Root Access

Now the real annoyance is that I need to resize the partition because I'm running out of space for backups. And those run daily obviously. So... will I have enough space left to backup until Friday? I better, because I sure as hell can't take the server offline sooner

Psycogeek

9:28 AM

ahh you never have enough space, and when you do it takes forever to manage it all.

Oliver Salzburg

That is problem A today. Problem B is that our BlackBerry app doesn't work. It only works when the web inspector (chrome devtools kinda thing) are connected.

Now I gotta inch my way through our git history to find the commit that broke the app. So it's checkout, build, enter PIN, wait for deployment to finish, wait to see if app starts properly, repeat

…fuck

Bob

@OliverSalzburg git bisect

Oliver Salzburg

@Bob Neat-o

But I think I might already be onto it. Mothereffing splashscreens

Bob

9:52 AM

O_O

Win8.1 added coloured fonts, and emojis: istartedsomething.com/20130628/…

And FF is the only browser to support them: bugzilla.mozilla.org/show_bug.cgi?id=889401

Even IE11 doesn't.

Chrome... Chrome's just absolutely fucking shit at font rendering.

Hm. Looks like it should work in IE11.

Journeyman Geek

🐕

For science!

Bob

XD

@JourneymanGeek what's your favourite song for testing audio?

Oliver Salzburg

@JourneymanGeek I want to see it too :(

Bob

10:07 AM

@OliverSalzburg Open FF, profit!

You get colour too if you're on Win8.1.

Oliver Salzburg

@Bob I'm on 7

Bob

@OliverSalzburg Then you get a black outline.

Oliver Salzburg

Bob

o.O

ooookay

most emojis I've tried work on 7

Journeyman Geek

looks different on android ;p

Bob

10:14 AM

I'm trying to figure out how I should test the fort properly :P

also, it's bloody loud at max

my headphones are as loud as most cheap speakers o.o

Journeyman Geek

@Bob: Thats normal

dropbox.com/s/wxvzryx0nd2rfy4/…

wierdly, this is the song I use to test audio gear. The first bit with the car starting sound especially

Bob

ok, let's try this

not flac? :P

not that it's gonna make a difference if encoded properly

Journeyman Geek

yeah

Its not a 1% test, Its a 'how does this sound to me' test

Bob

it probably helps that I can't understand it :P

focus more on the audio than the lyrics

Psycogeek

@Bob funny they had color font including color font editing ability for the user back in the amiga :-) imagine in 5 more years Apple will invent it and patent it too

Journeyman Geek

10:20 AM

yup

Bob

@JourneymanGeek I really need to find some way to do a blind test :P

I think the fort sounds better, but it's not night and day... so hard to be certain

Journeyman Geek

@Bob: Night and day is hard

Bob

@JourneymanGeek the engine starting is surprisingly good at highlighting the differences O_O

far more than the rest of it

Psycogeek

the problem with color font was the endless possibilites of variations, not only the color but creating ways they are embossed, outlined and all those other goodies (using color). the endless possibilites would make for a font folder that would take a single person just to manage. Having the variations in the programs using the font (IE Word) seems to be a better method, have 500 fonts and vary them after placemtn.

Journeyman Geek

IKR?

Bob

10:24 AM

@Psycogeek probably raster, though

you can still get colour raster fonts now

Psycogeek

@Bob I dont remember that well, but the editor was vector based. Not sure if the addition was.

Bob

huh.

Journeyman Geek

@Bob its kinda weird. Its not 'music' or lossless but that is the track I use to test stuff. Oh.. And a live bootleg of tori Amos covering total eclipse of the heart.

Psycogeek

Ya know they were even doing 3D digitising back in the amiga days, capable of creating a 3d model of a human for 3d import using this arm thing you would move to the shape of the item. Stereo eye viewing has been around long before there were lcds, and microsoft did not invent windows :-)

It is just that now most of that stuff is "usable" could have become something other than gadgets that work half assed.

Bob

@JourneymanGeek Well, I suppose it's best to test with what you like :P

tbh I normally listen on my phone now

Psycogeek

10:40 AM

Plus they actually had HD 1200 horizontal res (but in analog) using seperate hardware items that would attach to said computers, even before the computers themselves used those resolutions for desktops. A "video" card back then was for "video" :-) not games.

Journeyman Geek

@Bob: oh, with the tori amos one, its purely acoustic. and Its good for testing for noise in the silent bits

Bob

@JourneymanGeek Heh. Was trying to test for that with Gymnopedie No. 1... from a YouTube video.

Noisy as heck.

Then I tried via Spotify... ahhh much better.

