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

https://arstechnica.com/gadgets/2019/03/google-hardware-makes-cuts-to-laptop-and-tablet-development-cancels-products/

by cattlefarmer • 7 years ago

205 264 7 years ago

jasonvorhe• 7 years ago

Not very surprising.

I wonder if Google would benefit from spinning up new products and services in another brand or company. Later, they could migrate only what they plan on supporting in the long run.

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.

freehunter2• 7 years ago

That's what I figured they would do with Motorola, especially with the move to Alphabet. On the other hand, if it's under another brand name people may not associate it with "first party". When the Nexus line came out, one of the big selling features was "Android straight from Google themselves". If it was from Motorola, it'd be "Android straight from Google themselves because Google owns Motorola so hopefully their internal corporate structure means the two divisions have a close working relationship otherwise what's the point".

burtonator• 7 years ago

Google can't seem to do anything right anymore.

They're last place in cloud. They don't have a good search API product to compete with Algolia/Elasticsearch. They can't release new products. Can't build new hardware.

The only things they have going for them are Youtube, Google (the search engine), Android, Chrome,...

Basically everything they did 10 years ago. Even Android + Youtube were acquisitions.

Another good example of too big to fail.

ohaideredevs• 7 years ago

GCP is the cleanest cloud platform by far. But it's also the one I know only in theory, whereas I have worked with Azure and AWS enough to seem them fail miserably, so maybe I have a case of rose-tinted glasses.

zihotki• 7 years ago

I have experience with all three of them, indeed Azure fails miserably, AWS is a bit better but I didn't spend much time there. GCP is the most solid one and I had the least troubles there.

module0000• 7 years ago

I haven't experienced this failure with AWS. I've had good experiences with both GCP and AWS, and a handful of terrible experiences with Azure. I don't know your particular use-case, but I end up being clean-up for AWS disasters, and each one's root cause was poor planning and/or execution. Likely the result of an over-ambitious employee or contractor trying to implement by the seat of their pants.

hn_throwaway_99• 7 years ago

I've worked with GCP, and especially Firebase is a fantastic platform.

iampims• 7 years ago

GCP is worth it for Bigqueet, GKE & AppEngine + world class networking.

Everything else is not as good as AWS offerings.

dehrmann3• 7 years ago

I don't have Azure experience, but I usually look at it at AWS has better marketing people. On the tech side, they'll happily packing an open-source project into something managed. GCP has done some serious engineering for products like Spanner, Bigtable, and Bigquery, but there are gaps in the product offerings.

godzillabrennus• 7 years ago

Google is their own worst enemy in cloud.

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.

wstrange2• 7 years ago

Given that 3rd party apps can scrape sensitive data, I'm glad the barrier to entry is high..

xrd4• 7 years ago

So, Google makes (specifically subcontracts out) the Pixel (but it is their branded device). It definitely is not the most popular Android phone out there.

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. :)

bilal4hmed3• 7 years ago

The problem with Chromebooks is that they have always been touted as a way to get a computer very cheaply. Years have gone by with the marketing being that Chromebooks are cheap. When you get pixelbooks that are $600 to $1500 people are going to have sticker shock, no matter how great the product is.

The pixel phones on the other hand have been selling better ( in fact grew quite a bit numbers wise as well ). The phone line has a better chance of turning in to the Googles surface line esp. since the marketing around that has been "Its the best camera ever" and people are willing to pay a premium for that.

pjmlp9• 7 years ago

Chromebooks are only successful if one constrains the world to the US school system.

Outside that they are even less than the worldwide GNU/Linux desktop market, hardly a measure of success.

fvjft2• 7 years ago

The Pixel could've been a success if they hadn't sold it at such a high price. They think they are Apple, but they obviously are not.

scarface745• 7 years ago

How would Android be called a successful strategy? It came out during the Oracle trial that Android has only made Google $38 billion during its entire existence - less than what the iPhone has made Apple during its last quarter.

pjmlp9• 7 years ago

Even Android is not done right.

- 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.

dman• 7 years ago

Building performant UIs has become a lost art across the entire industry. As a junior engineer I was given a hard time if any screen did not load in 250ms saying that was the cliff beyond which humans perceive things as slow. Now I have had multiple bugs I filed for a web based tool being too slow (> 30s load times) closed because the tool is working as intended.

therockhead2• 7 years ago

Also Google also has two GUI technologies for Android, the original Java based one and Flutter, which replicates the material UI in full.

sequence7• 7 years ago

The only thing they have going for them is:

* 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

s3r3nity2• 7 years ago

Some devil's advocate:

1) They bought Android - they didn't build it in-house.

