development

[warm sleet]

you should create a new table

[upbeat nymph]

I guess in general it's semi useless because I'll need to benchmark all libraries differently considering the others expect me to track it

[warm sleet]

with a pk: <player1, player2, match>

and then move all the player stuff away from match

player <--> playermatch <--> match

then you can just join on that really easily

[upbeat nymph]

In production sure...but I don't know how that level of polymorphism would help in this case?

[warm sleet]

this is just normalizing

[upbeat nymph]

ah

This is all I'm storing currently

[warm sleet]

https://i.imgur.com/eaqNjtI.png

[upbeat nymph]

I never dug into it that much

[warm sleet]

Yeah I am used to working in 4th NF

[upbeat nymph]

Why not 6th?

[warm sleet]

because I am not google

@upbeat nymph these layers of normal forms are just to create consistency in your database

but way out of proportions for this case

@upbeat nymph you don't really need that id

wait.. no you do

[upbeat nymph]

yeah, I thought that too, until I realized this program is just counting lines

Again, me being stupid

And needing sleep

[warm sleet]

SELECT matches.id, matches.score, p1.rating AS rating_p1, p2.rating AS rating_p2
FROM matches 
JOIN players ON matches.player1 = players.id AS p1
JOIN players ON matches.player2 = players.id AS p2

@upbeat nymph how about this ^?

[upbeat nymph]

The problem is other libraries (like this: https://www.npmjs.com/package/glicko2) expect you to track the instances of the players yourself...so that wouldn't work on all...therefore giving advantages to one over the other...although based on the results, I could make it work...but I'll have to tinker with the results and how that query is returned in this environment..my concern though is that I'll be doing "unfair" steps for other libraries that I'm not for mine (go-glicko).

I may just end up splitting it into two individual queries in the effort to be fair

[warm sleet]

:3

I've written queries that were >50 lines

querying two dozen tables

returning a single result

becomes a headache real quickly

right

this seems straightforward

you pull the players you need to calculate for

query their entire match history

and then calculate

you don't want to be crunching the entire database in one go

https://github.com/mmai/glicko2js/blob/master/example/index.html

[upbeat nymph]

Except this package does it differently: https://www.npmjs.com/package/go-glicko
And so do all the others

That's where everything gets messy.

So I have to try to be as fair as possible.

[warm sleet]

they use Periods

something the other lib does do

[upbeat nymph]

Yeah, except the way the players are tracked is inherently different.

[warm sleet]

they just provide them differently

interface bit different

[upbeat nymph]

same general concept though

[warm sleet]

players.sort(function(pl1, pl2){ return pl2.glicko.getRating() - pl1.glicko.getRating(); });

is what

period.Calculate();

probably does

[upbeat nymph]

nope

period.Calculate runs the algorithm which runs about a 7 step process to up mu, phi, rating, rd, sigma, and things like that based on the difference in the score between player a and player b and then sets the post rating based on that.

The first on is just a sorting based on the rating

But you need all the players because even if a player didn't perform in a rating period they need to updated to reduce the confidence in the player's performance by increasing sigma and rd.

[warm sleet]

https://github.com/MatthewSH/go-glicko/blob/main/src/Period.ts#L29

[upbeat nymph]

what about it?

[warm sleet]

the other library

does not even do this

[upbeat nymph]

yeah, they do it differently. But same basic principal

I just split the calculations into more functions and a different class because .Touch uses the same functions as well

[warm sleet]

Player.prototype.update_rank = function(){

I hate javascript

[upbeat nymph]

yeah....

it's ugly

That's also 6 years old

https://github.com/ReedD/glicko-two

[warm sleet]

    Player.prototype._variance = function (){
        var tempSum = 0;
        for (var i = 0, len = this.adv_ranks.length;i<len;i++){
            var tempE = this._E(this.adv_ranks[i], this.adv_rds[i]);
            tempSum += Math.pow(this._g(this.adv_rds[i]), 2) * tempE * (1 - tempE);
        }
        return 1 / tempSum;
    };

think anyone understands this

[upbeat nymph]

this is more comparable

https://github.com/ReedD/glicko-two/blob/master/src/Player.ts#L121

[warm sleet]

one mechanic

3 different kinds of interfaces

and this one is different yet again

[upbeat nymph]

You talking about the packages?

mhm

[warm sleet]

yeah

this one expects you to make a playerFactory

[upbeat nymph]

Yeah, the reason I converted go-glicko is because it focuses more on the rating period

[warm sleet]

this algo is so simple

[upbeat nymph]

Which is where the matches and calculations should live

IMO

The player should just be used to track the player ratings...not track matches in a rating period..etc.

And instance of rating period should not be directly tied to a player...but a player should belong to a rating period

IMO

[warm sleet]

you should just have a List of matches that you feed into the algo

not care about anything else

all this weird player nonsense they do

[upbeat nymph]

Which package?

[warm sleet]

all of them

https://github.com/ReedD/glicko-two

[next igloo]

var gunTimer, bulletMove, ammo = 20, cursorX,
  cursorY;

$(document).ready(function() {
  $("#stage").click(spawnBullet);
});

$(document).ready(function() {
  gunTimer = setInterval(bulletMove, 1000 / 60);
});

function spawnBullet() {
if (ammo >= 1) {
    ammo--;
  let bullet = new Object({
  setup: function() {
    this.css("top", ($("#gun").position().top = $("#gun").position().top - 4  + "px"));
    this.css("left", ($("#gun").position().left = $("#gun").position().left - 4  + "px"));
    this.css({transform: "rotate(" + (Math.atan2(cursorY - ($("#gun").position().top + 4), cursorX - ($("#gun").position().left + 20)) * 180 / Math.PI) + "deg)"});
    function bulletMove(e) {
    this.move($(this).css("left", ($(this).position().left = $(this).position().left - 5  + "px")));
        }
     
    },  bulletmove: function() {}
  })
}
};

[upbeat nymph]

https://github.com/MatthewSH/go-glicko
https://github.com/mmai/glicko2js
https://github.com/ReedD/glicko-two
https://github.com/Googlrr/Glicko2-JS

So yeah. I'm actually, and this is gonna sound stupid, but probably create a more "scalable" package which actually says fuck the previous interfaces and makes it less "friendly" to the other packages but is purely meant for scaling to higher levels of datasets with a focus on clustering and stuff using redis

[next igloo]

what is wrong with this code

[warm sleet]

https://i.imgur.com/D2j31cz.png
ez, let me just get my calculator

[next igloo]

particulally the function spawn bullet, the rest of it works

[upbeat nymph]

That's just one equation

@warm sleet, there's about 6 more

[warm sleet]

@upbeat nymph my deep dive into sewer flowrate calculations based on geometry and sensor values was bad enough

I think I'm going to leave it as is, good luck with your.. assignment?

[upbeat nymph]

It's not even assignment. This is just for fun

It's taken me months. And this is just one part

Because I'm abstracting it for a team based game

which is...different

[upbeat nymph]

@warm sleet answer whenever you feel because I'm going to bed...but I'd love to know your thoughts on how you would do this: https://github.com/Glicko2/benchmarking/blob/main/data_seeding/index.js

This is essentially moving well...slow. Maybe it would be best if I generate the data once and dump it...or? Thoughts? Really I just need to seed data. No real structure of it. Just need to seed it with random values.

[warm sleet]

@upbeat nymph if you want to create reproduceability between individual runs

use a seed

to generate all your data

and then just store the seed and the result

[upbeat nymph]

I guess in theory the file size would be semi-small. I'm just thinking about seeing 10m rows.... :/

That would take a while.

