What can we do to incentivize hardware manufacturers to be less "evil"? I have an iPhone, and Apple has screwed me over; this is my story: http://www.anderson-net.com/~nathan/apple-broke-my-phone (also see http://pandodaily.com/2012/11/23/apples-stick-in-the-mud-routine-is-getting-old). I know, I know...you can say "I told you so" if you want to.

As a customer of theirs, I'm sure I'm well in the minority in terms of how I use my devices, and as long as most of their customers have no problem with how they do business and they continue to rake in money hand-over-fist, Apple losing me as a customer is a mere drop in the bucket for them. If the loss of my money and goodwill as a prior customer is not enough, and other people continue to desire and to buy their products, how can we communicate to companies like Apple that the "open" way is a better way, and do so in a language they can understand and respond to?

-- Nathan

You actually weren't TOO far off the mark...even though the whole OS was publicly sold as "version 3.0" and "version 4.0", the kernels from those two products were internally versioned as (and self-identified as) 2.3 and 2.4. Kind of like how Windows 2000 was NT version 5.0, XP was 5.1, Vista was 6.0, and (bizarrely) Windows 7 is 6.1 (and Windows 8 looks to be 6.2).

If you ran "ver /r" in an OS/2 CMD window, you'd see this, so perhaps that's what you are remembering.

-- Nathan

As far as SOM being ugly, I'll admit it had its cumbersome points, but what exactly has replaced it today that is better and is being used at the core of a platform to drive its UX?

I'm sad that you posted this anonymously -- I was a big OS/2 fan back in the day, and would have loved the opportunity to pick an IBM/Boca engineer's brain.

Thanks for all the fish,

-- Nathan

...that is, TRUE PtP is expensive. Most rural wireless providers are going to be running most of their residential & SMB customers off of PtMP systems: one antenna, multiple customers. Even better, these systems are all half-duplex.

True PtP is a really cost-prohibitive option. It doesn't scale well for the provider since there is only so much spectrum and tower space for antennas to go around, so the customers that actually need a dedicated, high-bandwidth option with an SLA are the only ones who are going to be willing to pay what it costs for an actual PtP connection with an antenna dedicated to them on the ISP's tower.

That's not to say that PtMP cannot work, or work well. But it is important for everybody to keep in mind that it is a shared-resource kind of connection, not unlike DOCSIS.

-- Nathan

Listen to this guy. I work in the industry, too, for a regional ISP in a very rural area, and I have a couple of things to add.

To begin with, I know it hurts to hear this, but sometimes reality bites: the residential ISP business model is BASED on oversubscription. Period. Anybody else who tells you otherwise is lying or doesn't know what they are talking about. When an ISP sells a residential or SMB customer a 3Mbit/s down asynchronous connection at under $100/mo, it's guaranteed they don't have the bandwidth to back this up for you and everybody else they have sold a connection to. All of the usage models for scaling up bandwidth are based on bursty usage by their customers. They simply cannot afford to have every single customer of theirs pulling down their 3 megs all simultaneously, 24 hours a day, 7 days a week. Thus all of the "up to" language that was surely a part of your "contracted" rate.

But thanks to recent inventions such as Bittorrent and Netflix, certain customers *are* constantly filling their pipe 24/7. And people wonder why ISPs are in such a hurry to institute pay-per-use models...I mean, what other industry that resells scarce, shared-resource services sells it to you at flat-rate all-you-can-eat pricing? Electricity? Water? Telephone? Fuel? None of those.

Now, it absolutely could be argued that if you are seeing 100-200kbit/s down at certain times of the day, they aren't keeping up their end of the bargain because they haven't scaled up their upstream bandwidth to cope with increased demand (especially if they are continuing to install new customers). A successful last-mile ISP will be watching their usage, constantly running the numbers, and making sure that they still have enough capacity to meet demand at any given time. Of course, this is all still done within the assumption of "bursty" usage models, so if they have 10 customers each provisioned at 3Mbit/s down, their models are not going to suggest to them that they need to have 30Mbit/s of total capacity available. So if all of those customers are filling their connections 24/7, then that creates a problem.