2) They bought Double Click - they didn't build it in-house

3) They bought YouTube- they didn't build it in-house

4) Third most popular cloud platform is a bit disingenuous; it's like saying Apple Maps is a top mapping platform: sure, you're kind of right, but the gap at the top is fairly wide.

(You might even make the case that the best parts of Chrome weren't built in-house, but taken from open-source contributions via Chromium, but I'm not sure I quite buy that.)

I think the overarching thesis deserves merit: what was the last in-house built product that became a powerhouse with true sticking power, besides Gmail / Google Apps?

fvjft2• 7 years ago

Some of those purchased (OS, video, ads), some of those getting considerably worse as years pass (search engine, maps). And don't forget adtech will eventually blow up.

But well, if they can keep buying stuff, then they are okay. It's what Facebook is doing.

dylan6042• 7 years ago

You're leaving out their most important product: AdSense.

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

spaceflunky• 7 years ago

Building hardware is difficult. Like really difficult. More difficult that anyone expects it to be.

dylan6042• 7 years ago

Sums up most of Kickstarter or any crowd funded hardware product that fails to launch. People get super excited about the prototype they built, but then come to realize that mass producing is way more difficult than anticipated. The money raised in crowd funding gets burned through faster than anticipated with the various iterations with each of their vendors.

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

therockhead2• 7 years ago

This is so true. Apple/Samsung make it look easy, with great reliable hardware every year. At work we had multiple google Pixel C tablets for dev reasons, and they were plagued with issues.

cc439• 7 years ago

Everyone assumes that designing a product for mass production must be easy since "industry" is a part of the "old economy" and thusly must have been so figured out that anyone can do it. The reality is that the past century of modern manufacturing development hasn't resulted in amazing amounts of simplicity just the same as how the "tech" world has only become more complex. The difference is one has a century of exponential complexity growth behind it while.the other has ~30 years of complexity growth in its modern era.

on_and_off• 7 years ago

It seems that the company is entirely unable to have a vision about anything.

Too big to dare might also be a good way to describe them. They are not going to dare making e.g. one great messaging product (or even 2 .. there are after all some niches like office communication). Instead they do 7 and see which one works.

Apparently launching a new product is also the only sure way to get a promotion at Google, not improving an existing one.

s3r3nity2• 7 years ago

I am not an Apple fan-person by any means, but Steve Jobs had a Product philosophy that stuck with me many years after he described it:

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.

eikenberry• 7 years ago

I think Google is bad at taking this advice because their original big success was a cool technology that they built a product / user experience around (their search tech).

dehrmann3• 7 years ago

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_.

That's another way of saying don't build-out solutions looking for problems.

itbeho• 7 years ago

Well damn... I have my dev environment running very nicely in a Pixelbook via Crostini and love it.

gkfasdfasdf• 7 years ago

Same here...crostini is awesome and I absolutely love the portability and keyboard/trackpad of the pixelbook.

carlosdp2• 7 years ago

Yea I would very much like Pixelbook development to continue, I'm loving it as a main dev device.

ringaroll• 7 years ago

Sounds like a good reason not to buy new hardware from Google. They discontinue products too fast.

pgeorgi• 7 years ago

They discontinue products too fast.

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

xvector3• 7 years ago

I'm sorry, but I definitely don't see this behavior from Apple or Microsoft.

kevin_b_er• 7 years ago

I'll take this as the sign that entirety of Chromebook is at high risk of being canceled.

akhilcacharya4• 7 years ago

Isn't it still wildly successful in the edu market? This would be shocking if they went this route.

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

pjmlp9• 7 years ago

US education market, hardly noticeable anywhere else.

Question is if one country is enough to keep it going, even one of the US's size.

dragonwriter4• 7 years ago

I'll take this as the sign that entirety of Chromebook is at high risk of being canceled.

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.

drdeadringer• 7 years ago

I may be soon in the market for a new Android tablet. Does anybody have suggestions?

c256• 7 years ago

Wait.

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.

Marsymars2• 7 years ago

I gave up and got an iPad.

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.)

surds• 7 years ago

Have a Nexus and iPad at home. Would recommend the iPad over other Android tablets anytime.

post_break• 7 years ago

There's really only one game in town that isn't iPad and that's the samsung.

readittwice3• 7 years ago

Google just recently abandoned Android and pushed ChromeOS for tablets. Now this, there is just no continuity with Google.

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

cbhl• 7 years ago

Agree on the iPad recommendation.

That said, 2-in-1 x86_64 Chrome OS devices are bearable. Unlike Android; all the drivers are upstream.

