Apple’s culture of secrecy wore down Swift creator and Xcode lead Chris Lattner | Hacker News Reader

Apple’s culture of secrecy wore down Swift creator and Xcode lead Chris Lattner

9to5mac.com

234 pointsitg104 comments

The 'extreme secrecy' would be an advantage if they actually managed to surprise their customers. The reality is, over the last couple of years they haven't really created something groundbreaking. Why would you want your own employees to not know that the next big thing they're working on is a minor product update. Also, despite the extreme secrecy, every product announced since the iPhone 5s has been leaked in advance. Factory photos, hands-on reviews of cases, photos of chips, etc. In other words, their employees suffer from the secrecy, but the outside world (who they want to surprise) aren't that excited about their launches and already know in advance what's coming up. So... what's the point really.

While at Apple, I once had to get director approval to get an unreleased version of iTunes for some feature I was working on. It took a couple weeks to plinko up the management chain.

When I finally got the build the interface was a slightly lighter shade of grey. It was fucking stupid.

Swift itself was a huge surprise.

Secret by default - they never had to share it with 100,000s of people building it all.

A low value secret though, most Apple customers don't know or care about swift/llvm/etc.

Swift is not a consumer product though, a leak would have had very low impact for them.

I sometimes get the impression that all the secrecy, and leaks, and rumors, etc. are a game they play to generate anticipation and excitement about their products. There is no surprise when they announce, but we are all awaiting the big event anyways.

It also generates billions of dollars of free marketing as everyone creates stories and pictures and live blogs at the keynote.

That has tangible value to Apple.

> It also generates billions of dollars of free marketing as everyone creates stories and pictures and live blogs at the keynote.

Also the hundreds of thousands of dollars or millions of dollars that this speculation generates on tech sites (that follow Apple news and rumors) through ad revenue and affiliate revenue.

Apple is a victim of their own success. It's a dilemma really. Their culture vs public interest.

I am sure they are exploring new spaces like VR, AR, cars, etc but they haven't finished a decent product yet so they don't announce it. They are the opposite of Google, where they announce every new project only to kill it later (Ara, Fiber, Glass, etc).

They don't really like to announce a product before it's ready. The root is their hardware background. They need to nail the product in the first iteration or it will fail. Move fast and break things don't apply to them.

Don't Forget Wave !

... I really liked Wave...

On Apple's culture: Ship has sailed. Tim doesn't have the gravitas to ensure a new 'Wow!' product is going to ship. He's a damn good manager, but not a superb innovator. Unfortunately, Apple is at it's heart, a 'Wow!' company. I suppose Apple has 'topped off' now and is a mature business state. This is a good thing though! More room for innovations!

And GOOG-411 (didn't have a smartphone yet when they killed this, it was still useful), Google Labs (public access to experimental features in Gmail and other services), Buzz (jk nobody misses that), Reader, Code Search, Q&A, iGoogle portals, and plenty of others.

Google Questions. I know it's been long ago and there were problems but when you needed it for real it worked wonders.

Didn't know that Google Labs was shut down. Some of the Gmail Labs were useful - and some have probably gone mainstream by now. The one that reminds you if you use a word like "attached" in your email, but have no attachment, comes to mind. I tried at least half a dozen of them, may have only kept using one or two. Still, the Labs idea was good.

Undo send originated there as well.

Right, good point. That has saved me a few times from accidentally sending an incomplete email - luckily not anything worse :)

I've been using it for years, so I did not realised they had closed Labs in the interim.

"Wow!" products usually don't happen overnight though. Take the iPhone for example! Apple was playing around with ideas for tablets, handhelds, PDAs, and other stuff for decades before they launched that. A lot of "what if" prototypes were made and dead end projects ensued.

Google kinda had the right idea with 20% time, I think. In order to make truly ground breaking ideas, you have to have a little bit of margin for playing with ideas. If you make your entire business about trimming fat, making synergies, and maximizing the efficiency - squeezing it out of everyone involved, you're no longer an innovator, you're just an assembly line.