Because it's already a fraction smaller than a json file is

ehhh, fuck it. I'll see multiple tables raning from 1k - 10m. Let's see what happens

bed time

Huh, didn't even make it...ran out of memory

time to pool the connections

Still running out of memory. Huh. I'll have to tackle this issue tomorrow

[next igloo]

im getting an error that this.css is not a function
this.css("top", ($("#gun").position().top = $("#gun").position().top - 4 + "px"));
what would i have to change at the start for the code to work?

[warm sleet]

@next igloo https://i.imgur.com/MmqMEgy.png

?

its uppercase?

[next igloo]

i have other code that uses $(this).css

so i would not think it would be that

oh

[warm sleet]

thats jquery

[next igloo]

i forgot $()

maybe thats why it wont work

[warm sleet]

@next igloo https://i.imgur.com/flFEtDn.png

look

this is jquery, not normal javascript in the browser

https://jquery.com/

@next igloo you can only retrieve values through the API

https://api.jquery.com/position/

https://i.imgur.com/1AEb6yn.png

if you set a style to have absolute positioning

[upbeat nymph]

Why not just...use JavaScript?

Like fuck the CSS

And shit

Use something like bye

Vue*

Wait. Never mind

This is my sleep deprived ass

[warm sleet]

what you could do is $('selector').css("top", newvalue)

[upbeat nymph]

I’m a fucking dumbass. I need to get off discord and go the fuck to bed

[warm sleet]

https://i.imgur.com/TSgYIdJ.png

these are the fields you could set for absolute positioning

@upbeat nymph https://mithril.js.org/

https://i.imgur.com/6625PsO.png

Quite sweet with the react extensions

[late plank]

Personally I use tailwindcss and I really like it!

[nocturne galleon]

This is always mentioned by my colleagues, very interesting

[bright hound]

so the problem is classically refered to as traveling salesman

[warm sleet]

@bright hound traveling salesman problem is something else

thats the question: how can a person travel to all nodes, with fewest moves possible

but thats not computable

[bright hound]

so it's brute forcable for small graphs

[warm sleet]

@nocturne galleon they may also know A*, (A-star) a variant of Dijkstra's with heuristics.

@bright hound its not computable

there's no one definitive answer

or at least, no way to consistently calculate

@bright hound https://en.wikipedia.org/wiki/P_versus_NP_problem

this gets philosophical very quickly

[bright hound]

it's np hard. so it's computable as a brute force with O time (O(n^4), O(n^3*2^n)) Tra

brute force computation is still computation.

[warm sleet]

there's no way for you to verify if it is indeed the shortest path

[bright hound]

brute force.

[warm sleet]

yeah but you never definitively know

[bright hound]

test all paths. add their lenghts find shortest

[warm sleet]

there may be another route, equally as fast

[bright hound]

this is a terrible algorithm and it's time expensive to run but it is an algorithm that can be run. it scales horribly as n increases and you will run quickly into calculation times that get ridiculous quicly. it is howeverl quite parallel friendly pick a start for each node have it calculate all paths starting and returning to that node etc. then collate results and do a sort and find shortest.

[hollow basalt]

but the traveling salesman is my favorite problem in algorithms

[bright hound]

no.. how can they take the shortest path overall. while visiting all the nodes

[wide ravine]

does anyone know of a decent guide on configuring apache and permissions related to it? all the ones i've been through are way too basic or fragmented

[pliant siren]

They don't. They work backwards.

[hollow basalt]

you can ask here though

[pliant siren]

If you work from the starting point, you need to evaluate way too many options to see if they even yield a result. But if you know your target, and work backwards, taking only completable paths and knowing the cost it took to get from each previous node, you vastly reduce the number of paths which must be evaluated, and can quickly eliminate portions of paths which produce a cost exceeding what options you have already evaluated.

This is similar to how raytracing works - instead of casting light from the source, you cast a ray from the viewport to the light sources.

Effectively, you could use RT cores to produce very fast A*, Djikstra, and similar derivative implementations

[hollow basalt]

Agreed

why cast the light [ray trace] from the source if some of the objects [surface] won't even show up in the screen

[mystic coral]

Bringing you the entire Roadmap for Web development in 2021.
If you are beginner who wants to get hands on web development, then this video is for you 🙂. Wanna be a Front-end, backend or Full stack developer, now is the right time to start 🔥🔥
I have explained the entire work flow for web development, what all technologies you should learn & have also provided the resources to start your journey 😀

Watch the video and don't forget to subscribe to the channel ✌🏻
https://youtu.be/AZ3zc0YLTsY

[warm sleet]

@mystic coral mongodb is only used by amateur web developers

and wordpress is a bad meme

[midnight wind]

is mongodb bad? I've never really done web dev that deeply, just my little bad project

[warm sleet]

@midnight wind its schemaless

document database

not a huge fan

doesn't do well with relationships and 'ACID'

its mostly used by javascript developers

[midnight wind]

I thought there are document schemas?

[warm sleet]

because you can store entire json documents

and then query them

but it encourages bad data structures

@midnight wind not just that, the guy in the presentation

has this chart of all the different technologies

[midnight wind]

oh

[warm sleet]

https://i.imgur.com/MTkTl2j.png

look at this ^

he completely skips the fact that, there is no single best technology

it all depends on your architectural constraints

[midnight wind]

yeah

[warm sleet]

and that graph makes it look like javascript is purely front-end

even though there are many serverside javascript projects

[mystic coral]

@warm sleet I'm not saying that who uses it. My motive of this video is to make newbies aware of the technologies that they can work with.

[warm sleet]

@mystic coral I think newbies ought to go to school

Web development is something you get a degree for

[mystic coral]

Brother I have also stated node na in backend. See the chart carefully

[warm sleet]

node isnt a language

rather a runtime framework

[mystic coral]

JS framework though

[warm sleet]

its a runtime

@mystic coral there's a lot more to architectural synthesis than this talk makes it look like

[midnight wind]

node runtime is chrome v8 engine right?

[warm sleet]

honestly, the whole server side stuff he points out can be completely skipped

[midnight wind]

or did they modify it

[warm sleet]

serverside technologies, from a front-end perspective, you only care about APIs

so REST, GraphQL, gRPC, etc

[mystic coral]

honestly, the whole server side stuff he points out can be completely skipped
@warm sleet why so?

[warm sleet]

^

[upbeat nymph]

Morning peeps

[warm sleet]

Frontend devs only call upon facades

They don't need to know that some backend service is running on python + flask

[mystic coral]

I wanted to throw light on all the technologies in this video. So I talked about server side languages as well > serverside technologies, from a front-end perspective, you only care about APIs
@warm sleet

Full stack developers need to > They don't need to know that some backend service is running on python + flask
@warm sleet

[warm sleet]

Sure

But I am pretty sure no newbie will be full-stack ;)

I've done fullstack development and I hate it

fuck frontend development xD

[mystic coral]

Yeah that's right 😅

[warm sleet]

the javascript/css/html ecosystem is garbage

browsers are not standardized

[mystic coral]

But after sometime they might want to switch to full Stack

[warm sleet]

HTML5 and CSS3 gets you only 90% there

the last 10% is webkit vs applekit

and whatever microsoft put into their browser

[upbeat nymph]

Chromium

IICR

For Edge

[warm sleet]

chromium browser engine is based on Gecko

webkit is based onthat too

but yeah; tldr web development is a mess

APIs are not mature

[midnight wind]

so is C/C++

that's annoying

[hollow basalt]

no, it isn't

[upbeat nymph]

Isn't Firefox running Quantum which is based of Gecko too?

[midnight wind]

Ok, I said that in a wrong way

Like between windows/linux/mac

[upbeat nymph]

Or is Quantum just their Rust version?