I really like my Samsung Chromebook Pro, but would recommend looking for something that's on a later Linux kernel (at least 4.4) because not all the new features get backported. (See http://dev.chromium.org/chromium-os/developer-information-fo... for a list of all the options.)

supernovae3• 7 years ago

I have iPad and Surface Pro 4. iPad is nice if you just want "an app for that" but I find I use my surface much more. iPad was an amazing piece of tech, but it just feels suck in 2012. We got retina and then all the innovation stopped and went to mobile phones. Meanwhile, surface pro just keeps getting better and better and it's cool having steam, linux, vs code and all of that on my SP4

gst• 7 years ago

Google just recently abandoned Android and pushed ChromeOS for tablets. Now this, there is just no continuity with Google.

It's GTalk/Hangouts/Allo/etc. all over again, just this time with Android and ChromeOS. I wouldn't be surprised if they announce that they're going to abandon ChromeOS and that they're going to only use Android instead. How can the different product teams at a company be so completely disconnected from each other?

morsch• 7 years ago

Bought a 10 inch Galaxy Tab S2 (T813) mostly because Lineage OS is available. It's fine.

rchaud• 7 years ago

Samsung and Huawei are pretty much unchallenged in the tablet space since every other major OEM has now abandoned the market.

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.

snazz3• 7 years ago

Anything recent-ish that allows easy installation of an Android distribution that isn’t covered in manufacturer bloatware: https://wiki.lineageos.org/devices/

Zigurd• 7 years ago

I'm delaying a decision on an Android tablet:

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.

pjmlp9• 7 years ago

Like others are saying, Samsung or Huawei.

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.

akhilcacharya4• 7 years ago

The way I see it the only good pieces of Google Hardware are the Chromecasts (which are absolutely amazing devices, might be near perfection in terms of simplicity and price) and the Google Home (though that's still 2nd in the smart speaker market). Everything else is either niche (Google WiFi, which I forgot existed even as a techie) or a hanger on (Nest).

rtkwe• 7 years ago

The Pixel/Nexus phones have been my only phone for the last several years. I'd put them on the good list too just for giving a good phone without having to use Samsung's OS and avoid their attempt to force Bixby on the world.

akhilcacharya4• 7 years ago

I loved my Nexus 5... but the pixels have been incredibly disappointing in all respects save for the camera.

Marsymars2• 7 years ago

I do wish they'd give their Nexus/Pixel devices some longer support periods though.

The Lumia 950 was released at the same time as the Nexus 5X, and despite Windows 10 Mobile being dead, is receiving security updates for a full year longer than the Nexus.

carlosdp2• 7 years ago

Pixelbook is actually really solid.

ChuckMcM2• 7 years ago

The move comes after the group received pressure to turn Google Hardware into "a real business" from higher-ups at Google/Alphabet -- the squeeze continues unabated :-)

Its too bad, since the Pixelbook is pretty nice and with actual investment it could have been something much more than it is, but alas it would not have the margins search advertising does. Back in the 90's I met the head of Xerox's PARC after he had given a talk on innovation. One of the challenges he saw was that gardening innovation was like gardening crops, which was to say it was very hard to know when sprouts had one leaf out of the ground which were weeds and which were crops. Ideas were the same way, you need to nurture them to at least adolescence before you can reasonably decide if they are going to go on to be great or not so great.

From the outside, Google appears to be at that stage where they have not learned how to nurture an idea to see if it will be great or not before killing it.

jpm_sd3• 7 years ago

As an Xoogler, I can tell you that promising products that "will never be a billion-dollar business" get killed all the time. $100M in projected revenue? Yawn. Kill it.

ChuckMcM2• 7 years ago

That was true when I was there as well. Organically Google needs to figure out how to create 100M businesses within itself that pay for the organizations that enable them, and staff them for long term success. Doing so would incrementally add a small bit of margin for each one. I expect though they are caught up in scarcity thinking, which leads to bad choices overall.

dv_dt2• 7 years ago

The odd part is that this doesn't select for billion-dollar businesses, it selects for billion-dollar businesses that can be developed with revenue growth that doesn't pass through $100m for more than x-years. Essentially billion dollar rev with discontinuous growth through lower revenue - this seems unnecessarily constrained for any business that plans to be around for the long haul.

defen• 7 years ago

$100M in projected revenue? Yawn. Kill it.