Stupid YouTube.

Oliver Salzburg

11:04 AM

Neat, I just got my first infrared thermometer

That is pretty entertaining for a while

Psycogeek

what is really cool is the low priced IR thermomitors work pretty good, at those prices everyone should have one.

Oliver Salzburg

@Psycogeek Yeah, it was like 15EUR. I always assumed they would be much more expensive

Psycogeek

I assume they used to be, just like IR cameras with color coding of the different temps still are ?

Bob

@OliverSalzburg heh, reminds me of getting a laser pointer

shined that shit everywhere

@OliverSalzburg the cheaper end is about half that?

Oliver Salzburg

@Bob It was actually 11EUR. I assume you'll get one for 5 on DX. But I wanted it now not next year

Bob

11:10 AM

lol

Bob

11:33 AM

@JourneymanGeek fort gets a tad warm on the bottom :P

surprising for such a low power device... but it has been overh alf an hour

Journeyman Geek

not too surprised

Bob

@JourneymanGeek for the purpose of avoiding Beats, works perfectly

Journeyman Geek

;p

Bob

for sound quality, it's not worse and probably slightly better than the builtin card

but I can't tell normally anyway :P

it might be a lot better and I jsut don't know

Journeyman Geek

lol. It probably is!

slightly better that is

jokerdino

12:20 PM

@allquixotic DUUDE. Haha

Journeyman Geek

12:39 PM

@allquixotic hm. Needs mips. ;p

Bob

@JourneymanGeek Don't forget PowerPC and ARM and AVR.

Journeyman Geek

AVR is a microprocessor ;p

and ARM is ~~granted~~ common and hipstery.

and I suppose we need z80 and m68K ;p

Bob

@JourneymanGeek So?

Journeyman Geek

you generally don't have computers running on em?

Bob

(I think you're looking for micro*controller*, but that doesn't mean you can't run Linux on it anyway.)

Journeyman Geek

12:48 PM

yes and I guess so ;p

Journeyman Geek

1:06 PM

@Bob oh. I remember this one. It is pretty crazy

Simon Sheehan

1:48 PM

@Bob 4 hours to boot... oh lord.

Oliver Salzburg

@OliverSalzburg dx.com/s/dt8380 18USD

allquixotic

hai

Bob

2:11 PM

@allquixotic 'lo

Journeyman Geek

mid

Bob

here we go again! :P

Journeyman Geek

2:52 PM

;p

that's the puppydog emoji I used earlier.

Bob

o.O

doesn't seem like somethng that needs to be deleted, @JourneymanGeek

Journeyman Geek

@Bob: needs cropping ;p

→ 1 message moved to Recycle Bin

allquixotic

3:18 PM

wait, what?

what's the puppydog emoji and how do I get it?

Bob

@allquixotic Unofficial Unicode codepoint, available in Win8's Segoe UI Emoji, and on Android's system fonts. And maybe iOS and OS X.

Actually, I'm not sure if it's part of some Unicode standard.

5 hours ago, by Journeyman Geek

🐕

Works in FF/IE on Win8.

Don't even think about Chrome.

Other (face) emojis work on Win7 as well, though uncoloured (that requires 8.1).

allquixotic

hah

that's cool :D

running Chrome on Win7 in the office, so no go

Bob

If you ask a question about font/Unicode support in Chrome, the answer is generally "!!no"

It's ridiculous they haven't managed to sort this out yet.

Darth Android

What's the expected behavior? SuperUser is specifying the Verdana font in their css, which renders the emoji as a box.

Bob

@DarthAndroid Unicode fallback fonts.

On Windows 8, it should fall back to Segoe UI Emoji.

allquixotic

3:25 PM

@DarthAndroid it's supposed to display a ~~wolf~~ puppy, like this:

28 mins ago, by Journeyman Geek

Bob

That's the Android one.

Here's the Win8.1 one:

5 hours ago, by Bob

allquixotic

hah

Bob

(it's not great - the face emojis are much better)

@DarthAndroid Oh, you must be using FF or IE.

Darth Android

Hmmm, seems chrome only falls back to the fonts listed by the website

Bob

Chrome doesn't do ~~Unicode fallback~~ anything remotely sane or smart with fonts.

allquixotic

3:27 PM

!! s/ with fonts//

ChatBot John Cavil

@allquixotic Chrome doesn't do Unicode fallback anything remotely sane or smart. (source)

allquixotic

grc.com/revocation/crlsets.htm

Darth Android

Adding it to the font: tag causes it to fall back correctly in chrome: Verdana,Segoe UI Emoji,Arial,sans-serif

Chrome just doesn't seem to inject fallback fonts automatically if the website doesn't specify them

where would it look up that information (if character X is missing, fall back to font Y) ?