[warm sleet]

I haven't used firefox in forever

mostly because if 1 tab freezes, the entire browser dies

[midnight wind]

really? never had that issue

[hollow basalt]

me too

[warm sleet]

Chrome splits each tab into a seperate process

[hollow basalt]

Never had that issue in firefox, mostly I blame windows

[warm sleet]

uses more resources, but its more reliable

@midnight wind it was especially bad when coin.js was common on the net

it would completely lock up the browser

[upbeat nymph]

Oh ew, I remember those days

[warm sleet]

for those who don't know: coin.js is a browser based bitcoin miner

that secretly tried to use your CPU to mine coins

[upbeat nymph]

unethically too IMO

[warm sleet]

well yeah xD

but do you think advertising networks give a fuck?

[upbeat nymph]

If it were me, and the coin was valuable....I'd give users an option personally. Ads or CPU power

[warm sleet]

@mystic coral I skipped over majority of the presentation, but one thing I feel that is missing

there's no general explanation of underlying technologies

like, if you go to teach someone the basics of web development

start with HTTP

explain how documents and content is retrieved in a stateless manner

and the mechanism that provides these documents can either be static, for things like images and resources

or dynamic, with server side programming

[hollow basalt]

good thing my DNS blocks that automatically

so I only experienced that in UNI

but then again

[warm sleet]

pretty sure chrome blocks coin.js

[upbeat nymph]

I try to modify my DNS and use things like Warp but Hulu and them block it 😦

[warm sleet]

I tried using pihole

and then the twitch launcher fails to start

because it can't connect to their advertisement server

[hollow basalt]

I love those guys

But then again

it's their business

[warm sleet]

I just want to play minecraft :(

[midnight wind]

don't smart TV's nowadays ignore dns and use dnssec

[hollow basalt]

ikr

[warm sleet]

smart TVs have built in backdoors now

[hollow basalt]

but why tf does modding have to go thru tiwtch

[warm sleet]

samsung smart TVs are leak af

[hollow basalt]

I miss old launcher days

without twitch

[warm sleet]

@hollow basalt because CurseForge

and Curse is owned by Twitch

[hollow basalt]

they deserved the name Cursed

[warm sleet]

and Forge mods are provided primarily through Curse

dubbed

CurseForge

[upbeat nymph]

And there's some TVs that have video ads literally on start....

[warm sleet]

@upbeat nymph fuck really?

[upbeat nymph]

yeah

let me find it

brb

[warm sleet]

@upbeat nymph time to get a screwdriver and hunt for a UART port

[hollow basalt]

Your TV is weird lol

[midnight wind]

mine TV doesn't because it's a dumb TV, but the roku has an ad on the home screen

[warm sleet]

I've got this one project I've been pushing back for so long now

[mystic coral]

@mystic coral I skipped over majority of the presentation, but one thing I feel that is missing
@warm sleet I know. I haven't explained anything like that in this video. I'll make seperate videos explaining each technology.
But thanks for your reviews, I'll definitely keep this in mind while making next video.

[warm sleet]

I have an old seagate NAS with a dead MMC

but I can still get UART output of the UBOOT loader

so in theory I could load a kernel over serial onto the NAS

and flash it with my own OS

[upbeat nymph]

https://www.flatpanelshd.com/news.php?id=1583755244&subaction=showfull
https://www.thedrum.com/news/2019/09/10/the-first-thing-you-see-lg-smart-tv-now-ad
https://www.rtings.com/tv/tests/ads-in-smart-tv

[hollow basalt]

Yikes

[warm sleet]

@upbeat nymph wow. thats rude

[hollow basalt]

Time to install your own OS to that

[upbeat nymph]

https://www.reddit.com/r/assholedesign/comments/co5aw4/unremovable_ads_on_my_2500_samsung_smart_tv/

[warm sleet]

I have the facebook app on my galaxy A50

and I cannot delete the app, only disable it

I don't use facebook

[midnight wind]

same

can't get rid of it

[warm sleet]

And rooting the device is not an option right now

because it has an eFuse

so until warranty expires, I wont be rooting

[midnight wind]

doesn't rooting disable like that google security thing that I forgot it's name

[upbeat nymph]

https://piunikaweb.com/2020/06/02/block-ads-on-samsung-lg-roku-amazon-fire-stick-too/
https://www.reddit.com/r/OLED/comments/btcwd7/i_just_got_a_clickable_ad_on_my_lg_while_watching/

etc

[warm sleet]

@midnight wind Samsung has their own controversial tech

'Knox'

[midnight wind]

yeah

[warm sleet]

it says that on boot

'Secured by Knox'

as if vanilla android isnt secure

@upbeat nymph we have a 4k samsung TV here

but it runs outdated firmware

I actually isolated it completely from the rest of the network

only smartphones on the wifi can communicate directly with the TV

so they can still do streaming

[upbeat nymph]

That's probably a life saver here tbh

[warm sleet]

netflix runs

prime video works also

what more do you need

I also have a chromecast for when all else fails

[upbeat nymph]

HBO, Hulu, Peacock, Netflix, Prime Video, Disney+, YouTube, Boomerange, Twitch, Funimation/Crunchyroll

You clearly need all of that

[warm sleet]

@upbeat nymph some of the more annoying configs

https://i.imgur.com/C42wxkI.png

to get IPTV to work

the ISP gave me a manual on how to configure it

but the manual was for "DrayTek" routers only

draytek ofcourse uses way different jargon for the same functionality

multicast is so weird on v4

https://i.imgur.com/QIM97Nu.png

[upbeat nymph]

So I think I'm gonna ditch the NodeJS driver for seeding data and go straight SQL.

[upbeat nymph]

@warm sleet this may sound weird, but I think I like doing it this way better.

Our conversation(s) inspired me to do it in straight SQL instead of using middleman language

[warm sleet]

@upbeat nymph oof that loop though xD

[upbeat nymph]

And this, hopefully, removes some of the limitations I've been having with memory too

Okay look, I've never done this before

I'm trying

[warm sleet]

you can treat a SELECT result as a table

[upbeat nymph]

What do you mean?

[warm sleet]

@upbeat nymph well, it is a test setup so it dont matter

but things like your parameters, you can feed into the procedure as an argument

@upbeat nymph you are iterating over a list, to increment the primary key.. mh

[upbeat nymph]

I could set the primary key too.

[warm sleet]

tbf looping like that to create keys for records is perfectly fine

@upbeat nymph if you'd want to make it a little more special, what you could do

is use a subquery to generate a sequence from 1 to 1000

and then SELECT INTO your actual table, using the sequence to generate keys

sometimes I miss oracle DB

[upbeat nymph]

Why is that?

[warm sleet]

you could just SELECT * FROM my_sequence WHERE rownum < 10

generates 10 fresh numbers from a unique sequence

@upbeat nymph just an example on input params:

DROP PROCEDURE IF EXISTS proc_ban_ip;
CREATE PROCEDURE proc_ban_ip(
  IN addr     INTEGER UNSIGNED,
  IN reason   VARCHAR(128),
  IN duration BIGINT
)
  BEGIN
    INSERT INTO ip_address VALUES (addr, NULL, NOW(), NOW(), duration, reason)
    ON DUPLICATE KEY UPDATE
      banreason = reason, banned = duration;
  END;

IN keyword

oh, interesting you even use the while keyword, didnt know that was a thing

I'm used to using the GOTO method xD

https://i.imgur.com/jNZ04WQ.png

you can then just ITERATE <loopname>

and END LOOP <loopname>

[upbeat nymph]

Wtf. Can you not call multiple procedures D:

that's new

[warm sleet]

you should be able to

[upbeat nymph]

hm

[warm sleet]

a PROC can call another PROC

procedures also have a return type

if the last statement is a SELECT

it will return a table

[upbeat nymph]

Oh so I could in theory just make a proc called "seed" and call the two others within?

