Google Launches Free Tool To Let You Run Your Own Online Courses | Hacker News Reader

Google Launches Free Tool To Let You Run Your Own Online Courses

edudemic.com

153 pointsAnon8434 comments

I'm a very good programmer. Self taught. Built some popular tools, top percentile on StackOverflow. My blogs get a few thousand hits a day and some nice fellow translates them into French. I love self-learning, and dislike most organized learning.

However, I don't get a lot of replies if I apply for a job. HR ladies don't care if you're good, only if you check an item on a checklist.

I get by on my own products and on my rep, which is fine, but sometimes I need a gig to pay the bills.

So, my point is, it's great to have all these online courses, seriously, but how is this going to help people get jobs? I can see it helping them make their own job, but some people just want a regular old day job.

If I'm missing something, and there's a good program with some sort of certification, then please point me that way.

I am just like that, but I got one thing - if you want to live within existing ecosystem, you have to apply certain rules to yourself as well. Yes, these rules are stupid sometimes (education requirements,) but we live in real world, not ideal one. If you want to get hired (i.e. become part of the "system") - get education and piece of paper to show to HRs, otherwise - start on your own.

Very true. it's too bad though. Some of the best programmers I know were self taught or are electrical engineers

With all these online courses going on, the recruiting will change. Not now, but it will.

I'm really into this kind of teaching, and I hope it helps the poor, cause there's lots of job openings but not enough talent.

This is just one step, a big one, but not the hole solution.

I agree, it would be amazing if some forward-thinking recruiting firm went by your codeacademy score, stanford online course mark and SO rep.

Um, what? Here we have a product serving the needs of teachers, built for "webmasters". How many course-builders out there want to familiarize themselves with 3 coding languages (HTML, Python, and JavaScript), when a easily conceivable web UI could accomplish the same goals?

Google is either throwing AppEngine fluff or a pump fake.

Having worked in academic IT for most of my career, it's also really hard for me to imagine many professors using this considering how many of them have called me about problems due to WYSIWYG editors and whatnot. I guess this is mainly for an IT/coding audience, not for teaching history or something like that.

Quite a lot of CS professors could handle this, and and even higher proportion of their grad student Research Assistants. Y'know: the people who work at Google later and the people that train them.

This isn't iMovie; it's a courseware building platform. They didn't claim it would be useful for everyone.

Yeah, but are most of us who are capable at code going to want to deal with Appengine? Definitely not my own first choice.

It seems to be a very early release just like the article says. Additionally it is Apache Licensed. So even if Google is unable to deliver a brilliant and easy-to-use tool, others can probably create a layer on it. It's a win either way.

There's no reason a nice GUI couldn't be thrown on to the front end of this.

Just post the link of the course builder https://code.google.com/p/course-builder/

I'm surprised that they seem to require course creators to build courses using HTML and Javascript, when most online learning environments focus on browser based course builders.

As someone who is currently forced to use Moodle to build courses, I honestly say I'd rather just build the thing with HTML and Javascript rather than a web-based interface.

Speaking of which, what tool would you suggest instead of Moodle?

Why not use your favoured programming tools, then include the appropriate metadata to talk to the Moodle database if that is what local admin want you to do?

If you find Moodle frustrating, you're going to be suicidal if you try any of the major competitors.

Do they supply a course in how to use it? :)

You remember how I've been telling people they need to learn to code? If you can then you can look at their code and see that they actually don't have much there, and what they do have is kind of crap:

https://code.google.com/p/course-builder/source/browse/#git%...

There's really not much there, and a lot of the code is kind of junk code meant to get people stuck on GAE. Here's my criticism of it so far:

* Most of the controller code looks slapped together and kind of hacked on rather than designed with any thought. https://code.google.com/p/course-builder/source/browse/cours...

* There's hardly any other services other than a model for Lesson/Unit/Student and controllers that are hardcoded for their own course. For example, their entire Assessment system in lessons.py isn't backed by any model and is just dumped into the Student model, which means students can't be assessed more than once or on multiple courses or lessons? Hopefully I'm reading that wrong.

* The model isn't even really right. There's no Course or Instructor model and all the other things that should be models are just tossed into lessons.py controllers and dumped into memcache.

