Scrivener Temporarily Withdrawn from Sale on the Mac App Store | Hacker News Reader

Scrivener Temporarily Withdrawn from Sale on the Mac App Store

literatureandlatte.com

50 pointsMaysonL47 comments

"2.4.1 was rejected on Monday night for a reason that didn’t make sense (the reviewer said it wasn’t sandboxed when it was)."

I've had some incredibly fun (sarcasm) experiences with the review team for the MAS over the last few months. Reading that sentence got a belly laugh out of me.

The first major WTF that I had with the review team was their insistence that an app I had submitted was calling a private method. A method with the same name was present in an included library but rather than explain that to them I found it quicker just to remove the library (a workaround was simple enough). The hair pulling started when they again rejected the app for exactly the same reason. As the app relied on a heavy 3rd party framework it took a lot of investigation to understand what (wasn't) happening. Finally I took out a $50 tech ticket and after a week the engineer explained that the review team's tool for checking API usage was broken and it had nothing to do with my app.

The next time the review team rejected an app (seemingly within minutes of it going into review) because it was "producing garbage output". This time I just did a cursory review of the app before using another $50 tech support ticket. The app was sometimes writing files with the format somefile.ext.sb-randomstring. I suggested to the reviewer that the app was messing up an atomic write and given the sb- part of the filename had something to do with sandboxing. A week later the engineer explained what it was to me -- which was exactly what I had suggested.

My advice to anyone dealing with this is to spend zero time talking to the review team. Don't resubmit (unless it's something really obvious) and don't try to reason with them. Filing a $50 tech support ticket will basically guarantee that someone capable of passing the Turing test (have a conversation via iTunes Connect and you'll understand what I mean) will look into the problem. The tech support team will usually ask the review team to expedite your review once the problem has been solved -- something you'll rarely get if you ask yourself.

  the reviewer said it wasn’t sandboxed when it was

I'm afraid I'm utterly ignorant about Mac programming, but how is that even possible? I thought sandboxing was built into the execution environment, not the app.

The app requests sandboxing. That request then causes it to be run in a sandboxed environment. Apps can refrain from requesting it and run non-sandboxed, but such apps are not accepted in Apple's store.

Ah. Somehow I assumed app store apps were built into a different format than "regular" apps.

I can see why one might think that, but it's not the case. The core requirements for the store are optional things that can be added to any app: sandboxing and code signing. Apple just enforces those by refusing to accept any app that doesn't have them.

Finally I took out a $50 tech ticket and after a week the engineer explained that the review team's tool for checking API usage was broken and it had nothing to do with my app.

Did Apple at least refund your $50 when they discovered it was a bug in their own software?

No, neither time. You do get a certain number for free per license so neither of those were actually out of pocket.

To me, the most interesting part of the story is the parenthetical comment at the end:

  (Incidentally, because of some of the changes we are being
  required to make, we cannot guarantee that Scrivener on the 
  Mac App Store will have as good support for .docx, .doc and 
  .odt formats in 2.4.1 and future releases. This does not 
  affect the version on our site.)

I wonder what's behind this, how it actually affects Scrivener and whether it will hit other apps in the future.

This is why I'm (slowly) weaning myself off of Apple platforms after being an Apple user and enthusiast since 1984. A computing platform where the vendor is the ultimate arbiter of what software the user can and cannot run is fundamentally broken.

iOS has always been this way and, though I used it for a while, eventually the stories of apps being rejected for all sorts of reason were too significant to ignore. When Apple tried to dictate the set of acceptable _programming language_ for apps I gave up my iOS devices.

Though I can still run software on my macs without permission, I'm now looking for a suitable replacement for Mac OS as well. It's difficult to imagine a day coming where the Mac would be "app store only" like iOS, but it's not impossible.

It's not quite how you make it sound.

Apple is the ultimate arbiter of software that is sold through their store. They do not control what you build and run on your device through Xcode. You just can't distribute the end result. It's still a computing platform, and you can still write and run code. You can make your iOS device compute things.

There is value in being able to sell and distribute the end result — but that's not what computing is, that's marketing.

Edit: Why not make your same complaint about Sony's PlayStation 3? Do you not use that because it is a closed environment, one where you can not run your own software?

The PS3 isn't a general purpose computer like a Microsoft or Apple PC.

But an iOS phone is a general purpose computer? (The discussion was relating to iOS devices, in addition to Macs.)

A PS3 is just another computer in your home. It just happens to be locked down and branded.

