00:27
yewscion left
00:28
yewscion joined
00:46
LainExperiments4 left
01:15
yewscion left
01:24
LainExperiments joined
02:10
hulk joined,
kylese left
|
|||
jdv | [Coke]: perladvent.org/2024/2024-12-19.html | 02:17 | |
that was fast | |||
02:21
yewscion joined
02:26
yewscion left
03:15
hulk left,
kylese joined
04:02
LainExperiments left
06:41
Aedil left
06:51
Aedil joined
09:02
Sgeo left
09:12
dakkar joined
09:50
[Coke] left
10:19
sena_kun joined
11:20
LainExperiments joined
11:49
[Coke] joined
|
|||
Geth | advent: librasteve++ created pull request #125: 23 scheduled |
13:40 | |
advent/main: 185344bd48 | librasteve++ (committed using GitHub Web editor) | raku-advent-2024/authors.md 23 scheduled |
|||
advent/main: 72aaabdece | librasteve++ (committed using GitHub Web editor) | raku-advent-2024/authors.md Merge pull request #125 from librasteve/main 23 scheduled |
|||
14:09
LainExperiments left
14:39
yewscion joined
14:44
Sgeo joined
|
|||
antononcube | Has anyone worked on programs / apps that cheat employee monitoring software? | 14:49 | |
Or the other way around -- has anyone worked on employee monitoring software? | 14:50 | ||
Preferably, using Raku. | |||
lizmat | TIL there's such a thing | 14:57 | |
antononcube | 🙂 Here is relevant discussion: www.reddit.com/r/overemployed/comm...?rdt=54750 | 15:00 | |
15:17
vrurg_ joined
|
|||
El_Che | antononcube: in the old days, I just wrote a small mouse randomizer. Nowadays I say fuck it, and install Linux :) | 15:20 | |
no idea how I get away with it for decades in different places | |||
15:20
vrurg left
|
|||
antononcube | It is most likely illegal in EU. (If you live in EU...) | 15:21 | |
El_Che | it was used in the old days | ||
but not as extreme as in the link | |||
mostly something that recorde inactivities | |||
d | |||
antononcube | Also, to quote Fat Boy Slim: > If you walk without rhythm, you won’t attract the worm. | ||
@EL_Che In what language did you write that? (Perl?) | 15:23 | ||
El_Che | java, there was some basic windowmanager integration, like moving the mouse and get the resolution. It was trivial | 15:24 | |
antononcube | There is a need for two agents: (i) a employee minitor and (ii) a cheater. | 15:30 | |
Otherwise, the cheater agent might get too easily detected by the monitoring software. | 15:31 | ||
To rephrase: > The AI I made for doing my work is going to be monitored by another AI . | 15:34 | ||
holmdunc | I wonder with those web CAPTCHAs that ask you to do nothing more than click a checkbox or drag something to the right, is it looking for human-like mouse cursor movement | ||
antononcube | They might be! I strongly suspect they deliberate put images they do not know about in order to get human classifications. And/or cross-validation of their ML algorithms. | 15:36 | |
15:50
yewscion left
|
|||
librasteve | lizmat: I think that employee monitoring was one of the main features of Crowdstrike AV software btw | 16:31 | |
antononcube | I should have advent-blogged about data obfuscation (with Raku.) | 16:35 | |
From Machine Learning perspective. | |||
librasteve | anti AI monitoring agents that use AI - now that's a real use case for AI | 16:36 | |
17:32
dakkar left
|
|||
Geth | advent/main: 82003fbb2f | ab5tract++ (committed using GitHub Web editor) | raku-advent-2024/authors.md Update authors.md |
18:05 | |
antononcube | I have this "old" REPL code: $!compiler := nqp::getcomp("Raku") || nqp::getcomp('perl6'); $!repl = REPL.new($!compiler, {}); | 19:20 | |
How should I change it in order not to get the error: "Invalid REPL environment: Unable to initialize REPL outside of a TTY" ? | |||
ab5tract | antononcube: checking | 19:21 | |
lizmat | ab5tract might know | ||
antononcube | That question is in relation the the issue of "Jupyter::Chatbook" reported here: github.com/rakudo/rakudo/issues/5726 | 19:22 | |
I tried to figure out what to do looking at the code here: github.com/lizmat/REPL | 19:23 | ||
ab5tract | one option is to have whatever code is starting the Raku process include `--repl-mode=process` | ||
antononcube | I just tried: $!repl = REPL.new($!compiler, {}, try-mode => 'process'); | 19:24 | |
Will try that suggestion now... | |||
ab5tract | it needs to be set in the compiler already (as currently implemented) | 19:25 | |
antononcube: keep in mind that I'm not sure what the approach should be for when changes to Rakudo internals breaks user-space modules that are opting to use nqp | 19:26 | ||
as it's not really a bug per se when things change on layers where fewer/no guarantees are made | 19:27 | ||
antononcube | Ok, this should be a problem with "Jupyter::Kernel" too. I will verify that it is and file an issue. (For Brian Duggan to figure out. 🙂 ) | 19:28 | |
ab5tract | Wherever the original "start the raku process" is happening is where the `--tty-mode=process` needs to be added | 19:29 | |
as that's a setting in the compiler object that you are retrieving via `nqp::getcomp` | |||
you might be able to get away with just creating a new compiler object :) | 19:30 | ||
antononcube | Well, I cannot figure it out, for now. I have to postpone the investigation for later this week / month. | 20:09 | |
20:22
orangebot joined
20:54
Aedil left
22:17
zetaaaa joined
23:08
sena_kun left
|
|||
ab5tract | antononcube: unless I'm mistaken, this should be enough github.com/antononcube/Raku-Jupyte...ook/pull/6 | 23:55 |