[warm sleet]

ye

lol some of these procedures

DROP PROCEDURE IF EXISTS proc_bungee_check_login;
CREATE PROCEDURE proc_bungee_check_login(
  IN ip VARCHAR(15), IN playerid BINARY(16)
)
  BEGIN
    DECLARE adr INT UNSIGNED DEFAULT INET_ATON(ip);
    IF EXISTS(SELECT ''
              FROM player
              WHERE uuid = playerid AND banned IS NOT NULL)
    THEN
      SELECT
        'PLAYERBAN' AS state,
        banreason,
        banned
      FROM player
      WHERE uuid = playerid;
    ELSEIF EXISTS(
        SELECT ''
        FROM ip_address
        WHERE address = adr AND banned IS NOT NULL)
      THEN
        SELECT
          'IPBAN' AS state,
          banreason,
          banned
        FROM ip_address
        WHERE address = adr;
    ELSE
      SELECT
        'NONE' AS state,
        NULL   AS banreason,
        NULL   AS banned;
    END IF;
  END;

[upbeat nymph]

oooo it works

[warm sleet]

Checks ^ if a player is banned on my minecraft server

[upbeat nymph]

huh

[warm sleet]

checks both user ID and ipaddress for a ban

xD

[upbeat nymph]

nice

[warm sleet]

IF EXISTS (query)

isgreat

returns true if the query has any records

@upbeat nymph this is why I love procedures;https://i.imgur.com/Z5tRPER.png

makes it so simple to have complicated rules in your db, but code is still clean

[upbeat nymph]

Probably quicker too

[warm sleet]

@upbeat nymph single round trip yeah

its key

[upbeat nymph]

Well. The 11 scripts are all running and seeding the databases successfully. So far no memory issues or anything.

Half of them are done because they were small databases

[warm sleet]

noice

[umbral saffron]

GG

[upbeat nymph]

So I guess the next thing I’ll need to decide is how to do the queries efficiently. I know for a fact selecting all 1 million-10 million rows will not perform well. So I might chunk it out and limit it to sets of 100k. So count(*) % 100000 or something.

Bro. The fuck. TikTok knows what I’m doing.

[next igloo]

can you convert scratch to javascript?

[vestal glen]

according to a quick search in the internet, yes.

[next igloo]

intresting, becuase ive trying to make a gun mechanic for a game in JS, but i still cant figure out how to do it. but i have made it work in scratch

are they free?

[vestal glen]

dont know

I didn't click on any of it

but in general, you should be able to apply the concepts from scratch to JS, too

unless you're talking about using the scratch drawing vs. a HTML canvas

[next igloo]

just the scripts

because scratch 3.0 is written in js i think

the thing worked except for the "step" function, since it has a lot of complex trig formulas i believe

.

what is a formula that will make something move 10 pixels in 360 degrees
preferably in JavaScript but if you have it in another language its fine too

[vestal glen]

well, there's scratch to js converters out there, go and search for yourself, given that "scratch to js" returned one as first result seems to me like it should be rather simple to find

and it may be a good help at teaching you how to transfer the concepts between the two languages.

[next igloo]

yeah i tried one of them now and it works except for the "step" function

i know how to code it in native JS

i just dont know the math formulas

[vestal glen]

So there's multiple ways forward here. You don't know the maths that you need to program what you want? Or what exactly is the issue you're facing ("not working" is a very vague description)

[next igloo]

i dont know the maths

for example, i want to move 10 pixels at 22 degrees

[vestal glen]

so 10 pixels is the length of your hypotenuse.

[next igloo]

correct

[vestal glen]

so you need cosine and sine

(to calculate the deltas for x and y coords)

[next igloo]

would that be angles?

[vestal glen]

no, those are functions that help you turn your hypotenuse and the angle into the other two sides of the triangle

(trigonometry)

[next igloo]

i didnt get to learn trigonometry since schools got shutdown so sorry if i get confused

[warm sleet]

@next igloo I guess you have to brush up on your trigonometry

anyone should be able to do basic calculations in a 2d or 3d plane

[next igloo]

i know as far as soh cah toa, the rest i never learned

[warm sleet]

ZERICO2005: what is a formula that will make something move 10 pixels in 360 degrees

this doesnt make sense btw

position of an object is always relative to a corner

so 0,0 is usually topleft of your screen

if you rotated the image, the position also has to change

depends on the dimensions of the shape

if your object has a dimension of 10x20

the relative position changes if you rotate it

but this is stuff you should figure out yourself

its not that hard

just play around

usually helps if you draw it on paper

and do the math like so

and then try to do this in code

thats why game engines are great

cus you can just translate the position relatively xD

[raw geode]

just wondering if anyone uses R or has experience with it? need some help with parallelization (or even GPU computing) ping me or dm me bc im offline 99% of the time

[nocturne galleon]

any1 have suggestions for an Android Coding App?

Thats like Virtual Studio Code?

[hollow basalt]

[nocturne galleon]

hey, was asking for suggestions, aight

i know there are lots of apps on the playstore

just wanna know which is good

the reason being is imma be out on a date tomorrow and i was gonna code while on my freetime on my phone hahaha

[hollow basalt]

don't

[upbeat nymph]

Then that relationship won't last long. If you're spending your entire date thinking about code, then you shouldn't even go on the date tbh. You should be focusing on your date. If you're not, then I would just avoid going out all together. I'm not saying that you couldn't take a break and check messages and such, but your date should be your main priority. If you have an idea subconsciously, which happens to me personally a lot, then write it down or something.

[nocturne galleon]

omg, it is a priority date, don't worry

[nocturne galleon]

But thanks for the words Zencep

Even tho my GF was the one who brought up the idea of helping me finish the project

I'll consider that

[orchid ferry]

Does someone know how to create a Contact form for a site

?

[warm sleet]

https://www.w3schools.com/html/html_forms.asp

@orchid ferry forms when submitted send all the fields from the form to the server

how it sends this is declared in action= and method= fields of the <form>

by default it uses POST

[orchid ferry]

?

Is this ok?

[upbeat nymph]

Does it work?

[warm sleet]

action has to point to the url of where it sends the data

action could be something like

[upbeat nymph]

If it works, it's okay. If it doesn't work, it's not okay

[warm sleet]

/submit

or whatever resource on the server is responsible for handling it

[orchid ferry]

no

[warm sleet]

@orchid ferry just put action="/" method with get

then press submit

should see your values in the address bar

[orchid ferry]

it shows them

[warm sleet]

yeah

if you were to use POST

it puts those in the http headers

server has to do something with this if you want it to have functionality

[orchid ferry]

How can i make it so that when someone submit it tit send a mail to me

?

[midnight wind]

you would need a backend server

[warm sleet]

You'd have to write a program that sends an email

that program would be called by sending an HTTP POST to a specific url

thats what your form submits to

[midnight wind]

I only know how to use node for a backend

[warm sleet]

@midnight wind I was taught this shit in php lol

$_POST['myform'] xD

[midnight wind]

idk php at all

[warm sleet]

php has global parametrs with request info

like $_SERVER $_GET and $_POST

$_SERVER['method'] returns http method

its crude

and error prone

[orchid ferry]

Do you use dreamweaver for html?

[midnight wind]

I wouldn't

wayyy to complicated

if you want a nice frontend use a framework like vue or react

[warm sleet]

I use chrome dev tools lol

test straight with a browser

[midnight wind]

same

but with firefox

vscode firefox debug extension

[warm sleet]

intellij has a tool for chromedebug

[midnight wind]

I can add breakpoints in vscode and it carry that data to firefox

[orchid ferry]

Should this thing work?

[warm sleet]

mail() in php requires some setup to function

[orchid ferry]

So this wont work?

What should i add

?

