p-lang thread: (now (have you (problems two)))

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > p-lang thread: (now (have you (problems two)))

«‹›1784 »

tef: May 30, 2004; -> some l-system crap ->

Mr Dog posted:

now you have to write a condescending and bad article explaining it to the mundanes

and then they can go "oh poo poo i think i wrote some java i didn't understand so it must be monads, right".

# ? Dec 12, 2014 18:45

Adbot: ADBOT LOVES YOU

# ? Jun 12, 2024 00:08

Bloody: Mar 3, 2013

Nontechnical woman on this train: my Web design class was easy, but python was hard

# ? Dec 13, 2014 02:26

Deus Rex: Mar 5, 2005

sarehu posted:

PL (Programming Language) people: Should I use U+0305 as the syntax for bitwise complement?

Example:

x = a̅ + 1

yeah

# ? Dec 13, 2014 16:22

Zaxxon: Feb 14, 2004; Wir Tanzen Mekanik

tef posted:

and then they can go "oh poo poo i think i wrote some java i didn't understand so it must be monads, right".

Ask your doctor. It might be monads!

# ? Dec 14, 2014 19:50

MononcQc: May 29, 2007

Raft's been reviewed a bit: http://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-857.pdf

Particularly, they review the choices of timeouts picked by the Raft authors and assumptions about the speed of their leader election algorithm, which in practice turned out to be slower than reported (due to generally healthy clusters having more good candidates than the original paper had assumed)

The best bit is this one on asymmetric netsplits or cases like mobile devices leaving and joining networks:

p49 posted:

The mobile devices will often be disconnected from the majority of other devices, thus running election after election without being able to gain enough votes. When these devices rejoin the network, they will force a new election as their term will be much higher than the leader. This leads to frequently disconnected devices becoming the leader when reattached to the network, thus causing regular unavailability of the system when they leave again.

[...]

It is also interesting to consider the impact of a single node being behind an asymmetric partition such as a NAT box or firewall (Figure 4.6(b)). This node will repeatedly re-run elections as it will not be able to receive incoming messages, each RequestVotes will have a higher term than the last, forcing the leader to step down.

[...]

In particular, a cluster can be rendered useless by a frequently disconnected node or a node with an asymmetric partition (such as behind a NAT).

# ? Dec 15, 2014 00:36

tef: May 30, 2004; -> some l-system crap ->

awww yiss

# ? Dec 15, 2014 03:52

the talent deficit: Dec 20, 2003; self-deprecation is a very british trait, and problems can arise when the british attempt to do so with a foreign culture

the best bit of that paper is that the recommendations for improving raft make it harder to reason about than paxos

also this:

quote:

In my experience, all distributed consensus algorithms are either:
1: Paxos,
2: Paxos with extra unnecessary cruft, or
3: broken.
- Mike Burrows

# ? Dec 15, 2014 06:23

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

MononcQc posted:

The best bit is this one on asymmetric netsplits or cases like mobile devices leaving and joining networks:

This is the sort of stuff that is absolutely obvious in hindsight, but I would not have been able to see it in all my life.

# ? Dec 15, 2014 06:58

MononcQc: May 29, 2007

Suspicious Dish posted:

This is the sort of stuff that is absolutely obvious in hindsight, but I would not have been able to see it in all my life.

# ? Dec 15, 2014 14:11

Captain Foo: May 11, 2004; we vibin'
we slidin'
we breathin'
we dyin'

the talent deficit posted:

the best bit of that paper is that the recommendations for improving raft make it harder to reason about than paxos

also this:

that's a p fun quote

# ? Dec 15, 2014 14:18

Blotto Skorzany: Nov 7, 2008; He's a PSoC, loose and runnin'
came the whisper from each lip
And he's here to do some business with
the bad ADC on his chip
bad ADC on his chiiiiip

Suspicious Dish posted:

This is the sort of stuff that is absolutely obvious in hindsight, but I would not have been able to see it in all my life.