"They do not control what you build and run on your device through Xcode"

Late response: There is some software I want to run that I do not have the source to. There is some software I want to run which is not being written because the person interested in writing it is not convinced that they will be able to distribute it to anyone in a convenient or profitable fashion.

I don't own a PlayStation, but of course have many other computers that are "appliances" (mp3 player, digital camera, etc.). I don't expect or even want all of those to be general purpose computers. For a long time I thought of my phone as an appliance, but with modern smartphones it's pretty clear that being a general purpose computer >> being an appliance and so I want to use platforms that are open in that way.

I go with direct download over Mac App Store every time when it's available. This story illustrates the reason pretty well.

Interesting. Many people (including me) will choose the Steam version of a game if its available, to get all the niceties of auto-updating, easy reinstalls, a central list of games & achievements, and so on. Shows the difference implementation can make.

OS X has awesome updates, though. Almost every single OS X app uses Sparkle [0] to check for updates on startup. I feel bad giving the credit to OS X, but Windows has no such prevalent counterpart.

0: http://sparkle.andymatuschak.org/

It's funny too since I actually find Sparkle updates far less intrusive than App Store updates (I'm tired of logging in and out of iTunes over and over and over again).

Why do you log into and out of iTunes? That makes no sense to me.

Having old apps on a different iTunes account than everything else can force this.

Why would you feel bad about it?

I agree fully – the danger of the walled garden platform has been illustrated quite quickly. Though, part of me wants it to happen more, in bigger ways, so that everyone notices how bad it can be.

Relatedly, I think it's somewhat amusing – I find myself using the Mac App Store the way I sometimes use brick and mortar stores – to look and see my options and then I go see if I can find a "better deal" from a website – in this case, the software publisher's site.

Similarly, the stuff I have actually bought from the App Store is either items that are only available there or are so cheap/single use items that I don't care about shopping around.

One person's receipt validation bug hardly means the "walled garden" (so sick of that term) is dangerous because his update review is taking a week. He's the one who made the mistake. Other computing platforms, such as consoles, have had centralized approval over software titles for decades. It's only become an issue to desktop users who are used to the Wild West and are experiencing a transition to the software model that has existed everywhere else for a long time.

By the way, anyone wanting automated Apple receipt validation code should look at Receigen on the Mac App Store. It generates a validation solution with each new version number.

I do have to agree that the term walled garden is completely inaccurate to models companies like Apple are pushing towards. More accurate would be calling it a cell block; you gotta do what the warden wants, when the warden wants, regardless of how arbitrary it is. Apple's push towards centralization is why my mac mini is now sitting unplugged on a shelf -- it's becoming harder and harder to do neat things under the apple system because you have to ask for permission every five minutes.

The initial mistake certainly was the software author's but that's hardly the most important point.

There are at least three key issues brought out by the original post, IMHO:

1) By moving to the App Store distribution from a previous, independent distribution system, publishers have clearly taken a step backwards in terms of being able to push out bug fixes. This is a huge issue and it's irrespective of what other platforms have to live with. It's a clear downgrade of a valuable feature for end-users and that's a problem – particularly because it's largely driven by policy and not technical requirements.

2A) Even worse, Apple's gatekeeping is horribly broken – approval is so opaque as to almost seem arbitrary and therefore cannot be relied upon. This is not the author's fault.

2B) The standards for gatekeeping, such as they are, are not constant. It seems clear that features that were previously acceptable can become no longer acceptable if Apple deems them to be competitive with Apple products. Or, worse, if Apple alters their products in such a way as to overlap with the existing features of a product in the App Store, a program may no longer be approved – essentially without warning and with no grandfathering due to prior approvals.

So, yes, this story says a lot about the controlled environment of a walled garden ecosystem.

I buy computers, because I want programmable tools that I can use for my purposes. I don't buy game consoles. As Apple pursues the game consolization of the Mac, many of us who are "experiencing the transition" are finding the Mac less and less appealing, but Apple is after the general consumer electronics market. They're making it clear that they can happily do without people like me--there just aren't enough of us to matter.

And ironically, despite caring less and less about users like me, they still get my business, because the other computer makers are degrading so badly.

If only someone were interested in making a computer for the rest of us [sigh...]

On a related note, Scrivener is the best app I've ever used on any platform. My wife and I use it for legal writing, and I've never found anything that can handle our workflow as well as Scrivener.

Scrivener is awesome, yes!

I really don't get on with Scrivener and I don't completely understand why.

I've gone back to Voodoo pad for writing even though it doesn't quite fit all my needs.

Apple's slowly strangling the life out of their golden goose with their capricious and heavy-handed control of their platforms. Their app stores are full of crap already and Apple is just increasingly in the way of developers' and users' best interests.