Bob

@DarthAndroid Oh, I don't know, maybe the OS APIs?

Uniscribe

Uniscribe is the Microsoft Windows set of services for rendering Unicode-encoded text, especially complex text layout. They are implemented in the DLL USP10.DLL. USP10.dll became available to the public with Windows 2000 and Internet Explorer 5.0. In addition, the Windows CE platform has supported Uniscribe since version 5.0. Although Uniscribe continues to be maintained, with Windows 7 its intended replacement DirectWrite was introduced, which has more features. == USP10.dll == USP is an initialism for Unicode Scripts Processor. The main purpose of Uniscribe includes the following: arranging input...

Chrome already uses the API, they just apparently think it's a better idea to display boxes everywhere rather than substitute in a glyph from a font that actually has it.

allquixotic

yanno what? screw Chrome. I'm uninstalling it. sure, it provides a bit better perf (especially startup time) but I'm fed up with the way they treat users

*ragequits from Chrome*

Bob

3:35 PM

@allquixotic ...that's just... -______-

@allquixotic Mozilla's kinda going the same way :\

They're more responsive than Chromium devs/bugtrackers, though. For now.

IE... yea, there's another brick wall to bang your head against.

No real good options. Mozilla's probably the most open of the three, if only barely.

Safari and Opera are irrelevant.

ekaj

Where space is concerned over a long period of time, is it better to keep incremental or differential backups? I assume differential, because you can delete backups that are not used after a period of time

Darth Android

aren't incremental and differential backups the same thing?

an original backup, and then just storing the changes since then?

A good backup solution should allow you to reduce the backup resolution as needed to stay within space requirements

ekaj

3:54 PM

no

Kinda

allquixotic

no? boy, he sure told you!

ekaj

Incremental backups are like VM backups... you have the main image (a full back up), and then the next incremental backup records what have changed since the full backup

But the next incremental backup records the changes to the previous incremental backup, so you have to have ALL of the incremental backups in the chain, even if there are 500, for it to work

a differential backup has the full backup, then for each of the differential backups all changes are recorded, and you can delete previous backups that were not used

allquixotic

OK, and how would differential backups solve this problem? By violating information theory principles?

Er... the way you described it, differentials are just comparing whatever's the latest to the original backup, which after time will come to look similar to a full backup as the data starts to grow and differ. If you're taking "differentials" (using your terminology) repeatedly, they'll start to stack up, and grow as fast in size as full backups.

I've never heard that term used in that way, anyway. And it sounds like a really weird scheme for backing up.

If you're not going to keep incrementals, you might as well just do full backups, and keep the most recent 2 full backups (or 3, or 10, or how ever many you need).

Bob

@allquixotic standard backup terminology

differential = delta since last full

ekaj

Correct me if I'm wrong, I didn't say I was right. It's a system I don't have regular access to, so I can't go in and delete backups every so many days once 1TB fills up

Kate

3:59 PM

@ekaj what SQL server flavor are you backing up

Bob

incremental = delta since last increment

ekaj

With the differentials I can go in and delete all of the backups in between

Darth Android

@ekaj Most backup solutions that I'm familiar with use incremental backups, and allow you to delete the incremental backups in the middle by merging it with the next incremental

ekaj

Even if there are like 400?

Darth Android

And I'd recommend that since that is the "best of both worlds", in that incrementals will be smaller than differentials, and you can still delete ones in the middle to free up space if you don't need them.

Bob

4:00 PM

@DarthAndroid could be quite slow though

ekaj

What I'm using is Seagate BlackArmor

Darth Android

Even if there are like 400, but it does take time.

Bob

@ekaj normally you'd want a full backup at least once every dozen or two

allquixotic

you'd have to manually make the decision of when to merge the incrementals, no software can intelligently do so in a 100% safe way

Bob

too many increments means longer backup and restore times

and greater chance of failure

ekaj

4:01 PM

hence differential

Bob

differential backups are generally safer, but you still need to take full backups now and then, otherwise the delta becomes quite large

ekaj

I'll have to look into the merging with BlackArmor

Bob

it doesn't suffer as badly as incremental though

Darth Android

@allquixotic You can get decently good with automatic. Crashplan does incrementals every 15 minutes, and then merges down to hourly after they're a month old, and then merges down to daily after 6 months or something like that

ekaj

The previous guy who set the incremental backups did that

but after two years..

allquixotic

4:02 PM

@DarthAndroid ahh, didn't know they had such a complex schedule going on.. I just "use" crashplan, never investigated the details of how it's done

Bob

@ekaj My current backup scheme is daily incremental with weekly full

Hennes

Ah, that is what mine should be

Bob

@allquixotic I just finished reading that.

