Slashdot Mirror

Actually the iPhone plans cost the exact same as their "Nation" plan for the same number of minutes plus SmartPhone Connect Unlimited, which is effectively the same thing.

iPhone plans (the image name is actually "plans_nation.gif")

Nation Plans

SmartPhone Connect Unlimated

GSM coverage is definitely not excellent in the central U.S. plains or the Rocky Mountains. There is coverage now on interstate highways and in cities and larger towns, but smaller towns and highway routes have spotty coverage or none at all. There were a handful of small telephone companies that sprung up to fill the void in the past few years. Two of the largest such have been recently acquired.

Western Wireless was acquired by Alltel.

AT&T recently acquired Dobson Communications (which was the largest vendor licensing the Cellular One brand). I think that the "new AT&T" realizes that Verizon will kick their ass if they don't start competing on the coverage front. Nobody can touch Verizon in the U.S. for actually getting a wireless signal in more places.

The AT&T coverage map is, ahem, optimistic, regarding its coverage in the plains states. The Dobson/Cellular One acquisition helps with this, but not with the rocky mountain states.

Here is a relatively honest assessment of GSM coverage in the western US: Cellular One coverage map (includes the Dobson network and all the other mom-and-pop licensees of the Cellular One brand).

The Alltel coverage map is also a little deceptive, because it mixes their GSM network in with their other network, but they provide much better GSM coverage in the west than anybody else, thanks to their acquisition of Western Wireless. Both T-Mobile and AT&T were asleep at the wheel and either should have acquired Western Wireless. It will be more difficult for AT&T to assemble a network which can rival Alltel or Verizon in the west. Fortunately, they are able to lease time on the Alltel network for their customers, as does T-Mobil. Oddly enough, Alltel does not provide GSM service to its own customers.

It's not clear that AT&T can fix this problem via acquisitions. Alltel already bought the company that AT&T should have acquired. What's left now is a whole bunch of relatively small players. AT&T will need to go on a major buying and building spree if they are really serious about providing nationwide cellular GSM coverage. They have been cherry picking the best (highest population density) markets for years, but they'll need to get serious about providing coverage to the rest of the country if they don't want the "I can't get an iPhone because AT&T doesn't offer service in my area" blogger chorus to drown out their efforts to promote the iPhone beyond the initial product launch. The stock valuations of AT&T and Alltel are interesting, too. AT&T has brand recognition, but they are themselves possibly an acquisition candidate.

http://www.wireless.att.com/coverageviewer/ disagrees. The majority of the country in terms of shear area IS GSM and the overwhelming majority of the country as far as population is GSM. Now, their 3G service coverage sucks, pretty much 800K+ pop metros and very select other places.

(1) It's not widely deployed (a few dozen cities

A few dozen, a couple hundred... but who's really counting. HSDPA coverage isn't *that* bad.

the chipsets that support it presently consume too much power. Apple apparently wants a lower power chipset so that battery life of the iPhone isn't adversely affected.

Reminds me of the old joke:

During the space race back in the 1960's, NASA was faced with a major problem. The astronaut needed a pen that would write in the vacuum of space. NASA went to work. At a cost of $1.5 million they developed the "Astronaut Pen". Some of you may remember. It enjoyed minor success on the commercial market.

The Russians were faced with the same dilemma.

They used a pencil.

Nokia, when faced with the same dilemma, included a battery door.

;-)

Well, the specifications, and typical consensus amongst people on the internet say EV-DO is many times faster than EDGE. Though, I don't use EV-DO for Slingplayer, I use it for work... VPN connectivity, file transfers, remote administration tools; which are typically bursty traffic. I perceive it to be much faster, especially considering my situation is the opposite of yours... I don't even have GSM coverage at all in my home, unless I stand in my driveway, where I get 1 bar, whereas I get a full CDMA signal. We have 2 of each type card here at work that the sales people take when they travel, they prefer the EV-DO cards, and fight and bitch over who gets them.

It's hilarious because AT&T won't even show you where they have 3G/HSDPA coverage unless you zoom into the city level! Hmm... wonder why that is?

You must be comparing 3G/HSDPA to 1xRTT (that would be faster). Because EDGE is not faster than EV-DO (it also has higher latency) nor is it faster than EV-DO rev. A.

There are plenty of smartphones and PDA phones running on AT&T's network that have built-in Wi-Fi. I own an HTC 8525 which has Wi-Fi. I had an 8125 this time last year that had Wi-Fi. And there were ones before it that had Wi-Fi. This is absolutely nothing new. Additionally, AT&T is offering Wi-Fi hotspot coverage in their highest-end consumer data package now -- it's not all too surprising. It seems to me like chatting up Wi-Fi is just a way to hedge the 3G issue, which has already been discussed ad nauseam.

See:
http://www.wireless.att.com/cell-phone-service/cel l-phone-plan-details/?q_sku=sku70005&q_planCategor y=cat1460003

The other issue with 3G rollout is that AT&T has their data network throttled back to 1.8mbps max speed (actually: they may have some places throttled up, but I don't know of at the present time).

The service itself is supposed to max out at 3.6mbps, and those of us using it keep wondering when/if they'll throttle it up.

