Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
CloudKit JS (developer.apple.com)
162 points by mindrun on June 8, 2015 | hide | past | favorite | 35 comments


The link to JavaScript reference is 404 right now, but after fiddling with the URL a little bit, these seems to work:

CloudKit JS: https://developer.apple.com/library/prerelease/ios/documenta...

CloudKit Web Services: https://developer.apple.com/library/prerelease/ios/documenta...

Edit: They've fixed the links.


Heh. I'm happy to let other people try this out - iCloud hasn't exactly had the best record for uptime and little things like this reinforce my worry that Apple doesn't do backend services well.

That said, I look forward to being proven wrong.


iCloud has better uptime than Amazon Web Services.

People bash Apple and iCloud and claim they can't do backend software, but they ignore that iTunes has been trooping for over a decade under tremendous load without a pickup. Contrast that with another store on the web- Amazon.com-- which was (at the time I worked for Amazon) generally partially down %40 of the time. (Though not all of it and it does a good job of faking it when it is down.)

Look at AWS where companies like Netflix and Heroku have had days of outages because the AWS network went down or whole datacenters were taken offline.

I've never seen that kind of an outage in iMessages or iCloud.

Apple just keeps trooping not making any noise because it's always working... but because Apple's business model is selling hardware, and people give Amazon a free pass for some reason (I mean why is AWS popular when it's down all the time while Google is rock solid with its hosting services yet less popular?)


Every post you find a way to bash Amazon with half truths. It's really tiresome and inaccurate.


I don't think that iCloud has a better uptime than Amazon Web Services:

https://cloudharmony.com/provider/aws#aws:ec2


iTunes is a specific Apple service, not a platform upon which 3rd party services run. It's one thing to run a product; it's another to run a platform. As a provisioner of PaaS, Apple are not seen as proven.

As for AWS, they're almost ubiquitous. Half the world seems to run on AWS. Regions go down time to time, but half the point of AWS is that it's a distributed architecture that allows apps running on that platform to stay up - usually. Extreme examples (such as dependent PaaS Heroku being taken down) are the edge cases that make it to the press.


It seems like you're only claiming Apple's services to have higher up times than Amazon's because you were able to look at Amazon's up times more closely. Who is to say that Apple's services don't also often have partial downtimes and "faking it when it is down?"


Cool, thanks! Much appreciated! :)


If I'm reading correctly, that means I can build an app without any explicit auth on OS X/iOS with backend which allows access to personal data via AppleID username/password.

That's neat.


This is great for some people, certainly. I won't knock it for people who's lives are better with this.

But let's not kid ourselves. It's still their system; you're still playing by their rules and doing things their way. Their history in this space is not good at all. Maybe things have changed for the better; but it doesn't change the fact that this only brings you deeper into their walled garden.

One of the reasons apple doesn't like the web is they can't make money from them like they can from apps. Now, they have a way. They still have a bad track record with infrastructure; numerous instances of failures, but I imagine they're getting better.

Thanks, but no thanks.


Sorry Apple. You are many years of good behavior away from my trusting you with this big a piece of my infrastructure.


I absolutely don't understand why you got downvoted because i absolutely feel the same. Apple has a proven record of not understanding anything related to the web, and having product and technology issues with anything related to it.

Just yesterday evening, you couldn't add an app on itunesconnect because of some angularjs error ( which they apparently used to refactor their site), and you could trace the debug log on the console, and follow the link on angularjs documentation ( a "bindonce" directive not included error).


Can you elaborate on this? Why? And why many years?


They have been providing rock solid internet infrastructure, and the tools for it, going back to the acquisition of NeXT. WebOBjects, for instance, powered Dell's online store, and selling direct was Dell's entire purpose of being in business in the 1990s. It's been powering iTunes since 2003 or so with less than a tenth as much downtime as Amazon Web Services... yet lots of people use AWS.


I don't think the OP was talking about the quality of the product.


Is this something similar to parse/firebase?


This was my thought as well. It looks like `(parse|firebase) + apple ID authentication`.


Yeah I heard that Apple considered buying Parse before deciding to build CloudKit instead. It's a direct competitor, although Parse is still much further ahead. CloudKit is a lot cheaper though.


Parse was acquired by facebook more than two years ago (April 2013), so they had to plan it for a long time then.


Free storage for 1PB - can that be right? Am I misunderstanding when I think petabyte?


There's a "Capacity scales with your users" section that says you have to have 10M active users for that.


Correct. You cannot just sign up and get a free 1 PM cloud storage account. The more registered users you have, the more storage space you are allocated. If you go over your allowed storage space you start moving into the paid tier.


Yep, you're probably right. I mean, dragging the slider from left to right clearly increases the TB-number to 1000 (where it stops) and 1000 TB = 1 PB


What are the chances Apple will pull your app if you start accessing these web services from an Android app?


Yesterday I'd have said 110%, but I dunno... I'm getting this dreamy feeling that something wonderfully developer-friendly might be going on here. Not to mention owning that infrastructure that let folks also -run- back Android apps seems like it'd have value in itself, beyond just the dev PR.


From the flash message on: https://developer.apple.com/library/prerelease/ios/documenta...

"Apple is supplying this information to help you plan for the adoption of the technologies and programming interfaces described herein for use on Apple-branded products. "

'for use on Apple-branded products.'


I think think they will do anything to prevent that. I mean, it's made for the web, so it's technically made for every device from every provider.


0% It would kill any chance of adoption in favor of parse or firebase.


Powered by FoundationDB?


No it's powered by the tears of former FoundationDB clients that got fucked in the process.


When they say $100 per additional 10 reqs/s - do they mean each time you exceed it?


I'd guess it's billed monthly as http://en.wikipedia.org/wiki/Burstable_billing so spikes don't kill you.


I hope this is a precursor to Apple Pay on the web.


I assume the cost goes up with users but the js on this page doesn't work on my kindle.


It doesn't, apparently. The cost seems to be per-user, so

250MB asset storage / user

3MB database storage / user

50MB transfer / user

10 req/s per 100k users

is all free. I don't know how the requests part scales though, 10/s for 100k users is each user hitting it once per 3 hours. I can't complain about the amount given away for free, but I don't know quite how this would scale with costs for different types of apps/loads.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: