Computer Science Study Groups?

Would anybody be interested in forming / joining a computer science study group? I don't know if we'd all simultaneously study one topic / book together like in other groups, but ideally members could ask each other questions, share what they are currently studying and network with others. I'm trying to relearn some typical subjects such as discrete math and algos (I've forgotten most of it since I graduated a year ago - also due to bad instructors) and it'd be great to talk to others starting out or in the same boat.

I would actually be very interested in this. I am a physics student, but I would like to learn more formally about computer science over the summer. I'll let others respond to this, as I do not really prefer one topic over another.

i'm very interested in this group

I like this idea. I can share the books we use in my CS major, although they're not too difficult to find.

>>13533 i would like to partecipate

ye

You guys remind me of those ants in the Amazon that stick together to form a raft and don't see that everything is flooding around them.

>>13533 I'd be interested too. Graduated several years ago, trying to go back through all my math and such so I don't forgot it, and working on learning programming better.

>>13533 i would be keen dont know much about the math side of programing and trying to learn

I'm interested in a simulated collaborative project where we assign tasks for different people and can practice the git workflow.

OP here, it's great that many of you are interested in this. As for the platform this group would be hosted on, I'd like to gather some recommendations / opinions. Discord? IRC? Likely I'd just use Discord as it's probably a given most of you are on it already.

>>13550 I think IRC would be a better idea than discord. I'm not too familiar with discord, but wouldn't they crack down on us for sharing things like PDF's of books? Also, irc is more universal than discord (no account needed, all FOSS software, lower system requirements), so more people could join.

>>13551 or we could use both platforms and set up a bridge between them https://github.com/qaisjp/go-discord-irc >sharing things like PDFs they won't crack down on that soykaf I have seen full lenght movies (very low quality obviously) uploaded to discord and never had issues sending pdfs/epubs to other people

>>13533 We already have several of these, see #wired-students on Lainchan IRC for the IRC channel that was first created when this was suggested. The reason these study groups don't tend to stick around is because Lainons find self organization and deciding on a curriculum and repeatedly showing up consistently to be difficult. Lainchan has a calendar and an IRC for study groups, please feel free to use them.

i need to learn some computer networks. Anyone wants to also learn it? Anyone klowledgable to be mentor for a study group?

only few study groups survived long enough to be useful on this board and those groups had very narrow topic & tight time table to keep peers engaged (right now I can only recall one lisp study group and one DIY lisp interpreter implementation group). Faster people can settle down on specific topic & book to share, better chance of actually learning something this summer. anyone wants to study SICM with me? My scheme does not suck that much anymore and I want to relearn physics

>>13553 Ah, ok. I wasn't aware of that. An idea popped into my head just now though. Given that organization and consistency seems to be a concern, perhaps instead some of us could arrange lightning talks or even one-off lectures on whatever we are studying? Has that happened before? The talks don't have to follow a consistent schedule or anything but I feel that could put the #wired-students channel to good use.

>>13545 isnt the floating ant raft a winning strategy as long as the queen survives?

>>13549 im keen to join that

>>13547 In a similar situation myself. I'd also like to take part in group studies.

I'll join if it's competitive and you compulsory have to do your part in a project at the end.

>>13555 I'm in a course of communications and computer networks right now at uni, I need to study it too. I can give you the outline so we can study together. The third lecture was just yesterday, it was an overview of TCP/IP layers and protocols.

>>13569 I am also in networking course, but labs (with cisco routers and all this fluff) are hectic, i have problems with static/dynamic routing, and i am probably laughing stock for guys that do it every day at work

>>13570 >>13569 I just finished a networking course. My advice is to memorize everything. Dedicate time each week to just memorize stuff. As for practicals just actively develop your own workflow. Decide what you do first; do you connect everything up or setup the IP and default gateways on the machines first. Then double check stuff you expect to be working is working at each stage, double checking you didn't typo an IP address is a life saver btw. After that it's really just practice with entering commands. Oh and if you have access to machines then use the machines avoid packet tracer at all costs. Packet tracer is very easy and you'll pick it up in like 2 hours. If you use it instead of the machines you'll end up like most of the people on my course who couldn't find the console port on the switch for the exam.

>>13557 It another interesting idea is if someone starts a thread about a topic they know lots about, write about it, common pitfalls, advice, Q&A, assignments. No time line required, just keep it alive so it doesn't 404 Basically it's like how things are now, anyone shares any input they have, but it's more guided by the wise OP, who watches the thread for the long long. Ignorant people might start teaching threads too, but that's OK if discussion, and discovery is done. Learning still will happen.

