Ask Ubuntu General Room

Mateo

00:53

:(

this isn't a good sign

the information page was deleted yesterday for the api

Mateo

01:18

well, maybe I'll try and find a different idea...

Nathan Osman

o/

Mateo

hi, they are taking down the api I was using

so far just the docs disapeared

Nathan Osman

That's terrible.

Isn't that the same API that Amarok uses, though?

Mateo

lyrics.wikia.com ?

Nathan Osman

It's this one, right?

Yeah.

Mateo

01:28

oh dear

Nathan Osman

I wonder if they're aware of the problem.

Worst case scenario, you could screen scrape.

Mateo

well, the actual api seems to still be there

Ubuntu Questions on U&L

0

Q: How to send all keystrokes to a screen session?

I am starting a screen session with screen /dev/ttyACM0 115200. The system I am opening a session to is running Ubuntu, and I would like to pass it Ctrl-Alt-F1. How can I make screen capture those keystrokes and send it to the system?

Mateo

@NathanOsman well, if I knew how...

Nathan Osman

It's fairly straightforward in Python, not sure about Qt, though.

There's always libxml :P

@RPiAwesomeness not quite true - there's this: symbolhound.com

It lets you search for any character.

Mateo

01:32

python yeah, I saw that it looked nice there

Ubuntu Questions on U&L

0

Q: Understanding the automated mounting of USB Thumbdrive and doing it myself

I can break this down into two subcomponents: Why/how does this automated mounting procedure create (and destroy) it's own mounting point? Why do I have to manually create my own mount point when doing it myself (or how can I jump on the automated way of doing it)? I am not clear on the exact...

ubuntu mount usb-drive

Mateo

well, I'm publishing what I have while it still works

Nathan Osman

01:54

I'm going to give this a shot (running Apple Music through the Chrome runtime).

Doesn't work :(

It just hangs on the screen.

...and this shows up in the console:

ChildPlugin backgroundpage:0:umareporter(so_executable) Run: 2001.973ms
main.js:370 timeout to spawn ChildPlugin process: umareporter(so_executable), umareporter bool ArcRuntime.DexoptInterrupted false

Well, back to Android x86, I guess.

Chan-Ho Suh

02:13

@Serg you like brain teasers?

Serg

02:30

@Chan-HoSuh depends on whether or not a grade depends on it and whether or not I have any additional motivation :)

Chan-Ho Suh

hahah

when you go for interviews, some people may ask them

it's good to get those neurons firing early on... no use practicing them right before an interview

Serg

@Chan-HoSuh That's actually quite true. I've not gone to much interviews whatsoever. Well, try one on me right now :) For the sake of boredom

Zacharee1

@Chan-HoSuh Hey

Haven't seen you here in a while

@NathanOsman What's this?

Nathan Osman

What's what?

Zacharee1

@NathanOsman what are you trying to do? (TL;DR)

Serg

02:39

Well that's a new one. Chocked on honey

Zacharee1

?

Serg

I thought i'm gonna flippin suffocate

Zacharee1

lol

choked

Chan-Ho Suh

@Zacharee1 hey, yep been busy

Zacharee1

understandable :P

Serg

02:40

@Zacharee1 thanks . . .grammar is the last thing on my mind right now but thanks

Nathan Osman

@Zacharee1 Run Apple Music on my desktop.

Zacharee1

@Serg Well I had no idea what you meant. That is something new though

@NathanOsman What OS?

Chan-Ho Suh

@Serg here's a classic one: suppose you have 12 coins. They look identical. However, one coin is fake, it's a different weight than the other 11. You have a balance you can use to compare weights. It doesn't tell you the exact weight, you only know one side is more than the other, or that the sides are equal weight. Using at most 3 weighings, can you identify the fake coin and whether it is lighter or heavier than a real coin?

Nathan Osman

@Zacharee1 Ubuntu.

Zacharee1

@Chan-HoSuh wouldn't it depend?

@NathanOsman ah

Nathan Osman

02:43

If it were Windows or OS X, I'd just use iTunes :P

Zacharee1

lol

I guess

Chan-Ho Suh

@Zacharee1 what do you mean depend?

Zacharee1

@Chan-HoSuh I didn't see the ID part :P

@Serg You OK?

Serg

@Zacharee1 still alive, writing answer for @Chan-HoSuh

Zacharee1

ah

Serg

02:49

@Chan-HoSuh well it's obvious one could place half the coins on one side of the balance and the other , on the other half (so 6 and 6, first measurement). At this point we don't know which half contains the fake coin. Then we can split both halves in halves themselves (so 3 and 3). Measuring one of the 3 & 3 sets will tell us whether those are real coins - the 3 and 3 will be equal mass if they are real. So we can deduce fake coin is in the other 3 and 3 set

Conversely if we get unequal 3 and 3 set we know the fake coin is among them

So for measurement 2 we don't need to know which 3 and 3 set to pick, just infer

Chan-Ho Suh

that's very good... I think you can get it to work in at most 4 that way

Serg

So . . . If in the first measurement we took heavier set, and it turned out to be equal mass in second measurement, then we know real coins are heavier, fake is lighter. And conversely heavier fake coin will outweigh real in second measurement.

So . . .now we just need one more measurement . . . to identify real one

At this point we should know whether fake coin is lighter or heavier

Chan-Ho Suh

I think for your approach, the hard case is: suppose you split 6 and 6. You take one group of 6 and weigh 3 and 3 and find they are equal

so you are down to the other 6 and have one weighing left ... you know if it's heavier or lighter, so that will help

but you have 6 possibilities for which coin is fake, with one weighing that tells you a trit of information

so log base 3 of 6 is approx 1.6 which is > 1

anyway, for a fundamental reason, you will not be able to determine which of the 6 coins is fake from one weighing

Serg

good point

Zacharee1

split into 4 groups of 3, weigh two...'

if they weigh the same, switch one out

keep switching it out until there's a difference (so three tries at most)

then you can see if the fake coin is heavier or lighter

still figuring out what could be done if they had a different weight

if they don't weigh the same, switch one out. If they weigh the same now, that was the group with the fake coin, so you know if it's heavier or lighter

if they still don't weigh the same, you know that the constant group has the fake coin and you can go from there

so 3 tries for the first one, 3 tries for the second, 2 tries for the third

Serg

03:04

It's quite interesting. I can see it done in 4 steps or so.

Choose one group of 4 as standard/ pivot . . . For group in i do : measure(i). If weight(i )!= weight(pivot) : divide in half and

But 3 steps . . . very interesting

Zacharee1

@Chan-HoSuh ^^^^

Chan-Ho Suh

@Zacharee1 that's a good start, but it takes too many tries on the first step to reduce to a group of 3

Zacharee1

@Chan-HoSuh Can't you just split them into 4 groups of three from the beginning? Only three weighings

Serg

@Zacharee1 you need one more weighing or two on the suspected group to find out exact coin

Chan-Ho Suh

right, but you want three weighings total

Zacharee1

03:08

@Chan-HoSuh @Serg I forgot about the ID part again.... -_-

guh

Serg

ID ? user ID ?

Zacharee1

identification :p

What if the scale had four bowls? ;P

Serg

@Zacharee1 then it's cheating :p

Chan-Ho Suh

for this problem, the numbers are picked to make it quite tough so yeah if we started fiddling with the parameters a bit, it's gonna get simpler

Zacharee1

Is it possible?

Chan-Ho Suh

03:09

it is... but most people, including me, find there's a solution by just trying at it

recently I found out, there's a quite nice general approach to seeing there is a solution

I was joking about it with someone else actually ... because most people that ask this problem probably don't know the general approach

so you could be weird in an interview, and start enumerating codewords, and they'll be like, what on earth are you doing?

anyway, there's a general approach using the idea that what you are doing is encoding the starting state of 12 coins

Zacharee1

encoding?

Chan-Ho Suh

you have 12 coins, one which is lighter or heavier than the others... that's 24 possible configurations

you can think of the problem in this way: you have an algorithm where you are sorting the coins and making three measurements on them

Serg

Ah . . . 0 or 1 . . . .each coin could be that . . .so 12*2

Chan-Ho Suh

the three measurements give you a code

0 means they balanced, 1 means left side down, 2 means right side down

so for a given configuration, your procedure of sorting the coins around and using the balance, gives a code like 201

the problem is asserting that it's possible to go back from a code, i.e. decode, and get the original configuration

now if you think about this a bit, if you could really do this decoding, there are some constraints on the encoding algorithm

for example, you can't have two coins that are following each other around

if you have say coin 1 and coin 2 always being on the same side of a scale, or not being on the scale, for each measurement

then there's no way you can distinguish the cases where coin 1 is fake from when coin 2 is fake

anyway, I'll let you think about it ... I had a lot of fun working this out... there's a complete explanation on the Internet if you google, of course :)

hmmm, I can give you the complete algorithm though

how you would write the code and use it... anyway, ping me if you want it

oh, the nice thing about this approach is it explains why you have to start with dividing the 12 coins into 3 groups of 4 :)

I never understood that

Nathan Osman

03:45

If anyone is interested, I'm looking for beta testers for NitroShare.

2

Just install it from this PPA and let me know if anything is broken.

Vishwa

Hi all, does anyone know whats the highest bounty awarded in askubuntu history?

just curios :P

Serg

Highest that can be set is 500 rep, but there's no limit on how many bounties an answer may receive . . .

so it's not known what is the highest awarded

but highest available is 500

Vishwa

@Serg i c , thanks . I didn't know that, coz I thought Only I cant set more than 500 as I am having only 530 :)

Ubuntu Questions on U&L

04:19

0

Q: How to Upgrade LinuxLight from Trusty to Wily

I'm giving LinuxLight 2.8 beta a try, and have noticed it's completely based on Ubuntu Trusty. How can I go about upgrading it to Ubuntu Wily? Is it possible to replace my current repos with the newest Wily Repos? Perhaps this cannot be done with the current state of the beta? If that's the c...

ubuntu upgrade

Nathan Osman

04:43

Well, well, well.

Look what I managed to get working:

Chan-Ho Suh

Windows 3.1?

Nathan Osman

Chan-Ho Suh

Windows 95

97

Nathan Osman

Excel 97 on Windows 98.

Chan-Ho Suh

my god... why???

Nathan Osman

04:45

For nostalgia.

And curiosity :P

This is what I grew up with.

Chan-Ho Suh

I didn't use much Excel in those days... if I did it at all

I remember Wing Commander

Nathan Osman

How about that?

Chan-Ho Suh

I like seeing the navigator icon

did you get that working?

Nathan Osman

Sure did.

You want screenshots of that? :P

Chan-Ho Suh

sure

Nathan Osman

04:49

Hrm... can't seem to get any websites to load :P

At least not in Navigator 3.

Trying 4.7...

George's website works:

:P

But here's the thing that continues to completely shock me.

Google still works perfectly in IE 5:

^--- that's probably the most modern browser I can install

What the... I can't enter a network password longer than 14-characters in Network Neighborhood???

And then there's defragmenting...

Chan-Ho Suh

05:08

good ol' defragmenting

should make that a screensaver

Vishwa

05:28

What is she/he asking askubuntu.com/questions/724537/i-love-my-arm-ill-lead-content

ok, deleted

Serg

@Vishwa that was spam

Vishwa

@Serg How can we know that? by its content ?

Serg

@Vishwa yes. Because it tries either to sell something or is sheer nonsense or has weird links to fishy sites

i need sleepz

bye

Vishwa

guys, Whats the difference b/w hot and active?

based on no. of views??

Seth