ChromeOS

[schild]

This shit:

http://www.youtube.com/watch?v=0QRO3gKj3qw

By 1 minute in, I want to skullfuck this narrator to death.

At work, assuming that's the elaborate video. I liked it.

[fuser]

I'm still a little skeptical about why many people would want netbook+chromeos instead of netbook+win7+chrome, which can browse the web *and* run WoW or whatever.

Same, a Dell 10v/1Gb ram runs 7 Pro fine and that's not the stripped down starter edition they offer now. By the time ChromeOS is out the next generation Atom's will be out with more power speeding this whole process even more.

I have tried to sway some people on using Google Apps vs other services with no success, I cannot image what a whole Google/cloud PC would be like.

[HaemishM]

No, that's the 3 minute video you linked first with the faux handdrawn diagrams and shit. Infuriatingly smug.

[Yegolev]

Assimilate, you bastard!

[MahrinSkel]

I've become fond of Asrock mobos because of their "Instant On" feature, when you shut down the computer it reboots and then enters Hibernate mode, which means it the loading overhead comes when you don't care.  The reality is that most of what most people use computers for as tools (wordprocessing, photo manipulation, even spreadsheets/budgeting software at the personal/small business level) doesn't *need* a lot of computing power or memory.  ChromeOS is probably too stripped as-is for most people, but with the appropriate add-ons everything else people need from a computer comes from connectivity, not the hardware.

--Dave

[Venkman]

I'm about as skeptical of this as I am of netbooks in general. I was there in the mid/late 80s with the VAX system my university had. Cloud computing works well as long as it's a) ubiquitous; and, b) so redundant it removes human error.

We are probably at least a decade from both. There is no such thing as an anytime/anywhere connection in the US and EU. There is no such thing as 0% downtime. And there's no such system that survives some human screwing something up. Look how people freaked the two times this year Gmail went down (along with a bunch of other services tied to the login account). 99.9% uptime is only good when that 0.1% happens at 3am local.

And I use "human error" here very broadly, because it's not about forgetting to plug something in. It's about the humans making decisions across all sorts of corollary spectrums. Like the idiots who decided they either didn't need a backup solution or didn't bother using it for that "Danger" mobile phone.

Yea, I'm old. And yea, many of us have seen this tried many times before. And yea, sometimes that's a crutch that prevents open minds. But this is still today more promise than reality. It'll probably catch up first with kids, because they're not stuck in the world of installed programs, local files and email. But it's going to take that generation growing up and replacing old farts in corporate IT groups for cloud computing and netbooks to truly take off.

Which is where I got my decade estimate from

[Yegolev]

Age does not seem to be a factor in "I want my own server", which happens a lot in corporate IT.  People don't want to share servers.  I'm sure they would have dedicated ethernet cables run to their desks if they could.

[MahrinSkel]

Server for what?  Only thing I ever insisted on having my own server for was a database, and rather than argue about it I just frankenboxed one together out of old parts (mobo and CPU were from my old gaming rig, case, power supply, and drives from the reject pile at the office).

--Dave

[Yegolev]

Server for what?  Only thing I ever insisted on having my own server for was a database, and rather than argue about it I just frankenboxed one together out of old parts (mobo and CPU were from my old gaming rig, case, power supply, and drives from the reject pile at the office).

--Dave

I don't mean individuals since those guys have zero chance, I mean projects/apps and the managers who control them.  Like anything that needs an DB under it, those guys all want their own server... seems like you also wanted your own DB server.  Sometimes this is warranted but for small things it is not, not in our shop anyway.  Sometimes it is warranted for security or "security", but I'd argue that all of our machines are nearly equally at risk.  Specifically known to me, we have some SAP applications that really could be rolled into the existing giants, but the project leaders want their own "hardware" which calls for another HA setup and backup nodes, etc.  Kinda silly for a 200-500GB DB, in my opinion.

