|
Kallikrates posted:Right now FRP, and MVVM are popular and you can see parts of them within in MVC depending on where you draw the boundaries of responsibilities. Speaking of which, does anyone know the idiomatic way of working with ReactiveCocoa? My experience thus far has it working for simple stuff, and for more complex usages written by a coworker, occasionally crashing on the dealloc of...something. I haven't yet shaken the feeling that there's some elegant way of applying it - possibly. That said, the "complex usage" I speak of has it doing a lot of really stateful things in the middle of a workflow, which doesn't seem right. Off the top of my head, it seems like FRP needs some sort of adapter to manage updates to stateful/long-lived objects in order to interact meaningfully with UI.
|
#
?
Sep 19, 2014 19:59
|
|
|
|
| # ? May 16, 2024 18:15 |
|
|
Doctor w-rw-rw- posted:That said, the "complex usage" I speak of has it doing a lot of really stateful things in the middle of a workflow, which doesn't seem right. Off the top of my head, it seems like FRP needs some sort of adapter to manage updates to stateful/long-lived objects in order to interact meaningfully with UI. I'm not sure how complex my RAC stuff is but I tend not to do stateful things until I'm actually subscribing to a signal (usually in a ViewController). The stateful things I do before that final subscribe usually involve writing data received over the wire to disk in a map or flattenMap before transforming said data to one of my models. Glimm fucked around with this message at 04:23 on Sep 20, 2014 |
|
#
?
Sep 19, 2014 20:22
|
|
|
For the status bar in screenshots, I've been replacing my carrier's name with the name of the app.
|
|
#
?
Sep 19, 2014 20:23
|
|
|
Filburt Shellbach posted:For the status bar in screenshots, I've been replacing my carrier's name with the name of the app. Is there a handy way to do this, or are you just doing it in photoshop? I was poking through the UIStatusBar API briefly but couldn't make any sense of it.
|
|
#
?
Sep 19, 2014 21:01
|
|
|
Our iPhone 6 test device came in today, installed our app. Interesting that they just scaled up the apps this time to fit instead of the black bars like they did with the 5. I imagine theres some new config/launch image we include in order to support the new resolutions? Google, I know, 
|
|
#
?
Sep 19, 2014 21:07
|
|
|
Doh004 posted:Our iPhone 6 test device came in today, installed our app. Interesting that they just scaled up the apps this time to fit instead of the black bars like they did with the 5. I imagine theres some new config/launch image we include in order to support the new resolutions? Google, I know, Yep, or a launch xib/storyboard (warning: will turn your app universal if it isn't already).
|
|
#
?
Sep 19, 2014 21:09
|
|
|
pokeyman posted:Is there a handy way to do this, or are you just doing it in photoshop? I was poking through the UIStatusBar API briefly but couldn't make any sense of it. Photoshop's how I do it. Back in the day, like iOS 3, you could do it with a jailbreak. Maybe there's still some private API for it, dunno.
|
|
#
?
Sep 19, 2014 21:11
|
|
|
Filburt Shellbach posted:Photoshop's how I do it. Back in the day, like iOS 3, you could do it with a jailbreak. Maybe there's still some private API for it, dunno. IIRC the status bar's view is accessible by name without a jailbreak, but would obviously get caught before you reached the app store.
|
|
#
?
Sep 19, 2014 21:14
|
|
|
Since it's just for app screenshot's that's fine.
|
|
#
?
Sep 19, 2014 21:25
|
|
|
Doctor w-rw-rw- posted:IIRC the status bar's view is accessible by name without a jailbreak, but would obviously get caught before you reached the app store. I couldn't figure out how to mess with the signal strength indicator and such. I got at the views ok though.
|
|
#
?
Sep 19, 2014 22:13
|
|
|
We've used doctored screenshots before (like straight from PS/AI, not even from a device) and Apple didn't give a poo poo.pokeyman posted:Yep, or a launch xib/storyboard (warning: will turn your app universal if it isn't already). Good to know! We're already universal so this should be interesting. Hopefully my old code isn't poo poo and things aren't hardcoded (they probably are  )
|
|
#
?