Firstly, I'd like to say WHAT THE FUCK ARE YOU DOING, GOOGLE?

Darth Android

@allquixotic That's merely the default, you can customize it however you want

ekaj

What do you do that with?

@Bob

Bob

4:03 PM

@ekaj Acronis True Image. It works well enough for my purposes.

I'm not really going to recommend it though.

It might suit you. It might not. I don't know.

ekaj

I was just trying to stick with Seagate BlackArmor as that's the NAS being used

Bob

I've had good and bad experiences with every backup solution I've touched.

allquixotic

@Bob Secondly, would you like to say, HOW CAN ANYONE BUY THE BULLSHIT THEY'RE SELLING?!

(like that Larry Seltzer guy)

Bob

@allquixotic Well, on the surface, it doesn't sound too bad. It's not immediately obvious that it's so utterly broken.

allquixotic

a lot of people seem to be of the attitude that "CRLs and OCSP are unreliable because servers can't keep up with the load, so we shouldn't even try to worry about certificate revocation at all"

Bob

4:05 PM

Initially it just sounds like a slight delay in CRL distribution.

Apr 13 at 0:08, by Bob

So Chrome failed to check CRL/OCSP :P

allquixotic

@Bob the 24k limit is hilarious though :D

Bob

Apr 13 at 0:08, by Michael Frank

@Bob there's a link further down that explains why

Apr 13 at 0:09, by Michael Frank

https://blog.cloudflare.com/certificate-revocation-and-heartbleed

Apr 13 at 0:10, by Ramhound

remind me to enable that setting in Chrome....LOL

@MichaelFrank, @Ramhound, reminder: if you care about security, GTFO Chrome.

See: grc.com/revocation/crlsets.htm

allquixotic

I'm back to the latest stable Firefox, myself.

Bob

Or you can try ticking that "check server certs" checkbox

But since the defaults are so... I don't have words strong enough to describe the stupid.

Apr 13 at 0:12, by Bob

Insecure defaults are no better than a lack of security.

@allquixotic FF OCSP checking doesn't hard-fail by default though.

Still far from ideal.

(if an attacker blocks the connection to the OCSP server, then it'll be silently allowed, IIRC)

Apr 8 at 14:29, by allquixotic

i think chrome doesn't check the CRL (because performance? idk)

No. Because whoever is in charge of that decision is a moron.

allquixotic

@Bob Must-Staple should take care of that, but I don't know if it's been implemented yet client-side

pretty easy to add server-side, seems like

Darth Android

4:11 PM

@Bob 10 years from now we find out that the NSA forced them to do this.

allquixotic

@DarthAndroid don't attribute to malice...

"I work at Google! I'm SMART! What I say is good enough for EVERYONE! :D"

!!hv

ChatBot John Cavil

Bob

@allquixotic Somewhat ironically, MS/IE is more secure.

allquixotic

my trust in and appreciation of Android is lower and lower every day considering the stupidity I see out of Google

not that I'd trust iOS more, but...

(Apple is just good at hiding their problems from users)

Darth Android

@allquixotic I only use that quote in situations where malice has not already been clearly demonstrated. My cynical comment is about the malice of the NSA, not the ignorance of Google.

Bob

4:14 PM

> Secondly, and more fundamentally, when issuing certificates a CA validates ownership of a domain by sending an email, or looking for a specially formed page on the site. If the attacker is controlling the site, they can get new certificates issued. The original owners might revoke the certificates that they know about, but it doesn't matter because the attacker is using different ones.
The true owners could try contacting CAs, convincing them that they are the true owners and get other certificates revoked, but if the attacker still has control of the site, they can hop from CA to CA gett…

imperialviolet.org/2014/04/19/revchecking.html

allquixotic

@DarthAndroid still, it's hard to imagine how this vulnerability could be used for malice... if the NSA hijacks some site or forces them to use a new cert, chances are the replacement cert they use is going to get signed by the CA and trusted just fine, so revoking the old cert (even if the browser checks) won't help.

Bob

...and did they just happen to forget about the case where the private keys are stolen without further access to the server (or email or domain)?

You know, in this blog post two weeks after Heartbleed?

What the fuck are they smoking?

"if the attacker is controlling the site" => what if they aren't?

