Maximum PC

Im having difficulty on trying to figure what is next to learn in terms of programing. I mastered the basics and GUI programing with WinForms and WPF. I look around and I see a ton of homebrew apps out there on the internet. Examples would be, handbrake, paint.net, autopatcher and several others. Ive ask around how to make these kinds of applications. The only answer I get is ‘you gotta do research’. That’s very vague, and research what exactly. That answer dosent help me very much. I don’t believe people have written the above apps I mentioned or any free app you find off the internet from complete scratch. Theres gotta be some library they’re using to make these applications.

Im not trying make huge applications like Photoshop are something like that. I just want make apps than one person can handle. I need a better answer than just research. That’s like looking for something that you have no idea what to call, or what it looks to even begin researching it. Languages Im familiar with are C++ and C#. Im also a student at Westwood College studying game programing with direct X.

That's because research is how you learn what to call. One of the things in my 'Things to do' list is to modify handbrake's video encoding to support CUDA. Parts of handbrake is written in .NET, so I spent a few days looking for the encoding components (I had to stop because I forgot how much I enjoyed playing Bad Company 2...and I had other things I needed to do). Part of mastering programming (or at least becoming very good at it) is to go and think the problem through and figure out what you need, then researching it.

Oh, and a lot of these things were built from scratch. Ruby on Rails was written from scratch by one guy (David Heinemeier Hannson). Ruby was written by one guy (Yukihiro Matzumoto). Linux was written by one guy (Linus Trovalds). The reason why they can write these things by themselves - from scratch - is because either 1. there were no existing libraries or 2. the ones that existed, sucked. It's not vague for someone to tell you to research it, what it means is that you need to learn how to think through your problem domain better.



Why don't you focus on working on algorithmic challenges through TopCoder? Building an application is one aspect of programming, the answer is solving puzzle questions. It's a great mental exercise, plus it'll harken back the things you learned in school. Learning to use a library like WPF or WinForms is cool and all, but if you can write a string munger then it's like knowing how to build a house but you don't know how to measure and cut.

Im having difficulty on trying to figure what is next to learn in terms of programing. I mastered the basics and GUI programing with WinForms and WPF. I look around and I see a ton of homebrew apps out there on the internet. Examples would be, handbrake, paint.net, autopatcher and several others. Ive ask around how to make these kinds of applications. The only answer I get is ‘you gotta do research’. That’s very vague, and research what exactly. That answer dosent help me very much. I don’t believe people have written the above apps I mentioned or any free app you find off the internet from complete scratch. Theres gotta be some library they’re using to make these applications.
Im not trying make huge applications like Photoshop are something like that. I just want make apps than one person can handle. I need a better answer than just research. That’s like looking for something that you have no idea what to call, or what it looks to even begin researching it. Languages Im familiar with are C++ and C#. Im also a student at Westwood College studying game programing with direct X.

I look around and I see a ton of homebrew apps out there on the internet. Examples would be, handbrake, paint.net, autopatcher and several others. Ive ask around how to make these kinds of applications. The only answer I get is ‘you gotta do research’. That’s very vague, and research what exactly.

Linux was written by one guy (Linus Trovalds).

Stallman actually deserves quite a bit of credit for starting, not one, but two significant OSS projects: emacs and gcc. While Emacs had the support of several other 'hackers' in the MIT AI Lab, I believe he worked on GCC on his own for a number of years.

Speaking of Linux, Git was written by Linus. He wrote the protocol in two weeks. Amazing.

Yes. He does deserve all of the credit for Git. Have you used Git? I haven't had a chance to try it out, but I did watch a Google Tech Talk video on it a couple of years back.

That's because research is how you learn what to call. One of the things in my 'Things to do' list is to modify handbrake's video encoding to support CUDA. Parts of handbrake is written in .NET, so I spent a few days looking for the encoding components (I had to stop because I forgot how much I enjoyed playing Bad Company 2...and I had other things I needed to do). Part of mastering programming (or at least becoming very good at it) is to go and think the problem through and figure out what you need, then researching it.

Oh, and a lot of these things were built from scratch. Ruby on Rails was written from scratch by one guy (David Heinemeier Hannson). Ruby was written by one guy (Yukihiro Matzumoto). Linux was written by one guy (Linus Trovalds). The reason why they can write these things by themselves - from scratch - is because either 1. there were no existing libraries or 2. the ones that existed, sucked. It's not vague for someone to tell you to research it, what it means is that you need to learn how to think through your problem domain better.