Sep 19, 2014 22:27
|
|
|
Any recommendations on iBeacon hardware/SDK to play around with? The estimote looks pretty slick. Or maybe the kontakt? Apprently beacon companies are unable to spell anything correctly...
|
|
#
?
Sep 19, 2014 23:07
|
|
|
fankey posted:Any recommendations on iBeacon hardware/SDK to play around with? The estimote looks pretty slick. Or maybe the kontakt? Apprently beacon companies are unable to spell anything correctly... We ordered a bunch of Estimote beacons and they died in about ~3 months  They're replacing them but I'm not sure I'd recommend them based on that.
|
|
#
?
Sep 19, 2014 23:37
|
|
|
Doctor w-rw-rw- posted:Speaking of which, does anyone know the idiomatic way of working with ReactiveCocoa? My experience thus far has it working for simple stuff, and for more complex usages written by a coworker, occasionally crashing on the dealloc of...something. I haven't yet shaken the feeling that there's some elegant way of applying it - possibly. All of the state involved should be stored in models and viewmodels, and the RAC stuff should be stateless wiring between the two, and between the viewmodels and the views. If you have any state in the RAC stuff beyond maybe some memoizing/caching[0] there's a very good chance you're doing things wrong. [0] But I tend to be very dubious of caching anything short of a network request without very good evidence that it's useful regardless of context.
|
|
#
?
Sep 20, 2014 03:46
|
|
|
Plorkyeran posted:All of the state involved should be stored in models and viewmodels, and the RAC stuff should be stateless wiring between the two, and between the viewmodels and the views. If you have any state in the RAC stuff beyond maybe some memoizing/caching[0] there's a very good chance you're doing things wrong. Okay, that's what I'd expect. Definitely not what is being done in the code I'm looking at, though.
|
|
#
?
Sep 20, 2014 04:32
|
|
|
Plorkyeran posted:All of the state involved should be stored in models and viewmodels, and the RAC stuff should be stateless wiring between the two, and between the viewmodels and the views. If you have any state in the RAC stuff beyond maybe some memoizing/caching[0] there's a very good chance you're doing things wrong. Stupid RAC question I've never managed to answer for myself by reading docs, articles, etc.: how do you deal with (do you deal with) a table whose rows change position (e.g. after a network response, or through user input), other than just calling -reloadData?
|
|
#
?
Sep 20, 2014 04:40
|
|
|
pokeyman posted:Stupid RAC question I've never managed to answer for myself by reading docs, articles, etc.: how do you deal with (do you deal with) a table whose rows change position (e.g. after a network response, or through user input), other than just calling -reloadData? The aborted 3.0 version has a magic thing to do it automatically. With 2.0 you can use KVO a bit more directly to get the information (as that PR does), diff the old and new arrays, send update events over the signal rather than just the full new array each time, or just say gently caress it and use reloadData. One of the apps I worked with went the route of using RAC for all the trivial views that would otherwise be 90% boilerplate, and just did stuff the old fashioned way any time using RAC would require any sort of cleverness due to that some of the people involved really didn't grok FRP, and that actually worked out reasonably. Using RACSignals as if they were just promises isn't ideal, but mostly works.
|
|
#
?
Sep 20, 2014 05:00
|
|
|
Interesting, thanks. Nice to hear that a hybrid approach is workable too.
|
|
#
?
Sep 20, 2014 05:33
|
|
|
Somehow they've managed to make the new iTunes Connect even worse than the old one.
|
|
#
?
Sep 20, 2014 18:35
|
|
|
Lots of really bad widgets. I went to the apple store and tried to load up the two widgets their display phones have prepackaged, and both of them didnt load content.
|
|
#
?
Sep 20, 2014 23:58
|
|
|
Boy, I sure did miss when breakpoints would jump around your code randomly in old versions of Xcode. Thankfully, they brought that back in Xcode 6! edit: I take that back. My breakpoints are being placed just fine, but none of my code changes are being compiled no matter how many times I clean/fuxcode. Time to roll back to Xcode 5 for a few months. ManicJason fucked around with this message at 07:51 on Sep 22, 2014 |
|
#
?
