Post: #301

(03-29-2017 07:22 AM)roid Wrote:  anyone here has remote programming jobs experience?

Yes.

Post: #302

So, I've fumbled my way through getting a remote server up with phoenix running and accessible.

Now I'm playing with a small app, starting with html, css, and javascript and after I get a basic homebrewed framework I'm going to build the backend to drive it.

Lots of moving bits and it's all very interesting.

Post: #303

My company recently transitioned to Rails for our back-end, so I've been learning Ruby in my spare time in order to help out (and to get another programming language notch) whenever I can since we only have one Rails developer on our team. Anyways, I can see why it's often recommended as a beginner's first programming language. I'm just awestruck by how beautiful and simple its syntax is.

Post: #304

@rafaeld

how much is ur hourly rate?

Post: #305

Got a basic backend working with a basic frontend. Now I have to spend some time determining the data I want to push around and what I want to do with it.

One step at a time.

Raw data in Elixir -> Phoenix -> Displaying basic graphics with a basic homebrewed javascript framework.

Lots of troubleshooting in how phoenix, elixir, and javascript all dance together. Phoenix wraps javascript files into modules and placed them by default at the bottom of the page. Turns out that makes separate template pages not run scripts properly because the overriding script wasn't loaded first.

Post: #306

(03-30-2017 03:26 AM)roid Wrote:  @rafaeld

how much is ur hourly rate?

I'm not working remotely any more. Last year I was doing full time remote Rails work for a company based in the US and charging a nominal $65 per hour, works out to a little over $10,000 per month.

I usually worked about four hours a day but I charged for eight.

Post: #307

(03-30-2017 03:26 AM)roid Wrote:  @rafaeld

how much is ur hourly rate?

I do as well. Don't charge by the hour, charge by the day (if you can get away with it.)

Don't call yourself a freelancer, that puts you in a lower market. You're a consultant.

This of course depends on just how green you are.

Also, go here:

Double Your Freelancing

That guy has a lot of good podcasts about remote freelancing.

Post: #308

Been making some great progress.

However, switching from javascript's more procedural, object-oriented style to elixir's functional style is a mindfuck.

Have to completely switch the mind around every time I jump between the back end and the front end. But, I'm building up a nice suite of tools. Going to be doing some interesting heavy lifting.

Post: #309

When it comes to web development I am mostly back-end focused. However I have been experimenting with VueJS for front-end the past few days and I must say I love it.

Post: #310

(04-04-2017 08:43 PM)FretDancer Wrote:  When it comes to web development I am mostly back-end focused. However I have been experimenting with VueJS for front-end the past few days and I must say I love it.

Huh. I see so many JS frameworks I'm totally lost. React, d3, etc. That looks interesting though.

Been coding plain JS with manipulating SVG and decorating with CSS. Have the backend handling physics and playing with static JS delivery or sockets.

Eventually I'll get enough of the core framework down to do more fancy things. I really like what this website has in their collection of designer blogs:
https://tympanus.net/codrops/collective/

Post: #311

I'm going in the exact opposite direction of all this JS heavy web stuff. I built a static site generator and my own simple markup language (self indulgent, but whatever, it took me an afternoon and I'm not a fan of markdown) a couple weeks ago. I'm trying to keep my sites as lightweight and fast as possible, and it literally took me less time to build a basic site generator than to get one of the many crappy site generators already available to produce what I wanted. The only JS I use is for form suggestions in certain web apps.

I fucking hate web work, gotta say. CSS is a complete and utter disaster. HTML is half arbitrary, hacked together shit best handled through a sane preprocessing layer. JS... the less said the better. Halfway kidding, JS is okay I guess, but I strongly prefer Python and I would never use JS for anything if it wasn't the default browser language. And while I know about the various ways I can "turn Python into JavaScript", the last thing I want is another abstraction layer between me and what my code is actually doing.

Even Python has been pissing me off lately. I'm getting close to just saying fuck it and writing everything I can in C. It would be painful for a couple weeks but probably not significantly slower in terms of development speed once I built up the basic codebase I need to replace the Python tools I already have, whether that meant finding actually good libraries or rolling my own.