I really like Apple's technologies but I absolutely despise their idiotic policies.

Hmm. I find it annoying too, but I can't help feeling that for my parents (and most other non-techie users) this level of control, with a single source of approved apps, is just great.

A single source of vetted apps is fine but Google has already demonstrated that allowing expert users to use alternative app sources doesn't hurt. And I don't see how anybody can argue that required a full review cycle for every minor bug fixes benefits users. I see more and more mobile shops iterating on Android first because there are just fewer hassles.

Also, Apple's refusal to allow developers to do lower-level things like create alternative keyboards means Android users get to use modern input technologies like SwiftKey while iOS users are still pecking away at a 2007 keyboard.

I see more and more mobile shops iterating on Android first because there are just fewer hassles.

Hmm, more like different hassles, I'd have thought: having to support thousands of screen size/hardware/OS combinations must be a huge pain, and less of the UI is given to you for free. Though I speak as someone who's done very little Android dev, and rather more iOS.

It's true that the Mac model (App Store PLUS third-party apps) is preferable for me. But then again, I do like the idea that my parents' machines won't run any unsigned code. I'm thinking of Windows virus support nightmares past.

I've been doing Android dev for the last six months or so after doing iOS for about two years and actually accommodating different screen sizes really isn't hard at all as long as you have a designer that understands it's not a fixed layout environment like iOS. Actually the dynamic layout tools on Android make things a lot easier in many cases, certainly far, far easier than AutoLayout on iOS (an API I absolutely loathe).

I like their review process. And I say this as a developer who has a had many rejections, and many extremely frustrating waiting periods (one involved the threat of legal action unless we updated our app, and Apple sat on the update for over a month!)

The rejections I have had have almost universally been for reasons that would benefit the end user. For example, one app we submitted used background GPS to track your running route. Apple told us we had to very clearly warn the user that it could impact their battery life. Another was a game that still had some placeholder art being used for achievements, Apple told us to resubmit with final art assets.

Their system is far from perfect, but after lots of rejections I can see its value.

Be really careful when requesting an expedited review. We learned this the hard way.

Your app update goes into a different review queue, and people who are not be familiar with previous versions of your app will review it. They will pick on things that previous reviewers had already come to terms with (either through discussion with you, or lengthy review).

Edit: more info, this happened with our app "Codea". Because it was a coding app for iPad, the initial review was extremely lengthy (over a month). When we had to urgently update due to a trademark issue, our "expedited" review also took over a month — as if it was being reviewed for the first time (despite having past updates reviewed quickly). It didn't seem to matter that the existing version was available all the while on the App Store, either.

Apple's review process is incredibly weird.

I nearly pulled my hair out when a sandboxing bug in 10.7.5 broke one of Apple's own apps.

http://imgur.com/F5roY1h

https://support.apple.com/kb/dl1599

So, for iOS, the app store and apple's possibly arbitrary gatekeeping is the only way to get an app to your users.

For OSX desktop... what are the benefits software publishers get, just marketting/visibility, basically?

Not having to run a shop.

How many users are still using 10.6? Couldn't the description be updated with a note specifying the temporary issue with 10.6? Seems like that would be a lot simpler than taking it down...

3 out of 4 of the macs in my offices were on 10.6.8 until last week (now it's only 2).

Edit: I figured this is a pretty useless stat. Here's some better ones from the Unity Web Player installations, Q1, 2013

http://stats.unity3d.com/web/os-mac.html

10.6 is just under 30% of the install base for these guys

Anybody with a single core intel machine (there's two of them in my house right now, an old Mac mini, and an old PowerBook that still works fine)

The Mac mini 1.5Ghz, a 7 year old model, is the only modern Mac with a single core Intel CPU. The PowerBook line had PowerPC microprocessors, and the MacBook Pro series never had single core CPUs. Apple did however release MacBook Pros with Core Duo CPUs -- OS X 10.7 requires at least Core 2 Duo. by 2007, all models had Core 2 Duo CPUs.

Both of my co-workers and my managers are still on 10.6 because of perceived concerns about 10.7, and the fact that our IT department wants to do a full reimage to upgrade to 10.7 or 10.8 rather than just running the upgrade.

I actually have more problems with 10.8 than I did with 10.7, so I understand some people's reluctance to upgrade.

reading through this, it all sounds familiar. then it hits me. this is like trying to get a grant reviewed by the NIH!

How's that freedom working out for ya?

What does this even mean?

apparently it means downvote me

Reply on news.ycombinator.com