Good job blacked out self.

I'm writing tests for every new feature on this app.

if you're serious, you'll write the tests first!

I didn't think angular was too bad to get started with. Once you understand a few key things that it does and what components handle what then everything just falls into place with experience as you use the framework.

Since I'm on a rails binge lately I haven't found much need to use a full blown JS framework as most of the JS is just minor enhancements. I might give marionette a shot though just because I want to try something new and backbone seems to be pretty popular with the rails crowd for when you need to do more than just sprinkle a little bit of JS around. Marionette seems to solve a lot of pain points that backbone has.

MEH

Can you explain what something like angular takes care of that overlaps with rails? Are you using pjax?

Marionette definitely helps backbone, but I still wasn't a huge fan. I've always heard that ember is supposed to be the "rails js framework" with a lot of community support, though I've never tried it.

I think this is how AngularJS is going to be perceived in the long run, as more of the use case is handled without leaving the AngularJS ecosystem, but as it stands, a lot of people have to bend AngularJS to work with large existing code bases and/or buggy foreign libraries (jQuery plugins, etc) to get the work done and this is where the complexity comes from. I've spent a decent amount of time helping people on the AngularJS google group and it's fairly clear to me that the learning curve is fairly steep you get past the basics (using existing directives, creating basic template-based directives, basic services/controllers, simple routes) to meet more complex requirements.

It's not so much overlapping, it's just that both don't really seem necessary to me for most sites/apps. This would go for ember as well I guess, but I can't say for sure because I never looked into it too much.

If you're developing an app that is "document-like" then you really don't need to make it a full blown client side app. pjax/turbolinks is going to provide a user experience that's very good and it also removes any extra work on your end to make it work for search engine indexing.

dhh said it best in one of his conf talks. They use backbone for an interactive calendar because it's very real time-y and has a few thousand lines of javascript. It's clearly something that needs more than a few dozen lines of JS and it would easily get out of hand if you sprinkled a ton of JS around.

They could have used angular for it and I don't think it would be a huge mistake to use angular in that case but it just seems like overkill to me.

I used to think I wanted to make really fancy purely client side admin panels or user interfaces with widgets and all that other **** in the end but it doesn't matter. The only thing that matters is the perceived snappiness of the site and that it doesn't fall over under pressure.

I'm really happy with the user experience of pjax/turbolinks. Things feel really snappy to me.

I also do not like full blown client side apps mainly because it always feels like I'm developing 2 apps instead of 1. You have the flesh out the server end points and implement user auth on the server as well as data massaging, but you also have to take care of a ton of stuff on the client too.

It just feels like a heck of a lot more effort for practically no gain at all.

I've only worked with angular from ground 0, as in... I went into a new project using only angular with no other js dependencies.

I think angular makes it nice to create a lot of those jquery plugins in angular instead. Like if I wanted a photo gallery in an angular app. My first instinct wouldn't be to try and shoe horn in a jquery photo gallery.

I would just use what angular gives you out of the box and write the photo gallery in the way it needs to be implemented. To me this is the point of these frameworks, it's so you can get the things you want done quite quickly.

It's sort of why there aren't 1000 different CMSs made in rails. You don't really need one because rails itself gives you the tools to build your "app specific" CMS in a very productive way.

If a gun were being held to my head and I had to develop a fully client side app I would certainly pick angular. It makes me happy to use it because I feel productive. Being productive is extremely intoxicating. When you don't get bogged down by bs and you can actually make what you want to make quickly... man, that makes my pants move.

Ya, I'm with you on the developing two apps instead of one thing. At this point turbolinks seem good enough for most of what I'm doing, and if I need a bit of heavier javascript I can just use a framework for that bit.

Speed of getting the site up is the main issue for me. It just takes longer and is more error prone when you have to double down, and I'm not sure the resulting user experience is worth the extra effort. It also makes me lose motivation when it takes an extra few weeks to get it done... I'd rather find a new project with new challenges, or move onto more interesting features.

SEO seems relevant as well.

Probably doesn't quite dovetail with current discussion of angular and turbolinks and other things I have no idea what they are, but I wanted to report back and thank you for pointing me in the direction of dojo/dijit. it's fantastic, and an old hack like me loves to be able to use the declarative syntax.

A little less than a month later I have an already very useful app, and lots more still to implement. Wasn't without stumbles (took me a while to get used to the asynchronous nature of the xhr calls which is obvious in hindsight). My work is also no doubt nasty and could be organised much better, it's basically one huge html page with lots of inline script and styles, and a handful of php files getting xhr'ed at various times to get/set stuff in the database. I'm not sure how best to fix that but I don't worry about it for now.

Yep, getting the site up asap is important for me too. If it's a difference between 80 and 40 hours then you're halving your hourly wage. 40 hours is no joke too, that's a lot of time.