The trash can Mac Pro was not leaked in advance (and by leaked I don't mean speculation that Apple was updating the Mac Pro, but actual pictures of it). Also I believe photos of the Watch didn't emerge ahead of time. Know why? The Mac Pro was built in the US, same for Watch prototypes.

People complain about leaks, but in reality the only things that leak from apple are products that are shipping immediately after announcement. New MacBook Pro, one model shipped day of. iPhone 7 - two weeks later (And pretty much every iPhone before it.) Apple Watch was announced months in advance of its actual release. Same with the original iPhone (And iPad as well IIRC.)

It is pretty much impossible to control leaks when you are manufacturing at that scale, there are just too many low wage workers in the chain that have no stake in the product. The Mac Pro also had the advantage of being a really small run in comparison to the iPhone.

Both of those are also not as mass-produced as the iPhone though; I mean they churn (churned?) out a million of those a day. That means there's thousands of people that get to see parts of it before the announcements, spread across multiple factories, and with people having financial incentives to smuggle pictures or even parts out of the factory and sell them to the media outlets.

The Watch was. Quite a bit. There were even photos of it on the wrists of various people before it was announced. It was another example of those not-a-secret secrets

The majority of leaks are by Apple themselves. The same goes for other smartphone brands.

When you can't control secrecy that much, you do exactly what they do - you submit multiple false leaks! And you are terribly not right that they do not surprise anyone. They do. The fact that you are technology guy that is somewhat informed doesn't change the fact that the ordinary user like my mother doesn't get surprised.

They could surprise your mother without the layers of secrecy, assuming you use as an example of someone who doesn't read tech blogs and finds out about products when they're advertised or discussed in broadsheets

> Factory photos, hands-on reviews of cases, photos of chips,

Someone forgetting it in a bar, leading to criminal conviction [1] of the person who found it.

[1] http://www.businessinsider.com/brian-hogan-lost-iphone-4-red...

You neglect to mention the part where the person who found it sold it rather than try to return it. Should that not be a crime?

selling abandoned property you found ? no thats not a crime. If this is such a serious "crime" , why wasn't Gizmodo charged for possession of stolen property ?

You know what's a crime ? The fact that Apple has expedient access to the police when they negligently lose a $400 phone at a bar, instead of being told to "fill out a missing property report online" like everyone else when they call to report something missing or "stolen".

> selling abandoned property you found ? no thats not a crime.

Is that true? I'm not a lawyer, but I'm pretty sure you have to own something in order to sell it legally... and I don't think you get to claim ownership of every "abandoned" item you "find".

If it wasn't a crime, there wouldn't have been a conviction.

And getting the police to act on information you provide yourself is not the same as filling in a report online asking the police to find your missing phone.

You neglect to mention the part where the person who found it actually tried to return it.

http://gizmodo.com/5520438/how-apple-lost-the-next-iphone

Chris Lattner on Twitter:

>My decision has nothing to do with "openness". The "friend" cited is either fabricated or speculating. Folk just want to make [apple_emoji] look bad. [angry_face_emoji]

https://twitter.com/clattner_llvm/status/819974025371787264

It's OK. We're in a post-fact era.

Even before reading his tweet, it seemed pretty clear it was speculation from the quote!

> "Eventually I know that can really wear people down."

That doesn't say s/he knows it wore Chris down, just that it could have.

If only the headline reflected that.

Blogs and shitty news sites like Business Insider don't practice standard due diligence like newspapers used to. They just print whatever they want, it's ground zero for fake news. There should be more consequence for sites printing things like this, but there isn't, and that's why we are suffering with misinformation.

There are consequences if the person affected chooses to take action and meets the legal definition of libel or slander.

Stop misusing the word "fake news."

Business Insider is not the New York Times, but even they would not openly fabricate a source [1]. My guess is they ran a story with an actual source, but the source was speculating.

[1] Outright making stuff up is the definition of fake news. 100% made up.

