Google Hardware makes cuts to laptop and tablet development, cancels products

https://system76.com/laptops

As for the so-called "Apple tax", I just don't buy into this. I mean I miss the glory days of the 2011-2012 Macbook Airs. These were so cheaply priced they destroyed the competition because no one could produce a similar spec laptop at that price point. But then the numbers men decided dropping the average ASP for Macbooks was "bad" so they had to invent new features nobody needed or wanted to drive up the cost (Touchbar being the poster child) but for the hardware you get, it's actually pretty good. Or it was.

> Is it impossible to make a laptop better than Dell or Apple?

It's clearly not trivial.

> Google: how much goodwill would you receive through developer love?

Based on the bubble that is HN the answer is clearly "absolutely none". Just look at Chrome adding DDG as a search engine option. Where in that thread do you see people saying "this is great"? No, the top comments are littered with theories about how this is just an effort to avoid antitrust or other such ulterior motives.

I guarantee you that if Google produced the laptop that you and 22 other people would buy you'd hear nothing but complaints about it.

An old laptop will run any Linux distro flawlessly and be equally snappy (as in, low-latency disk reads, fast boot times, many applications open) as a current-gen one if you put in a fast SSD and max out the RAM.

I know that this topic is controversial, but it strikes me as odd how many people seem to equate software engineering with a monster laptop, when many of us get lots of work done on smaller, less beefy client devices, offloading the real heavy lifting to a machine that can handle it.

Don't Chrome books provide you what you need for software engineering?

* I once started a set of experiments, drove off for a short 3 day vacation, and returned to see everything was still running smoothly. Which frankly I hadn't expected.

I don't know about impossible, but they've both been building laptops for a long time and they have large market shares, so if it was easy to build better laptops than theirs someone probably would have done it by now so it's probably really hard.

For me, wouldn't it be great if laptop shells had a universal connector you could plug the mainboard of choice via an edge connector, be that ARM, X86, MIPS,????. Certainly make laptops more upgradable and save upon production costs with a standard laptop mainboard form factor.

Thinkpad Carbon X1 6th gen (or 7th now?)

I only wish it was sold without Windows so I can save $30-40 on the win license.

Other than that, freaking perfect laptop that rivals a macbook (I will admit that the mac trackpad is slightly better... but this one is pretty damn good)

But that's nothing compared to the new stuff. It runs containerized linux with a great abstraction layer that connects GUI calls from linux into the beautiful hi-dpi Chrome OS desktop environment (I think it's still called "Aurora"?)

It's still in beta, but they're almost ready to add: - USB support (obviously important here, but they want you to be able to develop android apps on it) - Sound from linux container -> hardware speakers (I believe this is in beta channel now though?) - graphic acceleration (also is in early testing, canary branch maybe?) - backup/restore functionality (lose your laptop? get a new one and restore your container)

All of this in one of the sexiest laptops with the best keyboard of any of the ~5 machines I own for $999

I've been using it as my main development machine (a complicated electronJS app that windows substate on linux choked on) and all the backend/database work for over a year and I love it. You could get it on sale for $699, too.

Main complaint? If you need to do a lot of work in VM's. You currently can't spin up a Windows VM or anything very easily, but rumor has it they're working on letting you dualboot in the future.

I freaking love this machine. They should pay me to evangelize but I promise they don't :)

What does 'last' mean? Any laptop should 'last' 8+ years.

> have a high spec (i9 etc);

I really don't see the point of high-spec laptops for development use. Higher end, high power components just make the batter life worse - reducing the entire point of a mobile device. And you're paying a premium for the mobile version.

A much better option is to have a beefy linux desktop that you perform work on remotely. There is very little reason to locally compile code on a laptop. The economics work much better IMHE... If you have a $2000 budget for instance, you can build a $1300 desktop and buy a $700 laptop which gives you much more power and flexability than buying a single $2000 laptop. So yes, I do think many Google engineers just need a portable terminal to connect to beefy VM's remotely. If you are often working on a mobile internet connection, Mosh works very well for high latency connections.