I lose motivation too, especially when the code base sucks or I'm getting hung up on things that really don't matter to my specific app but need to be solved anyways.

Damn I wish I read this post and your earlier post sooner. I just started working today with a recruiter looking for contract opportunities, and of course I gave him a salary range.

Is it standard for agencies to ask prospective employees to take a skills test? My guy wants me to take some kind of online competency test for Java.

I want to give you an answer to that, but it really depends. Sometimes recruiting companies are very turn and burn and want to invest as much as possible in infrastructure (such as skills tests, etc.) to take the reliance off of the individual employees.

Other times firms will use it as a differentiator, and value add. But let's me honest what hiring manager gives a **** about some recruiting agency java competency test? I would guess not many.

Other agencies will make a rule that they only work with W2 American citizens and will not work with 3rd parties that sponsor a person's h1b visa. The 3rd parties can be sketchy as ****, and try and pull some of the more ballsy moves you could imagine (switching candidates and having fake candidate get VPN access to get the work done remotely during their shift, and literally sit there as a lame duck).

Different recruiting companies will have all types of different practices that they try and use to differentiate themselves and make them "better" than the others. In the end its pretty much 90-95% the same, and then there are outliers at either extreme (completely unprofessional, dont pay on time, float too much credit with companies that might not be good for it) vs. (more or less independently make hiring decisions for the company, give you great benefits, guarantee you some time at the beginning of the contract). There was 1 time where I knew about 99% that I would be basically choosing the candidate, for some people that is the norm.

Does anybody here know how much two programs would cost to make. My dad own's a business and he like 54 and still running it with brute force. He's done an amazing job running it by hand for the past 15 years, and has grown it to service over 20 different accounts. It's a janitorial business, btw.

The first one would be a website that inputs amount of stock and displays a table of how much stock is needed to be purchased.

The second one, and more complicated, is a phone-in system. Employess work on-site and check in at the beginning and check out when they are finished. Since it's a janitorial with 20 something workers he needs to be able to keep track and know that these places have been cleaned. Right now he does it all by voicemail, and that is a lengthy process, so I want to somehow be able to input phone calls from a google account, and display the status of each account.

These should be vairly easy jobs, so I'm guessing it shouldn't cost me too much, right?

for the first question, you should look for inventory managment software. if the stocking needs are extremely simple, a google doc spreadsheet may be all you need. i wouldn't pay for anything custom in either case.

for the 2nd question, you should be able to hire someone to create a simple system using twilio. depending on who you hire, expect the cost to be anywhere between a couple hundred and a thousand dollars.

Thanks alot man.

Aw man, that wasn't intended at all.

I seriously wasn't saying anything bad about Lisp! It's just that the presence of many dialects that do different things makes it easier to compare them. It's a sample size thing.

The goal was do a bunch of crazy stuff so that I could essentially say, "using too many arbitrary idioms looks exactly like this to people who are learning". I'll have to think about it some more.

It's possible. It already happened once but it may need to happen again.

Well yeah, I wouldn't write the second one in Python either. But I would write:

For some reason, this particular pattern doesn't bug me so much in languages where you can access undefined variables (e.g. JavaScript).

Hm. I do agree with this and I probably could make it clearer. Maybe I should move the "reasonable" examples to the end so it doesn't look like I'm lumping "while (1)" with abominations like "[(print(x), x)[1] for x in L]"

Thanks all. btw, did anyone else see Bootstrap 3's out? All the more incentive to toy with the thing a bit tomorrow:

http://getbootstrap.com/

I find the ternary syntax very verbose and difficult to read for something simple like this. Especially with longer variable names:

I'm not overly concerned either with how other falsey values behave. It's obvious that this API is expecting something like a list.

Carmack's keynote speech is up:

http://www.twitch.tv/bethesda/b/439369577?t=15m

It's 3 hours of him talking about next gen game console development and functional programming.

If anyone is capable I'm looking for a quote on both programs, am serious about getting this done.

If anyone is capable I'm looking for a quote on both programs, am serious about getting this done.

Could probably do the start/stop check easier with a web form...

The businesses employees are paid a lowish wage, so everyone doesn't have a smartphone. That's why the input has to be either call-in or text-in. That Twilio seems like it would cut it, but I also know that, with a Google account, you can get phone calls and voicemails sent to email.

I was thinking it may be best to just write a script that searches the email account, and looks for which phone numbers have already called.

Then I just need to display a simple table with rows represtenting each account, and columns that displays when they phone-in and when they phone out.

i guess this is kind of old but it came up today and seems oddly appropriate for this thread:

https://twitter.com/molovo/status/36...259456/photo/1

You think the janitor site should be written in git?

That's BRILLIANT!