[warm sleet]

https://www.quackit.com/php/tutorial/php_mail_configuration.cfm

@orchid ferry RTFM

all of this is documented

[orchid ferry]

ok

Thanks

[upbeat nymph]

So at some point my cat unplugged my PC 😦

And I lost the 1m player and match data and 10m player and match data

So the 1m variation is almost done seeding, the 10m will have a long time to go

[upbeat nymph]

It's gettin' there

[umbral saffron]

why cant i install pyaudio

[upbeat nymph]

Could be anything

Is there an error output?

[umbral saffron]

yeah

oh i saw something saying it might be i need to be on an earlier version of python

i fixed thaat issue (I think)

but running ```py
import speech_recognition as sr

listener = sr.Recognizer
try:
with sr.Microphone() as source:
print("Go ahead...")
voice = listener.listen(source)
command = listener.recognize_google(voice)
print(command)
except:
pass```

nothing happens

@upbeat nymph

[upbeat nymph]

I don't know Python and I've never written in Python before

[umbral saffron]

hmmm

i dont know if i have the right imports

theres no issues

but it cant register the audio

so an import got messed up but no errors

so its running

but nothing happens

so now

i find out what

[swift spoke]

Never used the library, but could it be that you need to create an instance of the Recognizer class? So, basically, what you're currently doing is assigning a reference to the sr.Recognizer class to listener, and not an actual instance of that class. Using listener = sr.Recognizer() will initiate an instance of the class.

[umbral saffron]

What I think the issue could be is that I tried importing pyaudio (the thing that actually takes the input and sends it to speech_recognition) and I tried to run it anyways and I thought it worked

So what I think I need to do

Is find out how to import pyaudio Bc I can’t do it the normal way

[umbral saffron]

so uhhhh

pycharm wont open

[hollow basalt]

sure

[umbral saffron]

wdym sure

[nocturne galleon]

I'm working on a ping command, and am new to discord.js scripting.

Im using;

`${Date.now() - message.createdTimestamp}`

To get the ping time, however getting a negative ping, from what I've found is because off the system clock. So I looked up what to use instead, and I got;

resultMessage.createdTimestamp - message.createdtTimestamp```

`resultMessage` doesn't seem to be a property, so what do you use instead?

[devout wave]

Anyone know why Android studio would now require me to build twice in order to propagate all changes?

It's always scary when the logs you just added are not showing

[upbeat nymph]

I woke up to all data seeded

[warm sleet]

@upbeat nymph took a day generate 1m records?

sounds like terrible perf

[waxen quarry]

does anyone know the best linux distro for programming?

[upbeat nymph]

20 million records, but yeah. Probably not that long because I fell asleep at about 8m/20m.

It was generating about 280 records a second

If it did it in 20 hours

Although, I'm really surprised how small the database dump is for the 2 million records.

[warm sleet]

280records/second

@upbeat nymph I bet if you used batches

it would be more like 5x as fast

[upbeat nymph]

I really only needed to do it once

[warm sleet]

RTT to database becomes significant

[upbeat nymph]

Well, would it be faster to restore a database from a dump or to generate new data in batches?

[warm sleet]

restore from dump is faster

[upbeat nymph]

Yeah, that's all I needed

[warm sleet]

@upbeat nymph SELECT INTO from a foreign db is even faster

you can copy records between databases

[upbeat nymph]

Just only providing these if people want to run benchmarks across multiple systems so all benchmarks are done off the same dataset.

So that's why I wasn't caring too much about the time to generate data at first because I knew I would dump them all and restore from that in the future.

[warm sleet]

can probably compress those sql files

to 1/20th size

[upbeat nymph]

Oh? How would you suggest that?

I just did a DB dump from DBeaver

[warm sleet]

gzip -k file.sql

that will keep the original, and create a file.sql.gz

when you go to importing it, you just uncompress and pipe to mysql

like zcat | mysql

zcat uncompresses and prints content

[upbeat nymph]

lets see

doing it now

[warm sleet]

but ur on windows

so I doubt gzip is available

[upbeat nymph]

7zip

😄

[warm sleet]

yeah but gzip is not a container

its not like zip file

there's no directory structure

[upbeat nymph]

7zip supports gzip as well

[warm sleet]

hopefully

typical gzip archives are tarballs

wrap files into a tarball structure, then compress that

yields .tar.gz

[upbeat nymph]

Yeah, I work with tarballs a lot when moving files from linux to linux machines

transfer.sh is a great site omg. I love it

[warm sleet]

heh

@upbeat nymph I have a little webserver and a shellscript on laptop that puts a file onto the webserver via ssh

its basically like transfer.sh

[upbeat nymph]

yeah but with transfer.sh I don't have to host it myself lmao

[warm sleet]

yeah I am always privacy concerned

and don't like to litter files around third parties

except imgur, they can have all my screenshots

[upbeat nymph]

Interesting. 🤔

[warm sleet]

ok maybe not 1/20th

more like 1/3rd

[upbeat nymph]

Probably would be more on Linux based

[warm sleet]

@upbeat nymph repeated characters or sequences of characters saves a lot of space when compressed

INSERT INTO

can just be 1 byte

[upbeat nymph]

Now, would it be more compressed if I gzip a gzip file?

[warm sleet]

nope, those algos are actually capable of recognising archives and ignoring them

[upbeat nymph]

D:

[warm sleet]

if you add a .zip to another .zip its a 1 to 1 copy

[upbeat nymph]

Fuck gzip doesn't like multiple files.

[warm sleet]

gzip does not support multiple files

[upbeat nymph]

That's good to know

[warm sleet]

its not a container

zip, rar, 7z and such are all container archives

and gzip only with tar

though I guess you could gzip a zip file

xD

[upbeat nymph]

Hmmm. Let's try.

[warm sleet]

gzip takes bytes[] input and gives bytes[] output

it dont care if its a file, datastream or whatever

browsers use gzip as well :)

[upbeat nymph]

although 7z seems to be better.

[warm sleet]

most html documents are compressed before transmitted

[upbeat nymph]

For archiving multiple files

Oh wasn't that added with http2?

[warm sleet]

@upbeat nymph tar czvf myarchive.tar.gz directory/

@upbeat nymph nah http 1.1 has had this for a long time

https://i.imgur.com/d2CdrGA.png

https://i.imgur.com/5LhD0b5.png

deflate is zip

[upbeat nymph]

what was added with 2.0 that everyone was freaking out about? I can't rewmember

[warm sleet]

@upbeat nymph concurrent requests over single connection

https://i.imgur.com/75XwOpt.png

the idea is less requests, less latency and wait time, is faster load times

Its ideal for mobile devices and networked software

[upbeat nymph]

not too shabby

[warm sleet]

@upbeat nymph one of the reasons that I like gzip is because you can easily use it in pipelines

since it can dump its output in stdout

[placid cedar]

[umbral saffron]

yeah pycharms still not doin anything

[spring pond]

@umbral saffron is it doing anything in task manager?