In any case, I think we are on the same page: people don't really want their shit in someone elses hands.

[caladein]

Yea, I'm old. And yea, many of us have seen this tried many times before. And yea, sometimes that's a crutch that prevents open minds. But this is still today more promise than reality. It'll probably catch up first with kids, because they're not stuck in the world of installed programs, local files and email. But it's going to take that generation growing up and replacing old farts in corporate IT groups for cloud computing and netbooks to truly take off.

Maybe I'm one of those kids (turn 23 in January) but hell if I want local files at all.  They just make it a hassle to get things on my phone, or office computer, or friend's laptop I borrowed for five minutes.

As cliché as it sounds, I do live in my Chrome window.  Ignoring games (or something ultra-nerdy like WinSCP or SyncToy/Dropbox), the only app I would miss from my current Windows 7 set-up is Notepad++.

[Ironwood]

Some of you old timers are really surprising me.

Caladein is not alone in his thinking, nor is it too hard to apply his thinking to bigger and bigger businesses.

There was a time that factories had their own Power Generators also.  Look what happened to that.  The Cloud computing model IS going to be a reality, like it or no.  Already there are huge economical factors driving it and they're only going to get harder and harder for any business to resist.  Crowdsourcing is also a massive factor in driving this to reality, in almost any sector.

Downtime isn't a factor.  Downtime is NEVER a factor.  Any company looks at downtime as a reality of doing business, not as something that you can ultimately avoid.  That's fantasy. 

Also, the guy that wants his own DB ?  He gets fired because it's easier and cheaper to leverage your business on someone elses architecture, remove the cost of administration and then FIRE that guy who's fucking leaching thousands off you as the only fucker who 'knows how it works'.

Surprised.

[Engels]

I can see both sides of the arguement. In a very generalised way, Ironwood's prognosis is probably accurate. Pervasive computing tools, such as office suite programs, SQL, Oracle, CAD and widely used mathematical programs will probably be entirely 'leaseable' from a cloud structure. Deeper within a scientific enclave, however, tools need to be tweaked and specialized and that will invariably need a more hands on approach and a level of control that can't be left to the some cluster managed in India.

[K9]

No, that's the 3 minute video you linked first with the faux handdrawn diagrams and shit. Infuriatingly smug.

++

[MahrinSkel]

I don't mean individuals since those guys have zero chance, I mean projects/apps and the managers who control them.  Like anything that needs an DB under it, those guys all want their own server... seems like you also wanted your own DB server.  Sometimes this is warranted but for small things it is not, not in our shop anyway.  Sometimes it is warranted for security or "security", but I'd argue that all of our machines are nearly equally at risk.  Specifically known to me, we have some SAP applications that really could be rolled into the existing giants, but the project leaders want their own "hardware" which calls for another HA setup and backup nodes, etc.  Kinda silly for a 200-500GB DB, in my opinion.

In any case, I think we are on the same page: people don't really want their shit in someone elses hands.

When I program DB stuff, I need a snapshot or proxy database I can do whatever I want to without worrying about.  It doesn't need to be fast, or complete, or even real data necessarily, I just don't want to have to triple check every line of code before I start debugging.  When my PHB's wouldn't give me that, I created my own.

Anyway, on the larger issue: I've been hearing people say that everyone having their own CPU, RAM, and storage was wasteful and inefficient since client-server first started taking over from mainframes and dumb terminals.  Somehow, because the flow-chart diagrams have replaced big iron with "the Cloud", things are supposed to be different now.  I'll believe it when it happens.  I have no particular bias against it, I'm just skeptical.

For the foreseeable future, bandwidth will continue to be the major limitation, and most methods for minimizing bandwidth will continue to involve putting the data on the client and leaving it there as much as possible.

