code logs -> 2016 -> Mon, 31 Oct 2016< code.20161030.log - code.20161101.log >
--- Log opened Mon Oct 31 00:00:11 2016
00:00 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [[NS] Quit: Quit]
00:30
<&McMartin>
http://wccftech.com/nvidia-gtx-1080-1070-evga-cards-dying/
00:31
<&McMartin>
Man, I remember back when electronics only caught fire when you configured it *improperly*
00:32
<@Azash>
McMartin: I remember a guy at my dept wrote a paper on green ICT and using DC exhaust air to heat the building
00:32
<@Azash>
Only natural to take it to the next level for private consumers
00:35 * Alek had a graphics card burn out in like 09/10, at default card configs.
00:36
<@Alek>
burning caps, ew. I SAW it burning. :P
00:36
<@Alek>
had to replace the mobo, cpu, and ram too, because even with the card replaced the machine no longer worked.
00:37
<&McMartin>
Ugh.
00:37
<&McMartin>
Yeah, my original comment was in reference to igniting CRTs by overdriving their horizontal refresh rate
00:42 catadroid [catalyst@Nightstar-gi5itv.dab.02.net] has joined #code
00:49
<&McMartin>
Aha, I see, yeah, this is an overclocking issue and also just a high-ambient issue, it seems
00:49
<&McMartin>
https://www.reddit.com/r/pcgaming/comments/58p5z2/evga_response_to_gtx_1080_1070 _high_memory/
00:49
<&McMartin>
This seems pretty mensch-y
00:52 catalyst [catalyst@Nightstar-bt5k4h.81.in-addr.arpa] has quit [[NS] Quit: Leaving]
03:37 Alek [Alek@Nightstar-cltq0r.il.comcast.net] has quit [Ping timeout: 121 seconds]
03:45 Alek [Alek@Nightstar-cltq0r.il.comcast.net] has joined #code
03:45 mode/#code [+o Alek] by ChanServ
04:23 Derakon is now known as Derakon[AFK]
04:38 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds]
04:45 Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code
04:45 mode/#code [+qo Vornicus Vornicus] by ChanServ
06:38 catadruid [catalyst@Nightstar-vdpa1t.dab.02.net] has joined #code
06:41 catadroid [catalyst@Nightstar-gi5itv.dab.02.net] has quit [Ping timeout: 121 seconds]
08:01 Kindamoody[zZz] is now known as Kindamoody
08:02 celticminstrel [celticminst@Nightstar-5boimd.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
09:12 Kindamoody is now known as Kindamoody|afk
10:03 catadruid is now known as catadroid
10:41 catadroid [catalyst@Nightstar-vdpa1t.dab.02.net] has quit [Ping timeout: 121 seconds]
10:59 catadroid [catalyst@Nightstar-vdpa1t.dab.02.net] has joined #code
12:08
< catadroid>
C++ really despises having to copy any data
12:50
< catadroid>
Well, our classes do anyway. I guess C++ actually rather perversely enjoys it in general
13:08 catadruid [catalyst@Nightstar-tt8rht.dab.02.net] has joined #code
13:11 catadroid [catalyst@Nightstar-vdpa1t.dab.02.net] has quit [Ping timeout: 121 seconds]
13:19
<~Vornicus>
yeah, C++ is all about dat copy unless you tell it otherwiser
13:48 gnolam_ [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has joined #code
13:48 gnolam [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has quit [NickServ (RECOVER command used by gnolam_)]
13:48 gnolam_ is now known as gnolam
13:48 mode/#code [+o gnolam] by ChanServ
14:39 macdjord is now known as macdjord|wurk
14:49 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has joined #code
16:27 Alek [Alek@Nightstar-cltq0r.il.comcast.net] has quit [Ping timeout: 121 seconds]
16:30 Alek [Alek@Nightstar-cltq0r.il.comcast.net] has joined #code
16:30 mode/#code [+o Alek] by ChanServ
16:37 gizmore|2 [kvirc@Nightstar-digmob.dip0.t-ipconnect.de] has quit [Operation timed out]
16:37 gizmore|2 [kvirc@Nightstar-jrjsqj.dip0.t-ipconnect.de] has joined #code
16:43 celticminstrel [celticminst@Nightstar-5boimd.dsl.bell.ca] has joined #code
16:43 mode/#code [+o celticminstrel] by ChanServ
17:10 Derakon [chriswei@Nightstar-5mvs4e.ca.comcast.net] has joined #code
17:10 mode/#code [+ao Derakon Derakon] by ChanServ
17:14 Derakon[AFK] [chriswei@Nightstar-5mvs4e.ca.comcast.net] has quit [Ping timeout: 121 seconds]
18:37 catadroid [catalyst@Nightstar-tt8rht.dab.02.net] has joined #code
18:37 catadruid [catalyst@Nightstar-tt8rht.dab.02.net] has quit [The TLS connection was non-properly terminated.]
