Chrome OS and IT platform longevity | Hacker News Reader

Chrome OS and IT platform longevity

marco.org

56 pointsduck35 comments

I work at a bank, and while what the article says is true, we will never use Chrome OS. Why? Because the IT department can't easily make it miserable.

For example, does Chrome OS have a $1000/workstation program to pop up a message saying, "OMG you're going to be fired for charging your phone from the USB port"? Nope? Then they won't use it.

The goal of corporate IT is to make using their machines as miserable as possible while costing as much money as possible. ChromeOS is limited, but friendly and cheap. So no sale, sorry. (Also, can you unleash a team of 10,000 C# developers on it to deliver full-screen messages from our CEO? No? Double no sale!)

You say that but the irony is Chrome OS holds the potential for IT departments to make your life even more miserable. One example...

It's still reasonably difficult to document every single thing a user does on a PC. It can be done with key logging but it's a pain and even key logging solutions can miss things. With Chrome all you need is a browser extension.

That same extension can be used to detect your phone on the USB port and deliver your CEO's full screen messages. So having everything in the browser actually provides IT departments with a level of control and monitoring they'd never have dreamed of before.

No extension. Block 443 , handle it at a network level.

>>irony is Chrome OS holds the potential for IT departments to make your life even more miserable

That is, when IT Teams in those places themselves are aware of playing around well with those systems.

IMHO, banks (like where jrockway is employed) or BigCorp Inc (where I am) have IT teams comprised of subcontracting teams with MCSE types, hired precisely because their CTC is far less than average. Am pretty sure most of those teams would never (or even allowed to) take the pain of hacking up and altering OSes for that company's needs (blocking USBs/CD-ROMs, limiting privileges to ridiculous levels etc), when they could do it the easy way in Win XP or Win 7 installations.

That article is 100% wrong. My aha moment was when they showed the citric 'app'. A google chrome computer is a modern day dumb terminal, and the citrix server is the mainframe. Corporations who spend millions of dollars supporting computers (and most of whom already host their business critical applications on a citrix server) will jump at the chance to replace their costly (in terms of money and tech support) pc's with cheap dumb terminals.

But IBM etc supported their cheap dumb terminals.

From the server side -- sure you can continue to patch/update your citrix apps etc.

If Google drops Chrome in 5 years time and in 5.5 years there is a massive flaw (some form of root-kit that saves every username/password and grants you access to the citrix system or the ceo's cloud based email account). If google decides that it wants to focus on Android and drops Chrome -- the pain could be immense. // This obviously doesn't apply for smaller faster moving companies.. But if you roll out 10,000+ Chrome systems -- Long Term Support plans are a good thing.

Otherwise - I agree. If you can replace a complex win/mac system with a cheap dumb chrome terminal -- the benefits for support are immense.

It seems most of these corporate "dumb terminals" are running Windows 2000 or Windows XP. In a few years (or already?) Microsoft will stop releasing security updates for these operating systems. The argument to switch to a new system, especially one that is free and open source, more secure, faster and better in almost every possible way, will be clear and obvious.

See OP Footnote: Win 2000 had 10 years of support. Win XP has until 2014 (13 years). The point is not: What is the best option at this point in time, but What is the best option for 5 years from now.

// Something like Ubuntu LTS might fit the bill perfectly so please don't think I'm just advocating Windows. I'm just considering that Google already produce a very successful light-weight linux-based OS (That has shipped far more units than Chrome likely will) and has a history of killing off products that don't get much traction (see Wave for a recent example)

IIRC, XP was EOL'ed several times, and to M$'s dismay, they've lengthened the support date a couple of times.

Why do you care about security updates for a dumb terminal?

Well, you're probably typing credit card numbers, passwords, and other private information into the dumb terminal. It seems absolutely critical to make sure nobody is running a keylogger.

Except that Chromium OS is open source so you or the community can patch it if such a thing were to happen.

Will Google allow you to install a patched version on a native Chrome OS notebook? It seems like some of the value lies in the hardware/software integration, and those machines are going to ship with locked-down bootloaders.

Otherwise, I agree that the community aspect does differentiate it from past legacy products like Win2K.

> Will Google allow you to install a patched version on a native Chrome OS notebook?

The answer is an unqualified yes. According to Google, jailbreak mode is a feature of Chrome OS hardware. On the Cr-48, it requires flipping a small switch under the battery.

If it's a hardware feature, who knows what the other hardware vendors will do with that?

You don't have to buy that hardware. The point is Google is all in favor of you having control.

"You can do it yourself" is not what enterprise IT people want to hear either.

I predict that somebody will ship chrome notebooks to corporations that have the 'jailbreak button' removed.

Yeah. In some sense, it's an advanced thin client with local persistence, portability, with additional security and better performance than a traditional thin client.

The hardware will be cheap, and the software will be ridiculously easy to maintain. It will be easy to adopt, and comparatively easy to abandon, if needed.

