All Devices Enough’s enough

[simonsi]

I think the amount of bugs average users experience and perceive, has gotten relatively speaking completely out of control.

How many actual data points is that based on, really?

 

[newellj]

And half way through office there would be a single disk with a corruption

I remember back in 1987 (I think - we had just moved into a new house) using Quicken on a dual-floppy Mac, the data disk failed. There wasn't really any such thing as a backup at that time. The reason I remember it is that we weren't keeping paper records (the computer was magic!) and it took three months to reconcile the checking account because we had to wait for all the checks to clear. So the good old days weren't necessarily quite as rosy as they sometimes look through rose-tinted glasses.
[automerge]1571830131[/automerge]

How many actual data points is that based on, really?

That's a very fair question. It makes me wonder: does Apple have any grasp on these numbers? Which makes me wonder in turn how many users turn on or off analytics reporting to Apple (and does that even report ordinary bugs)?

 

[newellj]

i don't even want you to elaborate on this because i have a pretty good guess on where will it go, and its gonna end up with one of us being muted

Not to mention that even with the most generous interpretation, there are no facts to support that contention.

 

[secretk]

Why should it be that way? That's been answered repeatedly now. It's an extremely complex system written by humans that make mistakes. Just because they're Apple, or have a public beta, or have lots of money doesn't give them access to an exclusive pool of perfect humans who make no mistakes. Software development works the way it works and neither extra people or money can change that. It takes one woman 9 months to make a baby. How long would it take with 9 women? 9 months. Not every problem can be solved with added resources. It takes one guy 6 months to build a house. It takes a crew of ten just a month to build the same house. It takes a crew of 10,000 people many years to build the same house because there are just too many people. More people isn't always the answer. Apple being richer than pretty much every other company out there doesn't mean they get to skirt around these basic ideas.

That I agree with . More people is not always the answer. Especially when more people means more new people that would have to be first trained and onboarded.

Why have a beta at all? Why are you assuming having the beta doesn't help? Wouldn't it make more sense to assume Apple wouldn't spend the money putting it all together if they got nothing out of it? I presume that they get valuable feedback or they wouldn't waste their time. They love secrecy, so they're not doing it for fun. It seems more natural for them to do things that make sense rather than doing things that don't make sense. Companies generally don't become as successful as Apple by making bad decisions. Just my opinion of course.

Beta versions are not actually for end users to do tests. Beta versions are needed for Developers to get a sense of the new iOS version and to be prepared to adapt their apps to the new iOS verison.

What release updates that aren't fully ready? Business reasons. Also, bugs are inevitable. If they wait till it's bug free, we'd get updates every 10 years. Then 10 years worth of features would be added which would take 20 years to debug. So where do you set the bar? How many bugs are acceptable and still release the software? There are always tradeoffs. Do you keep the new feature from 99% of your users because it's broken for 1%? Things aren't black and white. Neither are they simple questions to answer when you get down in the trenches. We just did a major upgrade to our claim processing system this past weekend. It got put off several times because of bugs. But there were other deadlines pushing in the opposite direction which meant it absolutely had to go this past weekend, ready or not. It all went well in the end. But it's a lot more complex than: bugs = no release. It seems like a simple question, but it's far more complex when there are dozens, if not hundreds or thousands of things that need to all come together and pass before you're ready to release. We probably had a dozen or two departments involved in our upgrade. Each of them probably had their own go-live checklist of things that needed to be done. Some of those tasks may have taken multiple people to complete. The IS department had probably 15-20 different teams in the department doing the same. Getting 100% of the people to be 100% done with all their tasks doesn't always happen. Then the question needs asked: do we go live or not? There were probably over 200 people ultimately involved in one fashion or another to make things go smoothly. The sites my team launched will be used by thousands of people. Apples software will be used by hundreds of millions of people. I bet their releases are more complex than what we just went through.

Hope that helps you see things from the other side and helps answer your questions.

I agree bugs free version does not exist. That being said there are also quality gates. I mean in terms of how many severe/critical bugs are allowed in a version to receive GO status for release. Should there be any critical/serious bugs. What about regression tests? Did they pass or not. If they did not pass what kind of regression bugs are open.

I can see that you have background and expertise and in Software Development. So do I. While I understand that iOS is a complex system I also see something really concerning. Different people report different bugs for the same software. This means that somewhere down the line the architecture went under the drain seriously. A system that has that many specific bugs that you need to know the specific steps to reproduce shows technical debt. And ultimately this is what Apple needs to work on. Otherwise we will have releases like this one in the future too.

 

[TheLocNar]

How bout if the industry just not release new products until the bugs are worked out?
Not a good idea?

We’d still be communicating by smoke signal.