* Very tied to GAE it seems, and to set this up you'd better be damn good at that and python. There's no way in hell this'd ever be useable by anybody but a programmer.

* The assessment stuff is either going to be useless or a major time sink for them. Instructors are notorious for having batshit crazy assessment systems making it impossible to create a universal one they'll all use. I'd ditch that stuff or hook it into google docs spreadsheet instead.

* Seems to have almost zero Student->Instructor communication, and only a placeholder for the Forums. If they plan on doing the classic Google Supportless Tech Support then they better fix that, or hook it into G+ better. Problem with G+ is they'll need to make it so students can't monopolize instructor time in private.

* They'll very definitely be tying this to all their services, which means it'll only be slightly useful outside of the google ecosystem. If they plan on getting into the K-12 hosting then that'll be a problem for them since many school systems can't host student data on outside computers because of student privacy laws.

It's easy to rip apart code like this, and most likely when I release the code to my stuff people will do the same (and should). But, if they post a breathless press release touting it, then I expect it to be much more capable than this. This to me looks like classic corporate FLOSS white elephant code where they hope some poor suckers will jump in and make it great for them so they don't have to work on it anymore.

I'm surprised to see you of all people nitpicking an initial release like this. Peter Norvig could fart in a crowded room and it would lead to breathless press releases - don't hold it against these few developers that their modest repo has been published early on and already gotten more press than your newish competing project. I'm sure there will be plenty of room for both of you in the marketplace.

I don't agree that this is entirely nitpicking. Sure, ripping apart some code in a controller might be considered coming down hard on a brand new project, but a lot of his criticism is directed towards the foundation/model behind the system as opposed to "You missed a semi there." Hell, he even mentioned ways they could integrate existing services (G+) into this project.

Oh well, at least they shipped.

Keep working on inculcate.me and show them how it's done.

re the "get people stuck on GAE" comment, I could likewise make a comment about how most dev stuff shown on HN "get people stuck on OSX".

I use GAE and feel it's an incredible boon to teams who want to focus on app logic and not infrastructure.

sure it might not be your preference, but it's a great technology, extremely easy to use, and very useful. You should recognize it as such.

It's good to see the education space getting so much attention, it sorely needs it! Within ten years, I think we'll see a big shift towards online learning, or at least a much heavier lean on it.

Shameless plug, but for those looking for more of a turnkey hosted solution, check out Pathwright: http://www.pathwright.com

Looks interesting, but why do I have to create a username and password even if I sign in with Facebook? The whole reason for allowing people to sign in with Facebook is so they don't need to manage yet another login and password, or so I see it.

Google is touting a tool they used to build a course called "Power Searching", that was supposedly successful. I signed up for the course, only to find it is not ready/available until Sept 24th. I am completely confused.

That must be the next time they're planning to offer it.

Installing the example code and running it indicates that there was a previous course that started July 10.

There's way too much nasty FIXME configuration stuff scattered all over the code (course name, institution home page, etc.) All that should be in (e.g.) a JSON file that gets read at load time.

without compliance with open standards this kind of tool won't go any further expect than experiment. It supposed to be integrated learning management systems easily with http://en.wikipedia.org/wiki/Sharable_Content_Object_Referen...

Nice, just too bad it forces you to be on AppEngine, which has some very controversial pricing model...

Controversial? That is an interesting word choice. I've heard some describe it as expensive, but what is the controversy surrounding how they have modeled their prices?

I think they're getting at the way Google hiked their prices which resulted in some peoples bills rising by a factor of 100. GAE has every right to make money - it was the speed and direction of the change which caused 'controversy'.

It's FREE until you get traction, and then you just need to know how to use it efficiently. It's a tradeoff, like many things - put in more effort to understand the animal, never have to fiddle with the server, never have to hire ops people.

Most of the complaints about pricing were fixed when users looked more carefully at their code. For some it's not worth it to do so, but that's not Google's fault. And before that, their inefficient code was costing Google money. You'd have done the same thing.

How would one go into running this code in your own server?

AppScale (http://appscale.cs.ucsb.edu) or Typhoonae (http://code.google.com/p/typhoonae/)

Reply on news.ycombinator.com