Why don't you focus on working on algorithmic challenges through TopCoder? Building an application is one aspect of programming, the answer is solving puzzle questions. It's a great mental exercise, plus it'll harken back the things you learned in school. Learning to use a library like WPF or WinForms is cool and all, but if you can write a string munger then it's like knowing how to build a house but you don't know how to measure and cut.

In my point of view,the reason why the author did not focus on working on algorithmic is that we can get some more milestone or achievement easily on programming app, don't you think so?

It depends. In the case of new to moderately experienced programmers, I believe the desire to write an application is somewhat addicting, but ultimately futile because their knowledge of the various aspects of CS/SE are so immature that the task is simply too overwhelmingly. A lot of time is wasted trying to learn CS writing (and rewriting and rewriting... 10x) ONE application which could have been used more productively learning algorithms, data structures, a programming language or new language features which can be applied to EVERY application.

As you said, the EVERY application is not realistic project, it is useless for getting progress. Did I got the point ?

The inevitable outcome of this process is garbage; They simply don't have the requisite knowledge to create an interesting or complex application. They should spend their time developing a solid foundation.

The fundamental is more important than the EVERY application ?

Personally, I think that algorithm competitions are a great teaching tool: They challenge you with problems that are unexpected broadening your knowledge base; You have to learn algorithms and data structures (and to a certain degree computational complexity);

Yep, actually I study mechanic as major, not CS. So that I didn't learn too much about algorithms, data structures or compile theory.

You learn to debug and find bugs in software -- The TopCoder competitions, for example, include a challenge period where you can earn additional points by finding bugs in other competitors software; They promote problem solving -- exercising your mind; They also promote unit testing.

Does software mean that the code of open source software ? I have never doubt the source from internet, it is my weak point - always believe in others
It may be good for me to doubt some code and try to find bugs from it and solve it.

Why does this happen? Unfortunately, both software and software development tools promote this time wasting behavior because so much complexity is hidden from the user. Ask yourself how many freshman physicists take one or two physics courses then decide they're going to create a particle accelerator or a nuclear reactor? None. Why? Even if they could get their hands on some nuclear material, it is justly plainly obvious they can't successfully complete a task of this complexity.

Do you mean the realistic requirement is more complex than the project we learn from book ?

The complexity isn't hidden behind a little icon and UI created with modern development tools. Now granted, none of the newbies here are attempting to write a compiler, driver, theorem prover or some other relatively complex undertaking; Most of them really just want to create something that "looks cool".

I agree with you in this point. I supposed this situation is just happened in China before. It looks like almost every newbie is facing this problem. They always wanna "loos cool" application and blow it up.

However, even with these underwhelming ambitions (ie I'm going to use the GUI development tool that someone else created to create a UI that is almost as cool as something that someone else created ten years ago), they still fail because they don't take into account the 80% to 90% of development that lies below the surface of the UI. Here is the typical dialog:

(Inexperienced developer) Hey, can you help me with my [chat; talk; message; text; picture] app?
(Experienced developer) Why do you want to write that? Aren't there already 10,000x apps that do that?
Yeah, but we have an idea for one that is way better. It's a picture/chat application! The UI is going to be killer!!!
Alright, what are you having trouble with?
Well, for starters, what programming language should I use?
Ugh... huh... what programming languages do you know?
Well, I took [Java, C++,C#,Obj-C] for about two weeks, but my school dropped me from the course because I hadn't taken a prereq math class. That stupid math class... what the hell do computers have to do with math. The school is so dumb.
Maybe you should take a few more classes then write the app? I don't think you're going to be able to do it now.
Why not? I was going to use the XYZ gui tool to build the thing. I'm really good at this stuff -- I can do it if you help me some.
Alright, are you planning on using a client-server or peer to peer network topology?
Topology... that's one of the damn math classes that I have to take. I had no idea topology was about networking.
{oh dear}
Well, IRC does both, right? So it'll have to do both. Plus pictures, video and internet radio.
That sounds like a lot of work. Have you ever worked with image or video data?
Sure, I rip movies from dvd all the time. I've been using multimedia my whole life. I uploaded a video to YouTube over four years ago way before anyone else could.
Never mind the media stuff. So you're not planning on using IRC or some existing chat protocol?
Uh... I didn't think about that... Maybe I should find a programming language already have the protocol. Hmm... I don't know. I've spent all my time designing the UI in Flash.
Also, are you going to store the user data? Were you planning on using a database or using your own data structures to store the data?
The only data structure that I know about is an array. Can't I just put everything in arrays?
Well... yes, but that would be a bit unusual. Plus, it would make writing the application really difficult.
Yeah, maybe I should use objects instead. Well, I'll worry about the data later. I need to work more on UI. The data isn't as important.
{Frustrated} Yeah, that sounds like a good idea. Good luck.