I have an iPhone, but I assume it works the same as an iPad, both Image Capture and iTunes pop up for me. Image Capture is also available in the Applications folder if she somehow set it to not automatically load when the iPad is plugged in, try opening it from there.

Never heard of Dojo before right now.

The problem isn't the image uploading library (though I did struggle with it a year ago while angular was newer, and I was newer to it)... it's more the rails end. If you're background processing images (totally mandatory in rails), what sort of hooks are given to you to display when something is done processing, what links on S3 are valid, etc... I'd really rather not handle it if I can offload it to an awesome API for $20/month.

It's not amazingly hard, it's just a pain in the ass. The existing rails tools kinda suck. That being said, I haven't explored new options in about a year, and was mainly trying to refactor some old code on a site I had written last year in angular/rails, and upgrade the rails version from 3.2 to 4.1.

And if you google stack overflow for "multifile upload rails nested attributes" you'll find some weird questions that have like 12 upvotes and 3 answers that have 1 or 2 upvotes. Rails can't handle batch uploads, so each ajax call on a multifile upload is hitting the controller once. This makes it a pain if you're trying to associate the screenshots with an item that hasn't been saved (i.e. a form for a guest user). I don't actually know if this is different in other languages, but you'll often have to either pre-save the form object to associate screenshots, or just run a cron job or something later to clean up unassociated images. As I said, not amazingly tough, but annoying.

What background job gem are you using? You can read out the progress of a task.

Example if you're using sidekiq:
https://github.com/mperham/sidekiq/w...b-has-finished

Not sure about the others but it's likely possible.

I'll agree that file uploads in general are lame tho if you want accurate progress, batching, etc..

https://github.com/lardawge/carrierwave_backgrounder for carrierwave, which ties into delayed jobs. Delayed jobs does not have a lot of the features of sidekiq, so I should probably start using that or resque for future sites.

Yeah I would use sidekiq or resque for sure. I like sidekiq because it seems way more well maintained and the performance is solid. I haven't found anything about resque that makes me want to switch to using it but I haven't looked in a few months. Haven't had a reason to.

I guess it's like, how do you value your time in a project? I often spend way too much time implementing stuff that a lot of gems will do for you pretty quickly. This was a big problem initially, and I have slowly becoming lazier and lazier as a developer... more because my time has become more valuable, and I've already tried to implement most features from scratch.

We've had discussions about things like heroku (which I think is awesome). I could probably spend two weeks learning dev ops well enough to just provision my own aws servers with a bunch of add ons, or use one of the other newer providers for a new website... but $50-$150/month instead of say $15-60 is not going to make or break anything. I don't want to do dev ops full time, or even most of my time at any company going forward. If a start up takes 3 months to build a beta, does the extra hour or two a week I'm spending hacking through my own chef/capistrano scripts (or w/e is used to provision stuff these days) offset the difference? My guess is I'm wasting way more time/money trying to do it myself.

Recently I've been playing with websockets, which aren't amazingly complicated... but when there's something like the pusher API out there, why bother introducing buggy/possibly insecure code at the expense of a small cost? I'd rather just use something that people are working on full time to make my life easier at a reasonable price.

So I like finding stuff like GM linked that makes my life easier. Obviously, there are always points in the lifecycle of an app where you have to evaluate whether it's worth it to roll your own solution and remove a dependency/cost.

I just love tools that save me time and make me more productive. This is just me thinking out loud.

By the way, rubymotion is totally awesome. Within like 3 days of fulltime-ish monkeying I could build a legit iphone app with a rich ecosystem of gems to make my life easier. $200 is a small price to pay to work in ruby if that's what you're used to. I will always be pro spending money to save me time and frustration.

Depends. With heroku, I do seriously think the difference is much more than $100 vs $50 a month. A $10 or $20 a month unrestricted VPS can do a lot.

I try not re-invent too many things from scratch. I use a bunch of gems but I'm often not willing to pay per month for something unless it's a big deal.

Also, for example let's say you landed a client for a 2k job or something. How do you explain to him that it'll be an additional $150/month for a few third party services?

Lastly the "wasting time" side of things only applies when the time you spend working on something like devops could have been spent on something else that's generating money. I think this is a big misconception with time. Example, hiring an employee on its own doesn't help you grow your business because if you hire him to save yourself time but you don't do anything productive with your new found free time then you're really hurting yourself because now someone else is getting paid the money you used to make.

I find this topic pretty interesting tho. It's a bit OT for programming but maybe it applies to us since we do freelance/consulting work and like any contractor our wage tends to be associated directly to how much time we work.

I found myself to be miserable when I associated everything to time == money because I felt like every time I did something other than work I was losing money. You can imagine how crappy this could make you feel if you really believed that.

Just doing something as innocent as going to dinner with a friend was "wasting" 2 hours and on top of that you spent money so it was like a double loss.

The reality of the situation IMO is unless you could have spent 2 hours billing someone at that exact moment in time then you can't say you lost money. I feel like if you're at the point where time really is your limiting factor when it comes to income then you could consider doing things that save your time but doing it any sooner than that seems like a mistake because what else are you going to do.

It's also why you need to actually enjoy what you're doing because you will have points in time when you don't generate income and then you spend your time doing things you like or want to learn more about. If it comes down to watching a movie or learning devops and you choose watching a movie 100 out of 100 times then you'll never learn dev ops so it has to compete with other things you might want to do and win occasionally.

I mean if I'm billing out at $100/hour on a project, it's pretty easy to explain that an extra $50/month ($150/month is a lot of add ons, which would mean it's way more than a $2k job) or whatever is not worth me budgeting in another 5 hours. It's also way easier for someone to pick up where I left off if they bring on someone full time to expand on the app. Most high end rails consultants I know just use heroku for everyone's sake.