# ? Dec 15, 2014 14:29

Blotto Skorzany: Nov 7, 2008; He's a PSoC, loose and runnin'
came the whisper from each lip
And he's here to do some business with
the bad ADC on his chip
bad ADC on his chiiiiip

all the worst bugs are like taht ;_; porgamming is a vale of tears

# ? Dec 15, 2014 14:29

Axel Rhodes Scholar: May 12, 2001; Courage Reactor

Progressive JPEG posted:

if i liked twilight princess should i try skyward sword??

im a weirdo that has a gamecube but not a wii

you should try and play it on an emulator and then when you hit bugs in the emulator you fix them and then after a while you notice you have more fun fixing the bugs than playing the game and then you're me trying to play old ultima games

# ? Dec 15, 2014 16:16

suffix: Jul 27, 2013; Wheeee!

MononcQc posted:

Raft's been reviewed a bit: http://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-857.pdf

Particularly, they review the choices of timeouts picked by the Raft authors and assumptions about the speed of their leader election algorithm, which in practice turned out to be slower than reported (due to generally healthy clusters having more good candidates than the original paper had assumed)

The best bit is this one on asymmetric netsplits or cases like mobile devices leaving and joining networks:

this is cool but lol @ making it byzantine tolerant in future work

# ? Dec 15, 2014 21:27

sarehu: Apr 20, 2007; (call/cc call/cc)

Just learned today that there was a bug in RethinkDB where any index-using descending-order key range query, such as r.table('foo').between(1, 5).orderBy({index: r.desc('HP')}), will return values in an exinclusive interval (1, 5] instead of the inexclusive interval [1, 5).

(I'm really making this post to spread the idea of using the word "inexclusive" instead of "clopen.")

# ? Dec 18, 2014 07:50

FamDav: Mar 29, 2008

sarehu posted:

Just learned today that there was a bug in RethinkDB where any index-using descending-order key range query, such as r.table('foo').between(1, 5).orderBy({index: r.desc('HP')}), will return values in an exinclusive interval (1, 5] instead of the inexclusive interval [1, 5).

(I'm really making this post to spread the idea of using the word "inexclusive" instead of "clopen.")

clopen should be reserved for sets which are both open and closed, so agreed.

# ? Dec 18, 2014 08:01

sarehu: Apr 20, 2007; (call/cc call/cc)

Hrm. It's really disturbing, then, that I've heard the term used for half-open intervals.

# ? Dec 18, 2014 08:59

Symbolic Butt: Mar 22, 2009; (_!_); Buglord

I remember reading once about how every basic notion in topology has a correspondent equivalent notion in logic

so someone who knows about this please tell me the correspondent notion in logic for a clopen set

FAKE EDIT: I just found it. "Note that this gives a natural interpretation to clopen sets: they're precisely the properties whose truth and falsehood can both be verified in finite time."

this sounds so loving bullshit

# ? Dec 18, 2014 09:23

sarehu: Apr 20, 2007; (call/cc call/cc)

They're probably the same thing in category theory, or affine category theory in the context of affine logic spaces.

# ? Dec 18, 2014 09:26

PleasureKevin: Jan 2, 2011

power botton posted:

what is so mission critical + urgent that you both had to install it on ubuntu and upgrayedd at 5 AM

nothing. I just get up really early because my day job sucks and I need to get as much useful time out of the day as possible.

also I didn't even try upgrade, I just power cycled and that broke it.

# ? Dec 18, 2014 11:00

Dylan16807: May 12, 2010

Symbolic Butt posted:

I remember reading once about how every basic notion in topology has a correspondent equivalent notion in logic

so someone who knows about this please tell me the correspondent notion in logic for a clopen set

FAKE EDIT: I just found it. "Note that this gives a natural interpretation to clopen sets: they're precisely the properties whose truth and falsehood can both be verified in finite time."

this sounds so loving bullshit

okay I think I halfway understand that after staring at it for a while

The hard step is defining 'open'. In this case it's a semi-decidable theory. If an instance is true, you can algorithmically prove it true. If it's false, you might never be able to prove an answer. The blog post gives a physical analogy of checking if someone is over six feet tall with ever-finer-grained rulers. You can never disprove the idea that they're taller but only by a miniscule amount. More relevant to your question, checking if one statement implies another in first order logic is semi-decidable.

