Samsung Buys US Cloud Services Firm Joyent

An anonymous reader writes from a report via VentureBeat: Samsung has announced Thursday that it has acquired Joyent, a company with public cloud infrastructure and private cloud software, to help beef-up its software and services around its smartphone business. While terms of the deal weren't disclosed, Samsung did say Joyent will continue to operate as a standalone company. "Until now, we lacked one thing. We lacked the scale required to compete effectively in the large, rapidly growing and fiercely competitive cloud computing market. Now, that changes," Joyent chief executive Scott Hammond wrote in a blog post. With Samsung's brand name and money to invest, Joyent may become more popular and challenge some of the top cloud infrastructure providers like Amazon Web Services, Microsoft Azure, and the Google Cloud Platform. Joyent was the original steward of server-side JavaScript framework Node.js and helped to establish the Node.js Foundation in 2015.

The bubble bursts

[Anonymous Coward]

Notice the lack of a dollar figure. Companies are running for the exits. Same thing happened in the last months of 1999 / the first few weeks of 2000. Joyent ran out of runway, having spent tons of VC money with no viable revenue to make the company float on its own. LinkedIn was bigger, but had the same perpetually money-losing business model.

Sun alumnis

[zdzichu]

Joyent is mainly known as a refugee for brilliant Sun engineers after Oracle takeover. They've built their platform around Solaris, ported linux' KVM to it and further enhanced using ZFS and DTrace.
The question is: will Samsung integrate those pearls? Or would this Solaris platform be shelved?

Re:

[geek]

Or would this Solaris platform be shelved?

One can hope. Solaris is terrible.

Re:

[BitZtream]

If Solaris is terrible I'm guessing you've never used anything other than Linux and don't actually realize how shitty your favorite OS is.

Just curious, what do you think is so awesome about your favorite OS that solaris didn't do first and did it better?

Re:

[jamesdwi]

yeah lets rip out decades of tried and tested code, and replace it with systemd that breaks many tenants of unix. and protect logging info by using a format that is more likely to get corrupted. Well its safe no one can read it, its corrupted. SMF didn't replace init, it just controls services in a more intelligent way that didn't break historic behavor. How often does a user have to deal with the XML files? almost never, and even developers are limited to just design phase, after that its not used.

Re:

[armanox]

SMF is what systemd wishes it could be. If the GNU guys weren't so hellbent on it's GPL or nothing they could have even used SMF if they believe that they needed a more intelligent init system.

Re:

[brambus]

The question is: will Samsung integrate those pearls? Or would this Solaris platform be shelved?

Shelving the Illumos core of SmartOS (Joyent's cloud OS platform) would essentially completely destroy all value in Joyent. Its distinctive cloud technology is intimately tied to Zones.

Re:

[E-Lad]

You're missing an important distinction - SmartOS is based on illumos, which was forked years ago off of OpenSolaris when Oracle decided to shut that down.

SmartOS is thus illumos, and illumos and Oracle Solaris have diverged a good bit in the intervening years.

Nooooooo duh.js

[null etc.]

Oh, look, node.js is mentioned. Quick, let's start talking about how JavaScript is a terrible language.

It's good to ridicule bad technology like JS.

[Anonymous Coward]

There's absolutely nothing wrong with ridiculing bad technology like JavaScript. In fact, it's something we should do as often as we can.

We don't even need to get into the numerous problems with JavaScript. They're obvious to anyone who has used any real programming language.

But when somebody says, "But JavaScript supports prototype OO!", we should tell that person to sit down and shut the fuck up. Prototype OO is a failure, and that's why we repeatedly see it used to replicate real class OO, but in a half-

Re:

[K. S. Kyosuke]

But when somebody says, "But JavaScript supports prototype OO!", we should tell that person to sit down and shut the fuck up. Prototype OO is a failure, and that's why we repeatedly see it used to replicate real class OO, but in a half-arsed way.

No, prototype OO is a failure because JavaScript failed to replicate the Self environment. (BTW, what's so great about "real class OO", unless you assume generalized predicate dispatch to help you fight ontological inadequacies of cast-in-stone assumptions of most commonly used languages?)