18:42 catadruid [catalyst@Nightstar-tt8rht.dab.02.net] has joined #code
18:42 catadroid [catalyst@Nightstar-tt8rht.dab.02.net] has quit [The TLS connection was non-properly terminated.]
19:35 catalyst [catalyst@Nightstar-bt5k4h.81.in-addr.arpa] has joined #code
19:38 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [Connection closed]
19:38 catadroid [catalyst@Nightstar-7uddmo.dab.02.net] has joined #code
19:41 catadruid [catalyst@Nightstar-tt8rht.dab.02.net] has quit [Ping timeout: 121 seconds]
19:42 catadroid [catalyst@Nightstar-7uddmo.dab.02.net] has quit [Ping timeout: 121 seconds]
21:07 * gnolam !s.
21:07 * Tamber ?s
21:10
<@gnolam>
I have been given API specs for a new machine to implement... /and it's sane/.
21:10 * Azash &amp;s responsibly
21:10
<@Tamber>
...lies.
21:10
<@Tamber>
and damned lies
21:10
<@Tamber>
gnolam: Clearly, if they're sane, then they're wrong.
21:11
<@gnolam>
This is an industry where the only open standard is based on, I shit you not, first negotiating a protocol version by reading and writing to set locations in the Windows registry and then writing and reading data and signal files in magical folders.
21:11 * Azash has yet to encounter an API that pleased both the server and client
21:12
<@Azash>
gnolam: W-what
21:12
<&McMartin>
Azash: So, like
21:12
<&McMartin>
gnolam is a real scientist
21:12
<&McMartin>
It turns out though that we computer "scientists" actually have made progress that we haven't properly, fully shared
21:13
<@Azash>
McMartin: I mean when you negotiate a protocol version by manipulating the registry the only scientist I can think of is that guy from ATHF
21:14
<&McMartin>
Right, so, um, hm
21:14
<&McMartin>
This reference is a little dated, but...
21:14
<&McMartin>
You know the joke about how the most terrifying thing to hear from your mom/uncle/etc re: 'halp my email doesn't work' is 'wait n/m i solved it'?
21:15
<&McMartin>
Because it invariably turns out that their 'solution' was to copy paste images out of the web browser through microsoft word and then export as email or something
21:15
<@Azash>
Yeah?
21:15
<&McMartin>
ANAICT this is the entire world outside of professional IT, including about half of CS academia
21:15
<&McMartin>
(I am literally a CD PhD and I can tell you right now you do not want me within 100 meters of an LDAP server)
21:16
<&McMartin>
*CS
21:16
<@Azash>
Painful memories of a highly lauded networking pioneer publishing 1-2 times a month since his MSc
21:16
<@Azash>
"How does this projector work"
21:16
<&McMartin>
Modularization: Not just for program structure~
21:17
<@gnolam>
McMartin: not doing SCIENCE anymore. Now I'm an industry whore!
21:17
<@Azash>
McMartin: Also for posterity, this guy https://www.youtube.com/watch?v=W36pKI6qNKo
21:17
<&McMartin>
gnolam: Strictly speaking, I didn't say you were *doing* science, but rather that you were a scientist
21:18
<&McMartin>
Also, I would like to note that the last time I was helping out a bioinformatician, they had in fact named all the variables after their favorite movie and videogame characters
21:18
<@Azash>
Amazing
21:18
<&McMartin>
So I think some of these things are baked in but we get them out of our system when we were hacking BASIC at age 9
21:18
<@Azash>
McMartin: When I started my CS bachelor I spent some four or six weeks going back and forth about whether Horgan or Denning was right
21:18
<@Azash>
I think you've given me the gift of realization
21:19
<@Azash>
It's simply a way to protect impressionable younglings from an academic career
21:19
<@gnolam>
(I got the choice between "continued awesome SCIENCE, but on a project by project basis" and "full-time industry employment with a 100% pay raise to boot". I chose the latter.)
