|
UIApplication posted:Lol my company is so hosed, like 90% of our ipad implementation doesn't use autolayout Yep.
|
# ? Jun 8, 2015 23:38 |
|
|
# ? Jun 6, 2024 20:01 |
|
quote:Added HKCategoryTypeIdentifierCervicalMucusQuality It's great that they addressed the women's health items that were lacking previously but how the hell did they ignore all the other items people have been requesting over the last year? Particularly, there's still no way to track medications in HealthKit. At least the shortcomings of WKWebView seem to have been addressed!
|
# ? Jun 8, 2015 23:40 |
|
Choadmaster posted:At least the shortcomings of WKWebView seem to have been addressed! Custom protocols?
|
# ? Jun 9, 2015 00:54 |
|
Thoughts on the BitCode thing: The talk about automatically gaining support for new architectures makes it sound like BitCode is really just meant to help with the transition in case Apple ever switches iOS devices to Intel. Otherwise, compiling for an architecture the developer never tested on seems like a good way to introduce bugs, and I can't see what "new" architectures they could have up their sleeve other than a switch to x86. I don't see the argument about getting "free" performance benefits later on when Apple compiles your app with a newer compiler actually playing out, either, since any app that's still being updated will get these performance improvements anyway when the developer ships the next update built with the latest version of Xcode, right? And if a developer stops updating their app, you should probably switch away anyway. And the only optimizations possible at that point would be code generation, since the source code would've already been parsed and compiled to intermediary binary. Or am I just completely misunderstanding?
|
# ? Jun 9, 2015 01:02 |
|
Bitcode could also help in auto-optimizing for future ARM variants. That was my assumption, anyway.
|
# ? Jun 9, 2015 01:06 |
|
Well yeah, but if the developer is testing on those variants then they would also be shipping binaries for those variants, right? And if they aren't testing on them, why would you want Apple compiling to them for you?
|
# ? Jun 9, 2015 01:11 |
|
Because they differ in only small ways, like register count or having some extra SIMD instructions or such. Do you test on all the CPU families that are currently released?
|
# ? Jun 9, 2015 01:22 |
|
Almost (A4, A5, A6, and A8). I see your point, but I'm still wary of users getting what are essentially untested binaries. Guess it depends on the likelihood of Apple optimization changes exposing new bugs.
|
# ? Jun 9, 2015 01:38 |
|
Also, isn't LLVIM IR at least a little bit platform-specific? Is x86+ARM on the same bitcode possible?
|
# ? Jun 9, 2015 01:45 |
|
Possibly. They did say that submitting 64-bit-only iOS binaries would be allowed now, so it at least sounds like it's "bitness" specific.
|
# ? Jun 9, 2015 01:53 |
|
Doctor w-rw-rw- posted:Also, isn't LLVIM IR at least a little bit platform-specific? Is x86+ARM on the same bitcode possible? Yeah, you can run llc on a bitcode file and generate x86 or ARM or whatever. All the instruction-selection optimizations live below bitcode iirc.
|
# ? Jun 9, 2015 02:13 |
|
LLVM IR encodes almost every detail of the ABI except the actual assignment of arguments to registers/stack, and even some amount of that. In particular, you would not be able to effectively switch from a 32-bit to a 64-bit ABI. Bitcode submission really is mostly about microarchitectural updates and theoretical hardware workarounds.
|
# ? Jun 9, 2015 05:16 |
|
Subjunctive posted:Bitcode could also help in auto-optimizing for future ARM variants. That was my assumption, anyway. Yeah, I was imagining it'd be for the next time there's another armv7s situation where there's minor improvements, but nothing huge. It would certainly help for apps where including that extra slice is basically "lol no" if you're up against one of the various size walls.