For Chrome OS to be considered by any reasonably large business, their IT decision-makers are going to want to know that Chrome OS is going to be around and supported by Google many years from now.

Actually, they just want to know that it will be around and supported by someone.

The change over to things like Chrome OS will be partly generational, much like it was with web applications. Young people brought up on Chrome OS will start businesses which will use Chrome OS to run their businesses fast and lean. Larger firms will have a lot of cultural inertia to overcome.

You bring up a good point, actually: since Chrome OS is almost entirely open source as Chromium OS, couldn't just about anybody step in and provide a reasonable level of support for a fee, in the unlikely event the Google decides to quit?

Absolutely. To me the key "risk" really is not so much the OS going unsupported but any Google services that it depends on doing so. Which is more about the buy in to things like Google Docs, Calendar, etc. The risk is all the higher because the services are "live". If Microsoft gives up on MS Exchange, your exchange server doesn't immediately stop working. But if Google turns off Google Calendar one day because it's not turning a profit ... oops.

The problem is that, for me, a huge part of the value of ChromeOS is provided by Google's suite of apps. Those apps are the reason it is viable in the business space. I would be looking for a way to run hosted versions of those in house and then we would have a really interesting situation.

Surely Google Apps for Business goes some way to counter this concern.

This issue was the main reason that we decided against Google Apps for our email system. If we're running an Exchange or Domino server then even if MS/IBM stop supporting it we can continue to use everything and plan a migration. Google could turn off decide GMail was unprofitable and turn it off tomorrow, and everyone would be screwed.

I went to a Google Enterprise sales event and asked them about an internally-hosted version, but they were very clear that this wasn't something they were considering.

You could protect yourself against this by building an internal IMAP server and syncronising it with Google regularly. But that's a lot of work, and maintaining your own email server negates many of the benefits of outsourcing to a cloud service.

turn it off tomorrow, and everyone would be screwed

Could Google stop offeringing Google Apps at some point? It could happen. Tomorrow? No way. They have shown that they give people time and options when they no longer want to provide a service. Google Apps would be no different and they would probably even work a lot harder at making that easy.

Just anybody could step in, but that will not solve the trust problem the enterprise has, unless that anybody is in the IBM/Oracle category. If such a trusted company steps in, chances are most of the financial arguments to move to Chrome OS would lose credibility.

A big point that the article doesn't bring up is the ongoing upkeep costs that already exist even for smaller organizations. Even if the computers are all running windows 2000, the easiest way for most companies to license the software is by paying a yearly site licence to Microsoft.

If Chrome OS costs less than than that yearly upkeep cost, and comes with perks of low maintenance and remote connectivity, I could see a lot of companies being very tempted to switch.

I'm pretty sure Google has shown the longevity of Gmail. That's another large IT platform which they are trying to get enterprises to use.

At least that's a start. Plus, it's not like Google OS is a from-scratch endeavor. It is just a highly customized Linux, is it not? So at least enterprises could be reasonably comfortable with getting support in that regard. However, I suspect that businesses that might be inclined to such migrations may have already moved in that direction.

You can't go wrong by doing nothing : I think that the most prevalent behaviour in bank. Anything new is dismissed as too unsecure, so only the most essential things get ever done.

I've been reading the comments here and I think most are thinking of this the wrong way. Chrome OS is simply not going to be a viable corporate environment out of the gate. That's not even a question (the incremental updates that are automatic and can't be turned off would prevent that all by themselves).

The question is whether the computing model will be enough to entice CIOs and IT Managers. The dream of every IT manager has always been the dummy terminal it was just never viable (Though Citrix solutions and Terminal Server get much closer). If Google can prove they're serious about this they could start to make serious in-roads by the time Microsoft releases Windows 8 (around 2013 I'm guessing)

I support a small business with 3 PC's (XP, Vista, Win7). One of them has an intermittent problem that I have spent 50 or 60 hours tracing. Chrome OS would solve 99% of the support problems that they have - reducing all of the support to just "Are you connected to the internet?" I so wish that all the apps that they use were available as SaaS.

2 words - desktop virtualization.

then the deployment nightmare goes away, support costs go down, and a ChromeOS appliance which is a head for a PC in a data center can justify itself.

For Chrome OS to "work" in the enterprise, I suspect you need to have working AD-authentication or the IT-department will make it a no-go.

So far my attempts to make Linux do AD-authentication has all been failed and miserable with immense amounts of work compared to just "join domain. done." in Windows.

Not saying it can't be done, but I seriously doubt Chrome OS was made for a corporate environment or that corporate environments see Chrome OS as a viable fit for their needs. Yes, it may be cheap, but a cheap OS alone doesn't make for a cheap total.

http://technet.microsoft.com/en-us/library/bb742516.aspx

> So far my attempts to make Linux do AD-authentication has all been failed and miserable with immense amounts of work compared to just "join domain. done." in Windows.

Have a look at likewise-open. It's pretty much that easy to join a domain and allow domain auth. Integration is where it gets interesting.

Reply on news.ycombinator.com