The $2000 laptop makes sense if you want to game while traveling and/or are using Windows where offloading work to a remote machine is a bit more clunky than just using X11 forwarding and/or tmux over SSH.

I have two still serviceable MacOS laptops but I don’t think I will buy any more. But, I love Apple Watch and iPhone - probably be a customer of those products for a long time.

Admittedly it's about 6 years since I interviewed there, but when I did, there were a number of engineers in my loops who were using chromebooks.

They found it more than sufficient for their needs.... I can see that, I guess. I'm assuming the bulk of the "heavy lifting" happened outside of the context of the chromebook.

I passed up the opportunity to buy the new Pixelbook as the original Pixel still works nicely and I don't need Android on it particularly.

However, I think that Google have gone wrong with ditching the Pixelbook. There is a definite market for a decent developer machine that you can run Linuxii on that isn't a hack. Homebrew, Virtualbox, Ubuntu in Windows isn't ever native and the instructions are always less straightforward than with a stock Ubuntu or other regular distro.

From recent news there seems to be a small army of pissed off Mac users that want simple things like an escape key. There are also plenty of developers that don't really need Windows. The Dell developer XPS specials are very nice but the original Pixel and its descendants get a few things right, the 'no expense spared' 3x2 screen, nicest keyboard ever, best speakers ever and best trackpad going works for me.

Google should accept that their take on a 'developer machine' is going to be low volume but there is no reason why they can't make it pay its own way, really it is a matter of specification and not nickel-and-diming things like the keyboard.

If anything they should make the machine exclusive. Anyone can have a posh Apple computer but it is still just an Apple to people like the clients that I met today. Break out the Pixel though and even if you are just scrolling through a web page it has some impressive statement value to people who know nothing about computers.

I think the hardware division went the wrong way trying to make the Pixelbook 'for everyone' (rich). If they made it super cool with the ability to run a linux distro native as well as having ChromeOS/Android then they could have made something that Pixel laptop owners like myself would buy. Instead I run a regular PC for development.

I also think that if their developer machine was the de-facto thing for Google to use in house then that would make the machine aspirational to developers. If one were to be using the same hardware that Google designed for their developers then you would feel that you had the best going for Android/web development. A Dell machine doesn't quite have that kudos because nobody at Dell writes a lot of code.

I'm also the owner of a Google Nexus Player - remember that device? I liked the idea of an Android TV / console for Android games. I bought 4 gamepads they were selling for it.

It was buggy. It was rarely updated. They stopped updates after about a year. Then it died unexpectedly - it didn't even last 2 years. Now I'm left with 4 bluetooth gamepads and a dead Nexus Player in a box in the basement.

It saddens me to say this, but I would never buy a device from Google again unless it's the second+ iteration in a successful line of products, and even then I would have to factor the chance that they'll discontinue it at any time.

You can't start with a cool technology and build a product / user experience around it; you have to start with a vision of a coherent user experience, and build the technology around _that_.

I think that's my biggest frustration with Google: they have interesting and really cool tech that is just way too fragmented. The structure of controlled chaos via launching many products as experiments and culling down to the "best" ones makes it difficult to provide consistent, cohesive experiences.

As guess could be said of any startup though. Not just hardware.

As near as I can tell from the outside, there aren’t any good Android tablet options, and the problems are baked into the OS at a fairly deep level — deep enough that OEMs can’t really address them, and Google hasn’t shown any recent signs of being to do so either. (For example, read any review of the Pixel Slate).

If you already have a software/app/etc niche you like, I hear the new Samsung tablets are decent (not ‘good’) and not too expensive.

I have a Fire HD 10 tablet that I use for international travel without personal data, but Fire OS is pretty unpleasant and I was hoping to replace it with a new Pixel laptop from Google. Not sure what I'll being doing about that now. (Also, I expect the number of people cross-shopping Fire tablets and Pixelbooks is vanishingly small.)

Unless you absolutely need Android for some reason, I would recommend the iPad.

