Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Chromebooks outsold Macs for the first time in the US

theverge.com

Post reply on HN
Print view View original

251–256 of 256 posts

Re: Chromebooks outsold Macs for the first time in the US

#251

Post by marssaxman »

rmrm wrote: post #100
marssaxman wrote: post #93

Earlier quoted context omitted.

Makes sense that my specifics would be out of date. I'm not arguing that network infrastructure will shrink; I'm observing that the scaling factors for network infrastructure investment are not nearly so favorable as the scaling factors involved in CPU or storage capacity. We can already build hardware with capabilities greatly in excess of our actual needs. When network capacity becomes a bottleneck, it will be much…

Network capacity is a continual bottleneck, and many billions are spent to upgrade the transport network yearly. Transport signals have gone from 2.5G to 10G to 40G to 100G to 200G per wave, and from 8 waves to 16 waves to 40 waves to 96 waves to 100+ waves per fiber. The bottleneck has existed for a decade at least, and is invested in to keep pushing it out at the same time all this other stuff is happening.

Hmmm. Interesting that speeds at the edge don't appear to have changed much at all, in years. I didn't think there was much ongoing coverage expansion going on. What's driving the capacity investment? Greater utilization of existing connections? What's using up this capacity?

Re: Chromebooks outsold Macs for the first time in the US

#252

Post by xuejie »

millstone wrote: post #120
notacoward wrote: post #98

Earlier quoted context omitted.

Same here. Anybody who repeats the "Chromebooks are for kids who don't do anything sophisticated with their computers" schtick is just showing off their ignorance. A decent Chromebook (e.g. a Pixel or soon one of the HP/Acer models) plus Crouton is a more real development machine than anything Apple ever made, because it's more like the server where you'll eventually deploy. Then, unlike any of the other dozen Linux…

> Crouton is a more real development machine than anything Apple ever made, because it's more like the server where you'll eventually deploy Well, this assumes one particular type of development! I would not choose a Chromebook to develop, say, Chrome! I tried Crouton a few years ago, wanting to run xbmc. Audio didn't work, because it required a newer version of the Linux kernel. It turns out you can't just install a…

To develop things like Chrome, a mac won't be enough as well. Chances are you will need a powerful desktop beast anyway, in this case, a Chromebook can be a super handy addition for checking emails, browsing the Internet, not to mention you can also ssh to your desktop and work off the Chromebook.

Re: Chromebooks outsold Macs for the first time in the US

#253

Post by contingencies »

spb wrote: post #137
j2kun wrote: post #104

I think it's hilarious how little the HN community knows about Chromebooks. Every time it pops up, there's a wave of people saying "I don't know a single person who has one!" and "They seem useless without an internet connection!" and "I wonder how they compare to mac books." People forget that most of the world doesn't write code, and doesn't need native applications. They write email, take selfies, and write extrem…

Moreover, Chromebooks, when paired with a good cloud IDE like https://c9.io , are excellent for writing code - more so than any "traditional" OS, since the local system is only dedicating resources toward keeping the interface responsive, and not wasting them on background distractions like local search indexing or filesystem defragmentation. Meanwhile, if you accidentally forkbomb a cloud development server, your mu…

If you ever visit developing countries or places like China where internet just sucks you can't realistically use cloud services. Out of interest, I wonder what part of the experience of a cloud service aren't you motivated to replicate on your own system for added speed/security/control/learning?

Most of your arguments seem Windows-oriented (driver updates) or moot with decent CI/CD workflow practices (conflicting SDK versions).

Re: Chromebooks outsold Macs for the first time in the US

#254

Post by notacoward »

millstone wrote: post #132
notacoward wrote: post #124

Earlier quoted context omitted.