What if they're just in a position to MitM a limited set of connections, e.g. public wifi?

allquixotic

@Bob heh... funnily enough, even that crackpot site is recommending OCSP Must-Staple

Bob

> Adam Langley works on both Google's HTTPS serving infrastructure and Google Chrome's network stack.

allquixotic

OCSP + OCSP Stapling + OCSP Must-Staple is pretty much The Way Forward... once browsers implement support for it... which will be in 2016 for IE, maybe never for ~~Insecureme~~ Chrome, and probably early 2015 for Firefox.

Bob

4:17 PM

@allquixotic Yes. That's the way to go.

But their current stop-gap is laughable.

No. It'd be laughable if it wasn't so fucking serious.

bugzilla.mozilla.org/show_bug.cgi?id=901698

bugzilla.mozilla.org/show_bug.cgi?id=921907

allquixotic

@Bob We have no way to ensure that radioisotopes capable of producing chain reactions can be absolutely, 100% controlled and prevented from falling into the wrong hands; therefore, it's perfectly fine to ship enriched Uranium with the same security precautions as shipping a piece of office furniture.

...At least until we have a 100.000% fool-proof way to stop proliferation from ever happing.

Signed, Adam Langley.

Bob

code.google.com/p/chromium/issues/detail?id=365336

@allquixotic Bad thing is, FF defaults aren't much better.

@allquixotic More from the footnotes of that grc page: 'official' CASC response: casecurity.org/2014/05/08/casc-heartbleed-response

> The Google Chromium team have responded by disabling all external certificate revocation checking (CRL and OCSP) and then removing the checkbox to optionally allow the user to turn it on, using the justification that it confuses users.

lol.

so you can't enable it even if you wanted to

allquixotic

LOL

sounds like a cover-up to me! :D

Bob

4:32 PM

O_O

My FF's OCSP checking was disabled... might've forgotten to toggle it after testing. Or they new disable it by default?

@allquixotic mind checking yours?

allquixotic

where would I find the setting? about:config?

Bob

@allquixotic tools => options => advanced => certificates => validation

allquixotic

@Bob first box checked; second box unchecked

Bob

@allquixotic yea, that's what I expected

btw, if you want hard-fail, check the second box

eggo

(lurking, until now) - so do you @Bob run FF with "OCSP connection failed => invalid" checked or not?

Bob

4:37 PM

@eggo I do now. Let's see how that goes.

allquixotic

5:16 PM

TIL that the Spotify Web Player ships a custom font to your computer and has you use it for glyphs/characters representing a great variety of things, including many graphical elements on the page

I guess it's more efficient than using a bunch of pngs

Bob

@allquixotic s/pngs/svgs/

Though, doesn't the web player use Flash?

allquixotic

erm, it appears to use HTML5 >_>

Bob

o.O

allquixotic

it may have a little flash object hidden somewhere, though :)

well Google Play Music is working so I'm happy ;p

Bob

@allquixotic It requires Flash.

Probably some DRM-related thing.

allquixotic

5:26 PM

@Bob -_-

@Bob it seems that whatever transport it uses for the data isn't able to be shipped over an HTTP proxy.

if it were TLS, they could CONNECT then do their thing inside it, regardless of the protocol, but this system uses stunnel as its proxy...

meme scientist

Any moderators here?

Bob

we can ping one for you if necessary, though it would help if you could say why you need one (often, flagging for mod attention is the correct course of action)

...actually, I'm not sure who's awake and active at this time. Could guess, though.

meme scientist

0

Q: Are there symbols for newer Windows key icons?

Microsoft's Windows icon has changed multiple times throughout the past decade. In doing so, the Windows key's icon has changed as well, from Windows XP to the most modern iteration, Windows 8. Now I'm writing some documentation for some procedures for a friend of mines. I really want to use a...

would this work better on GraphicDesign.SE

it was a question regarding font

not sure if it'd be better off on superuser or graphicdesign

Bob

hm

@memescientist if you're not sure, you could ask on their chat if they'd like to accept it first

otherwise, just flag for mod attention saying you're the OP and would like a migration

meme scientist

well, i was hoping some experts here could figure out which is the better site

i'm not really too good with this stuff

:C

Bob

5:35 PM

I don't think it's really off-topic here, but I'm not familiar with this area.

You could also ask on meta if 'does unicode for X' or 'does a font containing X' is a valid question on SU. I'm not sure we have any precedent here. It's a tad on the software rec off-topic, but not really that either (font containing would be rec, if unicode exists not really but not necessary on-topic either, basically, I don't know.)

allquixotic

> i'm not really too good with this stuff

> 22.6k

!!meme okay

ChatBot John Cavil