob
14:47
it could still be the gpu being faulty...
@Burgi Sounds like other company are just trying to cover up their arses mistakes from joint customer.
agreed
@Bob I literally did the same thing last night while building my new rig
As it crashed after a week, I don't think replacing with a 8800gtx will help too much :( ... but, I got it from Amazon, I can see if I can return it
15:44
#ThereIsNoProblem
16:17
This guy HNQed because he didn't know what emoji were :/ superuser.com/questions/1173773/…
lol
actually its a little deeper than that
@JourneymanGeek ??
firefox handling emojis differently
and not knowing what they were
@JourneymanGeek He managed to create a document.title animation with them without knowing what they were... Seems very unlikely to me
@JourneymanGeek It's up to every engine to render their emoji however they like.
Firefox renders one way, chromium the other, android the other and whatsapp on android even another way
16:24
@Rahul2001 either he's copy pasting, or groked the HNQ lottery very well
@JourneymanGeek :/
Never put down to malice, what can be put down to stupidity
Applies to both the emoji question, and what happened at gitlab
@djsmiley2k What happened at gitlab?
wat
how can you not know?
It's like the actual news of the year
They deleted the db...
> TODO after data restored: Somehow disallow rm -rf for the PostgreSQL data directory?
LOL
Sid: try to undelete files?

CW: Not possible! `rm -Rvf` Sid: OK
i love how doing actual aerting is 4th
makes me think they have very little visability of the running systems
> out of 5 backup/replication techniques deployed none are working reliably or set up in the first place
You have backups, right?
a untested backup, isnt
and also doesn't produce "files only a few bytes in size." ;p
16:46
yup
The clues were all there ...
Operator Error + No backup == Disaster
feels like they trusted the software, but did not verify
It nice to see someone being so open about what happened ...
> we had an issue with the cluster handling requests for all 'S' sites on the cloud
wtf...
S is for Spam?
16:52
shrug
Backup is easy, restoration is the hard part
nothing but connectivity issues for all our customers today
are they all on gitlab?
@Burgi Did someone change the alphabet when nobody was looking? ;p
@Extrarius IMO the trick is to test your backups BEFORE crap happens
16:57
@JourneymanGeek Only if you really want to be able to restore from the backups. In that case, you should routinely restore backups to non-production hardware just to be sure the process is staying valid. Not to mention having scripts that run as part of the backup that make sure the backup is a reasonable size
@Extrarius "Only if you really want to be able to restore from the backups" Wat!?
@djsmiley2k no
What are backups for if you're not going to potentially use them to restore something?

« first day (2371 days earlier)      last day (2945 days later) »