And obviously it depends on what you can use your time for... but the few start ups I've worked at have literally been never short of productive things to do. Also, if it's 3 hours a week I can just take off for mental sanity, that's worth a lot to me as well. I agree that if there's nothing to do, you should be doing it yourself.

I think the reality here is that yes, if you're contracting and can get a steady stream of work, your time is worth a calculable amount of money. But my choices in life tend to focus more on utility (as an economics major). What's going to make me happy, 2 fun hours with friends, or $200? It's not even close (assuming I'm living comfortably working normal-ish hours)... I'll always go out with friends. Is spending an extra 10 hours a week building another boring CRUD app that I've built 20 times worth the money, or would I rather work on side projects to learn new projects and keep from burning out? Or read a book, watch tv, sleep, etc...

It's a skill set to find balance in situations like these, and if that's tough, then a salary position might be more up your alley.

Maybe it's just a difference of what we get paid per job.

$50 or $100/month extra on top of $100ish hosting for Heroku just seems mind boggling to most of the clients I work with because the entire job might be worth 2-3k as a one time payment.

Even at 5k telling someone their hosting bill is $200/month (third party stuff and heroku included) is a huge ass bill and could scare them off.

yeah that's a legitmately different use case where doing your own devops makes sense.

in eg a startup environment where you have any kind of decent funding, getting the product out on time and securely hosted is so much more important than $200/mo it's just a joke. it's like, if you bring that up with the business guys they will be confused and say, "why are you even wasting my time with that question, just pay the bill and do something i care about."

I know that there is utility in $$ involved, but dismissing learning any server programming or Dev Ops because it doesn't pay the bills is missing a large point, IMO.

I've never learned any individual thing without gaining a ton of proxy knowledge, and server ops offers proxy knowledge in spades.

Did you try trimming the clip in the Photo app where the video is stored on the iPad and then clicking the share icon in the bottom of the screen and emailing it to the Computer? I'm not sure why your mindset leads you to doing things the hardest possible way and then blaming Apple for how hard you made it on yourself.

Actually even easier since he wants it in YouTube and there's a share to youtube button. Like it's actually impossible to come up with an interface easier than what they offer.

This is true... once. On a side project.

As a developer, your time is more valuable than almost any other IT role. That's why you get paid more. If you're spending any time setting up servers other than because you need a peculiar setup either you need to change jobs if this is being forced on you, or your company needs to fire you if you're dicking around wasting time when you could be playing 2048 instead, or you need to talk to your management team and get a real infrastructure team in place to do that stuff if your management is unaware that you're doing all this work just to be able to do your job.

Yes, you learning is valuable to the company. But it's not as valuable as you delivering what the business wants.

Didn't know you could trim it in the Video app on the iPad. I will try that next time. It was something like 169MB – probably too much to email.

Still doesn't explain why it's so confusing to just get the thing off the iPad when I plug it into the computer. It's actually much easier on a PC because the iPad shows up as a drive. I really don't think it's that ridiculous to expect to be able to plug your device into your computer to get a large video off of it.

Yep but I needed to edit it, and I had no idea you could actually do that on the iPad. That's actually pretty cool now that I know it's there. Although I'm sure I'll forget in several years the next time I want to upload a video from my iPad YouTube.

But that conversation wasn't strictly about on the job stuff, though for certain, if you are focusing on building websites for businesses with a large bankroll... let's be honest, how many companies are looking to spend 70k for a site plus 200+ / month for their site who can't find their own talent?

And what kind of company do you work at that lets you play 2048 on the job wilst getting paid $30+ / hr?

I torn:

[php]create or replace fib (n int)
returns decimal (1000, 0)
as $$
declare counter int := 0;
a decimal := 0;
b decimal := 1;
begin
if (n < 1) then
return 0;
end if;
loop
exit when counter = n;
counter := counter + 1;
select b, a+b, into a, b;
end loop;
return a;
end;
$$ language plpgsql;

select fib(4);
3[/php]

In very basic economic terms (imo) the problem is that we subconsciously tend to conflate price and value. A dinner with friends provides more value than the X$ it cost (in economic terms I'd say that you usually expect an economic profit from going to dinner).

And like you said, 2h != guaranteed 2h income (playing poker is the closest I ever got to having a fixed $/h 24/7)

One that lets you work from home.

Ya this is more along the lines of what I was getting at. And it does depends on the scope of the consulting project, and what they're hoping to get out of it. If thoughtbot or some other huge dev shop is building something, it's going to likely be expanded on at some point... so having a lot of custom dev ops stuff is going to cost time down the road. If it's a one time deal, doing your own dev ops makes a lot more sense.

At small companies, getting a working product out fast and securely to either raise a big round, or because you raised a big round, is way more important than some small monthly cost. And $1000/month could easily be considered small depending on how much time you're saving and what you're getting out of it.

On another note, the job market in Boston is crazy. I'm moving back next Friday and have been flat out getting spammed since changing my linked in profile and joined some local ruby/angular meet ups (recruiters on those things kind of annoy me).

How does vBulletin handle the "users currently browsing this forum" feature at the bottom. I built something similar for a site and made a datetime field that'd get updated every time you did any action (like viewing a thread). I'd then just long poll (it was for an SPA) for users that have done something in the past ~5 minutes. I could add a field for what they were viewing to show who was viewing a thread or w/e. Or does php have a way to automatically handle that with sessions? I don't think rails does (or at least couldn't find anything).

I should probably just use websockets for that.