----------------------------------------------

Anyway, this is pretty interesting:

https://hero.handmade.network/episodes

Basically, Casey Muratori has been programming a game live on stream for a couple years now. It's mostly C with some C++. There's like 500+ hours of video at this point but there's an amazing amount of knowledge packed in there if you want to dig into it at all. The youtube channel has topic-specific playlists:

https://www.youtube.com/user/handmadeher.../playlists

You don't need to follow the series from start to finish to get stuff out of the videos. I don't have anything like the kind of time required to go through the series from the start, but I've learned a ton about solving specific problems from that series. Plus it's just interesting to watch, because Casey isn't some dry boring lecturer. He's a funny guy and very good at explaining complicated programming concepts. The only "problem" is he's a lightning fast programmer so it's hard to follow his actual code changes sometimes.

The series is geared towards developing on Windows, which is nice for me, since most things like this are done on Linux and I can't stand Linux for anything but servers. Stuff required for Linux version of Handmade Hero:

https://davidgow.net/handmadepenguin/default.html

And OSX:

https://github.com/itfrombit/osx_handmade

Well worth a look, IMO. There's also a decent short (well, like 7 hours of video) "intro to C" series at the first link. That plus skimming K&R is enough to get you going with C, if you understand the basics of programming.

Post: #312

I got a very basic app going with an Elixir/Phoenix backend and unimproved front-end. Started coding up more of the back end in Phoenix when I realized the effort to get something interesting to sell to users was high with that procedure.

Started poking around. Taught myself basics of Rust, enough to be dangerous. Wrote a small simulation program this weekend while the girl was cooking up a storm and am working on extending it so that it's actually usable.

I really like what I've seen in Rust. The compiler vomiting all over the screen because I'm abusing safety or sense or something, instead of compiling and doing things that it should be doing is quite nice. The IDE plug ins are a lot faster than elixir and overall it's a very performance oriented language compared to Elixir, at least when we're speaking of single threads. I'm not sure at all about concurrency with Rust, the idea scares me. Phoenix/Elixir gets that all going without any thought or fuss.

I'm working on plugging that into the Phoenix app and then wrapping a much better UI around the whole thing.

It's getting closer.

All of the tech I'm using right now is pretty bleeding edge and likable to me. Elixir and Rust are good at what they do. Javascript, with all of its warts, is much better than Android hell. But I figure before I make anything sellable I am likely going to have to outsource some UI once I get the basic features down.

Post: #313

Interesting.

Rust doesn't have the completed numerical libraries that I need for my particular problem. And I don't want to tie up my server on the backend with a bunch of numerical things that could be solved by the user.

So I've been playing around with speed comparisons for my problem between Rust and Javascript.

If I use the same numerical algorithm for a small problem on Rust compared to JS, the problem solves 10 times faster with Rust. Unfortunately, that's the slow algorithm.

Nearly doubling the problem size results in Rust only being 2x faster than JS due to how bad the algorithm is.

Switching the problem to the fast numerical algorithm available in JS, not Rust, resulted in the JS solution beating the Rust solution, sometimes by a factor of 2!.

That's by far the most intense part of what I'm doing.

But, this is one of those few spaces where this is an important consideration. This is a relatively small problem and my app will handle much larger, and so I'm going to have to set it up in Rust and kick it off to the user for their CPU cycles.

But, now I've indentified almost all the tech I'll be using from the backend to the front end, including the main libraries.

Now it's a matter of refining it and getting that blasted UI working.

Post: #314

(04-25-2017 12:13 AM)philosophical_recovery Wrote:  Interesting.

Rust doesn't have the completed numerical libraries that I need for my particular problem. And I don't want to tie up my server on the backend with a bunch of numerical things that could be solved by the user.

So I've been playing around with speed comparisons for my problem between Rust and Javascript.

If I use the same numerical algorithm for a small problem on Rust compared to JS, the problem solves 10 times faster with Rust. Unfortunately, that's the slow algorithm.

Nearly doubling the problem size results in Rust only being 2x faster than JS due to how bad the algorithm is.

How does that work? It doesn't make much sense to me that JS is scaling substantially better than Rust with the same algorithm unless your Rust compiler is fucking the dog somehow.

Quote:Switching the problem to the fast numerical algorithm available in JS, not Rust, resulted in the JS solution beating the Rust solution, sometimes by a factor of 2!.

That's by far the most intense part of what I'm doing.

But, this is one of those few spaces where this is an important consideration. This is a relatively small problem and my app will handle much larger, and so I'm going to have to set it up in Rust and kick it off to the user for their CPU cycles.