How is speculating not equal to fake?

Because it was clear from the article who the source was and that this was hearsay. Also the article updated itself with a rebuttal from Lattner.

Fake news would not talk about who their source is. Fake news would not publish Lattner's denial.

There is a difference between fake news and second-rate journalism. If someone doesn't think there is a difference, I'm guessing they also see the world in black and white.

Stop making up your own definition of "fake news." You're not some authority, you're just a random Internet user like everyone else.

Speculation (if it's clearly marked as such) - is just an "entertainment" piece - for example, you could write in your blog what are the next features for iOS without any clear sources - that would be just speculating about.

Fake news - news that has wrong info as facts or provide something as sourced when the original source never said that.

I think that's rather huge difference (speculation from some outsider can still by chance be true, fake is by definition false)

As a webdev it's occurred to me before that at almost every conference I've been to there have been employees from Google, Mozilla and Microsoft happy to chat about their browsers. They're also active online, via Twitter or whatever.

Apple, on the other hand... I've never once met a Safari dev. How does that help anyone - including Apple? Feedback with the people who develop on your product can be very valuable, and it seems Apple aren't interested in any of it.

Apple expects you to go to WWDC to meet employees and engage in dialogue. This way Apple controls the setting and the information being disseminated. (One could call it the "Mountain comes to Muhammad" approach.)

It worked pretty well in the 2000-2009 period when you could just buy a WWDC ticket like any other conference pass... But after the iOS App Store exploded, tickets became impossible to get because they started selling out in minutes.

It also seems like a downer for Apple employees. Do they ever get to leave Apple's campus?

They are around at conferences. They just don't make a big deal out of it and generally don't appear on stage.

Well since WWDC has been in Moscone for ages, and San Francisco isn't Cupertino, yes?

Hardly. When I worked there WWDC at Apple was like lunchtime in Italy: everything shuts down with an audible clank as everybody floods conference rooms to watch the presentations.

>It worked pretty well in the 2000-2009 period when you could just buy a WWDC ticket like any other conference pass...

It may have worked well for those who are happy to replace continued, in-depth, online discussion with travelling half way around the world to exchange a few words with someone after paying a ton of money for it.

you pretty much nailed down apple culture with that.

Nice straw man there.

Apple never replaced their existing online discussions with WWDC. You can continue to talk to WebKit developers using the mailing lists any many Apple developers continue to be present within the Developer program forums.

Apparently (or rather allegedly), some Apple developers feel constrained in what they are allowed to discuss publicly. I'm merely saying that the availability of WWDC conference tickets doesn't fix that. For me anyway.

I never said that Apple actively replaced anything. That's your straw man.

This has also bothered me for a long time but they've actually been doing a much better job of this over the last couple of years. I had never met a member of the Safari team either until this year where I got to chat with one, at Google's Chrome Dev Summit no less.

They also revamped the webkit.org site and started releasing developer previews of Safari every two weeks.

They still have long way to go in this area but they seem to be moving in the right direction so I'm hopeful.

> Apple, on the other hand... I've never once met a Safari dev.

Don Melton (creator of WebKit and Safari) is super cool -- if you run into him at a conf he is full of amazing lore and wisdom.

Apple is a hardware leader, but IMO their culture is toxic for building software.

Yes, they have great software achievements in OS X, iOS, Safari, etc. however those happened in spite of their toxic culture, for other reasons. As an inspiring market leader, Apple is able to attract a lot of talented software developers and they also have huge pockets, plus whatever faults they have, it's in Apple's DNA to ship polished products.

However I believe building software is their weakness, because building great software is also about collaboration, about building communities, about establishing standards, about reuse and composition. It's what Google and Microsoft are good at.

I think their current culture is toxic for both. Their hardware is polished but IMHO has prioritized form over function at an extreme level, and seems to be developed in complete isolation from the user base.

That's ironic, Apple IMHO is one of the few companies that takes UX seriously.

Such form over function, the current MBP is faster, slimmer and with much more battery life than any other.

The difference is that Apple can make it beautiful and funcional, and others just can't.

Is their software actually any better than anyone else's? Or is it that their software is only expected to work on their hardware so they can fully optimize it for their hardware and it seems better? I have a MacBook and an iPhone, but you would be hard pressed to tell me macOS and iOS are truly better than the alternatives.

One could write a book on that topic. My stance is that they prioritize what and how they do differently than most competitors, and that would be end user experience and polish.

E.g., the iPhone UI's smoothness was a result of their prioritizing smoothness over other factors, like ease of implementation. If you look at how they did it (view objects are backed by GPU rendered layers), it's not really developer friendly. Because developer friendliness is not their priority, at all.

So, what you should get from Apple is the most polished experience, and they exert a great deal of control to make sure that the designed experience actually arrives at the end user. They got this really right with hardware, but they still struggle with the software part. Sure, most software Apple produces is beautiful and works really well. But things are crumbling. Looking at the various fiascos - Maps, Me, then iCloud Syncing, increasing number of problems with OSX quality - this controlled approach does not seem to work that well on software. You can control all possible states and usage scenarios of a physical device, but good luck with that on a OS consisting of 4 billion bytes, or a massive neural network (Siri). IMO they need to let go and open up at the software side.

That's ridiculous, Views are backed by objects to the core.

You can even take a PDF screenshot of the whole screen and it will be pixel perfect.

Their API's are easy to develop and coherent, yes, they are different than other API's, but more and more other API's start to copy Apple, like Android did and does, and it's the norm right now.

I also have a Mac and a iPhone, and I disagree with you.

As someone who worked at Apple I can say you are talking rubbish.

Firstly, Apple doesn't pay all that great. Other startups and companies pay much better. And as a software developer the culture is actually pretty good but obviously depends a lot on what team you are in. It's an interesting mix of enterprise and startup with an amazing history. Teams are smaller and more cohesive. The frustrating part is the lack of collaboration between teams but that is hardly specific to Apple.

Secondly, Apple absolutely builds communities where required and contributes to standards. Where on earth did you get the idea they weren't doing this ?

If the secrecy wore him down, why is he moving to Tesla which is equally secretive? I'm not convinced by this argument.

It's not. the linked BI article speculates on other possible reasons.

>why is he moving to Tesla which is equally secretive?

In what way is Tesla secretive? They make all kinds of announcements about things that aren't actually "real" yet.

sure. but they share none of their engineering insights. As Mr. Lattner is an engineer, I would assume that sharing engineering-related stuff is more interesting than pre-announcing marketing information.

Didn't they open source [sic] all their patents? [1] I'd call that being pretty open.

[1] https://www.tesla.com/blog/all-our-patent-are-belong-you

Patents are open by design. "Open sourcing" a patent just means Tesla is agreeing to license the patent to anybody and won't sue you for using it. That's still a marketing/business decision, not an engineering one.

Wasn't that mostly supercharger patents? There's a shared network effect that helps every electric car maker if they all use the same charger standard...

I also not convinced. Chris was with Apple for more than a decade. He must have adapted to the secrecy culture after so many years.

Of course secrecy is not the sole reason he left. Swift was developed 100% in secrecy for the first 3 years of its life, and even after that, Chris was able to navigate the open-sourcing culture shifts required of Apple to make that a reality, no small feat.

Though boring and lazy, it's extremely popular for the hivemind to take aim at the throne. Note it's why any article critical of Apple (or every damn blogger's tbMBP/macOS -> Linux article) goes immediately to the HN frontpage, while interesting technical discussion re: Apple products or news showing they're still the market leader remains barely discussed.

> any article critical of Apple ... goes immediately to the HN frontpage

Considering that this article (as I write this) has 184 points, and is sitting one spot below a slightly older article with 35 points, there seem to be two sides to that story.

Tesla is much smaller than Apple. They may be secretive with the outside world but not between internal compartments/silos.

