« first day (1790 days earlier)      last day (3220 days later) » 

Bob
12:00 AM
hmm... actually, I might try sticking the rescue boot at the top to see what happens
that'll have to wait till I get home
...and get some sleep
 
ahh
New monitor ordered :)
 
12:20 AM
Just a sanity check....I should question an answer when the i contains statements like "then maybe try" right?
* it
 
no offense...
in your case, better not to question at all ;p
Link it here ;p
 
-1
A: How to clear the terminal under Cygwin?

XupicorIf you don't have clear then maybe try cls?

 
there, taken care of
(Sorry about that. People just react so... badly... to your comments)
 
can I just start using emoicons when people submit duplicates of questions with 80+ votes [sarcastic face]
In this case, reset is the command I am looking for, not clear. I know reset is the built-in command to use. — jww 2 mins ago
I am so confused...."Under Cygwin, it appears reset does not work"...."I know reset is the built-in command to use."
walking away from the desk, it has enough dents, from HOTS random matches.....
 
lol
nope. You can VTC as duplicate ;p
 
12:31 AM
trying to find the duplicate, its more or less answers over at SO, but can't believe somebody didn't ask on Superuser.
 
It's not a dupe of a SU question though.
12
Q: Cygwin alternate for windows?

ukanthI would like to have some of the most frequent commands which I use in Linux environment in my windows box , I know there is program called Cygwin. But I don't want to use such a heavy one . All I want is a simple ported unix commands for windows. Is there any Utilities like that ? PS: I've cre...

Check out the PS at the end of that question.
That's basically exactly what the user is asking
4
Q: Is there a command to clear a Cygwin console?

erikvoldIs there a command to clear all of the history in a Cygwin console window without losing the .bash_history?

 
there we go :-)
 
Interesting that he claims reset doesn't work, when it's quite clear that he ignored the error about requiring an additional parameter to function.
 
I missed that.
Odd i can't find a single example of reset....[sad face](google.com)
 
Yea... he also acknowledged it in a comment that I missed.
 
12:53 AM
Do people here generally A)store all their data on an internal hard drive, or B)have lots of data stored on external hard drives?
 
yes
one is none
 
A or B?
 
Networked internal drive for primary backups + Networked internal drive on a different system for secondary backups and a external drive for third level backups
 
^^ I just added A and B labels for clarity!
So it looks like everything is internal?
then backed up from there
 
its on different systems tho
 
12:56 AM
yeah, the backup is remote..
I get that..
 
My desktop and laptop backup to a shared drive on a fedora system.
THen that's backed up back to my desktop
 
And the original that it is backing up, that is internal
 
which is then copied over to an external drive
 
Is that all the data you ever use, stored internally?
 
Annoyingly, I am having troubles on the laptop -> storage box backup
yup
:(
3 week lead time on my monitor.
 
12:58 AM
What capacity do you backup?
like how big is the original
 
There a different between internal and external hdds?
 
@Ramhound: speed, I guess. It just use an external since I wanted one more layer
internals are less... messy
 
How many Terabytes are you backing up?
 
1:20 AM
Maybe Grandpa Terminator will put a smile on my face...
right now i just want to build one....<sigh>
 
Interestingly, compressing a large file as lzma2 produces uneven spiky usage across eight cores, bzip2 does a 100% uniform usage.
 
@HackToHell it's a many-stage algorithm, and it does multi-cores in some parts and not others
imagine it like a sieve whereas bzip2 is more of a conveyor belt
 
1:37 AM
Ah that explains it
 
Bob
1:50 AM
Speaking of compression... I have LZ4 enabled on my ZFS pool :)
@allquixotic I just hope ZoL is stable. And lxd, I guess, but it doesn't even work yet :P
 
There's an ongoing series of spam posts promoting various male enhancement supplements. Can we track it down in some manner?
I lack the tools to effectively fight this spam beyond flags.
 
@DragonLord Like some kind of Internet Police?
We could shorten it to InterPol! :o
 
2:06 AM
@MichaelFrank This (10k only) is the most recent. I'm afraid we might be seeing Bayesian poisoning attempts on the global Stack Exchange spam filter.
 