They should at least tell us what those projects are so someone can pick up the scraps.

threatofrain2• 7 years ago

Maybe that's opportunity cost working.

pjc50• 7 years ago

How on earth does this make economic sense? Normal business operates on return on capital or margins, not overall size.

vondur• 7 years ago

Too bad they never had a fully supported way to install Linxu on it. I vaguely remember reading that google internally used a heavily modified Ubuntu for themselves. (or maybe just some custom repos)

mav3rick5• 7 years ago

There is Crostini now. Basically a Linux VM that is integrated in the OS.

netinstructions2• 7 years ago

You can go into the Chrome OS settings and click the button for Linux... after a few minutes of downloading/installation you're ready to go. I use it for development.

https://www.aboutchromebooks.com/news/chrome-os-69-stable-re...

https://chromium.googlesource.com/chromiumos/docs/+/master/c...

sriram_sun• 7 years ago

To me it feels like Engineering really delivered on the Pixelbook. That it didn't sell well was a marketing failure (not the first). So higher ups should take a real hard look in the mirror and work on developing a team that can brand and sell good products.

willtim• 7 years ago

The Pixel slate didn't fair so well in all the reviews though.

Theodores• 7 years ago

I got my 6 year old Pixel laptop out the bag in a client meeting today, used the touch screen to quickly scroll through some stuff and the non-tech crowd wanted to know what the machine was. Despite its age it still had wow factor. "You should get one of these!" - if only they knew it was 6 years old...

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.

discordance• 7 years ago

My 6 year old Pixel laptop has been end-of-life'd by Google. I got notifications last year that they would not be releasing any further OS or security updates.

I also have a 2011 Macbook Air, and am still able to get OS and security updates.

remir• 7 years ago

In the grand scheme of things, I think it is a good thing that Google, or any other big tech company, isn't dominating every product category.

Think about it: Google is dominating search and online advertising, they have the most popular web browser in the world, the most used mobile OS in the world, they have Youtube, Maps, Gmail, Chromebooks.

Now imagine if they also ruled messaging, social media, had the most used laptops and tablet in the world, were dominating on cloud, etc...

I cannot see that as being a good thing for consumer.

lainga• 7 years ago

Or for Google: it would just build an antitrust case against themselves.

abrowne• 7 years ago

Any impact on Fuchsia, I wonder? Or a much more short-term change?

jpm_sd3• 7 years ago

I'm looking forward to the upcoming announcement of a Google game console in 2019 [0], and the subsequent announcement that it has been discontinued in 2022.

[0] https://arstechnica.com/gadgets/2019/03/all-signs-point-to-a...

Balvarez• 7 years ago

Agreed. There's no way I'd buy a Google gaming console for this reason.

xvector3• 7 years ago

This is the issue with most Google products/services, period.

Literally nothing outside of Google's core services is trustworthy. Even their mail isn't entirely trustworthy! I've come to rely on Inbox and it's being killed for an inferior product.

stevenwoo• 7 years ago

There was some hiring mini surge around 2006 or so where Google seemed to hire a lot of my game development coworkers and they already cancelled whatever they were working on within three years.

akchin• 7 years ago

Ha - I think you might be thinking about 2009-2011 or so. That was the Google+ effort (under VicG) and thinking that games (like it did with Facebook) would be a key driver of growth. I think this was also around the time of the bad $180M+ acquisition of Slide to get Max Levchin. Lots of folks used this "social games" and the Google+ track to launch stuff and get promoted quickly and then after it shuts down, find another place within the org where they can now hang out with the nice salary that comes with the promotions they got. :-)

optymizer• 7 years ago

I used to believe in Google's ability to deliver quality devices: I bought the G2 when it first came out with its hardware keyboard, I've only owned Nexus and Pixel phones, and I've got Chromecasts in the house.

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.

robocat12• 7 years ago

Please build me a Linux laptop for software engineering, Google.

For engineers only. No software support beyond ensuring the BIOS and the drivers work and are updated.

I hate Dell, but I got a Dell XPS 15 because it was the only laptop I knew would: boot Linux reliably; last two Ubuntu LTS upgrades (4 years); have a high spec (i9 etc); and was reasonably priced (Apple tax was 50% higher for worse hardware).

Is it impossible to make a laptop better than Dell or Apple? Google: how much goodwill would you receive through developer love? Why do I have to pay for Windows when the OS division seems to be actively working against making my work easier? Google, you have the skills and size to make a secure laptop.

Google, maybe it is because most of your engineers internally just need a good terminal, instead of a fast laptop?

jpm_sd3• 7 years ago

I have used a few different system76 laptops with Ubuntu LTS pre-installed (you can opt-out of Pop!OS, whatever that is!). They've been just fine for my purposes.

https://system76.com/laptops

robocat12• 7 years ago

I am in New Zealand (we are far away from everything!) and I wanted something that at least had the potential to be supported locally.

cletus• 7 years ago

I'm not sure I understand what your issue with the Dell is. I mean I have a Dell XPS 15. It's... fine, I guess.