The rule of thumb for me is to buy the device with the highest screen resolution you can afford. Those tend to have the beefiest specs. Common high-res tablets have screens either at 2048 x 1536 (4:3) or 2560 x 1600 (16:10). I prefer 4:3 tablets so on Android, that really only leaves the Galaxy Tab S2 or S3 (the latter having pen functionality).

I'm still getting an iPad because the productivity apps are far superior, purpose-built and actively maintained. And that happens because people actually pay for the app. On Android, everything is freemium, so people are incentivized not to pay, and developers are incentivized to jam up the UX with ads and "unlock this with Premium" CTAs.

Yes, I can write webdev code on my Android tablet with Droidedit, and connect via SSH to my web server. But it's a hassle to do so, especially when the app in question feels like a scaled up phone app instead of a purpose-built tool.

1. This news plants doubt about whether the ChromeOS/Android hybrid in the Pixel Slate is The New Way or a dead end.

2. Google has botched the evolution of Android's tablet capabilities and tablet support in app frameworks so badly there may not be a future for Android tablets outside of Samsung's proprietary extensions to Android. Let that sink in: Samsung may be, by default, leading the Android tablet future.

3. I don't like what Samsung does to Android. This is a matter of taste. But I know I don't like it.

However when tablet dies, I will rather go for a Surface like tablet.

Eventually iPad Pro as alternative.

Getting fed up of Android J++, and a NDK experience that seems like the outcome of a 20% project.

With that said, the Pixel Slate was an unmitigated failure so hopefully the only impact this has is on the Chromebook Pixel line.

Google backing off on first-party work means that third parties that have always been the backbone of the Chromebook effort have less concern that Google's going to snatch the valuable segment of the market away from them, so I would say that its at least as likely to be a positive sign for Chromebook as a continuing effort as a negative one.

They raised pricing on Google Maps and have a reputation for dropping services.

Now they require a $75k third party audit to have an app use Gmail API’s.

Google makes the Pixelbook. It is definitely not the "best" laptop when you consider less expensive alternatives, and it is not the best seller.

But, Android would be called a successful strategy.

And, Chromebooks would be called a successful strategy.

What I'm trying to say (and would love to hear comments suggesting otherwise) is that these forays into hardware might just be to guarantee a baseline level of quality for vendors of the products to aspire to. And, in that light, even if the hardware divisions aren't making a dent in the Google's reported revenue, they still serve and were effective in their goals.

These cutbacks are just evidence that Google wants to optimize the way things are done, not evidence that they were failures.

GCP is a different beast, however. :)

- Android J++

- Kotlin, the successor, still faces lots of tooling issues, regarding incremental builds, code completion, thin APKs,...

- NDK users keep being neglected, after 10 years there is still no alternative to manually write JNI boilerplate, handle native libraries, a proper C++ API in the NDK, the ones that exist like Oboe are either plain code dumps on GitHub, or use Blaze, which isn't part of the standard SDK tooling, IDE tools if they get added always feel like an afterthought to their Java/Kotlin counterparts

- Stable support library and Android Studio releases have as much bugs as canary ones

- Android Studio feels like it needs a rendering workstation to work properly

The last two points have escalated so high that they finally created Project Marble to improve the current state of affairs.

OEMs love Android only because it is free beer.

* They have the most popular operating system in the world

* They have the most popular browser in the world

* They have the most popular search engine in the world

* They have the most popular advertising network in the world

* They have the most popular video platform in the world

* They have the most popular mapping software in the world

* They have the third most popular cloud platform in the world

It seems to me that they're doing at least something right

After that, every other product is just a mechanism to feed the beast including search.

The cost of deprecating products every couple months has to be expensive - if only we could measure it.

Is anyone aware of a good response to this concern?

I understand that Google would rather deprecate something that not support it well but that does not mitigate the apparent risk of adopting / deploying Google products.

Fashionable trope, but apparently this is more about unreleased products not being released? Every company has these, software, hardware or otherwise.