And the problem is a real economic problem. The previous poster was correct in saying that the ISP is not making money off of you hand-over-fist. If they are strictly an ISP, I guarantee you they are barely squeaking by. (If they are a regulated incumbent telco with an ISP side-business, like a Verizon or Frontier or CenturyLink, that's another whole story...) ESPECIALLY if they are a rural ISP. Verizon/Frontier sells, what, 3Mbit/s DSL for around $30/mo to residential users? That's great. I *guarantee* you that IF we are talking about a rural ISP, they are LUCKY if they are paying a rate of $30-per-MEG-per-month to their upstream. That would be CHEAP. And you expect them to turn around and sell you a 3 MEG all-you-can-eat circuit for $30/mo? That would mean you are paying them 1/3rd of what that kind of bandwidth actually costs them to get for you. That's called a money-losing proposition.

So, to the OP: by all means, complain to your ISP. For all we know, your problems are not related to constrained throughput as a result of peak usage, and are instead being caused by a physical problem with your circuit. But keep in mind that there's a reason that to this day, getting a connection with an actual contracted SLA is not cheap. There's a reason why you can still find yourself paying $300-500/mo or more to a telco for a T1 (~1.5Mbit/s synchronous) circuit where that throughput is guaranteed. If you actually need 3 megs down guaranteed to you 24/7, then you're going to have to pay dearly for it.

The problem is that nobody is willing to pay what it actually costs.

-- Nathan

I'm not exactly the most vocal contributor here, but as you can probably see, I've been around for a while, and believe-you-me your presence will be missed.

Thanks for /.

-- Nathan

How do you know that he did this, out of curiosity?

-- Nathan

You misspelled "now."

Ah. Sorry. I misread the beginning of that second sentence as you offering a second point in favor of the Nexus S (as in, "they're the most dev friendly, AND they're supported pretty well, too").

Nexus S is Samsung, not HTC.

'nuff said.

(...alright, troll: I'll bite...)

Microsoft wrote OS/2 1.x, yes. Calling anything above 2.x a recompile is pretty silly, though. I'm sure that there is still Microsoft code that exists in IBM-branded versions of OS/2 after their "breakup" (most likely at lower levels, in the kernel or filesystem layers), but MS had absolutely no hand in IBM SOM or the Workplace Shell. WPS came to be after Microsoft decided to wash their hands of IBM.

-- Nathan

This is a non-story, at least how it is written.

As part of iPhoneOS (now iOS) 3.0, in June *2009*, Apple announced that hardware manufacturers would be able to have their hardware directly interface with their iPhoneOS applications, either through the dock connector OR through bluetooth. They have an official set of APIs built into the OS specifically to facilitate this.

I think it was cool that they did this over a YEAR AGO, but hey, that story doesn't make for as sexy a headline as "OMG Apple suddenly loosening their Death Grip on their iPhone hardware?!?!?!"

-- Nathan

P.S. -- No Apple apologist here; in fact, I'm generally very critical of the locked-down nature of the iDevices. But come on...let's strive for accuracy here.

First, I don't think it's about BES (at least exclusively). If BES doesn't also go through RIM's servers (As in: your provider's Exchange server BES gateway Blackberry servers Blackberry phone), then why do all of the articles about this India scuffle to-date talk about "RIM's network"? (Yes, I read TFA.)

Second, if it was about BES and BES worked the way you imply it does, then it seems to me that RIM still shouldn't need to get involved. Sure, the communication over-the-air is all encrypted, but since BES (according to you) is simply a gateway to your corporate Exchange/Notes/Groupwise server, why not simply ask the owner of the server for direct access to its contents rather than dragging RIM into this? Or are you trying to tell me that IMAP4+SSL users or encrypted Exchange users on other phones that don't need to be tied down to a proprietary network will give the Indian government the same kinds of headaches?

-- Nathan

See, this is exactly why device manufacturers shouldn't be making devices that are entirely reliant upon an external "cloud" service that is also controlled by the device manufacturer. If Blackberry was merely making devices that could be configured to talk to any server(s) using industry-standard protocols, they wouldn't get themselves into the kind of situation where 1 million deployed devices could have been turned into doorstops overnight. (Maybe my understanding of the way that Blackberries work is misinformed, and so my rant here could be completely groundless -- and just for the record, I'm open to correction -- but I am under the impression that Blackberries need to be in constant communication with the BIS servers that Blackberry themselves run in order to function.)

This is also why the whole push notification system that Apple came up with for the iPhone is stupid. If something goes wrong with servers that Apple controls, then suddenly that feature across every single phone that has shipped to-date is dead. Device features should not be wholly reliant upon a service that the device manufacturer controls...all you are doing is making a single point-of-failure when you do that.

-- Nathan