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.

robocat12• 7 years ago

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.

Counterpoint: I know plenty of happy developers that own multiple Nexus and Pixel products, without bagging them. Engineers want good hardware and many are willing to pay for quality.

As for the so-called "Apple tax"

For as close to equivalent as I could get (specification, reliability, etc) the equivalent Apple cost 50% more than the XPS 15. Paying 50% extra for an OS I don't need makes no sense for my situation. Paying for an unused Windows license for the XPS was unavoidable (I have our own licenses for the Windows VMs we need for testing).

It's clearly not trivial [to make a laptop better than Dell or Apple]

The Google hardware I have experienced has mostly impressed me (even if third party, they have mostly enforced quality constraints). Although I have owned a Nexus with a design flaw, I would consider buying a Pixel (although at present I find the Nokia gives me better bang for my hard earned buck).

mav3rick5• 7 years ago

Hey most Intel Chromebooks now support Crostini i.e. you can run native Linux programs . It runs in a VM under the hood and it's well integrated with the outside OS. Give it a shot.

robocat12• 7 years ago

I would love to, but how many Chromebooks have an i9, 32GB of RAM and 1TB SSD?!

(I wrote another comment for why I use a high-spec laptop: basically multiple VM's and greedy development tools).

danharaj• 7 years ago

I've had good luck with the HP zbook studio line. The latest g5 version is quite polished and high end.

snazz3• 7 years ago

Do you need a fast laptop? No one I know actually does (although I’m sure some people do). Text editing and web browsing don’t require much horsepower. In my opinion, compiling shouldn’t happen on a laptop in the first place: this is why one of my friends has a mega-sized server than he just runs builds and trains ML models on.

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.

curry-castaway2• 7 years ago

one of my friends has a mega-sized server than he just runs builds and trains ML models on

I built my first desktop computer to do this recently. I can definitely recommend it, even for local server development. It's nice to be able to have a version of a codebase running on the remote machine so my local laptop stays cool while I'm testing.

robocat12• 7 years ago

It is simply that my time is expensive.

For the $1000 it costs for a faster laptop, work gets many thousands of dollars back due to more efficient development.

I could waste time optimising disk space, waiting for compiles (the developer tools we use need single-core CPU performance), enchancing the performance of development tools, waiting for VMs to boot, waiting for Windows VM to update, waste time shutting down that VM I am not currently using. Yes, I could waste time switching to other tasks while I am waiting for whatever it is I am waiting for.

I travel, and while travelling work gets some development time out of me because it is a laptop and not a workstation. A terminal is no use when I am in an area with poor cellular data connectivity (e.g. most planes).

Work definitely wins on this whole equation.

spectramax3• 7 years ago

Have you looked at Thinkpad? You could run linux on Carbon X1 and have ultra-mobile package with i7. $1500 but you have one of the best laptops in the market in my view - keyboard, screen, size, weight, feel, specs, battery life, etc. I own the latest Macbook Pro 15 and Carbon X1 6th gen. I want to pick up my thinkpad more than the MacBook. I find Aluminum not a good material for laptops - it is uninviting, heavy and cold to touch. Thinkpad Carbons use magnesium and carbon fiber body thats light and coated with soft touch paint. It is just fantastic.

mooman2192• 7 years ago

Mac comments aside, ThinkPads are fantastic for Linux. I'm running Arch on my T460. You can find a used ThinkPad in pretty much every price range on eBay.

stevenjohns• 7 years ago

How do you compare the battery life against the two?

bsder2• 7 years ago

You are overstating the price differential.

I'm seeing $2182 Thinkpad vs $2599 MacBook Pro--choices driven by WQHD display, 16GB of RAM and 1TB SSD. Processor on the Pro is significantly faster, the display on the Mac has slightly higher resolution. Maybe not $400 worth of difference, but certainly almost $200 or so.

And I still can't buy an i5 laptop with 32GB of RAM.

rayiner3• 7 years ago

The new X1 Carbon is not a good Linux machine. Some hardware just doesn’t work (built in LTE). Then there is a bunch of manual configuration required to fix various things, e.g. throttling. https://wiki.archlinux.org/index.php/Lenovo_ThinkPad_X1_Carb...

tannhaeuser2• 7 years ago

What is it about Dell that you hate? Anyways, Thinkpads are also held in high esteem for running Linux. I'm currently using a Dell XPS (2016 model) at home and a Thinkpad at my current customer project, both running Ubuntu 16.04, and I can recommend both. I chose the Thinkpad over a MacBook Pro, and would do again, but the display and touchpad is no match to Dell's, much less Apple's; that's however no problem because the thing is hooked on a docking station/two monitors/external keyboard.

snazz3• 7 years ago

And once you learn to love the TrackPoint, the touchpad won’t matter any longer.

kamarg2• 7 years ago

Three gripes with my Dell XPS13.