Sep 22, 2014 07:48
|
|
|
Is TestFlight borked for anyone else in iOS 8? Through actual TestFlight and not Apple. Can't load it at all on my iOS 8 devices, but can still view and install from iOS 7 devices just fine.
|
|
#
?
Sep 22, 2014 17:56
|
|
|
I ran into problems on some devices on friday. Couldn't load it on a 6, but loaded on a 6+ same network, same profile etc. Weird (shady?)
|
|
#
?
Sep 22, 2014 17:59
|
|
|
Looks like it still works on our 5s. *Edit* I lied, it works on our 5 but not our 5s.  Something must be hosed with the 6's. I'll be honest, I've had troubles using it the past couple of days with connecting to things.
|
|
#
?
Sep 22, 2014 18:15
|
|
|
I had trouble with legacy TestFlight on my 6 all weekend, but my other iOS devices had no issues even though they were on iOS 8. And then this morning TF is working just fine. 
|
|
#
?
Sep 22, 2014 19:17
|
|
|
So if you had problems Friday and Saturday, I guess Testflights status pages was intermittently red due to load. I didn't know they had a status page, and their support channels are dead as of the acquisition. But that might explain some of the issues.
|
|
#
?
Sep 22, 2014 21:14
|
|
|
Are we supposed to switch over to the Apple one now? I feel like everything has been super hush hush since the acquisition and it has me a bit worried.
|
|
#
?
Sep 22, 2014 21:16
|
|
|
Has anyone played around with the view hierarchy debugger, as previously seen in Reveal? I'm currently on a 80% success rate of it completely rebooting my laptop when I click a couple of views.
|
|
#
?
Sep 22, 2014 22:25
|
|
|
Doh004 posted:Are we supposed to switch over to the Apple one now? I feel like everything has been super hush hush since the acquisition and it has me a bit worried. Last I checked the Apple TestFlight only supports "internal testers", i.e. people with iTunes Connect accounts in your organization. If that describes you, you're good to go. If not you're stuck with the old.
|
|
#
?
Sep 22, 2014 23:00
|
|
|
Doh004 posted:Are we supposed to switch over to the Apple one now? I feel like everything has been super hush hush since the acquisition and it has me a bit worried. The Apple version of TestFlight seems a lot more restrictive. For internal testing, you're limited to 25 users, and they have to be identified in iTunes Connect under your team. For external testing (i.e. betas outside your organization), you can distribute to up to 1000 users, but the app still has to go through the standard review process first. Also, that part isn't available yet. Either way, it kind of sucks for my team's internal workflow. Our CI box builds an enterprise signed version of the app, and loads that onto ("legacy") TestFlight. Since it's enterprise signed, there's no restrictions; if you can get your hands on the .ipa then you can run the app. Our QA people use that, and it's also basically the standard dogfood build that a few dozen people in the company use. HockeyApp might be a decent alternative that keeps that workflow going. We've already been using it for our Android team since February, when Apple bought TestFlight and abruptly ended Android support.
|
|
#
?
Sep 22, 2014 23:07
|
|
|
pokeyman posted:Last I checked the Apple TestFlight only supports "internal testers", i.e. people with iTunes Connect accounts in your organization. If that describes you, you're good to go. If not you're stuck with the old. kitten smoothie posted:The Apple version of TestFlight seems a lot more restrictive. For internal testing, you're limited to 25 users, and they have to be identified in iTunes Connect under your team. Yeah that's what I figured. As long as they keep old testflight going then I should be happy. Also we do use HA for our Android apps. Worse comes to worse we bring iOS over there.
|
|
#
?
Sep 22, 2014 23:18
|
|
|
kitten smoothie posted:For external testing (i.e. betas outside your organization), you can distribute to up to 1000 users, but the app still has to go through the standard review process first. Wait what? I thought there was some kind of review process, but it was much less involved than the "standard review process".
|
|
#
?