Tesla doesn't strike me as secretive at all. Elon Musk is constantly tweeting about future plans, they frequently discuss new features before they're released, and they talk about new products and new product lines years ahead of time. For example the Model 3 was first talked about a decade ago, and they have explicit public plans (up to and including allowing customers to buy the option right now) for fully autonomous cars which will no doubt take years to actually come to fruition.

Tesla in many ways has the opposite problem that Apple does with secrecy: they talk so much about what they're going to do, and it doesn't always work out, so people start to mistrust what they say.

I don't know how secretive Tesla is or isn't at the moment, but with their aggressive approach to rolling out autopilot software, and the results that can already be observed, it seems likely to me that the autopilot division will see a significant amount of litigation over the next few years, and it would seem equally likely that this would lead to considerable caution in public communication.

BI's source: "someone in Lattner's circle of developer friends"

There is nothing to indicate this isn't some random guy who's met Lattner twice and has no idea what he's talking about but wanted to feel special by being a "source" to BI.

Apple wouldn't have a problem with extreme secrecy if their compensation went above and beyond the companies they compete with for engineering talent. But their compensation is basically equivalent, or even less.

So except for some particular affinity for an Apple product or team, or variance in interviewing success, I typically see undergrad/masters/PhD graduates vastly preferring Google/Facebook offers over Apple (I use these as an example because each company typically has a standard offer for each group).

and they work more hours... for that lower compensation

The source publication is a better read: http://www.businessinsider.com/how-apples-culture-of-secrecy...

And it's 100% pure bullshit.

Sorry, iHaters.

> BI also pointed to an incident in 2015 when Apple’s entire networking team quit over Apple’s refusal to allow them to participate in an industry-wide network security forum known as the Open Compute Product.

How big is the "entire networking team", how many people are we talking about here?

And most importantly, is this why my Mac crashes when using NFSv4?

Culture matters. In 1997 I had offers from both Apple and Sun. I was working at a Macintosh ISV which was very badly impacted by Steve cancelling the Mac clones. I knew from friends about the secrecy and siloed environment at NeXT and in Steve's Apple. While I had respect for Steve's accomplishments and vision I made a very conscious effort to choose a different work culture.

I ended up working at Sun (and eventually Oracle) for 16 years. It may not have been great for all those years but I can't imagine that I would have lasted at Apple that long and I certainly wouldn't have had the mobility between projects that I did at Sun.

Doing it again I would still choose the company with the culture that I felt was the best fit for me with that company's potential being secondary.

This seem more like speculation unless Chris actually says something. Being frustrated in a giant company is something I understand 100%. Moving up the food chain to be more in control is something I also understand which often means going elsewhere.

I'm not saying it wouldn't be a problem for some, but how would Apple's "culture of secrecy" affect Lattner? Apple didn't just release Swift as open source, they released the entire commit history even before it was open source, the future planned roadmap of Swift is discussed openly on a public email list, and the whole Tool chain down to llvm is completely open source.

FTA:

He always felt constrained at Apple in terms of what he could discuss publicly — resorting to off-the-record chats, surprise presentations, and the like," the person told us. "Similarly, I know he was constrained in recruiting and other areas. Eventually I know that can really wear people down."

Running an open source project with lot of contributors requires open communication. Recruiting constraints is self explanatory.

That's true for every company. I work at a private mid sized unknown company as the developer lead. I can't go out an publicly speak about what's going on at the company or their roadmap even to employees at other locations or sometimes to my own team. Even though I'm not officially management, I'm privy to information for planning purposes that the company may not want to get out

Do you think that Tesla is going to allow him to open source the self driving car software he will be working on?

Even though he's not working at Apple, do you think the head of the Swift project is going to refuse to take a pull request from him?

Well not every company. But I think you are oversimplifying it. Culture of secrecy in Apple's case most certainly means you can't do anything public at all without going through a bunch of legalese, approval chains and various other roadblocks. Apple for example did not allow AI researchers employed by them to publish papers - they hardly got any researcher interested and they have allowed publishing of one paper so far.