1) The docking station is horribly buggy. If I unplug it to take my laptop somewhere and then plug it back in, I have to do a full reboot or none of the USB peripherals on the docking station get power. Monitors, power, ethernet all still work through the docking station after being plugged back in. Dell support took 4hrs of my time installing/uninstalling drivers and firmware updates before finally giving up and sending me a new one. Which suffers the exact same issue. So I can buy a different brand of docking station, not take my laptop anywhere, or constantly reboot my machine.

2) Rebooting takes forever then fails. Every single time. After 8-10 minutes of sitting on the blue "Rebooting" screen it finally crashes and tells me something went wrong while rebooting. Every single time.

3) Sleep still doesn't work about a third of the time. If I put my machine to sleep through the Windows start menu and wait for all the various indicators to power off before closing the lid, there's a good chance that when I boot up tomorrow, the battery has fully drained and the machine has to boot back with all my applications closed down and context lost.

It makes me miss my Thinkpad.

MisterOctober2• 7 years ago

^+1 for Thinkpads running Linux. Switched from MBPs to these a year ago, never going back.

weew1211• 7 years ago

Also use a thinkpad, X1 carbon from 2015.

Run Manjaro with gnome-shell (actually gnome-shell-performance until 3.32 is released). Its lightening quick, reasonably stable, good battery life.

I also tried KDE, which works quite well, but the battery life was quite poor and I find the general KDE experience abrupt.

Now my 2017 mbp just gathers dust, never going back.

balls1873• 7 years ago

Not to start the flame war, but whats the appeal of Linux for software engineering, if you want the "it just works" approach of designed from the ground up experience?

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

kps• 7 years ago

What's the appeal of laptops for software engineering? Tiny screens and attached keyboards are a recipe for hand and spine problems.

0815test2• 7 years ago

whats the appeal of Linux for software engineering, if you want the "it just works" approach

For one thing, Linux-based OS's like Debian GNU/Linux absolutely follow the "it just works" mantra - if the OS installs properly in the first place and does what you need it to, you can expect it to be rock solid with few or no future issues. Plus, they tend to work for a lot more time than the 5 years max that Chrome OS does (or Mac OS, for that matter). That's simply invaluable in this day and age.

abhgh• 7 years ago

I recommend Dell's Alienware (AW) laptop if you need a machine with good specs that can reliably run Linux. I have been a happy Ubuntu user on AW for a while. Great options for hardware, built like a tank, and can run simulations/experiments for days without heat warping problems* (this was my primary requirement). Downsides are it is clunkier than the competition (personally I am now used to it) and the battery backup isn't great.

* 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.

geodel• 7 years ago

No goodwill. Only heap of comments on how that is another way to suck user information in to Google ecosystem.

turbinerneiter2• 7 years ago

Dell literally makes that laptop in the form of the developer editions. I had one and it was great. I have a Lenovo now and it sucks (work).

robocat12• 7 years ago

They make the XPS 13" in a developer edition, but literally Dell do not make an XPS 15" developer edition model. I prefer a slightly bigger screen when working while travelling (in fact I would probably go for a 17" if it only cost an extra 0.5kg).

You can get the Precision 15" with Linux. However I was quoted a price 2 times as much as the XPS 15 with a lower spec, less community support, and I was quoted it with 16.04 installed in Jan 2019 (presumably only 16.04 officially supported? Hard to get information when we are only a small business).

https://liliputing.com/2018/05/dell-introduces-2018-precisio...

ahmedalsudani2• 7 years ago

If you don't like Dell, three options are Lenovo, Purism, and System76. I'm personally quite happy with my thinkpad.

curry-castaway2• 7 years ago

Do any of these options have a case that feels sturdy? It seems a little silly describing it this way, but one of the reasons I keep coming back to macbooks is the aluminum case. I want good specs, but my lizard brain also wants a laptop which doesn't feel like a flimsy plastic toy.

cwyers• 7 years ago

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

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.

Zenst• 7 years ago

Alas whilst there are many who crave such a device, as a market, it is still niche for a sole target. Dell, offer you the option of Linux, more so, drivers. Yet, they still offer the laptop for the windows end user at the other end of the spectrum. That end also has larger market.