I can't view those. But since they're all deleted, I guess the system is working?
But I suppose you're more concerned with what happens when the spam is smart enough to flood the network instead of just handful a day ending up in the review queues.
 
Not just male enhancement supplements, but all sorts of faux personal enhancement products.
As a 10k user, I already have access to a fairly large set of tools to track down spam.
It's just that I can't do much more.
I might be able to do some sleuthing and add them to my spam blacklist, though.
It's been getting worse over the last few weeks.
 
2:29 AM
@DragonLord: Actually other than being able to see your flags, mod level tools arn't much different
 
WHOIS tracing isn't doing any good. They all seem to be on shared hosts.
A few are using Namecheap shared hosts, but I don't see any obvious pattern.
Likely a cybercriminal ring operating a botnet.
Loads of bad domains on each of these shared hosts.
On another topic...
Multiple questions about CPU microcode and micro-operations.
Can we try to help this user more directly?
I'm getting a bit suspicious here.
 
2:49 AM
Be advised that we have a network-wide spammer:
It's been brought up on the Tavern.
Can someone suspend this user network-wide?
 
3:06 AM
could report them to namecheap directly. They're pretty good about that sort of thing I am told.
 
@Bob ZoL has been extremely stable for me.
The more features you enable, though, the more chance you'll hit a bug.
"Basic" ZFS, maybe with RAID-Z, is probably the most heavily tested and stable path.
Layering on compression, dedup, crypto, whatever...? Risk.
 
3:40 AM
Morning.
 
Bob
@allquixotic Crypto isn't available.
Dedup is so memory-hungry I don't feel like enabling it, even on a 600 GB array with 64 GB RAM.
Compression is generally recommended... I'm uncertain.
 
LZ4 is a stable compression algorithm by itself (if done properly), so that's probably the least risky overlay to add
 
The guy on SF who does a lot of ZFS seems to think you need a ton of ram and a SSD for caching to do ZFS dedup.
 
Bob
3:58 AM
@JourneymanGeek Yes.
Dedup is recommended as something like 5 GB RAM per TB pool space.
But dedup takes up max of 25% of ARC, so ARC needs to be min 20 GB per TB
 
blehgoahgofdanbljfdanbofdab
 
Bob
My 600 GB pool is relatively minuscule. When you deal with 2 TB, 10 TB, 20 TB pools... ouch.
 
token size issues...
 
lol. work's hitting 1pb soon.
 
Bob
@JourneymanGeek Speaking of SSDs, though... my disks are SSDs :P
3x 300 GB (I mean "Go") Intel SSDs.
Actually, I should benchmark it sometime.
 
4:01 AM
0_0
What raid level?
 
Bob
@JourneymanGeek raidz1
(equiv. raid5)
 
ahh. The 3 disk layout threw me off
 
Very scary? ;p
 
That's really awesome.
 
Not any more :)
 
Also, insane ways to compress a large file in place:
0
Q: How to compress extremely large file in place

Ben SandeenI need to compress a 646.4GB file on a 1TB external HDD. This HDD currently has about 238.6GB free. However, since the straightforward compression methods create a new file, I'll not be able to compress this file, which is a .vhdx file (a Windows system image). I have access to both Windows (10 T...

It is one of those questions where I actually want to scream: Just use a second disk.
 
lol
So I did.
 
Back in 30 (cycling)
 