Tesla on the other hand seems to be at least making noises about Open Source spirit. It looks like Musk doesn't care about keeping things secret as much as Apple does - https://www.tesla.com/blog/all-our-patent-are-belong-you. It is quite conceivable that Elon Musk will allow Chris more openness to contribute to LLVM/Swift and whatever Self Driving car stuff he is working on. I mean what has Tesla got to lose if LLVM is improved or even if Tesla's self driving software is open sourced (not the data mind you!) and many others adapt it and improve it?

I don't know about Apple at all, but I've worked in environments where there was a heavy security culture and super tight protocols re: external communication.

In short, it's really stressful and frustrating. In my case, our external communications required high-level approval unless there is a specific protocol in place.

You get to a place where it's just too stressful.

Agree. Swift is (one of) the most open source project apple has ever released. If anyone could complain about apple secrecy, lattner was probably the last one. Something doesn't add up. I think this is all speculation, and that something else more concrete was the real reason.

Maybe something like lattner wanting to go server side / linux full speed with swift, and apple not wanting to allocate enough resources on that ? Or people inside Apple still sticking to objective c too much ? Anyway, i'm really curious to see how this story will evolve.

I imagine trying to run a consensus-driven open source project from within a super-secret topdownocracy like Apple is next to being an impossible task. That Lattner did it for as long as he did is impressive.

Apple made a mistake choking off iOS safari they're screwing up like MS did in 1999, they're I'm the same position now as MS was when Jobs came back. They're turning iOS into a second rate system by refusing to implement many html5 APIs like webrtc. Similar to the browser wars in 99, the original iPhone had the best mobile browser of its time but in 10 years they've lost that edge.

I think what people forget is that engineers are very curious by nature. They don't want to work on the same things forever. I don't know, but Mr. Lattner probably took Swift was far as he wanted to go with it and now it was time for a change.

I would still like to know who's replacing him and what the future is for all of this, particularly Swift given how much Apple was pushing it. As far as I know, it was Lattner's baby.

Who's running things these days?

From his leaving post on the swift mailing list: https://lists.swift.org/pipermail/swift-evolution/Week-of-Mo...

> I’m happy to announce that Ted Kremenek will be taking over for me as “Project Lead” for the Swift project, managing the administrative and leadership responsibility for Swift.org. This recognizes the incredible effort he has already been putting into the project, and reflects a decision I’ve made to leave Apple later this month to pursue an opportunity in another space. This decision wasn't made lightly, and I want you all to know that I’m still completely committed to Swift. I plan to remain an active member of the Swift Core Team, as well as a contributor to the swift-evolution mailing list.

Thanks very much indeed. I do wonder if there is a severe brain drain occurring at Apple.

It has long struck me as an intense, unforgiving place to work, with the upsides that you'll probably make a shit ton of money and possibly work on things that affect many people. Not the kind of tradeoff I would personally make, plus it's in the Valley which I wouldn't like either. I can certainly see it getting old after a time. Especially now, when the bloom is off the rose a bit.

I wonder what % of Apple employees hare secrecy culture and want to see it die versus the % who were hired due to a compatible attitude towards it and thrive in that kind of culture.

I'm very interested in things that make a massive ship very hard to turn. If they wanted a drastic culture shift, would they have to replace a lot of managers and leaders?

I don't think anyone likes the secrecy. I don't know anyone who doesn't want their name attached to their work.

It seems now without Jobs, Apple is just repeating mantras without asking themselves "why?".

When the company was working on iPhone, iPad, such secrecy made sense. For other stuff, no.

Or maybe he got bored and had a great offer from Tesla?

Secrecy has always been an advantage. If he doesn't like this, of course, he should leave. Nothing wrong with the two things.

Keeping your best people has always been an advantage. When you keep telling them "if you don't like it, leave", sometimes they take your advice and the problems they saw still exist after they're gone.

Reply on news.ycombinator.com