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.

bproven2• 7 years ago

IMO the best (or one of the best) for Linux :

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)

robocat12• 7 years ago

I will definitely look at getting a Thinkpad next time.

I needed a laptop (I try to refresh every 4 years in time with Ubuntu LTS releases), and the Dell met my requirements at the time, so that was what I got.

I looked at getting an Macbook Pro, but I was worried about the keyboard issues (I do use the laptop keyboard), and I had read about Linux driver issues.

iamgopal• 7 years ago

Would you be interested in open hardware doing exactly what you asked for ?

automathematics3• 7 years ago

Have you kept up with the Pixelbook development? Chrome OS is pretty great for general "cloud" work (AWS management for example). It runs containerized Android apps (which I use to run the native Gmail app to fly through my inbox).

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 :)

bubblethink• 7 years ago

The main downsides are the wimpy processors, soldered ram, storage and wifi cards, poor repairability, non upstreamed code.

robocat12• 7 years ago

As the article says "A report from Business Insider claims that Google has axed 'dozens' of employees from its laptop and tablet division."

I would love a Pixelbook, but I need a laptop if I want to work while travelling - I often end up with poor cellular data (or none at all!).

Does having both a laptop and a pixelbook make sense?

Work also bought a device to get satellite data - so that we can do operational support while in "remote" areas in New Zealand (plenty of areas we go to that lack mobile coverage) or when travelling overseas.

anonymous_fun• 7 years ago

I bought the low spec Pixelbook and gave it to my daughter. I replaced it with the high spec Pixelbook for myself. She since replace her Pixelbook with a 32gb 15 inch Dell XPS i9 for VM work specifically.

Prior to this, I gave my mom Chromebooks after she kept calling me for tech support for virus infestations, and it worked well.

Working with the Pixelbook, I totally agree with the VM bit. I would love to run Vagrant, MiniKube, KubeFlow, etc. The Linux emulation is good, but It's Google. 'We' can do it better. They probably did, but missed the adoption / backwards compatibility curve.

I hung out with a ChromeOS dude last weekend and he told me about how the Linux emulation is controlled by 'vsh' using their own replacement for QEMU, and inside that is an LXC container. Great for security, and I believe him. I entered one of the system vm's and saw LXC/LXD. The security seems great, but again, I can't do what I want to do. I felt dumb asking him about my 'emacs' in the ChromeOS Linux emulation. It always had a huge title bar on the top that took a bunch of the little screen real estate available. His answer was 'hit the full screen' button. I felt dumb. Usability for us dumb engineers will sell your product. Apple has abandoned the engineer market. The XPS flexible, but the Pixelbook is still sexy. I can't get my son to take my daughter's Pixelbook though, so maybe I am just old.

mehrdadn• 7 years ago

Is the world filled with that many 15" i9 laptops nowadays m that your problem is really Linux compatibility and the price? I had a hard enough time finding a reasonable quad core a year ago.

opan• 7 years ago

https://mntre.com/reform/ You may like the sound of this.

KingMachiavelli• 7 years ago

last two Ubuntu LTS upgrades (4 years);

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.

Dylan168072• 7 years ago

Laptops with GPUs tend to be power-hungry and expensive. But the middle ground of a laptop and an external GPU has a good shot at getting even more power per dollar, having sufficient battery life, and still being quite portable.

robocat12• 7 years ago

I do some of my work where there is no internet (except satellite). I happen to be CPU/memory/SSD bound for some of my development work. A four year old 4k laptop with equivalent speed and battery life!? - this one will pay for itself fairly soon by improving my productivity.

The 8950HK processor is fast[1] and NVMe is way faster than SATA.

I just use the integrated Intel GPU (I don't play games on my work laptop. If I were to play games, I would use a dedicated gaming rig or console).

I am expecting that in 4 years time I can get equivalent improvements to productivity by getting a replacement laptop (or even better, retire to a pacific island!).

[1] Benchmarks are not the answer, but anyway: https://browser.geekbench.com/processors/2145#family-64-sing...

mark_l_watson• 7 years ago

I bought a System76 laptop with a 1070 GPU last fall and so far I love it. Hard to explain but I just feel happy using it and having a GPU for machine learning is a big win.

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.

CurrentB• 7 years ago

As a developer who has repeatedly tried and given up on linux on laptops, a Pixelbook with crostoni has been the first that I have been extremely happy with. Everything I use just works pretty much as it should.

endtime• 7 years ago

FWIW, I just upgraded my Thinkpad for a newer one at work (Google, running GLinux/Debian). Seems great so far.

Twirrim• 7 years ago

Google, maybe it is because most of your engineers internally just need a good terminal, instead of a fast laptop?