|
# ? Jun 9, 2015 09:59 |
|
rjmccall posted:LLVM IR encodes almost every detail of the ABI except the actual assignment of arguments to registers/stack, and even some amount of that. In particular, you would not be able to effectively switch from a 32-bit to a 64-bit ABI. Bitcode submission really is mostly about microarchitectural updates and theoretical hardware workarounds. I thought it was more general than that, but I think I was confusing it with pnacl. Thanks!
|
# ? Jun 9, 2015 10:52 |
|
It's also a license to break binary compatibility if we need to.
|
# ? Jun 9, 2015 17:07 |
|
You could do that with ARM-to-ARM transformation as well, right?
|
# ? Jun 9, 2015 18:12 |
|
Excited about Address Sanitizer. There's enough C in my audio code that it's going to make a tangible impact.
|
# ? Jun 9, 2015 18:13 |
|
Ender.uNF posted:OK, the new plan is Wednesday for lunch. Who's in? I'm in. Where at?
|
# ? Jun 9, 2015 18:46 |
|
How about Garaje? If anyone has a better suggestion go for it.
|
# ? Jun 9, 2015 20:18 |
|
Subjunctive posted:You could do that with ARM-to-ARM transformation as well, right? It depends. Binary translation is always a bit fraught, because inserting instructions can break proximity constraints for e.g. short relative jumps. It's definitely easier to just compile the code the way you want it in the first place.
|
# ? Jun 9, 2015 20:25 |
|
rjmccall posted:It depends. Binary translation is always a bit fraught, because inserting instructions can break proximity constraints for e.g. short relative jumps. It's definitely easier to just compile the code the way you want it in the first place. I was thinking of something like edit-and-continue branch insertion, but yeah, might as well not make things harder on yourself.
|
# ? Jun 9, 2015 22:00 |
|
Ender.uNF posted:How about Garaje? If anyone has a better suggestion go for it. Looks as good as anything
|
# ? Jun 9, 2015 22:24 |
|
Garaje 's great unless you're looking for something remotely healthy.
|
# ? Jun 9, 2015 22:33 |
|
Subjunctive posted:I was thinking of something like edit-and-continue branch insertion, but yeah, might as well not make things harder on yourself. I would kill for this; I was never as productive when exploring new stuff as when using Edit & Continue. It's just so much faster to iterate when you don't have to recompile and get your app's state set back up. It's one of those things that never gets priority because it is a pain point extremely far removed from the people who work on dev tools. Anyway... Garaje it is, Wednesday at 12:15 (to give people time to get out of sessions and walk there).
|
# ? Jun 9, 2015 23:43 |
|
Ender.uNF posted:I would kill for this; I was never as productive when exploring new stuff as when using Edit & Continue. It's just so much faster to iterate when you don't have to recompile and get your app's state set back up. It's one of those things that never gets priority because it is a pain point extremely far removed from the people who work on dev tools. It's one of the best things about React Native. (Might come to SF tomorrow for a meeting and crash the lunch.)
|
# ? Jun 10, 2015 00:18 |
|
My 'app metadata' was rejected because the reviewers couldn't find the Restore Purchases button. I hid it real good at the bottom of the Downloads section! #sarcasm #fuckmylife So what now? I replied in the Resolution Center but the 'Submit for Review' button is active again. Do I have to resubmit? edit: yeah I guess I resubmit. Holy poo poo I'll be mad if it's another 10 day wait. lord funk fucked around with this message at 02:49 on Jun 10, 2015 |
# ? Jun 10, 2015 02:40 |
|
lord funk posted:My 'app metadata' was rejected because the reviewers couldn't find the Restore Purchases button. I hid it real good at the bottom of the Downloads section! #sarcasm #fuckmylife Aka: I haven't met my quote, quick clear some apps from the queue!
|
# ? Jun 10, 2015 03:19 |
|
It's pathetic how Apple is still so bad at this. Apps should be reviewed and approved in 48 hours or less. I know that it only takes me 2-3 hours to actually do it once it gets through the queue.