5:34 AM
ya always have to wait will your painfully out of space before doing something about it, then it starts with the really stupid methods getting it from 97% full to 88% full, but still not enough room for the next stuff.
Finally one pulls head out of buttox, and gets 3X more space than they presentally have , and proceeds to fill it up just about as fast :-)
This should be a "law" like Psycogeeks law of find a hole and fill it. This applies to more than just storage. Ex: put a married team of humans into a 3 bedroom house they will have 2 children (then one more, to 120% filled) But put the same married couple in a 5 bedroom house they will have 4 children (then one more).
If you bought a small boat, in less than 1 year you would find that it was to small for all the people you haul along, but no matter how large, it will always get filled again. Soon your driving the ferry and having to turn down passengers :-)
Make a cpu processor that is 3 times overall faster than the one 10 years ago, and by the time you get done the programs are 4 times slower and more beloted than they were. Create a GPU that can do massive resolution display in real time, and before you get done some person tries to put 4X the resolution in it, and some game can barely crank out 20FPS again.
Most man-made lakes are not dug-out like converted quaries, no they found a hole, blocked one end and filled it up :-) It is why septic tanks overflow, why the earth is 4/5ths water, and why colleges exist (regular school does not fill the hole completly:-)
It is why Twitter took off the message size restriction, how we went from 14.4k net bandwidth to 35meg bandwith and it is still slow. Why california highways look like huge parking lots during commute hours. And why before we effectivly have conquered the moon, they are already on the way to mars.
 
Bob
 
7:08 AM
The IBM T220 and T221 are LCD monitors with a native resolution of 3840×2400 pixels (WQUXGA) on a screen with a diagonal of 22.2 inches (564 mm). This works out as over 9.2 million pixels, with pixel density of 204 pixels per inch (80 dpcm, 0.1245 mm pixel pitch), much higher than ordinary computer monitors (which typically have about 100 pixels per inch) and approaching the resolution of print media. The display family was nicknamed "Big Bertha" in some trade journals. == IBM T220 (9503-DG0) == The IBM T220 was introduced in June 2001 and was the first monitor to natively support a resolution...
I'd like to put this out there and point out, we're just hitting that res and pixel densities on desktops again.
(we do have better refresh rates tho)
 
Bob
8:08 AM
Ahh, I like the shops where I can go up to the counter and ask if they have Pilot converters and not have them look at me like I've got three heads.
They don't actually have any, but at least they know what it is :P
(still waiting for the Massdrop delivery... why is it so hard to find those here?)
 
So... most shops? ;p
 
8:53 AM
@Bob Massdrop always take sooooo long.
 
0
Q: If Bounty questions are unanswered Rep should be returned to the user

David GoldingI a few months ago asked a question and after several weeks no one could provide a suitable answer so I put a small bounty on it for 24hrs, this didn't attract that many people so I then put on a larger bounty. After the bounty expired I still didn't have the answer to my question and I had los...

 
9:09 AM
I'm tempted to say "Its a risk you take"
 
Bob
@JourneymanGeek Nah.
 
(I've had times when I've gone into shops looking for what should be a COMMON cable and gotten odd looks)
I'm not talking 13W3 (with dip switches) to VGA. Talking IDE when IDE was common, and not called PATA.
 
9:25 AM
Or ATA :-)
Or mistress and slave drives :)
 
The last part may have some people confused.
 
So did asking for 13W3 cables for a big 21" Sun CRT for my i386SX (with ET4000 VGA)
 
lol. I just ordered it online.
 
10:23 AM
With wierd, esoteric crap its just so much easier
 
Bob
11:01 AM
Why is it so freaking hard to benchmark disk/FS perf in Linux?
I don't want to dive into a million and one benchmarking tools. Hasn't anyone written an easy script for basic results? O_O
 
Bob
@terdon Yea, I've looked at bonnie. The output is a nightmare.
sigh guess I'll just have to make my own tables
Nope. Completely bloody useless.
 
11:20 AM
@Bob Try asking on U&L. I'm afraid I've never had to do it so I can't really help.
 
Bob
I'm not going to sit through two hundred gigabytes of single byte writes 4+ times.
And there seems to be no way to disable the single-byte test even with the chunk size specified.
Who designed this pile of crap?
@terdon There's already a U&L question there.
The answers... are far from satisfactory.
Except: I'm trying to benchmark a FS, not a physical disk. dd is a kludge. The mentioned IOPS tool would be good, but it's read-only. The giant list of questions disguised as an answer is no help. The suggestion of some GNOME utility is pretty useless over SSH. The suggestion of FIO... well...
It's supposed to be flexible. Great. I compiled it. I ran it. Now what? Oh, I have to create my own workloads?
This is completely ridiculous.
 
What's wrong with FIO? I've never used it.
 