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.

ma2rten2• 7 years ago

Google has an internal cloud-based IDE which is quite popular, so it's possible to do all work in cloud. Otherwise you can ssh into your desktop. Either way Googlers are not allowed to store source code on a laptop.

There are some Engineers who use a Chromebook, but most have a Macbook. I think Linux is more common than Chromebooks.

But honestly, why would you choose a chromebook if you can also get a macbook?

shaklee3• 7 years ago

With WSL in Windows 10, a lot of the reasons for booting Ubuntu go away. Check it out if you haven't.

mruts3• 7 years ago

The big problem with WSL is that I can't run a different window manager, which is probably one of the biggest benefits of Linux anyways.

Personally, I can't live without focus-follows-mouse. That way, I can type and click on a window that is under another one.

ma2rten2• 7 years ago

FYI: Google employees can currently choose between a MacBook Pro, Pixelbook or Lenovo X1 Carbon with Linux as their laptop.

dehrmann3• 7 years ago

This doesn't fit with Google's product direction (charitably, the cloud; less charitably, a data vacuum), business direction (supporting the ad business), or the way they see the world (opinionated).

akhilcacharya4• 7 years ago

From what I hear most devs aren’t allowed to check out code locally (unless they do mobile development) and instead use a web ide.

ralmidani2• 7 years ago

I used a Samsung Chromebook 2 (Exynos processor, 4GB RAM, 13 in. screen) as my primary laptop for a couple years. It was fast and quiet, but getting GNU/Linux running via Crouton meant jumping through hoops every time it powered off and back on.

I now use a Macbook, and really wonder why any developer would buy a Pixelbook vs. a Macbook. If you want to run GNU/Linux, you can install a VM, and otherwise you still have a functional OS.

Edit: I see the value proposition in $200-300 Chromebooks. I'm specifically puzzled by the Pixelbook, which starts at $999 (not much less expensive than a new Macbook Air).

mav3rick5• 7 years ago

Chrome OS comes with Linux VMs now. You can do most dev work with it

ralmidani2• 7 years ago

Do you have to enable developer mode, and if so, does it stay enabled between boots?

saagarjha• 7 years ago

I bought a Samsung Chromebook 3 a couple of days ago and was able to get it to dual-boot Chrome OS and Linux with a little bit of effort. Overall, I'm reasonably satisfied with it, given that I put less than $200 into it.

izzydata• 7 years ago

Who do I have to kill to get a Nexus 7 (2020) model?

ggm• 7 years ago

Would that be without the bad eeprom hardware AND a headphone jack?

jmull• 7 years ago

"Dozens" of employees may indicate a fairly small course correction rather than a change in direction.

imroot• 7 years ago

The biggest issue that I have with google is product sustainability -- if something doesn't seem to have "Hockey stick" levels of growth and use, google kills it.

Do I want to spend $500 on yet another google branded brick?

threatofrain2• 7 years ago

Does Google kill hardware often? What's the previous Google brick you can remember?

dragonwriter4• 7 years ago

Does Google kill hardware often?

Its killed several hardware projects (Nexus Q comes to mind immediately, though that only got into anyone's hands as a prerelease version at I/O 2012), but I don't know that any of those have involved bricking devices already sold (though cutting off software support and associated services may have reduced the potential lifespan for some, but that happens for old models from most makers even if the project isn't killed; Nexus Q, for instance, had support dropped in Google Play Music and Google Movies & TV in May and June of 2013.)

hiven• 7 years ago

It’s not really worth investing in google hardware in my opinion- you get invested and they just can it. They have no staying power.

bepvte• 7 years ago

I really loved the design of the pixelbook, but I just didnt see who would be interested in it. When I buy a 1,000$ machine i dont go for that kind of OS and it just didnt stand out very much. I wish the best to those who worked on these.

vkaku• 7 years ago

Well, Google wanted to promote ChromeOS, and now it's killing it and slaughter all the engineers involved in the project.

Doesn't that sound hypocritical to say the least? Why axe those engineers? Why not get them to fix Android, Fuchsia or whatever needs fixing?

naner• 7 years ago

I don't think they're killing Chrome OS, Chromebooks seem to be as popular as ever with schools. Schools that can afford to buy Pixels probably would go for iPads instead, though. I doubt there is much of a market for high end Chrome OS hardware.

pjmlp9• 7 years ago

US schools, hardly used anywhere else around the world.

mruts3• 7 years ago

What's wrong with Google nowadays? They seem completely unable or unwilling to invest in anything besides search for longer than 3-4 years. Completely untrustworthy at this point. I would never buy another Google product ever again.

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

Related Stories