Yes, and it would be fantastic to use a Mac or Windows box to debug something involving Linux kernel drivers, right? There will always be special cases. Sometimes there's only one solution, and you take what you can get within those parameters. I've worked on "one of this exists in the world" kinds of hardware a few times myself. Nonetheless, the "Chromebooks are for n00bs" meme is still BS. For the majority of peopl…

Developing software for something other than a Linux server is not some bizarre edge case. I was reacting to your unqualified claim that a Chromebook is "a more real development machine." My "special needs" was running popular and well known Linux software. Many people in this thread have said that running Crouton on a Chromebook is just like having Ubuntu or Debian. But this is not so: 1. Google does not reliably up…

If you want/need to run a very particular version of Linux software, then maybe a Chromebook's not for you. But many thousands of people successfully run all kinds of Linux software that works just fine inside the Crouton environment. I didn't and wouldn't say you're a bizarre edge case, but xbmc is being far more picky about versions than it needs to be and that's not ChromeOS's or Crouton's fault. It's unreasonable to criticize the platform because it didn't satisfy your one use case, and it certainly does nothing to refute my original point that ChromeOS isn't just for newbies.

Re: Chromebooks outsold Macs for the first time in the US

#255

Post by spb »

contingencies wrote: post #253
spb wrote: post #137

Earlier quoted context omitted.

Moreover, Chromebooks, when paired with a good cloud IDE like https://c9.io , are excellent for writing code - more so than any "traditional" OS, since the local system is only dedicating resources toward keeping the interface responsive, and not wasting them on background distractions like local search indexing or filesystem defragmentation. Meanwhile, if you accidentally forkbomb a cloud development server, your mu…

If you ever visit developing countries or places like China where internet just sucks you can't realistically use cloud services. Out of interest, I wonder what part of the experience of a cloud service aren't you motivated to replicate on your own system for added speed/security/control/learning? Most of your arguments seem Windows-oriented (driver updates) or moot with decent CI/CD workflow practices (conflicting S…

It took me like five minutes to parse what you were asking about "replicating" the cloud experience locally - the way you phrase it, you're inherently assuming that a local development environment would be superior in a few factors, none of which end up being so:

- Speed: Local development environments are just as slow as cloud ones, when not slower (due to compositor hitches, background downloads, heavyweight UI toolkits, all that bullshit). Speed is part of the reason that I ditched local environments.

I can't emphasize enough how little there is to be truly gained from offlining a workspace in terms of performance. When looking beyond pure ideal-world benchmarks and taking the intersecting outliers that impact performance during actual usage into play, the net savings, as I'm trying to convey here, are negative.

- Security: The odds are greater that somebody's going to run a buffer overflow on one of the many undermaintained and/or proprietary services and applications running on your machine (when's the last time someone ran Valgrind against f.lux?) and slipstream a rootkit into it, than that somebody's going to attack your dev server with a zero-day against CentOS (and even if they did, you'd have been just as screwed anyway).

Put another way: which machine is more secure, the one with a dedicated 24/7 threat response team monitoring port access, or the one that gets left unattended on a table at Starbucks every Tuesday afternoon?

- Control: If your development needs more control than root access to the host filesystem, you're doing something wrong, and your codebase is going to be more fragile for it. (Hell, even using root access is a sign that you're being persnickety.)

- Learning: Learning time should be allocated toward the systems involved in what I'm actually building, not a http://www.xkcd.com/1579/ hodgepodge of perpetually-obsolete periphery.

I don't know WTF you're talking about with my examples being Windows-oriented - the driver example I cited was specifically inspired by an issue with my desktop's network adapter on Linux.

Re: Chromebooks outsold Macs for the first time in the US

#256

Post by tuxracer »

stephenr wrote: post #247
tuxracer wrote: post #243

Earlier quoted context omitted.

Web apps can work completely offline if the developer adds support https://codelabs.developers.google.com/codelabs/offline/ Not sure why this misconception about web apps requiring "always-on super-stable internet access" persists when this technology has been available for quite some time (in the form of App Cache before Service Workers)

It may be possible to host the ide component locally in the browser, but that doesn't give you the execution environment (ie you can't run a vagrant image or in the case of a server based ide, a container, in a browser)

There is a way to do this actually via https://copy.sh/v86/ Although this is straying from the main point. There are certainly things native apps can do that web apps cannot. But requiring "always-on super-stable internet access" is absolutely not required for web apps. This hasn't been true for at least 5 years now. It also seems to be the most often repeated (completely inaccurate) limitation about web apps. There are certainly limitations. Let's talk about the real limitations, not the mythical ones.
Post reply on HN