Bob
@terdon See above re: flexible.
There's a dir of example workloads, but the names are completely nonobvious, and there's no comments inside.
 
Yes, how is it not? I'm not saying it is, I have no idea.
@Bob Oh, great.
 
Bob
@terdon Basically, I don't want to spend hours trying to figure out how to use a benchmark tool.
This is something I could do a quick and dirty test of in 10 mins, max, on Windows. I assumed it would be similarly easy on Linux.
So far, I've been proven wrong.
(Granted, the Windows utils were graphical... but the philosophy of the OSes just differs like that)
 
11:27 AM
@Bob Well, it would be, I guess, if you had GUI access. You could use that GNOME tool. You can also do that over ssh, by the way.
 
Bob
sigh back to searching
@terdon It's a completely clean server I'm intending to use as a VM & container host. If I installed GNOME on it, I'd have to completely reinstall the entire server afterwards.
 
Damn
 
Bob
At this rate I might as well just go with dd, dirty as it seems.
Normally I'd be fine with a server reinstall.
 
What do you want to benchmark? read/write operations? If so, dd should be fine, right?
 
Bob
But this one is... special.
@terdon Primarily. And IOPS.
Problem is, dd probably can't sustain the IOPS required.
 
11:29 AM
You said that was RO, what do you mean? That you need to mount the FS in ro mode?
I'm talking about the IOPS tool.
 
Bob
@terdon That tool doesn't test writes. At all.
@terdon Details: I've got a raidz1 array over three SSDs.
I'm pretty sure dd is gonna bottleneck on a single thread.
For comparison, here's the iops tool results:
/dev/tank/bench-block, 214.75 GB, 32 threads:
 512   B blocks: 97644.4 IO/s,  47.7 MiB/s (400.0 Mbit/s)
   1 KiB blocks: 97891.5 IO/s,  95.6 MiB/s (801.9 Mbit/s)
   2 KiB blocks: 100023.3 IO/s, 195.4 MiB/s (  1.6 Gbit/s)
   4 KiB blocks: 104285.2 IO/s, 407.4 MiB/s (  3.4 Gbit/s)
   8 KiB blocks: 101878.1 IO/s, 795.9 MiB/s (  6.7 Gbit/s)
  16 KiB blocks: 96419.2 IO/s,   1.5 GiB/s ( 12.6 Gbit/s)
  32 KiB blocks: 87206.0 IO/s,   2.7 GiB/s ( 22.9 Gbit/s)
  64 KiB blocks: 71240.8 IO/s,   4.3 GiB/s ( 37.4 Gbit/s)
That's exactly what I'm looking for (though perhaps with fewer block sizes).
Except this tool doesn't do writes.
And I'm fairly certain those results are lying.
 
That's weird
 
Bob
Because it's not writing any (random) data, and I have FS compression enabled.
So... the simplicity of use and the output format are perfect. But the tool itself doesn't do what I need.
Other tools do what I need but basically require a three week course to understand the params and outputs.
 
Is there really no choice to write random data?
 
Bob
@terdon Not in iops.
 
11:37 AM
@Bob Yes, I have to admit that does happen in the *nix world.
 
Bob
I mean, look at this answer: serverfault.com/a/563805/174582
> On at least Linux, all synthetic benchmarking answers should mention fio - it really is a swiss army knife I/O generator.
I don't want a "swiss army knife" with a hundred or so options I'll never use.
It would be fine if those were hidden behind params and it had a default basic test.
But that does not appear to be the case.
> The list of features and output goes on and on.
urk
 
@Bob heh. Yeah, but that is usually a good thing since, after all, this is a tool for professionals and power users. Your average Joe will never need it.
 
Bob
oh hey this might be the right workload
 
This doesn't look too bad:
Have a look at the Simulating servers section
 
Bob
11:54 AM
@terdon Hmm...
I just went and tried the ssd one from the examples. Nope.
Must be the wrong engine.
fio: looks like your file system does not support direct=1/buffered=0
fio: destination does not support O_DIRECT
 
Dang. I do recommend you ask on U&L though. Explain that you're trying to run this tool and are looking for these results and ask how? There are quite a few experienced sysadmins there who should be able to help.
 