Has anyone created a room or channel to join?

>>13533 I hate to be that guy, but imageboards are pretty suitable for the study groups. You just make a thread about whatever you want to co-study and that's it. Anyone interested can write messages, post links, share pdfs or code. You don't need yet another platform, be it irc or discord, it's just a rationalization of your own weak will, so you tell to yourself "I will do X then Y happens", and then you don't do it regardless. A separate thread with well constructed OP post with links and some kind of FAQ is the only thing you need to start off. No excuses.

>>13680 You are absolutely right, in fact there are previous threads of study groups for SICP and lisp around here. If anyone wants to start a study group he should just do the same, create a thread with the material you're going to study and I guarantee some people will be interested

I am learning WebGL, if anyone is interested i can make a new thread?

>>13683 I would, all you need is to make sure is you don't start the thread with a one liner but with more information

>>13578 >>13680 >>13682 I'm wary of this idea. Threads are long-lived here and it would be a bad thing for many such threads to drown out other discussion, which I wouldn't allow to happen. My suggestion is that you try this idea in just this thread for now and see how it works with that. >>13683 You should instead have that discussion here or, better, start it in a thread that already exists for such a topic, adding to the discussion there.

>>13688 >Threads are long-lived here and it would be a bad thing for many such threads to drown out other discussion, which I wouldn't allow to happen. You already know you're a micromanaging cunt but I just want to remind you again.

>>13688 but there is no graphics programming thread in a catalog? most of vocal users are hardcore C low-level fanatics

>>13768 Complain in /q/ if you want to. >>13771 These are some related threads: >>5060 >>3026 >>8332 >>1403 >>3112 >>12077 If none of these quite fit, then by all means create a thread for discussing graphics programming.

I'd be interested in joining an IRC for this purpose.

Definitely on board for this. Pretty good knowledge of discrete math and would be willing to help, need to learn more about algorithms.

Anyone interested in working through www.buildyourownlisp.com with me? There was a group that did it about a year or so ago, but the thread is gone now and I can't find an archive. I think it would be neat to see what sorts of lisps everyone comes up with.

>>14199 Yeah this looks like fun

>>13533 don't we already do this on lainchan?

>>14199 The conclusion of the thread you refer to was that BYOL is a terrible Lisp book[0]. If another group implimentation is what you're after why don't we write a Forth? The parsing and evaluation semantics are dead simple and it's a cool language in of itself. [0] It's a nice demonstration of C programming though.

>>14199 I got to the part where we install mpc and build a language for doing basic maths. Where are you in the book? >>14243 What is lacking in BYOL as far as lisp is concerned? In what way is it a good C book? Do you think the book has any value besides C?

If I get some free time this summer (from the looks of it, it seems like I won't but a man can dream) I'd be down to a refresher of basic probability (maybe a book recommendation wouldn't hurt.) I'll be teaching some practical stuff in a basic OS course at my uni next (IPC, threading, schedulers, other stuff I can't remember right now) so maybe I can kind of tutor someone in here. That'll help me actually remember the stuff and practice how to explain the concepts (if the assigned prof is like kicking dead whales down the beach I'll be doing their job anyway.) At last, if someone is studying linear algebra (specifically linear regression) I could tag along.

linear regression and linear algebra aren't really the same thing. Linear regression is an equation solving mechanism irrc.

>>14247 Maybe create a thread where you teach the class

>>14254 Maybe I'm confused. I'm taking a linear algebra course and we're doing linear regression using least squares. >>14256 That would be a good idea but I don't want to kill a thread just to try to teach something that nobody cares about.