 

[digitalexplr]

Me too, DR Dos, Windows 3.0, the 3.1 then 3.11 for workgroups now that was a big jump
I really miss Office that came on a ton of 3.5" FDD's

If I remember correctly I believe Office came on 21 or 22 FDDs.
[automerge]1571842754[/automerge]

And half way through office there would be a single disk with a corruption

Almost without fail! And it always seem to be one of last 2 or 3!

 

[4492865]

That I agree with . More people is not always the answer. Especially when more people means more new people that would have to be first trained and onboarded.

Beta versions are not actually for end users to do tests. Beta versions are needed for Developers to get a sense of the new iOS version and to be prepared to adapt their apps to the new iOS verison.

I agree bugs free version does not exist. That being said there are also quality gates. I mean in terms of how many severe/critical bugs are allowed in a version to receive GO status for release. Should there be any critical/serious bugs. What about regression tests? Did they pass or not. If they did not pass what kind of regression bugs are open.

I can see that you have background and expertise and in Software Development. So do I. While I understand that iOS is a complex system I also see something really concerning. Different people report different bugs for the same software. This means that somewhere down the line the architecture went under the drain seriously. A system that has that many specific bugs that you need to know the specific steps to reproduce shows technical debt. And ultimately this is what Apple needs to work on. Otherwise we will have releases like this one in the future too.

I think that Apple's problems are exaggerated by a couple of things:

1. Their apps are too intertwined with the OS. Why does the Mail app need to follow the same release cycle as the OS for example? Or the Photos app? Decoupling them means that they can be released when they are ready, not because of a some hard deadline.
2. Release cycles are now 100% tied to the release of the next iPhone. Everything — iPadOS, macOS, iOS, watchOS — everything is linked to the iPhone timing wise. And again, because all services are intertwined with the OS, it has to be released all together. Otherwise, the new Reminders could have been released in, for example, May.
3. Stretched too thin. Apple wants to do it all, and resources are spread thin. More people doesn't solve the problem — the coordination of those people is actually even harder, especially if everything is still deeply integrated into the OS. Their services, like TV+ are still tightly integrated into the OS, forced the breakup of iTunes and created a whole new myriad of problems. Apple is turning into the sum of Dropbox, Netflix, Spotify, Microsoft Office, Venmo, MasterCard, Whatsapp... and that still excludes the OS which is massively complex as well. It's a lot.
4. Services not developed independently. The whole iCloud Drive saga could have been prevented by simply developing it separately, or giving it a separate release cycle.

I know it's easier said than done from the sideline, but... Apple is charging premium prices. And very full of themselves.

 

[secretk]

I think that Apple's problems are exaggerated by a couple of things:

1. Their apps are too intertwined with the OS. Why does the Mail app need to follow the same release cycle as the OS for example? Or the Photos app? Decoupling them means that they can be released when they are ready, not because of a some hard deadline.
2. Release cycles are now 100% tied to the release of the next iPhone. Everything — iPadOS, macOS, iOS, watchOS — everything is linked to the iPhone timing wise. And again, because all services are intertwined with the OS, it has to be released all together. Otherwise, the new Reminders could have been released in, for example, May.
3. Stretched too thin. Apple wants to do it all, and resources are spread thin. More people doesn't solve the problem — the coordination of those people is actually even harder, especially if everything is still deeply integrated into the OS. Their services, like TV+ are still tightly integrated into the OS, forced the breakup of iTunes and created a whole new myriad of problems. Apple is turning into the sum of Dropbox, Netflix, Spotify, Microsoft Office, Venmo, MasterCard, Whatsapp... and that still excludes the OS which is massively complex as well. It's a lot.
4. Services not developed independently. The whole iCloud Drive saga could have been prevented by simply developing it separately, or giving it a separate release cycle.

I know it's easier said than done from the sideline, but... Apple is charging premium prices. And very full of themselves.

It is obviously in the architecture itself. The fact that they were saying over and over again that support for external devices would be security risk shows you that they had no way to separate the kernel from the rest of it.

To me it seems that the initial architecture was not setup for the features they are doing currently. In other words they have not separated them because the code is not written in a way to separate them. It seems like they have to rewrite a lot of things and it seems that this would be big effort (in terms of DEV/QA time).

 

[MEJHarrison]

Beta versions are not actually for end users to do tests. Beta versions are needed for Developers to get a sense of the new iOS version and to be prepared to adapt their apps to the new iOS verison.

That is mostly true, but the fact that a public beta exists kind of does imply that Apple is hoping to get feedback from users.

 

[C DM]

Six Reasons Why iOS 13 and Catalina Are So Buggy - TidBITS

By most accounts, the release of iOS 13 and macOS 10.15 Catalina have been troubled, with numerous significant bugs making it past Apple’s internal testing and the public beta phase. Former Apple engineer David Shayer explains the underlying reasons these releases have had so many problems.

tidbits.com