Bob
> # This job file tries to mimic the Intel IOMeter File Server Access Pattern
hmm
oh hey that one works
or not
 
So this Adder KVM-over-IP device I ordered the other day arrived today
Pretty neat so far
 
Bob
@terdon Hm. I don't think this is important enough to warrant asking a question (I'm reasonably happy with the knowledge that it's "fast").
It was more a curiosity thing.
But I'm still pretty annoyed at just how hard it is.
It shouldn't be this difficult.
@terdon Looks like iozone might do what I want :D
 
12:17 PM
Pretty sweet. VNC all the way
Seems like I picked the perfect day to try this...
 
@OliverSalzburg Please change your password to password1, then change it agin to password2, then to .... pasword 10. Now for the final change ...
 
@Hennes I opted to just change the password policy
In this case, it's okay ;D
The mouse calibration on this thing is kinda off :\
 
12:59 PM
@Bob Yay!
 
Bob
                                                              random    random     bkwd    record    stride
              kB  reclen    write  rewrite    read    reread    read     write     read   rewrite      read   fwrite frewrite    fread  freread
       104857600       4   149214    63730   350793   348857
It's gonna take a while
 
@Bob Well, it isn't. For a simple test, dd will do perfectly well.
 
Bob
@terdon I'm still fairly certain dd can't cope with the iops :P
 
Don't really know.
 
hi
Oh boy. A pull request for ZFS, Bob? if you are thinking of adding that... you'd have to rebuild the kernel module :S
Not that that's hard, just might make it interesting if you want to keep up to date
 
Bob
1:12 PM
@allquixotic I'm not touching it :P
Just nice to know that it'd being considered.
 
looks like the original code was for OpenSolaris/Illumos, so the ZoL port is very new and experimental, comparatively
since they have to port a lot of stuff over to the Linux way of doing things
 
posted on July 01, 2015

The next Stack Exchange beta site: emotionallyscarredbyc.stackexchange.com.

 
1:31 PM
5
Q: Why is the Reciprocal Public License OSI-approved but not FSF-approved?

congusbongusThe Reciprocal Public License (RPL) is interesting in that it is the only license both OSI-approved and not FSF-approved. It is a copyleft license similar to GPL but with a "reciprocal" clause: Regarding deployment, under the RPL your changes, bug fixes, extensions, etc. must be made availa...

An interesting license - copyleft, OSI-approved, but not FSF-approved
 
1:47 PM
ooh
@OliverSalzburg looks shiny. and they have an office here.
How much do they cost tho?
 
Hi there :-)
I asked this question in SO (stackoverflow.com/questions/31157891/…), but now I wonder if I should have asked it in SU instead
Should have I? should I request it to be moved? (I didn't have much feedback in SO)
 
@Peque you asked it four hours ago - that's hardly conclusive as a sufficient time to get feedback
but if it's off-topic on Stack Overflow (I'm not sure if it is or not), you could request it to be migrated to SU
there's been a crackdown lately of migrating posts that are on-topic for the site they're originally posted on, except in rare circumstances
 
oh, ok, maybe I'm just used to really quick answers in SO (like in 2-5 minutes :-P)
I'll wait then, thanks :-)
 
don't cross post, though - if you really do want it migrated, flag it for a SO moderator and request it
link them to this next chat line I'm about to post:
I reviewed that question asked on Stack Overflow by Peque and I believe it would be on-topic and a "good" question to be asked on Super User. Signed, 23k rep SU user.
 
@Peque That would also be on topic on Unix & Linux and Vi and Vim. Pick only one though.
 
2:02 PM
sure, I wont create a new question on any other stackexchange site :-)
I'll wait in case I can get an answer in SO
if not, I may request it to be migrated
 
sounds like it's probably best to migrate it to vi.stackexchange.com
 
thanks for your help
 
never heard of that site before
 
/me neither
 
 
2:04 PM
@allquixotic It's newish. Still in beta.
 
