Games and Guis with Python (Kivy) Am I wasting my time?

[Wictorian]

I kinda like incremental games and as anyone who likes this genre I too would like to develop my version of these games. As I had some experience with python, upon some research I though Kivy had what it takes but halfway into learning Kivy I feel like it is not competent at all aand I lost my trust in Python guis and games. I almost lost all my motivation however then I thought no matter how incompetent they are guis might be useful for some pyhton apps. 

 

Now the question is, am I wasting my time?

 

[riklaunim]

There is limited demand in desktop apps in Python. Most jobs or projects are either web or some service/API. For actual games there are some bindings in Python to game engines/libraries (RenPy, PyGame...) but they are niche, majority goes into Unity and Unreal Engine via C#. Note that making a game takes way more than just coding

[Wictorian]

3 hours ago, riklaunim said:

There is limited demand in desktop apps in Python. Most jobs or projects are either web or some service/API. For actual games there are some bindings in Python to game engines/libraries (RenPy, PyGame...) but they are niche, majority goes into Unity and Unreal Engine via C#. Note that making a game takes way more than just coding

yeah Im not that proffessional

[cmndr]

While there's value in getting better at Python, you'll probably be better off... doing what everyone else in the genre is doing. Making RPGs? Latest version of RPG maker. Making something else... Unity or similar?

Game dev isn't my forte but it's DEFINITELY a field where planning (scope definition, writing, art direction, music, tool selection, etc) up front can cut total project time by 2-3x.

[Wictorian]

7 hours ago, cmndr said:

While there's value in getting better at Python, you'll probably be better off... doing what everyone else in the genre is doing. Making RPGs? Latest version of RPG maker. Making something else... Unity or similar?

Game dev isn't my forte but it's DEFINITELY a field where planning (scope definition, writing, art direction, music, tool selection, etc) up front can cut total project time by 2-3x.

The thing is, I don't wanna focus on game development. I just wanna make a simple game without music and stuff. Besides making guis can be helpful. So in the end I justified this. 

[cmndr]

If you're trying to do a fun project and maybe pick up some skills along the way, just do it.
See if you can find decent guides or resources first (which it seems you're partially doing with this thread).

[Wictorian]

1 hour ago, cmndr said:

If you're trying to do a fun project and maybe pick up some skills along the way, just do it.
See if you can find decent guides or resources first (which it seems you're partially doing with this thread).

yeah I have been watching the tutorial by codemy on youtube. I have watched 58 videos  of 67.

[cmndr]

videos only go so far. I'd actually start with interactive tutorials for the absolute basics.

After that... projects, such as what you're proposing.

 

There's also arguably SOME benefit in doing drills. Think easy level questions on leetcode or hackerrank. (also before doing TOO much look up a style guide and learn it - pep8 works, so does google's guide - https://google.github.io/styleguide/pyguide.html)

 

Also pick up some software engineering best practices (where possible, break code down into relatively small functions. For each function list out your expected inputs, expected outputs and the general process by which the function is to work as comments) such as doing unit tests and integration tests. It's often worthwhile to write unit tests BEFORE actually doing the actual coding. Yes this takes longer, it also kills bugs early (and debugging often takes EVEN longer).

[Wictorian]

6 hours ago, cmndr said:

videos only go so far. I'd actually start with interactive tutorials for the absolute basics.

After that... projects, such as what you're proposing.

 

There's also arguably SOME benefit in doing drills. Think easy level questions on leetcode or hackerrank. (also before doing TOO much look up a style guide and learn it - pep8 works, so does google's guide - https://google.github.io/styleguide/pyguide.html)

 

Also pick up some software engineering best practices (where possible, break code down into relatively small functions. For each function list out your expected inputs, expected outputs and the general process by which the function is to work as comments) such as doing unit tests and integration tests. It's often worthwhile to write unit tests BEFORE actually doing the actual coding. Yes this takes longer, it also kills bugs early (and debugging often takes EVEN longer).

Yeah I think I have followed the tutorial pretty interactively. 

 

I have done some hackerrank but I really dont like it. I will move on to leetcode after I'm done with the game possibly using another language.

[cmndr]

only really do HR or LC if you're having fun or you have an interview coming up.

For what it's worth HR+LC style questions get asked pretty often. I have actually found that it's made me a better data scientist since instead of "hacking something together" I actually think (a little) about how it'll work overall AND I have a better intution for how it'll all work AND I'm able to explain it almost effortlessly.