[umbral saffron]

its there

its taking up a little memory

[spring pond]

try killing it

[umbral saffron]

i restarted it and now it works but now discord wont open but i think everythings loading

(im using the browser version)

[spring pond]

im not sure what could be wrong with discord especially since youre running it in your browser

[umbral saffron]

the discord app wont open and the browsers mega slow

ill fix it later

but right now

import speech_recognition as sr
import PyAudio

listener = sr.Recognizer
try:
    with sr.Microphone() as source:
        print("Go ahead...")
        voice = listener.listen(source)
        command = listener.recognize_google(voice)
        command = command.lower()
        if 'computer' in command:
            print(command)
except:
    pass
  ``` this literally does nothing

@spring pond

[spring pond]

have you tried following the previous advice of others who have said to convert sr.Recognizer to sr.Recognizer()

[umbral saffron]

ah

yeah it just closes immediatly

[spring pond]

it seems that PyAudio is not blocking by default

[umbral saffron]

i dont know what that means

[umbral saffron]

i got tts to work

so now i need to fix pyaudio

and then the gui...

[spring pond]

basically the microphone doesn't make the program wait for input

https://realpython.com/playing-and-recording-sound-python/

[umbral saffron]

ohh

[deft pumice]

Hey, i have a question about Linux OpenSSL. I am studying Computer Science, and i have to show and save the x509 certficate of a website. Which of the console comands is used for that? openssl x509 (...) ?

[shy helm]

openssl x509 -text host:port iirc

Wait, nupe

@deft pumice openssl s_client -showcerts -connect linustechtips.com:443 | openssl x509

[fallen cedar]

without knowing anything about the library, shouldn't that be sr.Recognizer() instead of sr.Recognizer?

at least based on the example here: https://github.com/Uberi/speech_recognition/blob/master/examples/microphone_recognition.py

lol, just read someone else suggested the same right after your comment

[umbral saffron]

I mean if you scroll up that’s what squid told me to try

[sacred pumice]

Today, on weird things

with open("/tmp/file.txt", "r") as f:
    result = pickle.load(f)
...
with open("/tmp/file.txt", "wb") as f: # PermissionDeniedError, but only sometimes. Also we're root, also it magically fixes itself with no interaction, also it happens on multiple different machines
    pickle.dump(result, f)

[root@azelphur-server local]# python
Python 3.9.0 (default, Oct  7 2020, 23:09:01)
[GCC 10.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> with open("/tmp/external_ip", "wb") as f: print("Hello world")
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>

Yup my brain hurts...I'm root

[hollow basalt]

uhhh

[sacred pumice]

Most simple example, still root. My friend also has the same issue, it's like the file somehow gets into a broken state

[root@azelphur-server local]# python -c "open('/tmp/external_ip', 'wb')"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
PermissionError: [Errno 13] Permission denied: '/tmp/external_ip'

[warm sleet]

@sacred pumice try running the program prefixed with 'strace'

that debugs any syscalls your program makes

should also print the reason why the filehandle was unable to be created

you are opening the file with write permissions, so if another program already has a file-handle, this will fail

[sacred pumice]

Don't think it yields anything super useful

@warm sleet lsof says nothing is looking at the file, so no other program in theory.

[warm sleet]

openat(AT_FDCWD, "/tmp/external_ip", O_WRONLY|O_CREAT|O_TRUNC|O_CLOEXEC, 0666) = -1 EACCES (Permission denied)

[sacred pumice]

Yea, just tells me permission denied, which isn't super helpful anyway

[warm sleet]

@sacred pumice what happens if you open with readonly?

[sacred pumice]

@warm sleet that's fine, it works

[warm sleet]

@sacred pumice try replace the 'wb' with 'w+b'

[sacred pumice]

Permission denied

[warm sleet]

@sacred pumice create a directory in /tmp

then put the file in there, try again

might not be able to write directly to /tmp. But this is very odd

I've never seen this kind of behavior

[sacred pumice]

I can create files in /tmp, eg touch /tmp/waffle is no problem, I don't really wanna do anything to modify the file, I'm really curious to leave it in place and figure out why this is happening

As I too, haven't seen this kind of behaviour, there's no reason I can see why root wouldn't have write access to a specific file in /tmp

[warm sleet]

https://i.imgur.com/Wrovlm0.png

¯_(ツ)_/¯

[sacred pumice]

yea, I mean on my system the file already exists

which is obviously a part of the problem

[warm sleet]

what about file permissions?

if you don't have +w

obviously get a permission denied error

[sacred pumice]

I'm root, so I'd assume that doesn't matter

[warm sleet]

no

that does matter

if you don't have write permission on a file, even root can't write to it

root can change the permissions, and then write to it

[sacred pumice]

[root@azelphur-server ~]# ls -l test.txt
-r--r--r-- 1 root root 0 Jan  8 11:12 test.txt
[root@azelphur-server ~]# python -c "open('test.txt', 'wb')"

Works fine, with no problems.

[warm sleet]

yeah

you don't have write permissions

[sacred pumice]

but it works fine, because I'm root

[warm sleet]

what.

what about that /tmp file

[sacred pumice]

-rw-r--r-- 1 azelphur azelphur 113 Jan  7 08:51 /tmp/external_ip```

@warm sleet but you can see what I'm saying, root ignores file permissions, write flag doesn't need to be on the file if you're root.

[warm sleet]

ls -la /proc/$$/fd | grep external_ip

should print you any open file descriptors for that file

my guess, another process has that file open

[sacred pumice]

returns nothing

[warm sleet]

wat.

[sacred pumice]

lsof /tmp/external_ip also returns nothing

yes I agree, this situation is wat 😂

[warm sleet]

if lsof doesnt even list anything

probably not gonna matter trying to find anything with fuser

because that all uses lsof

fuser -v -m /tmp/external_ip

[sacred pumice]

                     USER        PID ACCESS COMMAND
/tmp/external_ip:    root     kernel mount /tmp```

[warm sleet]

PID: kernel

[sacred pumice]

so that means that the kernel currently has the file open for write?

[warm sleet]

not quite

that just shows us what process is responsible for maintaining the filesystem that /tmp/external_ip is in

which is the kernel, obviously

[sacred pumice]

ah, so nothing useful then really

[warm sleet]

try without -m

[sacred pumice]

returns nothing

[warm sleet]

yeah

I am.. at a loss

[sacred pumice]

me too, it's a weird one

[warm sleet]

yeet the file

and try again

xD

or perhaps try restarting the system

[sacred pumice]

I don't think it's some oddball condition, because my friend who is also running my program has the file in the same condition on his machine

@warm sleet if you're curious, https://dpaste.org/224k

It's really basic stuff though, the read on line 26 works, the write on line 47 fails (and of course, the examples I've posted above when the program isn't even running) this is just potentially interesting because it's how the file was created in the first place.

[warm sleet]

why do you even need an external ip reference

you could replace that with a shell oneliner that uses some online service

I have one of those ipv4.domain.com subdomains

that just echoes your ipv4

[sacred pumice]

@warm sleet it's a plugin for checkmk (service monitoring) it checks that the external IP address of my torrent client is not equal to my real external IP

[warm sleet]

to make sure you don't leak ur IP ?

[sacred pumice]

correct

[warm sleet]

why not check default route

[sacred pumice]

because route wouldn't contain my external IP, it's a machine on a NAT

It also does a few other bits (like returning current up/down speeds, and number of torrents, torrents in error state) which the monitoring system handles, but yea, I doubt it could be replaced with a shell one liner

the file is a cache so it doesn't hammer ifconfig.co

[sick dome]

I deleted the security tools package and the GUI is gone

This is a centos7 VM

How can I get the GUI back?

[midnight wind]

what gui?

cockpit?

or

KDE, MATE?

[sick dome]

GNOME gui @midnight wind

[midnight wind]

idk

this is over ssh?

or no?

[sick dome]

no, it's in virtual box - I think I may have fixed it

sudo yum -y groups install "GNOME Desktop"

used this command

I think it's installing

[warm sleet]

lol wat

@sick dome xinit probably failed

thats the program that launches the desktop environment

[obtuse night]

Are Azure Service Bus and RabbitMQ interchangable as they both use AMQP protocol? I am beginning developing microservices but do not want to be locked down to RabbitMQ. In short, can you swap out the message broker in an application by changing just the connection string