So closed means that the opposite of a statement is semi-decidable. If it's false, you can prove it false.

Then clopen means you can prove either case reliably. It becomes a pretentious synonym for 'decidable'.

# ? Dec 18, 2014 11:32

gonadic io: Feb 16, 2011; >>=

Isn't the only clopen set the empty one though?

# ? Dec 18, 2014 13:48

Brain Candy: May 18, 2006

nope, also the entire topological space

e: to actually be relevant instead of snarky, you can get non-trivial clopen sets by having subsets that aren't connected. you can see how that is applicable to boolean algebra

Brain Candy fucked around with this message at 14:16 on Dec 18, 2014

# ? Dec 18, 2014 14:04

AWWNAW: Dec 30, 2008

what the gently caress are y'all talking about

# ? Dec 18, 2014 15:21

gonadic io: Feb 16, 2011; >>=

something worse than monads

# ? Dec 18, 2014 15:42

Malcolm XML: Aug 8, 2009; I always knew it would end like ｔｈｉｓ．

Symbolic Butt posted:

I remember reading once about how every basic notion in topology has a correspondent equivalent notion in logic

so someone who knows about this please tell me the correspondent notion in logic for a clopen set

FAKE EDIT: I just found it. "Note that this gives a natural interpretation to clopen sets: they're precisely the properties whose truth and falsehood can both be verified in finite time."

this sounds so loving bullshit

sarehu posted:

They're probably the same thing in category theory, or affine category theory in the context of affine logic spaces.

Yeah probs, topoi are used to model alternate systems of logic

Open and closed not being mutually exclusive in the context of topology is one of the dumbest things

Thus clopen sets.

# ? Dec 18, 2014 16:30

Gazpacho: Jun 18, 2004; by Fluffdaddy; Slippery Tilde

cloacan sets

# ? Dec 18, 2014 16:31

Dylan16807: May 12, 2010

clopen is a bit weird, because closed doesn't mean it includes anything in particular

any time your border isn't in the space you're using you get a clopen set

'rational numbers between sqrt(2) and pi' is clopen

# ? Dec 18, 2014 16:54

Malcolm XML: Aug 8, 2009; I always knew it would end like ｔｈｉｓ．

Dylan16807 posted:

clopen is a bit weird, because closed doesn't mean it includes anything in particular

any time your border isn't in the space you're using you get a clopen set

'rational numbers between sqrt(2) and pi' is clopen

... no???

clopen has a very specific meaning

namely an open set whose complement is also open

# ? Dec 18, 2014 16:57

Dylan16807: May 12, 2010

Malcolm XML posted:

... no???

clopen has a very specific meaning

namely an open set whose complement is also open

"less than pi" in the space of rationals is open
the complement of "less than pi" in the space of rationals is also open
so it's clopen

in the space of integers it's clopen, in the space of reals it's not clopen

# ? Dec 18, 2014 17:00

sarehu: Apr 20, 2007; (call/cc call/cc)

Dylan16807 posted:

in the space of integers it's clopen, in the space of reals it's not clopen

You mean rationals, not integers.

# ? Dec 18, 2014 20:56

distortion park: Apr 25, 2011

clonads

# ? Dec 18, 2014 21:13

bobbilljim: May 29, 2013; this christmas feels like the very first christmas to me

cloppen

# ? Dec 18, 2014 21:25

prefect: Sep 11, 2001; No one, Woodhouse.
No one.; Dead Man’s Band

Dylan16807 posted:

"less than pi" in the space of rationals is open
the complement of "less than pi" in the space of rationals is also open
so it's clopen

in the space of integers it's clopen, in the space of reals it's not clopen

is this like when they would draw lines on graph paper, and you would draw a circle around a point if it wasn't included, or a solid dot if the point was included?

# ? Dec 18, 2014 21:26