[Wictorian]

8 hours ago, cmndr said:

only really do HR or LC if you're having fun or you have an interview coming up.

For what it's worth HR+LC style questions get asked pretty often. I have actually found that it's made me a better data scientist since instead of "hacking something together" I actually think (a little) about how it'll work overall AND I have a better intution for how it'll all work AND I'm able to explain it almost effortlessly.

tbh I dont get how HR works, would appreciate it if you can explain. 

 

I think these are helpful and fun, I dont really see a problem in spending time on them.

[cmndr]

HR, in this case I'm using as an abbreviation for HackerRank (as opposed to Human Resources)

 

The general idea for HR and LC is that you get a sample problem. You then work through it.

I usually add in comments

 

# Goal: do ____
# Inputs: A(INT), B(string), C (BOOL)
# Output: X(data frame)

function() {
  sdfasfsaf
}

 

You walk through each bit of what you're doing (verbally) while doing it. Also pretend there's a person and try to ask clarifying questions up front.

The goal during an interview is to show you have good communication skills, good code hygiene (i.e. it's easy to maintain by someone else LATER) AND that you have good logic (no/few unnecessary passes through the data, good asymptotic complexity, etc.)

 

[wasab]

11 hours ago, Wictorian said:

tbh I dont get how HR works, would appreciate it if you can explain. 

 

I think these are helpful and fun, I dont really see a problem in spending time on them.

If you want to actually practice doing these for technical interview, I suggest doing them with real person. Try pramp.com

 

there are good quality technical interview questions for software engineers on there and with actual people, most of whom are experienced software engineers themselves practicing to job hop. It is free. I used it back when I was job searching. The people there are great, I’ve met some ex faang engineers  who are very good at walking you through certain questions as well as giving you pointers.

[cmndr]

 

On 7/22/2022 at 7:51 PM, wasab said:

If you want to actually practice doing these for technical interview, I suggest doing them with real person. Try pramp.com

 

there are good quality technical interview questions for software engineers on there and with actual people, most of whom are experienced software engineers themselves practicing to job hop. It is free. I used it back when I was job searching. The people there are great, I’ve met some ex faang engineers  who are very good at walking you through certain questions as well as giving you pointers.

So this IS probably the best approach IF the other person is of high caliber (think passes most FAANG or similar interviews).

HR and LC are perfectly fine for doing 1-2 questions a day, critically assessing them and then learning. It's about doing fewer questions, more thoroughly and looking for ways to improve (check others' answers once you're done, see if you can improve algorithmic complexity or modularity or...)

[wasab]

1 hour ago, cmndr said:

 

So this IS probably the best approach IF the other person is of high caliber (think passes most FAANG or similar interviews).

HR and LC are perfectly fine for doing 1-2 questions a day, critically assessing them and then learning. It's about doing fewer questions, more thoroughly and looking for ways to improve (check others' answers once you're done, see if you can improve algorithmic complexity or modularity or...)

it isnt just about alogrithm though. People also generally like others who are easy to work with. A reference on linkedin that says an xyz person so and so is a pleasure to work with can go a long way.

 

Also, alogrithms and whiteboard questions are mostly used by the FANNGs who can afford to be picky and still attract top talents. Many smaller small-med size bussinesses or startups generally look at your projects and past work more than they care about how well you can solve alogrithms, which most engineers rarely used in their actual day to day jobs anyways. There are senior engineers who find leetcode style technical interviews downright insulting and simply refused to participate. I know a couple who would tell interviewers to stop wasting their time and end the interview if the interviewer ever tried leetcoding them. None have an issue finding job. 

 

As for op:

If FANNG is your goal, Amazon is by far the easiest to get into. They have been mass spam recruiting on linkedin recently for some reason, I have dozens of their recruiters email within the span of two months. Pretty sure their interviewing process is rigorius but it shouldnt be difficult to get an in person interview as long as you can pass their initial hacker rank screen questions. 

[Wictorian]

1 minute ago, wasab said:

As for op:

If FANNG is your goal, Amazon is by far the easiest to get into. They have been mass spam recruiting on linkedin recently for some reason, I have dozens of their recruiters email within the span of two months. Pretty sure their interviewing process is rigorius but it shouldnt be difficult to get an in person interview as long as you can pass their initial hacker rank screen questions. 