|
# ? Jun 10, 2015 05:39 |
|
I'm crashing lunch probs.
|
# ? Jun 10, 2015 07:10 |
|
Chiming in a bit late but goonmeets rock and I'd love to do lunch because the wwdc box lunches blow.
|
# ? Jun 10, 2015 07:58 |
|
wwb posted:the wwdc box lunches blow. What do you mean you don't like slicing the roof of your mouth open on extra-dry ciabatta bread turkey sandwiches?!?
|
# ? Jun 10, 2015 08:07 |
|
lord funk posted:My 'app metadata' was rejected because the reviewers couldn't find the Restore Purchases button. I hid it real good at the bottom of the Downloads section! #sarcasm #fuckmylife You can appeal rejections, but if you hit Submit again it might be too late. I once had a photo app get rejected because the reviewer couldn't figure out how to take a photo on the iPad version. Except the app doesn't actually support the iPad, so taking a photo with it was the exact same on iPad as on iPhone.
|
# ? Jun 10, 2015 09:16 |
|
Doc Block posted:You can appeal rejections, but if you hit Submit again it might be too late. gently caress. My. Life.
|
# ? Jun 10, 2015 13:52 |
|
Built 4 Cuban Linux posted:It's pathetic how Apple is still so bad at this. Apps should be reviewed and approved in 48 hours or less. I know that it only takes me 2-3 hours to actually do it once it gets through the queue. And half the time, once the app gets into the queue, my prod environment logs show nobody besides myself opened the app at all between submission and approval.
|
# ? Jun 10, 2015 14:09 |
|
Happens all the time.
|
# ? Jun 10, 2015 14:11 |
|
Few years ago we submitted a project and it was in review for about 2 minutes before approval. Think we got a reviewer on his/her last day or something
|
# ? Jun 10, 2015 14:12 |
|
Apple reviews are inconsistent. We white label apps and some but not all need Core Bluetooth entitlement for some features. I should be better in the build system by setting entitlements, but I was under a deadline. Apple approved 5 apps that didn't need the bluetooth entitlement, and then blocked the 3 that did. To their credit, they were quick to approve the apps once I gave an explanation. But I was still a bit confused, and panicked because of a tradeshow next week and the ridiculous 14 + day wait lately for submissions.
|
# ? Jun 10, 2015 14:17 |
|
Dirk Pitt posted:Apple reviews are inconsistent. We white label apps and some but not all need Core Bluetooth entitlement for some features. I should be better in the build system by setting entitlements, but I was under a deadline. Apple approved 5 apps that didn't need the bluetooth entitlement, and then blocked the 3 that did. To their credit, they were quick to approve the apps once I gave an explanation. But I was still a bit confused, and panicked because of a tradeshow next week and the ridiculous 14 + day wait lately for submissions. Is it still the situation too where if you white label apps you need each client to stand up their own developer account and submit under that? Just as everything else about reviews are inconsistent, it seems inconsistent as to what the actual guidance from Apple is on that.
|
# ? Jun 10, 2015 15:27 |
|
kitten smoothie posted:Is it still the situation too where if you white label apps you need each client to stand up their own developer account and submit under that? Just as everything else about reviews are inconsistent, it seems inconsistent as to what the actual guidance from Apple is on that. For some customers we use their corporate accounts and handle everything. Others we "sponsor" and give them a discount rate and use their images with our corporate account and name attached. Quite honestly it is a mess. One thing I thought was funny was that after iTunes connect come back yesterday, all the apps in our account showed a last updated as 6/9/2015.
|
# ? Jun 10, 2015 15:42 |
|
|
# ? Jun 6, 2024 20:01 |
|
ultramiraculous posted:What do you mean you don't like slicing the roof of your mouth open on extra-dry ciabatta bread turkey sandwiches?!? I didn't get that far, looked at it and headed down to Yank Sing.
|
# ? Jun 10, 2015 16:49 |