Watching People Code Is Becoming an (Even Bigger) Thing
itwbennett writes: Faithful Slashdot readers may recall the story of Adam Wulf, who spent two weeks live-streaming himself writing a mobile app. The phenomenon has quickly become thing, by which we mean a business. Twitch.TV, Watch People Code (which is an offshoot of the subreddit by the same name), Ludum Dare, and, of course, YouTube, are bursting with live or archived streams of lots of people writing lots of code for lots of different things. And just this week, Y Combinator-backed startup Livecoding.TV launched. The site has signed up 40,000 users since its beta went live in February, but unlike the other sites in this space what it doesn't have (and doesn't have plans for) is advertising. As co-founder Jamie Green told ITworld: 'We have some different ideas around monetisation in the pipeline, but for now we are just focussed on building a community around live education.'
This is even worse than a stream watching someone play a game. Who wastes their time with these things?
If you want to improve your coding skill you're better off practicing and reviewing code written by those more experienced than you, not watching someone "in the act" of doing it.
>> We have some different ideas around monetisation in the pipeline, but for now we are just focussed on building a community around live education
Translation: we are going to be ad-free to grab as many users as possible until we finalize the sale of the company to an appropriate advertiser. (That's pretty much how these start-ups work.)
As the OP said - if you want to learn to code then review and practice. Watching is pointless.
As a viewer, it's about learning technique and thought processes. Identifying issues, attempting a particular thought process, only those that provide a strong narrative to the work they are doing will be likely "stars". Watching how good programmers (assumption) deal with their environment and the typical problems they face. Seeing how people top down or bottom up write code is very interesting (within limits).
As a broadcasting coder, it takes a fair amount of personal confidence to do it, particular in this field. Having to verbalize what you are thinking and how you are considering the problems in front of you is actually quite challenging. Those that do well in the broadcasting scene will most likely be strong professionally as well.
That said, I personally don't understand the fandom about broadcast games to the level that it has taken. I get the benefits, but I don't get the market.
Arrive at office. Read email. Get coffee. Figure out what I need to code today. Start a for loop. Change CDs. More coffee. Flesh out for loop. Look up String API, find a method better than what I was after. Scrap everything. Lunchtime!
/.. Go home.
Collaborate with a colleague. Get a Mountain Dew. Change CDs. Write glue code to make shiney new String API do what's required. Waste an hour explaining something basic to some marketing dude in a different state. Get code to compile. Scratch butt. Test/debug. Change CDs. Check working code into git. Figure out what needs to be coded next. Manager enters office, informs me requirements have changed and what I just checked into git is now wrong. Read
I've actually done something similar to this in the workplace before and found it to be very effective. Basically I scheduled a meeting for the last hour or so of the workday Friday and invited the other developers to come jump on with me. I shared my screen, we had an open mic, but basically I talked through what I was doing and they had a chance to see how my thought process works and also to make suggestions or ask questions. It turned out to be a good chance for the more junior folks to learn from me and some of the other experienced developers and at the same time it made us think about what we were doing at a deeper level, which is something you can get complacent in after you've been doing it a while.
This is something I did for several weeks in a row, but the last time was at least two years ago -- and yet as recently as last month I've had some of the folks that participated bring it up and comment how much they learned from it. If you think about it, it's basically pair programming on steroids.
I spent 4 weeks thinking, and then I coded 22,000 lines of C++ in about two weeks.
Ended up in wrist braces.
The code was worth it.
Side Comment: why slashdot.org is the only site not allowing to edit your posts?
Errors are less of a problem than revisionist history.