Yeah. Well not necessarily Faang, for instance IBM would be okay too. I also have 2 years to prepare and I want to work in Switzerland. When I did some research I found out that the best options in Switzerland were Google, Microsoft and IBM. 

[Guest]

I recently went through a training program and found the 2-3 python developers were significantly worse off than those using languages like Java and other "C" languages. 

 

If your goal is to have fun making games you enjoy, use python. Another fun tool is rpg maker. Basically makes Gameboy Pokemon style games. SR dude on YouTube will teach you all you want to know about the program.

 

If your goal is to practice for a job, do Java. 

[Guest]

Just now, Wictorian said:

Yeah. Well not necessarily Faang, for instance IBM would be okay too. I also have 2 years to prepare and I want to work in Switzerland. When I did some research I found out that the best options in Switzerland were Google, Microsoft and IBM. 

Learn and practice java. C# is good if you use nuget. Unity teaches a lot of bad practices. 

[Wictorian]

4 minutes ago, fpo said:

Learn and practice java. C# is good if you use nuget. Unity teaches a lot of bad practices. 

I really hate Java. Not more than not working for Faang but I would only do it if it is the last resort.

[Wictorian]

6 minutes ago, fpo said:

I recently went through a training program and found the 2-3 python developers were significantly worse off than those using languages like Java and other "C" languages. 

 

If your goal is to have fun making games you enjoy, use python. Another fun tool is rpg maker. Basically makes Gameboy Pokemon style games. SR dude on YouTube will teach you all you want to know about the program.

 

If your goal is to practice for a job, do Java. 

I don't want to make games. I just want to make 1 game. Possibly a few more in the far future if I have the inspiration.

 

You have a point but as I said I hate Java. Are there no loopholes?

[cmndr]

8 minutes ago, wasab said:

it isnt just about alogrithm though. People also generally like others who are easy to work with. A reference on linkedin that says an xyz person so and so is a pleasure to work with can go a long way.

 

Also, alogrithms and whiteboard questions are mostly used by the FANNGs who can afford to be picky and still attract top talents.

 

It's a mix of things... One thing to note is that I emphasized earlier in this thread DESCRIBING your thought process.

 

Let's assume you're at the point where you can work through most problems like they're nothing (and you NEED to be at this level for easy and medium level questions). At that point most of the grading in an interview shifts over to how well you describe your thought process, your logic, your code structure etc.

Even if you don't have the most OPTIMAL solution, having a well structured bit of code that hints "this guy writes stuff that's easy to follow, maintain and troubleshoot" goes a long way. Little things like practicing variable naming, spacing, etc. all matter.

For what it's worth I've interviewed (final rounds) at all of the FAANGs except Netflix (never applied) and worked at Google for a few years.

 

It's also non-trivial to get into Amazon even with previous FAANG experience. Easiest doesn't mean easy.

 

8 minutes ago, wasab said:

Many smaller small-med size bussinesses or startups generally look at your projects and past work more than they care about how well you can solve alogrithms, which most engineers rarely used in their actual day to day jobs anyways. There are senior engineers who find leetcode style technical interviews downright insulting and simply refused to participate. I know a couple who would told interviewers to stop wasting their time and end the interview if the interviewer ever tried leetcoding them. None have an issue finding job.

DSA questions aren't perfect and there's definitely better things to check in my opinion but there IS very real value to having a FAANG on the resume and making FAANG income for a few years.

There ARE some real advantages to these questions: 1. They're relatively hard to fake - the smarter and/or more dedicated you are the better you do (test of will power/persistence) 2. Doing well in them means you know standard libraries at least somewhat well and can structure code well 3. they test communication and question asking skills


It's like getting an Ivy league university on the resume. Recruiters are more willing to take a chance on you. It get easier to get interviews and your experience isn't questioned as strongly. You still need to perform at an interview but...

And yeah there's some people who are insulted by it. They can have very good careers and often better work life balance. There is usually a salary cap though. You shouldn't expect to make 300,000 - $500,000 a year if you can't do leetcode style questions. In that instance, you end up settling for $100,000 - 200,000.

 

And yeah there are some OTHER things I wouldn't mind seeing more of. "here's some code, walk me through how you'd do a peer review of it" <- this is more structured, this also shows a bit more of how well people would collaborate and communicate.

[cmndr]

19 minutes ago, Wictorian said:

Yeah. Well not necessarily Faang, for instance IBM would be okay too. I also have 2 years to prepare and I want to work in Switzerland. When I did some research I found out that the best options in Switzerland were Google, Microsoft and IBM. 