[warm sleet]

@obtuse night doubt anything azure is compatible with amqp

ZeroMQ can also use amqp, though they have their own protocol spec too

And then there's MQTT

Which is a public standard for messenger exchanges

Specifically for embedded telemetry and IoT

[upbeat nymph]

TIL: You can use $npm_config_*flag* and use npm run *script name* --*flag*=*whatever* to use in your package.json scripts.

[sick dome]

So is it normal every single thing in my linux VM is all up to date?

I'm doing a linux course and the guy is updating a ton of things and they all seem to be updated on my machine

[lapis orbit]

@sick dome - depends - which distributions are you using?

[sick dome]

Centos7

[lapis orbit]

@sick dome - he is too? If so, perhaps perhaps you got as more recent image? You should check which versions of the packages you have compared to his.

[sick dome]

the course was recorded in 2017 and I downloaded it recently from a repo at my university - could that be why?

[lapis orbit]

Yes, that would be why. Distributions often make more up-to-date images for installing versions wit the bug fixes etc,. present.

[sick dome]

oh, okay. Thanks guys!

[velvet badge]

Any python programmer out there?
I need help adding music features to my bot, your help would be sincerely appreciated

[bleak breach]

I’m a python dev

[hearty edge]

yey

Have you ever had to choose between studing development or servers and comunications

xD

[warm sleet]

@hearty edge you mean software development, or managing software that other's have developed?

as a programmer, I learned networking skills on the side

learning linux also came early on, when you write code and deploy it on servers, linux feels incredibly natural

its all very streamlined and easy to use

developers have a much deeper understanding of how these systems work

network admins are only good at network stuff, maybe some linux as well

[hollow basalt]

No

[left drum]

hmm

[hollow basalt]

I was thinking that maybe i could just use the heroku database straight away.
What do you mean

[left drum]

Ahh never mind. I purged and reinstalled postgresql locally. Its working now. I explained myself poorly earlier

[umbral saffron]

how do i fix pyaudio

if i run a command

to use speech recognition

it immediately closes the console loop thing

like theres no code there

[spring pond]

take a look at the docs, there may be a way to make the thread wait

[kind sail]

Docs are your best friend

[next igloo]

01 function spawnBullet() {
02   $("#bullet").css({transform: "rotate(" + (Math.atan2(cursorY - ($("#gun").position().top + 4), cursorX - ($("#gun").position().left + 20)) * 180 / Math.PI) + "deg)"});
03   bulletAngle = (Math.atan2(cursorY - ($("#gun").position().top + 4), cursorX - ($("#gun").position().left + 20)) * 180 / Math.PI);
04   $("#bullet").css("left",($("#bullet").position().left = $("#gun").position().left + "px"));
05   $("#bullet").css("top",($("#bullet").position().top = $("#gun").position().top + "px"));
06   }
07 };
08 
09 function bulletMove() {
10   $("#bullet").css("left",($("#bullet").position().left = $("#bullet").position().left + (Math.cos(bulletAngle) * 10) + "px"));
11   $("#bullet").css("top",($("#bullet").position().top = $("#bullet").position().top + (Math.sin(bulletAngle) * 10) + "px"));
12   if (($("#bullet").position().left <= 8 || $("#bullet").position().left >= 776) || ($("#bullet").position().top <= 8 || $("#bullet").position().top >= 488)) {
13     //reset
14   }
15 };

is there an issue with lines 10 and 11, they do the math to find out where the bullet should move to

at the moment the bullets seem to fly off at random angles that are completely off from your aim (clicking on the same pixel gives the same angle)

the bullet <div> is at the correct angle, just not the movement

[warm sleet]

the definition of spaghetti code

cross cutting concerns? neverheardofit

CSS, math, abstract concepts, all in 1 method.

[hollow basalt]

the more the better

no need to obfuscate if the code itself is unreadable

[mighty wharf]

good lord do you want bologneese with that spaghetti?

and so the next one and a half years of my life begins

[gusty girder]

this might be a stupid question, but when using npm/yarn, do I just manually load the dependencies in my html file?
I'm used to PHP where you just load the single composer file that links everything

[nocturne galleon]

You put the packages you depend on in the package.json file

If you're bootstrapping your project, or simply cloned it from a git repo, run npm install

that will create a node_modules folder in which all deps are downloaded

[gusty girder]

No, I get that, I use terminal to download/install dependencies and the package.json stores them.

It's about how I access them within my html/javascript.

like <script src="/node_modules/somerepo/repo.js" /> can't be the right way to do it?

[nocturne galleon]

In the JS you do, at the top of your source file:

import myLibrary from 'myLibrary';

the syntax may differ:
for named exports you use import { NamedExport1, AnotherNamedExport } from 'myLibrary'

for default exports you use simply import aliasForMyLibrary from 'myLibrary';

then in the code you can do:

// named:
let foo = new NamedExport1(...); // NamedExport is a class
// default:
let foo = new aliasForMyLibrary.AnotherExport(...); // AnotherExport is a class

You can find out more here https://javascript.info/import-export

[gusty girder]

So I have to manually import each of my dependencies?
At that point I might as well just manually link to them, what's the difference between a single javascript file with multiple import statements and a few script tags?

Is there no way to automatically create an "import_all_yarn_packages.js" file and use it like: "<script src="import_all_yarn_packages.js" />

which would then enable me to use all the libraries/packages I imported below that script tag

[nocturne galleon]

You can do something like this if you end up using the same libs over and over:
📝 common.js

import _ from 'lodash';
import React from 'react';
import myLibDefault, { ClassA, ClassB } from 'myLib';

export { _, React, myLibDefault, ClassA, ClassB };

📝 anotherFile.js

import { React, ClassA } from './common.js';

def foo = new ClassA();

export class ComponentExample extends React.Component {
  ...
}

note that the actual way you export stuff in the first file depends on the way each lib exports what, but you get the idea

you're actually using common.js as a "dependency collector" for your project
though I've not seen this approach often.

[nocturne galleon]

you don't want to use script tags and expose deps if you're going to compile and minify your js

otherwise you'd have to make node_modules public, which will allow hackers to look in your package for vulnerable deps and exploit them

[nocturne galleon]

plus some deps might use an incompatible system, and the browser wouldn't know how to deal with the js file, resulting in console errors.

[gusty girder]

But javascript is client-side, so what packages I'm using gets exposed anyway?

I'm using yarn to manage packages for front-end development, I'm not using node.js, I'm doing front-end, not backend.

Thanks though, I'll look into that further

how does a random javascript file know that lodash is located node_modules/lodash/src/lodash.js?

I'd have to do import package from path/to/package and given js is client-side it'll be visible either way

[warm sleet]

@gusty girder isnt that what webpack is for?

generating an application entrypoint

I've just recently written an SPA with typescript

and webpack assembles the bundle.js including all the libraries you need

[nocturne galleon]

when compiling it gets resolved automatically

[nocturne galleon]

at least, you're not exposing the node_modules folder, with package.json files of each dep, with the version number inside

[warm sleet]

const path = require("path")

module.exports = env => {
  return {
    entry: {app: path.join(__dirname, "src/spa/app.tsx")},
    devtool: "source-map",
    output: {
      filename: "[name].bundle.js",
      chunkFilename: "[name].chunk.js",
      path: path.join(__dirname, "dist"),
    },
    module: {
      rules: [
        {
          test: /\.tsx?$/,
          loader: "ts-loader",
          options: {
            configFile: "tsconfig.client.json"
          }
        },
      ],
    },
    resolve: {
      extensions: [".tsx", ".ts", ".js"]
    },
    optimization: {
      chunkIds: "named"
    }
  }
}

This was the webpack config I used. This builds the index of your all dependencies automatically

[nocturne galleon]

the only things different from npm is:

• yarn add instead of npm install ...
• yarn.lock instead of package-lock.json

and a couple other minor things

[warm sleet]

@gusty girder https://webpack.js.org/