Why not just implement the fast algorithm in Rust?

You've got me all curious now. What is this algorithm you can't get in Rust that crappy old JS has?

---------

Sounds like you're having fun anyway... I've been wading through shitty documentation off and on the last few days trying to figure out a foolproof deployment scheme for my web apps. It makes me want to punch the shit out of something because the asshats who write the docs always leave crucial information out and the error messages I get are worthless.

$ (commands I pasted directly out of the documentation quickstart)
Server --> FAIL!

Wow, thanks a lot, fucksticks. I was following the tutorial right out of the docs on a brand new Linux VPS, so how the fuck did it just fail and why?

Post: #315

(04-25-2017 02:13 AM)weambulance Wrote:  Why not just implement the fast algorithm in Rust?

You've got me all curious now. What is this algorithm you can't get in Rust that crappy old JS has?

I could, but I don't want to spend the weeks/months doing it. It's not a trivial problem.

Post: #316

Been working with python and writing an algorithmic trading program for certain financial instruments. Exceptionally interesting. I'd recommend anyone starting with a language should just start building something, and learning as they go. It ingrains it into your mind as you build and learn how to tackle the next problem.

Post: #317

Does anyone have a startup idea they're looking for a co-founder to help implement?

I used to have my own startup idea, and looked around for a co-founder for a long time, but ran into the problem described at Stop Looking For A Technical Co-founder. The people I approached with the idea fell into three groups: (1) those who said "I'm not really much of a programmer," (2) those who said, "I already have a full-time programming job that I don't feel like quitting," and (3) those who said, "Sure, I'll work on that with you" and then flaked out before writing a single line of code.

So, I started to code it myself but after several months (and a few redesigns from scratch), finally threw in the towel, because it's not a lot of fun to work on such a major project solo. I find that when I have people reviewing my code, I learn a lot more and become a better programmer. Also, it's just nice to have someone to bounce ideas off of.

Eventually I decided, "Why not just help someone else implement their idea?" I'm mostly just looking for an intellectually stimulating project to work on, with someone who would be involved in planning, code review, etc. Most of my experience is with PHP and SQL backends, by the way.

Post: #318

Dammit.

I've been digging around in the back end of this program for quite some time.

Then I ventured into this thread and realized that I'm frustrated over my progress for a program 1) in a completely new, bleeding edge language that I don't know that well 2) part time while I have another concentration intensive job in an unrelated area and 3) in an extremely technical, specialized area where there aren't many good resources other than the books I already have laying around 4) in two weeks!

Looking at it that way, it isn't bad at all. The book I ordered specifically on Rust won't even be published yet from the pre-order and I'll already have a working website that integrates the features by then, at this rate.

It's just been a lot of refactoring and redoing things as I realize there are much easier ways to handle things that I was doing the dumb way. It's going to pay off when I get in the backend guts again, but I really want to get out of the nitty gritty and onto the UI. I just can't have a good UI without a functioning back end. But, once that is rolling I can play with a lot of different UI options with drawing graphics and dealing with user events.

Still better than writing an Android app. Hah!

Post: #319

I learned programming through simple Game Dev. You can learn C# fairly quickly following youtube tutorials on making flappy bird clones and the like. Then you can expand.
Highly recommend this for anyone who likes or needs quick gratification to keep going without quitting.

Post: #320

Just got rid of a bunch of dumbness in my code, although it could always be better I think the main parts are good enough.

Wrapping an enum in a struct is very powerful.

Did all kinds of wrangling with types in order to pass around information. Now I have more universal functions that I can actually add to later without pulling my hair out. Code is 2/3 as long. Builds an example problem in a millisecond.

Sooo close to wrapping up this bad boy and working on hauling it into the front end.

Post: #321

Put Rust down for now after refactoring.

Got Rust as NIF working with the examples given by the Rustler page. The Readme in autogenerated files conflicted with what was on the website, so I had to debug the code because it didn't quite work out of the box like it should. That's what you get for running code that's under version 1.0...

I was able to do a sanity check by adding a println! statement in the rust source code, typed in iex -S mix, and it compiled all of my elixir and rust files and ran the way it was supposed to.

Now, I'm going to push everything in my working directory for my web app and start plugging in the Elixir wrappers. Likely modifying the Rust code. And building the web interface to display graphics based off of basic user input.