>>14254 >>14258 Linear regression is assuming that you have some equation that "fits" the data and you are going to be solving for the constants of the equation by minimizing some loss function that describes the best "fit." The reason that "fit" is in quotes is because the best "fit" is unique to each application. If you are writing map software, maybe the best "fit" is using the residual sums of manhattan distance instead of the residual sums of euclidean distance. An example of a function that you might want to fit is y = ax + b. The constants a and b are the unknowns in your linear system of equations. The points in your data set are the constraints. You typically have many more points than constants, causing your system of equations to be overdetermined. However, this system is likely inconsistent since there are no constants a and b that can create a line passing through all of the points. For this reason, you try to minimize your loss function, finding some constants a and b that best "fit" your data set where the best "fit" is completely dependent on your choice of loss function. It's not exactly an "equation solving mechanism." It's an approach to creating an estimator for your data set. It's what Excel does on your graphs when you ask for a line of best fit. It can be extended to nonlinear cases as long as the system of equations is expressed in parameters that are linear. An example of this is finding a circle that best "fits" a set of data points by solving for the constants a, b, and r in (x - a)^2 + (y - b)^2 = r^2. I think that it is more important to understand that this simple ordinary least squares isn't just an "equation solver." It's the start of understanding what an estimator is and how we might find one. From here, you can start to think more broadly about what makes a good estimator.

>>14261 Alright that was really clarifying. I guess the professor didn't bother explaining the terms or I'm just dense. Thank you lain! I'm lacking on the math side of things because my CS curriculum is like kicking dead whales down the beach (plenty of math without a coherent framework to put it all in context) so I tend to just learn how to do the exercises and call it a day. Hasn't failed me yet but it feels soykafty.

I'm interested as well

Do you remember all of those times that you thought that learning about Taylor Series was a waste of time? It's not. All that math you learned is incredibly useful. Take your time to get it right.

Hello, I just wanted to share https://desu-cartes.org/ We're a group originally started with similar ambitions to the ones expressed in the thread, though with mathematical logic, not programming. Over time it's evolved and now we're mostly math and cs people. We've also moved from structured study groups, which we learned empirically don't really work, to everyone self-studying at their own pace and discussing whatever they're interested in or working on currently. Some people study together, but that's generally spontaneously organized. For example, we have a couple people working through SICP currently, but there was no real overarching plan at the start. Currently we have ~15 active members, so it's pretty comfy. A significant portion of that are programmers. We have a private git server and stuff too. If you're interested, email the address on the website for an invite. Unfortunately we currently use slack and probably will for the foreseeable future, it'd be nice if we used something FLOSS but it just didn't work out. Also, please don't spam the thread here: I don't know if the mods remember, but this isn't the first time we've looked for members here (some of our current members are from lainchan), and last time a mod got mad because we kinda hijacked the thread by accident.

>>14984 >it just didn't work out. There are 15 of you. How tf does it 'not work out'?

>>14989 We have around 1.2 million messages in the archives we'd like to migrate (we're a couple years old at this point), and operating a web 2.0 (aka totally bloated and utter crap to manage) service with a decent level of reliability requires time and money for maintenance. We evaluated a bunch of options like mattermost, etc but they all came up wanting for some reason or another. I won't get into it here. At any rate, there's not much reason to reject slack on freedom grounds, there are libre clients like the slack-emacs client which I use. It's understandable if you're worried about privacy, but you can always provide a disposable email and use a vpn.

>>14989 >>14990 Also, I should point out that while we have around 15 active members, we have a bunch more that only show up occasionally, so migrating would be a total pain to manage. >>14199 >>14243 BYOL also teaches bad C practice (I was part of that thread). I would recommend https://viewsourcecode.org/snaptoken/kilo/ or s9fos instead if you're interested in C.

>>14991 >>14243 hi, i made very simple text editor in JavaFX and i have this idea of embedding simple Forth interpreter in it, - emacs style (i write something on one line, evaluate with shortcut, result shows up at bottom). I dont quite know how to start with writing Forth interpreter, any of you guys would give me hint or start this project with me?

>>14261 Linear regression can be found in higher dimensions withv linear algebra

>>14991 What's the "s9fos" you mentioned?

>>15088 haha whoops I meant s9fes https://www.t3x.org/s9fes/

>>15098 Nice, thanks!

>>14296 https://flakyj.blogspot.com/2018/05/intechopen-intech-sciyo.html

>>14296 btw, you use IntechOpen often ? tell us about their books

>>14199 >>14991 Whatever, I'd like to join your C group(s) and maybe others. >>13553 Where is the calendar?

>>14999 Any interpreter that uses a text-based language will need a symbol table or similar. Start with a FORTH dictionary module that text-symbol names can be put into. It doesn't have to use any specific data structure or collection of data structures for as long as it gets the job done. FORTH interpreters have traditionally been extremely, built from the lowest common denometer... P.S. Don't bother with global vs. local scope at first, if at all.

>>15816 If you're the one who sent the email, an invite was sent. https://www.lainchan.org/calendar.html (it's in the top toolbar)