Switzerland is a good goal.

At least at Google if you're a mid career engineer (think $350,000 USD at around age 25-30) you make the same amount in Zurich as you do in a place like SF or Los Angeles. If you go to Dublin or London you're going down to almost half the pay.

I WOULD strongly suggest starting leetcode and HR style questions sooner rather than later though. 1-2 questions a week for 2 years is fine. Also be aware that getting into a top place usually means having a good internship or two so... you get tested on this stuff 1 year before graduation.

Also while IBM is a good name on the resume, the $$$ aren't the same (not that that's everything). With that said, careers are marathons and 1-2 years at IBM would set you up for interviews at other top places later on too.

[wasab]

31 minutes ago, cmndr said:

 

It's a mix of things... One thing to note is that I emphasized earlier in this thread DESCRIBING your thought process.

 

Let's assume you're at the point where you can work through most problems like they're nothing (and you NEED to be at this level for easy and medium level questions). At that point most of the grading in an interview shifts over to how well you describe your thought process, your logic, your code structure etc.

Even if you don't have the most OPTIMAL solution, having a well structured bit of code that hints "this guy writes stuff that's easy to follow, maintain and troubleshoot" goes a long way. Little things like practicing variable naming, spacing, etc. all matter.

well, that is true. These are quite valuble skills and top companies are more than willing to snatch up candidates that do well in them but these talents are rare and like i said, only the FAANGs and other fortune 500s can afford them. Smaller companies need to lower their bars, because there is abolutely no way they can compete with the faangs in the job market for these talents. This means many generally focus on industry experiences or onboard risky inexepreinced new grads through an internship programs and probation period in which they can properly train and minimize the impact of a bad hire. 

 

31 minutes ago, cmndr said:

It's also non-trivial to get into Amazon even with previous FAANG experience. Easiest doesn't mean easy.

 

DSA questions aren't perfect and there's definitely better things to check in my opinion but there IS very real value to having a FAANG on the resume and making FAANG income for a few years.

There ARE some real advantages to these questions: 1. They're relatively hard to fake - the smarter and/or more dedicated you are the better you do (test of will power/persistence) 2. Doing well in them means you know standard libraries at least somewhat well and can structure code well 3. they test communication and question asking skills


It's like getting an Ivy league university on the resume. Recruiters are more willing to take a chance on you. It get easier to get interviews and your experience isn't questioned as strongly. You still need to perform at an interview but...

And yeah there's some people who are insulted by it. They can have very good careers and often better work life balance. There is usually a salary cap though. You shouldn't expect to make 300,000 - $500,000 a year if you can't do leetcode style questions. In that instance, you end up settling for $100,000 - 200,000.

 

And yeah there are some OTHER things I wouldn't mind seeing more of. "here's some code, walk me through how you'd do a peer review of it" <- this is more structured, this also shows a bit more of how well people would collaborate and communicate.

Right, nothing wrong with trying to apply to faangs but what i have an issue is how many romanticized it like on so many youtube influencers day in the life of a software engineer videos in which all they do is petting cats and dogs and then drink a few cups of coffee as if it is a typcal work day. This is not very realistic. 

 

Ivy Leagues have prestige and can command high premiums yes but expecting everyone to graduate with these degrees is as unrealistic as companies expecting they can be picky and simply just fill up every positions with only top tier IVY league graduates. Supply is not there. Job market is competive and software engineers are already in low supply as it is. If they are gonna wait for the these schools to crank out enough graduates and for these graduates to sharpen up their leetcode/hacker rank skills high enough before hiring them, they would long be left in the dust by competitors. 

 

I dont know what the issue is with amazon but reading through the CSCareer questions subreddits, many others reported they got recuiter spam from amazon as well and from some insiders, i heard they have already exhausted the supply of candidates who can pass their initial bar and thus are now lowering it just to fill up the quotas. They still leetcode yes and still difficult yes but a lot easier than like said google who are more likely to throw your resume into the bin without ever contacting you. 

[cmndr]

17 minutes ago, wasab said:

well, that is true. These are quite valuble skills and top companies are more than willing to snatch up candidates that do well in them but these talents are rare and like i said, only the FAANGs and other fortune 500s can afford them. Smaller companies need to lower their bars, because there is abolutely no way they can compete with the faangs in the job market for these talents. This means many generally focus on industry experiences or onboard risky inexepreinced new grads through an internship programs and probation period in which they can properly train and minimize the impact of a bad hire.

Most fortune 500s at large don't pay FAANG salaries. Rule of thumb FAANG = 1.5-2.5x the pay of F500
You also have OTHER places that pay pretty well. AirBnB, DoorDash, etc.

Also, depending on the place smaller companies have HIGHER bars than a place like Google. Want to work for Quora? Higher bar than Google. Want to work for a 200 person startup with VC $$$$ also a higher bar. They can't afford to make a single hiring mistake. Some of these places pay about the same as a FAANG as well. These places almost all ask leetcode style questions and A LOT of their employees come from FAANG and Unicorn companies.

 

Minimizing the impact of a bad hire is usaully done through prevention.

 

 

17 minutes ago, wasab said:

Right, nothing wrong with trying to apply to faangs but what i have an issue is how many romanticized it like on so many youtube influencers day in the life of a software engineer videos in which all they do is petting cats and dogs and then drink a few cups of coffee as if it is a typcal work day. This is not very realistic.

They're not the be all end all and they're certainly not SUPER common. There are some very compelling arguments to making 2x the pay and having a gold plated resume.
Your experience at any company will depend on the team you're on.

 

 

17 minutes ago, wasab said:

Ivy Leagues have prestige and can command high premiums yes but expecting everyone to gradute with these degrees is as unrealistic as companies expecting they can be picky and simply just fill up every positions with only top tier IVY league graduates. Supply is not there. Job market is competive and software engineers are already in low supply as it is. If they are gonna wait for the these schools to crank out enough gradutes and for these gradutes to sharpen up their leetcode/hacker rank skills high enough before hiring them, they would long be left in the dust by competitors.

Companies don't hire almost exclusively from Ivy+ schools (unless they're McKinsey).
But there IS some value in having 2-4 BIG gold stars on your resume. Imagine you're a recruiter with 500 resumes to skim through in 8 hours... you take a big red marker and circle anything that looks good. You're circling top brands, high GPAs, strong test scores, etc.
Software engineers have it easier than people in many other industries but the same concepts still apply.

Second tier companies often take what they can get. Yeah you can totally get into Kaiser or Chlorox (have not confirmed) with 0 leet code. You can still grow a lot at a second tier company though (you're the most important person in your career) and there's some benefits to taking a higher profile role at a 2nd tier company (and then jumping back to a FAANG or similar at that level).

 

I'd generally recommend doing a bit of both. Maybe the first few years at an "elite" place just to get used to what "quality" looks like and then you can kind of coast somewhere else as a Principal Engineer. Pay cut but you get work life balance and aren't trying that hard since you're generally sharper and more polished than your peers.
 

17 minutes ago, wasab said:

I dont know what the issue is with amazon but reading through the CSCareer questions subreddits, many others reported they got recuiter spam from amazon and from some insiders, i heard they have already exhausted the supply of candidates who can pass their initial bar and thus are now lowering it just to fill up the quotas. They still leetcode yes and still difficult yes but a lot easier than like said google who are more likely to throw your resume into the bin without ever contacting you. 

CSCareers, TeamBlind, etc. are half-realistic and half memes.

 

 

Amazon hasn't really exhausted their supply of coders. They do have select markets where they're struggling to fill warehouses with people though. Those are meat grinders.
 

 

 

[wasab]

37 minutes ago, cmndr said:

snip

alright, i havent interview at FANNG yet so i guess i cant say much. When i was fresh out of school, i interviewed at quite a few startups and mid size companies, all entry level/junior level postitions. As I remember, 4 leetcoded me and two did not leetcode at all and would ask about school course works and projects. Two others had techical interview rounds but instead of writting alogrithms, they gave me short take home coding challenges and ask me to implement some certain features(it was some react projects, roughly one hour to complete) before handing them the next day. 

 

From what i can tell, leetcode is common but not necceasily a neccesity if you want to just get your foot into the door. For faangs, i am pretty sure all leetcode. for good or worse? i dont know. 

 

For amazon, i have never in my life see so many spams. I have been appoarch by technical recuiters((all for SDE I) from AWS, alexa, adverstisement, and some other xyz positions whose descriptions are so full of jargons i dont even know what it is. I was 1 year into my career at that point. I dont see this from any other fanngs, heck, they all ghosted back when i applied for their internships.