21:19
<@Azash>
(as in weekly essays)
21:19
<&McMartin>
I *think* I know what distinction you're trying to make here but can you spell it out?
21:19
<&McMartin>
Now, it turns out that some of this is our fault too
21:20
<&McMartin>
Said bioinformatician's co-workers actually tended to name their variables things like 'foo' and 'bar'
21:20
<&McMartin>
Because that's what all the examples did
21:20
<&McMartin>
And obviously they are savvily following the obvious best practice
21:20
<@gnolam>
(... unfortunately, in an industry where software is slapped together by chemists and mechanical engineers. ;_;)
21:20
<&McMartin>
I think that one's on us.
21:20
<@gnolam>
(Or, from the look of their APIs, literal witch doctors.)
21:20
<&McMartin>
Heh
21:20
<@Azash>
McMartin: Would you say the situation is foobar?
21:20
<&McMartin>
Azash: I seem to be recognizing the shit out of it, so I think it doesn't even manage that!
21:21 * Azash wonders if he's told the CCN story here yet
21:21
<@Azash>
Just as a tangerine
21:22
<@Azash>
So I used to be involved as a student representative in various faculty things
21:22 * McMartin does some googling, realizes that this the Horgan vs. Denning thing is different than what he thought it was
21:22
<@Azash>
One of those was being a member of the AP applicant review boards
21:22
<&McMartin>
On that particular front...
21:22
<&McMartin>
... I happen to consider CS to be Applied Philosophy
21:22
<@Azash>
In particular, I was in the group that reviewed demo lectures
21:22
<@Azash>
McMartin: Yeah we wrote essay after essay after essay
21:23
<@Azash>
It really does sap you of your will to write ever again
21:23
<&McMartin>
Well, it's not just that
21:23
<@Azash>
So I was on that board
21:23
<&McMartin>
I took a history of philosophy course as part of my breadth requirements
21:23
<@Azash>
And we had this one guy who did his lecture on his own specialty
21:23
<@Azash>
Content-centric networking
21:23
<&McMartin>
And, uh, yeah, actually, the 19th century analytic philosophy stuff
21:23
<&McMartin>
... that looked an awful lot like CS 70
21:23
<@Azash>
For those who aren't familiar, it's just the idea that you replace IPv\d+ addressing with resource IDs
21:24
<@Azash>
And he was very hyped about it
21:24
<@Azash>
"Finally, we can move past the outdated IP addressing paradigm" etc etc
21:24
<@Azash>
So he asked if there were questions and I meekly posed, "So.. In practice, how do you actually get your files and find them"
21:25
<@Azash>
"Well, that's the clever part"
21:25
<@Azash>
"It can just be built on top of TCP/IP"
21:25
<&McMartin>
... so, DNS?
21:26
<@Azash>
DNS needs addressing unless you somehow solve the problem of decentralized DNS broadcast systems that aren't just going to destroy everything beautiful ever in a way that would make James Mickens stutter
21:26
<@Azash>
That sentence came out as garbage and I apologize, I just got home
21:27
<&McMartin>
It's OK, I study German in my spare time >_>
21:27
<@simon_>
McMartin, so do I!
21:27
<@Azash>
Anyway yes to sum it up, he was advocating replacing IP addressing with a revolutionary new system that's built on top of IP addressing
21:28
<&McMartin>
welp
21:28
<@simon_>
my networking is rusty. why is it you want to replace IP addressing?
21:29
<@simon_>
surely there are reasons to replace TCP.
21:29
<@Azash>
simon_: https://en.wikipedia.org/wiki/Content_centric_networking
21:30
<@Azash>
simon_: Oh you reminded me of a fantastic Experience the other day
21:30
<@Azash>
I was getting hounded by a client manager who wouldn't listen to me saying that a particular problem was caused by a misconfigured simulator
21:30
<@Azash>
The reasoning?
21:31
<@Azash>
She had seen the logs and noticed in the networking statistics that in the process of sending 47 million UDP packets
21:31
<@Azash>
One had been lost
21:31
<@Azash>
Revealing a critical failure in our network stack.