This is what you are looking for to expose your code to a webserver

https://i.imgur.com/AhfXuYT.png

All you have to do is point to the main application codefile

and it will assemble a bundle.js, including all the code that your main file references

[covert axle]

impostor!

[nocturne galleon]

Yellow sus

[gusty girder]

thank you that looks very promising. Google said to use yarn for front-end which is what I did, but seems my question was the wrong one.

[warm sleet]

@gusty girder yarn is merely a package/project lifecycle manager

just like npm

they pretty much work same way

With webpack, your software is 'compiled' (minified) and gets written to dist/

you then expose this directory in the webserver that hosts your application, as a static asset

so you have /index.html

and /dist/bundle.js

Webpack also allows 'chunking' of code, to load code as it is needed

that's helpful if you develop large single page applications, with a lot of code

[ripe garnet]

I'm wondering if it's possible to get the current Display (xlib) in an application, as I'm trying to get all the pixels in an area of the screen

Nevermind, apparently I can just call XOpenDisplay

[viral parrot]

advice

https://cdn.discordapp.com/attachments/783511644524642344/798716079249031238/ErkaithU0AEa7sN.png

[hollow basalt]

not if I throw away my code faster

[latent estuary]

https://youtu.be/gEptbAfhpnk

[lament cave]

im trying to use nginx as a reverse proxy for hypercorn (with Quart), but it doesn't seem to be working.
there are no error messages or logs other than the "Running on" message

I have other subdomains (which is why im using nginx and not just hypercorn directly) and all of them work fine except for this one

attached is a screenshot of hypercorn:

nginx config file: https://i.cocobot.tk/iK7QhL0Q.png

nginx -t shows valid config file

[warm sleet]

@viral parrot good code is self-documenting

if you properly split your logic into functions, you can document those

but //inline comments like this one

that explain what a line does, are kinda stupid

[covert axle]

Everyone was saying that I was stupid not putting comments and now there is you saying that I shouldn't

jk, I think too that making the code easier to read with tabs and other things is better than comments

[warm sleet]

@covert axle

//writes buffer to output
out.write(buff.read(size))

I'm more talking about this kind of commenting ^

[covert axle]

yeah

[viral parrot]

well yeah but

if you have a bigger piece of code it is harder

[covert axle]

you only need comments before long functions or for loops so people can search for the part they wanna tweak

[warm sleet]

Not really

You just need to structure your code

[covert axle]

exactly

[warm sleet]

Spaghetti-o's are bad in any language

and no amount of commenting will make that sane

[covert axle]

I learned this when I moved from c++ to C#

[warm sleet]

I'll tell you an even worse habbit

Moving from java to nodejs/typescript

; everywhere

;;;;;;

don't need those in js

[viral parrot]

i see what your saying, but whenever you go to stack exchange to find what you need, do you really think the code you have will make you remember why its that way and not the other?

[covert axle]

yup

[warm sleet]

@viral parrot if you take an example from stack exchange and dump it in your code

it might not be a bad idea to just dump the URL into a comment above the function itself

[viral parrot]

no i mean

not taking an example but rather

[warm sleet]

all that you care about is that you can access the same source of information, when you revisit the code

code is self documenting, so only things that aren't, deserve comments

[viral parrot]

i understand what you say. personally, i think i need to improve on this though. I have ADHD and I don't know what i did an hour ago without looking at my web history so i need to keep more specific records

[warm sleet]

you and me both

Except I have Asperger's ontop of that too

@viral parrot lol you know what you should do

use git

I can just go through my application history

https://i.imgur.com/xUZVE3e.png

and see exactly what I did, when and what lines changed

[viral parrot]

oh yeah

btw, if you have adhd, check out adhd-alien on twitter and other places. very relate-able

[jolly arch]

Hello guys

does SFML multimedia library work on macOS big sur

[vernal marsh]

Hi! i was wondering if anyone can help me in a programming related question.... https://www.codechef.com/IARCSJUD/problems/EXPT this is the question .... I am trying to solve it in the following way lol its just paint so pls ignore that pic's details... I can have a 2d array whose elements are 1st row elements = array1 elements - array 2's 1st element as shown in the pic above .... then find the most occurrences of a single digit in 2d array then get the index of the occurences to then find back the elements in original given arrays....

https://gist.github.com/Sid911/ca18e27ff554169e796f229204ec6dd2

here is my very bad looking but still "should be working" code... I have checked it in many testcases and checked those same testcases with submitted codes of others... the answers are same but if anyone could tell where I am doing wrong it would be awesome

[south grove]

holy hell screw editing markdown tables

[next igloo]

$("#bullet").css("left",($("#bullet").position().left = $("#bullet").position().left + (Math.cos(bulletAngle) * 10) + "px"));
$("#bullet").css("top",($("#bullet").position().top = $("#bullet").position().top + (Math.sin(bulletAngle) * 10) + "px"));```

hello, can someone correct the math? its supposed to make a bullet travel at any angle
* 10 is the speed of the bullet

[pliant siren]

uhh... the assignment looks more of a problem than the math to me.

why are you setting the css value of top to an assignment statement, that seems odd and redundant unless there's some weird css nuance I'm not familiar with

[bleak breach]

$("#bullet").css("left",($("#bullet").position().left + (Math.cos(bulletAngle) * 10) + "px"));
$("#bullet").css("top",($("#bullet").position().top + (Math.sin(bulletAngle) * 10) + "px"));

[sacred pumice]

I just experienced recruiterception. The recruiter that got me the job at my current company called me to ask if I was interested at a job at my current company.

[umbral saffron]

so running

import speech_recognition as sr
import PyAudio

listener = sr.Recognizer
try:
    with sr.Microphone() as source:
        print("Go ahead...")
        voice = listener.listen(source)
        command = listener.recognize_google(voice)
        command = command.lower()
        if 'computer' in command:
            print(command)
except:
    pass``` nothing happens

so what i think i need to do is add a wait command to it

right?

wait but not even the print part works

so i dont get errors anymore

just nothing happens

[spring pond]

well you arent handling the exception

you need to see what the error is

except Exception as e:
  print(e)

imo you shouldn't pass in exceptions unless you are very sure you need to

@umbral saffron

[umbral saffron]

ok

[umbral saffron]

how would i find that if none come up in the console

[spring pond]

just replace the except code with what i wrote

[umbral saffron]

ok so I got rid of the if computer in command part

[spring pond]

wait what

import speech_recognition as sr
import PyAudio

listener = sr.Recognizer
try:
    with sr.Microphone() as source:
        print("Go ahead...")
        voice = listener.listen(source)
        command = listener.recognize_google(voice)
        command = command.lower()
        if 'computer' in command:
            print(command)
except Exception as e:
  print(e)

This is what your code should look like

[umbral saffron]

yes but I got rid of the
if 'computer' in command:
line because im just trying to test it now

[spring pond]

so what does it print

[umbral saffron]

'Microphone' object has no attribute 'get_pyaudio'

[spring pond]

uhh

welp i have no idea

try reading the docs again

[umbral saffron]

ok

[umbral saffron]

so what I think was the issue in the original code was that it wont wait for the input

@spring pond

[spring pond]

nah the issue (at least the current one) is that it wont even recognize your mic

that may still be an issue tho

[umbral saffron]

ok

wait lemme see something]

yeah i have no idea whats wrong

[spring pond]

try to add print(sr.Microphone()) somewhere

[umbral saffron]

still same thing

yeah still same message

[spring pond]

where did you put the print

[umbral saffron]

import speech_recognition as sr

listener = sr.Recognizer
try:
with sr.Microphone() as source:
print("Go ahead...")
voice = listener.listen(source)
print(sr.Microphone())
command = listener.recognize_google(voice)
if 'computer' in command:
print(command)
except Exception as e:
print(e)

[spring pond]

nah put the print right at the try