@JourneymanGeek It was pretty cheap. The switch + 12 VGA/USB connectors + rack-mount brackets = ~1800EUR
 
2:53 PM
Well, not cheap for home use, but cheap compared to other KVM-over-IP stuff I looked at
 
Oh, a single host one might be handy
 
3:38 PM
cmd.exe is horrible
 
Yes, it is
 
4:08 PM
that might not be SFW, GFSE. Can someone nuke that post from chat? Not that I care, but Bob might
don't have to remove it from GFSE itself, just the chat feed
 
Joe
On-topic clarification request: Want to ask a question about customizing gmail interface (in web-gmail). Am I correct that belongs on webapps and not on superuser?
 
4:24 PM
Question can be on-topic for multiple sites, but webapps might be the 'more appropriate'
Having said that: I did not read the FAQ for webapps yet.
 
Joe
Well, sure
most of what I was looking for was "not on superuser" :)
The tag wiki seems to suggest this, but it's not 100% clear to me - I am specifically talking about gmail on a desktop (not on a tablet/mobile), but it would be in a browser.
 
Office temp 84F (28.7C).
Not fun.
 
Joe
You can have my randomly dimming/brightening lights in exchange for that if you want. I wouldn't mind a bit warmer than the 66F we have right now...
 
But 19C is lovely
Shirt weather
 
Joe
It's entirely confusing. Summer in Chicago is usually nice and hot. I had to dress my children in long pants and jackets today.
19C is lovely in the fall or spring... but not when you've put away most of the wamer clothes ;)
 
4:30 PM
I'd take it. I tend to enjoy colder temps than most
pretty darn hot lately in MD... not so much today, but it has been
 
Joe
It's also keeping my peppers and tomatoes from growing properly. They need a nice warm climate to remind them to fruit.
Sigh.
 
4:46 PM
My peppers are growing nicely.
As is the coffee plant.
(See: I am prepared for the zombie apocalypse.)
 
Joe
Lol. Perfect.
I mostly am happy living in Chicago, but the one thing I'm not happy about is that I can't grow citrus, avocado, etc. trees here (without taking them inside for the winter, and I'm not organized enough to do that.)
webapps.stackexchange.com/questions/79829/… posted over there; we'll see if it's judged on topic. An older similar question in 2013 for a different chnage was closed as 'too localized', but I think that SE's moved on from that close reason and now considers that on topic (if not very useful after it no longer applies).
I'm surprised coffee grows successfully in MD, though - do you have to take it inside?
or is it frost-tolerant up to a point
 
are you also not happy with the local ball clubs?
Cubbies and White Sox :P
 
I never planned on coffee. I just got a cup a few years back with some sprouted seeds in it.
Those grew. And grew. And grew
 
Joe
5:05 PM
Ahh, in the window. Nice.
 
@allquixotic a little late but done.
 
Joe
@allquixotic As far as ball clubs, our hockey club makes up for any deficiencies in the ball clubs. Also, we did win a world series ten years ago, so we're good for another fifty or so.
 
5:27 PM
Time to fix dat dark theme!
 
5:39 PM
@OliverSalzburg Dark theme?
 
Aha
@OliverSalzburg I dunno if many people use fluxbox these days but I posted to my github a few themes I created for myself, since I use it on a daily basis but never really liked the themes that it came bundled with.
 
I don't use the theme right now. I'm only fixing it because of complaints ;D
 
:P
That theme I'm using right now works very well with transparency set. :)
I've come to like fluxbox a lot. At first, I just liked it being different and minimalist. But it's also super easy to configure. Harkens back to Windows 3.1 and how you basically just had a collection of a couple .ini config files that controlled all of its behavior.
Creating a theme basically just means making a text file with a list of options set to different values and dropping it in ~/.fluxbox/styles
Of course it can be more files than that if you include pixmap textures but you don't have to.
I haven't explored that just yet :P
 
5:56 PM
@Joe Not just in the window.
Kitchen:
Outside the kitchen:
And a few more outside aswell on the balcony
 
6:29 PM
There, the insane question is answered.
 
 
3 hours later…
 
1 hour later…
10:29 PM
The worst part about being thirsty at work... Getting back to your desk and the glass of water you just filled is now empty.
 

« first day (1790 days earlier)      last day (3220 days later) »