21:31 * Azash literally could not get work done before noon because he was being followed everywhere
21:32
<&McMartin>
Uninterruptible Datagram Protocol
21:32
<@Azash>
Exactly
21:32
<@Azash>
I should have promised to roll out the newest version of google maps onto the network to avoid packets getting lost again
21:32
<@Azash>
She might even have bought it
21:33
<@gnolam>
:)
21:33
<@simon_>
Azash, soo... content-centric networking = giant IP-less DHT where you basically ask "DOES ANYONE HAVE SHA512('Google Search')???"
21:34
<&McMartin>
I mean, my objection to these sorts of things is that as a rule I do not want $FIXED_ITEM
21:35
<&McMartin>
If I knew what content I wanted, I would have it
21:35
<@simon_>
the wikipedia page says that content-centric networking is not connection-oriented. but I don't get that - if TCP is in charge of the notion of connections, how does connection-centric networking make IP redundant?
21:35
<&McMartin>
I want to either interact with some service, or contact a specific machine.
21:35
<@Azash>
simon_: Simple
21:35
<@Azash>
The layered nature of networking means that any layer can be seamlessly switched out for another with no problems or complications
21:35 * Azash maintains a straight face
21:36
<&McMartin>
Hey, back in the 80s consumer networking ran on X.25!
21:37
<@Azash>
I've been yapping on a bunch now but there's one more thing I wanted to share before I tumble into bed
21:37
<@Azash>
I have to put up with systemd at ${CLIENT}, and was trying to figure out how to configure the log sync rate
21:38
<@Azash>
And as it turns out, google shares my rather partisan views of said software
21:38
<@Azash>
http://i.imgur.com/gw6xIET.png
23:05
<&[R]>
Ga
23:05
<&[R]>
Ha*
23:10 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has joined #code
23:14
<@simon_>
private static Tuple<Dictionary<StringTokenization, Dictionary<int, Dictionary<int, Dictionary<int, int>>>>, IEnumerable<AlignedSequence>> GetEditDistancesDistribution(string prefix, IDictionary<StringTokenization, IList<StringWithCompanyInformation>> substringTokenizationsDistribution)
23:16
<&McMartin>
And the most tragic part is that this symbol is probably 10x smaller than the full symbol name of an equivalent type implemented using standard library types
23:17
< catalyst>
I once broke the template name limit on the managed C++ compiler
23:18
< catalyst>
it's 1024 characters
23:18
<@simon_>
catalyst, sounds like an epic name.
23:18
< catalyst>
It was entirely generated from one that was about 50 characters long at most
23:18
<@simon_>
s/epic name/epic (as in the literary genre)/
23:18
< catalyst>
templates to select arities, and such
23:18
< catalyst>
Hey, it worked on the standard compiler
23:19
< catalyst>
(which had a limit of 4096 characters :s)
23:19
<@simon_>
what's a managed compiler?
23:19
< catalyst>
Managed C++ is .NET-ified C++
23:19
<@simon_>
oh
23:19
< catalyst>
basically not-quite-C++-that-has-horrible-.NET-things-in-it
23:19
< catalyst>
One of the things this lets you do is write templated generics.
23:19
<@simon_>
well, that's the microsoft mentality. "1024 characters should be enough for everyone!"
23:21
< catalyst>
I still don't know *why* we had anything written in managed C++
23:21
< catalyst>
but
23:21
< catalyst>
hey, nevermind
23:24
<&McMartin>
I think my personal record is around 2.5K
23:24
<&McMartin>
simon_: To be fair
23:25
<&McMartin>
The fact that 1024 characters for your type name is not enough is a reasonable critique of C++ =P
23:25
<&McMartin>
Remember, eventually you're going to need to do something with it
23:25
<&McMartin>
Which means you're going to have to build a symbol out of more than one
23:26
< catalyst>
^
23:41 gnolam [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has quit [Connection closed]
23:41 gnolam [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has joined #code
23:41 mode/#code [+o gnolam] by ChanServ
23:59 Kindamoody|afk is now known as Kindamoody
--- Log closed Tue Nov 01 00:00:27 2016
code logs -> 2016 -> Mon, 31 Oct 2016< code.20161030.log - code.20161101.log >

[ Latest log file ]