Re:

[K. S. Kyosuke]

Rigid class hierarchies make no sense. In fact, hierarchies as such often make no sense. In fact, even in the absence of hierarchies, fixed classes for an object after its creation (assuming the object is mutable) are problematic at times.

Re:

[K. S. Kyosuke]

When you say "Rigid class hierarchies make no sense.", you actually mean "Class hierarchies make no sense to incompetent programmers."

I guess Gerald Sussman must be an incompetent programmer, then! [github.io]

If you and programmers like you don't know how to use classes, inheritance, polymorphism, traits and the other features of real class OO, then it's a problem that lies solely with you. The rest of us know how to use these tools properly. They make sense to us.

Sounds like blub [c2.com] to me.

Instead of crying about how stuff you apparently don't understand "makes no sense", try to learn about it. You'll quickly see, like the rest of us intelligent programmers have, that your prototype OO hacks are just that: hacks. They're also completely unnecessary for people who understand class OO.

Fuck off.

Re:

[K. S. Kyosuke]

First, Scheme is irrelevant for the state of the world and the problems in it. Languages do not solve this problem (in fact, Scheme itself is woefully inadequate, despite being closer to solving many problems than most). Second, yes, for a programmer used to what you appear to call "class OO" (but clearly don't extend to its fullest possible extent, including for example CLOS/MOP), their language of choice is Blub because the inadequacies of it are not nearly as obvious (or obvious at all) to them as the in

Re:

[K. S. Kyosuke]

Also, it is apparent that "real class OO" is your only hammer. In that case, I understand that all problems look like nail hierarchies to you! Being a one-trick pony is tough in today's world, though.

Use same input validation on client and server

[tepples]

There's no reason to use JavaScript for server-side programming

If a web developer wants client-side prevalidation of input and server-side authoritative validation to use provably the same logic, what should he do?

• A. Write it in JavaScript
• B. Write it in another language that compiles to JavaScript
• C. Write an interpreter for the server-side language in JavaScript, as a workaround for the fact that what should have happened "years ago" ended up not happening
• D. Write it in both JavaScript and your preferred server-side language, thereby violating the DRY principle [wikipedia.org] and the

Re:

[tepples]

G. use a declarative approach to input validation, have the server side publish that information, have the front-end use this information.

I'll take that as a special case of "C. Write an interpreter for the server-side language in JavaScript", where "the server-side language" happens to be a declarative language specialized for input validation. This in turn requires that the declarative language be rich enough to express the language of what is and isn't valid input in a given case. Are you referring to something on the scale of XML Schema Definition (XSD) or something simpler?

Re:

[leptons]

Javascript is great, there are plenty of people that can see that - and you're a cranky old troll. Acting superior to others based on programming language preferences is a tired old thing to do. You may as well be huffing your own farts. While you're crying about how bad javscript is, millions of programmers are enjoying getting work done with it. Enjoy your obsolescence, neckbeard.

Re:

[frank_adrian314159]

-1, Redundant.

a possible angle

[epine]

I wonder if Samsung just bought a private test facility for their unannounced 3D XPoint rival technology? Joyent has first rate infrastructure visibility behind the scenes.

I doubt this was the main driver, but it might have been a consideration.

Re:

[epine]

A second thought just occurred to me.

If Samsung is a year or more behind on resistive memory, the short-term objective might be to pound the sand out of Intel's new 3D XPoint, so that they really know what they're up against.

Re:

[cdrudge]

I know if I was looking for a cloud provider I'd just LOVE to have one that uses my business as a guinea pig to test their unproven technology. /sarcasm

Re:

[MobileTatsu-NJG]

There is no ocean, only drops of water.

Joyent’s customers include

[pteddy]

"Joyent’s customers include Nerve, Paper Culture, Quizlet, Storify, and Wanelo."

Who?

Re:

[Psicopatico]

640 platforms should be enough for anyone.

Backstory

[jean-guy69]

The following post from Bryan Cantrill is worth a read, if only to know more about Samsung's motivations for buying Joyent.

Samsung acquires Joyent: A CTO’s perspective [joyent.com]