--Dave

[Quinton]

Numbers Everyone Should Know

L1 cache reference	0.5 ns
Branch mispredict	5 ns
L2 cache reference	7 ns
Mutex lock/unlock	25 ns
Main memory reference	100 ns
Compress 1K bytes with Zippy	3,000 ns
Send 2K bytes over 1 Gbps network	20,000 ns
Read 1 MB sequentially from memory	250,000 ns
Round trip within same datacenter	500,000 ns
Disk seek	10,000,000 ns
Read 1 MB sequentially from disk	20,000,000 ns
Send packet CA->Netherlands->CA	150,000,000 ns

Slide 24 from http://www.cs.cornell.edu/projects/ladis2009/talks/dean-keynote-ladis2009.pdf

[Ironwood]

Yeah, but that's technical shit and therefore doesn't matter.

Economic>Social>Political.

That's all you need to know.

Speed of access ?  Seriously ?  That's going to be your argument against Thousands of pounds of savings for a company ?

"I went into my CEO the other day and talked about Nanoseconds.  My ears are still ringing from the slap."

Look, we all recognise the limitations and we are ALL a little non-plussed by it all, but, at heart, that's because we're techies and we're far too invested in this shit.  If you look at it from an 'outsiders' point of view (as well as a historical point of view) the whole paradigm becomes irresistable.

Literally, like Resistant is Futile type irrestistable.

Are there tech limits RIGHT NOW ?  Yeah.  Security Limits ?  Damn Right.  Logistical Issues ?  You bet your ass.

But that only requires another small leap in the tech chain to solve.  We've been here before.  The information is outstripping the technological ability, but that's when people step the fuck up and SOLVE the problem.

Hell, right now we're back to square one, since one of the problems of cloud computing is apparently the energy required to do it right.  It's almost humourous how far around in a circle we've become.

I find it surprising, as I've said, that there will be people out there reading this on an Iphone talking to me about how Cloud computing isn't here yet.  Wake Up.

I now return you to your regularly scheduled Ironwood one liners.  I blame Righ.

[Quinton]

I don't believe I made an argument.  Just sharing some numbers which have an impact on computing in distributed environments.

[Ironwood]

Well, thanks.

 

[Yegolev]

Good to find that we can all angrily agree with one another.

[Ironwood]

Hey, it's me.  I can do anything with anger.

[dusematic]

Hey, it's me.  I can do anything with anger.

I hear you suck a mean cock when you're angry.

[Ironwood]

Um.

I'm fairly sure the two events would be connected, yes...

[Engels]

I feel naked, scared and alone when Ironwood says,

  You bet your ass.

Its somehow just not right.

[Ingmar]

Speed of access ?  Seriously ?  That's going to be your argument against Thousands of pounds of savings for a company ?

If the savings is offset by your employees sitting around waiting for files, then yes, that can be a pretty compelling argument against it. Remember, employee downtime is not actually free. A real cost analysis will take that into account.

EDIT: I think you may be underestimating the incredible conservatism of upper management in general. Speaking for my own company, it is like pulling teeth to try to get anyone to move off of old technology they're comfortable with. We won't see any cloud computing in this building until the current generation of management is all gone, I'm certain.

[Ironwood]

No, I get that.  But we're talking about seconds and, again, it's a technical problem that increases in tech will eventually solve.   Businesses need to start planning now for what's going to happen then or they'll be totally humped when the time comes.  Unless you have wooden shoes, you'll have very little option left to you.

As for the Management, I also concede that, mostly because that's exactly what history has taught us ;  these chaps won't innovate until they have to - and many, many, many will go under in the transition.

That's really sad and makes me cry great big tears of 'Fuck You For Not Thinking Ahead'.

I don't understand Engels.  I can live with it tho.

[Ingmar]

What makes me particularly  is our own core business is all about getting people to move onto our new technology from their old comfortable tech. But I guess finance departments and the like are the same everywhere, no matter what business you're in.

[Engels]

I don't understand Engels.  I can live with it tho.

Sorry, just an irrelevant quip about using what I think of as an Americanism. Carry on.