Sep 23, 2014 00:09
|
|
|
pokeyman posted:Wait what? I thought there was some kind of review process, but it was much less involved than the "standard review process". Maybe so. They said you're subject to the same guidelines as App Store stuff so I assumed they were applying the same stringency. https://developer.apple.com/app-store/Testflight/
|
|
#
?
Sep 23, 2014 00:22
|
|
|
kitten smoothie posted:Maybe so. They said you're subject to the same guidelines as App Store stuff so I assumed they were applying the same stringency. It says you have to go through a "beta app review" which I would assume is much lighter than a release app review.
|
|
#
?
Sep 23, 2014 00:56
|
|
|
Yeah, I don't think they're gonna reject your beta app for not being finished or whatever, they just want to make sure you aren't trying to distribute a "beta" app that's really just spyware etc to strangers.
|
|
#
?
Sep 23, 2014 02:38
|
|
|
Speaking of review, my app was rejected last night on the grounds that "2.21: Apps that are simply a book should be submitted to the iBooks Store", which is a bit rough after spending 8 months developing this app (a digital storefront/reader/library for comics created especially for phones). I've filed an appeal but christ, didn't sleep much last night. I'm really worried that what they're really saying is that all the comics I want to publish should really be published on the iBooks Store, experience be damned. That would be overreaching, right? There's already an app precedent with Comixology/Madefire etc. but I know that the Review Board pretty explicitly say they don't give a drat about precedent.
|
|
#
?
Sep 23, 2014 07:36
|
|
|
Froist posted:Has anyone played around with the view hierarchy debugger, as previously seen in Reveal? I'm currently on a 80% success rate of it completely rebooting my laptop when I click a couple of views. Hey, hard reboot buddy. On my Macbook Pro it always crashes if I drag / rotate the view debugger. Radar filed.
|
|
#
?
Sep 23, 2014 13:34
|
|
|
lord funk posted:Hey, hard reboot buddy. On my Macbook Pro it always crashes if I drag / rotate the view debugger. Radar filed. Ahh, it's drag to rotate - I couldn't work out how to go into the 3d mode at all. My crash always came when selecting different subviews in the breadcrumb bar at the top.
|
|
#
?
Sep 23, 2014 14:15
|
|
|
My iOS custom keyboard was approved today (the Unicode picker one I was talking about with Flobbster over in the Swift thread), and since I've learned a lot from you guys over the years, here's some coupons: P9LN3E73FAT9 X9YA46A4M9KA KJFT3EEWT43W 4A6KF9XKELP3 PNXYHFFXE97P Promo page: http://protonome.com/apps/glyf/ iTunes: https://itunes.apple.com/us/app/glyf-unicode-keyboard/id912177581?mt=8
|
|
#
?
Sep 24, 2014 13:46
|
|
|
|
| # ? May 16, 2024 18:15 |
|
|
Hmm. I have arrived at something I'm not sure of the best way to implement. The app I'm working on will allow in-app purchase of Sound Packs. The user can then select which Sound Pack they wish to use which will be in use until they select a different one. There are a few default sound packs to choose from provided. My initial thought is to store each pack in a SoundPacks directory in the app's doc directory, and keep the filenames for the sounds the same. So the breadcrumb trail would look something like: /<app docs dir>/SoundPacks/<pack name>/sound1.wav /<app docs dir>/SoundPacks/<pack name>/sound2.wav etc. with different packs being like: /<app docs dir>/SoundPacks/Default/sound1.wav /<app docs dir>/SoundPacks/Default/sound2.wav . . . /<app docs dir>/SoundPacks/MySoundPack/sound1.wav /<app docs dir>/SoundPacks/MySoundPack/sound2.wav Easy enough. So here's the "problem": The sounds are being used for localNotifications via: localNotification.soundName which simply takes the name of the file, with extension: code:I'm mostly thinking out loud here, but has anyone tackled anything similar who could provide some insight? TIA.
|
|
#
?
Sep 24, 2014 13:46
|
|