So, the user loads the html/js/css page and inputs data
-> Elixir (decides what to do with data and then fires it off to SQL database and Rust native code)
-> Rust (eats the input and spits out an answer)
-> Elixir (catches answer and delivers to user)
-> Javascript renders graphics based off what the backend Rust decided in combination with SQL data and management by the Phoenix Elixir framework.
???
Profit!

I've got the bare bones of a couple of the steps. The Elixir/SQL interface will need a schema definition, but I'm going to get that up after I build the UI and get some basic functionality down beyond what I had before, which was lacking in features because the backend wasn't developed.

Post: #322

(05-02-2017 12:19 PM)SnowHugger Wrote:  I learned programming through simple Game Dev. You can learn C# fairly quickly following youtube tutorials on making flappy bird clones and the like. Then you can expand.
Highly recommend this for anyone who likes or needs quick gratification to keep going without quitting.

Can you point me in the direction of the material/resources you used?

Post: #323

I have a programming project I've been thinking of since 2008, and working on desultorily ever since. It's a little like Infogalactic, but more ambitious.

I just have found it hard to stay motivated at times, because it's a solo project. I've done solo projects before, but it was usually stuff that I could finish and deploy in maybe a day or two. That makes it easy to stay interested, because the fun part of seeing how it works in practice is never too far away; and if the whole project turns out to be poorly-designed, it's no big deal, because I didn't sink too much time into it.

With this type of project, I have to organize the work into tasks, and sub-tasks, etc. The databases get complicated enough that I have to document them just to remember what I was doing and how it all worked.

It ends up feeling somewhat masturbatory, though, in that I check items off my task list and there's no one else around to care. I'm still not at minimum viable product, so all I've done is move a little bit closer to that point, while at the same time making the software a little more complicated so that it becomes all the harder to isolate and troubleshoot problems. Oh well; just gotta keep moving forward, I guess.

What always seems to stop my progress, however, is that I reach a point where I have to make a design decision, and I think, "If I had a co-founder, I would ask them, 'I'm leaning toward option A; what do you think?'" All the burden of the responsibility is on me, and if I decide later that I erred, then I have to think, "Gee, now I have to go back and do the work over again, the other way. I'm such a dumbass not to have anticipated this." I don't even have anyone to bitch to about it.

Last time, I ended up quitting because I reached a point where I said, "Well, I can either go with this harder option which is the more 'purist' option in that it adheres better to my basic principles of what I want to accomplish; or I can go with this easier option that gets me to MVP more quickly but with less functionality, and violates some of my principles." In the end, I let the perfect be the enemy of the good by setting the whole project aside rather than make an unpalatable decision, but not finishing the project at all turned out to be the greater violation of principles.

Now that my accounting career is in the toilet for the foreseeable future, and my wife has left, I guess I have nothing better to do than to resume this massive programming project. When I left off, I also was developing a testing framework, but probably not in the most optimal way. Hardly any of my programming is all that optimal; another downside of not having a co-founder is that there's no one to say, "Your code sucks; here's how you need to do it differently."

I backed up my last few days of work on copy.com (which has since shut down), and then my hard drive crashed. I got lazy about uploading my work to GitHub because I figured, "I probably won't pick up this project again anyway, and besides, what are the odds of my hard drive crashing?" Guess I won't be making that mistake again.

Post: #324

In less than 10 days, part time after my main job while enjoying the weather and doing things otherwise I:

*Made a basic Elixir web app (Phoenix)...actually more than one because I had to redo some tutorials and play around a bunch to figure out some things
*Figured out a javascript API for a relatively complicated outside service enough to get basic functionality
*Fumbled through security/SSL/https issues which I am completely unfamiliar with
*Figured out the https issues with the above javascript app when it defaults to doing everything over plain http
*Troubleshot said flying ball of spaghetti
*Spun up my server and got my phone to connect and access it

And my profession is not programming.

Post: #325

Anyone have experience coding a mobile android app and releasing it on the Google Play Store?

I want to create a simple app that works with Samsung's edge technology (S6/S7 edge & S8s). The issue with beginning development is that I have to use the Samsung SDK and I think this isolates me to the Samsung Galaxy store and disqualifies me from releasing on the Google Play store.

Not sure if there is a way around this...