Here's the link for the 3G coverage map for AT&T...
http://www.wireless.att.com/coverageviewer/

Since we are talking about the iPhone--which I'll repeat--is a GSM phone, I'll stick to my original comment. There's no 3G network to speak of for the iPhone to use, so we're right back at the beginning, where 3G is of no utility in 90% of the United States for the iPhone.

The AT&T 3G / BroadbandConnect network is currently available in most major metropolitan areas and is expanding rapidly. Click on a major metropolitan area below to view coverage for that area.

Jeez, you think I made up the prices and everything rather than copying them from the web site?

http://www.wireless.att.com/cell-phone-service//ce ll-phone-plans/smartphone-connect-plans.jsp

It's called SmartPhone Connect Unlimited with Xpress Mail. Unlimited data, $19.99 a month.

Actually, the iPhone plan is exactly the same as AT&T's regular price plans, to within a few pennies.

iPhone plan = 450 minute $39.99 voice plan plus $19.99 unlimited data smart phone plan with e-mail.

Actually, the iPhone plan is exactly the same as AT&T's regular price plans, to within a few pennies.

iPhone plan = 450 minute $39.99 voice plan plus $19.99 unlimited data smart phone plan with e-mail.

http://www.wireless.att.com/cell-phone-service/leg al/plan-terms.jsp contains the usual cellular crap language. Basically, you can't use it for video, movies, music, aka "any service we'd like to gouge you for because we don't believe in net neutrality." I suspect that this clause is legally unenforceable and eventually a suit will go forward on this. It seems like if you ran a VPN (which is allowed) that handled all your traffic, they wouldn't be able to discern the type of data you were transferring. Bits are bits GD it!

Actually, it's not quite that simple.

Here is who's carrying it:

- Apple retail stores
- Apple online store
- AT&T corporate stores (not corporate kiosks or resellers initially, but those may come in the future)

As to the other points:

Yes, it's always been generally said that iPhone would require a two year contract, but what was said at the keynote was that the agreement between Apple and AT&T was a "multi year exclusive", which has subsequently been revealed to be five years. Until today, it appeared that AT&T would be selling the phones with mandatory activation, whereas you could just walk out with a phone from the Apple store. So, a lot of people were hoping that the phone was hackable or easily unlockable, so as to use it with other carriers. If the iTunes activation thing is easily worked around AND the phone is unlockable in some way, then you might be able to move it to another carrier. The worst case scenario - assuming the phone is unlockable - is that you sign on with AT&T, have them unlock the phone after 90 days (assuming it's unlockable in this way), cancel the contract, and move to another carrier. Obviously, a lot of people are hoping they can just buy the hardware, unlock it somehow, and use it on the carrier of their choice. That may still be possible, but we really won't know until people start playing with these things.

So... for 450 minutes and unlimited data,

ATT is $39.99 + $19.99

and Verizon is $79.99

ATT Data
ATT Voice
Verizon Bundle

How sure are you the iPhone data plans will cost more than a Treo now?

So... for 450 minutes and unlimited data,

ATT is $39.99 + $19.99

and Verizon is $79.99

ATT Data
ATT Voice
Verizon Bundle

How sure are you the iPhone data plans will cost more than a Treo now?

Here's their CURRENT pricing, iPhone's will surely be different, but I doubt it'll be any worse than what they charge now. I'm guesstimating the unlimited data for iPhone will be $19.99-$34.99

IIRC, around 2000-2001, AT&T announced they were going to start purchasing GSM hardware to upgrade their network instead of switching to CMDA. As I recall, it was a big blow for the adoption of CDMA.

Anyways, AT&T still supports TDMA phones, just not for much longer.
http://www.wireless.att.com/learn/articles-resourc es/tdma-notification.jsp

"AT&T will be shutting down our TDMA and Analog networks in early 2008 and will begin that process by turning down TDMA service in about eighteen markets in 2007."

AT&T is not AT&T now, because the name was sold to SBC. My understanding is that the SBC trademark was worse than useless because the company is so abusive. So, the managers decided to use another name.

Those interested in how that happened can watch Stephen Colbert explain in a 1 minute 14 second video: The New AT&T.

SBC taking the name AT&T is, in my opinion, a kind of legal fraud, but fraud nevertheless. People are bound to be confused and misled. AT&T had a very good reputation. SBC-AT&T is a completely different company, and has no connection in its culture with the old AT&T. At the very least, the SEC should require the company to disclose in the first sentence of any prospectus for its stock that there is no connection whatsoever.

The growth in market capitalisation, in this case, apparently does not necessarily mean success. The growth is only because SBC is buying companies. (I have done little research on the history of SBC, so I say apparently.) So, there is a huge potential for investors to be fooled.

The recent history of the former AT&T is interesting. David Dorman, a manager who presided over the failure of his company made $30 million from selling the company to other managers who are reputed to be just as inept and corrupt.

The Wikipedia article says, "Dorman's management finesse can be ascertained by tracking the value of AT&T stock during his tenure." Dorman became president in 2000, Wikipedia says. The stock performance tanked beginning just before that, and continued down until the AT&T name was sold to SBC.

Is this a correct summary of the story? The suspicion is certainly raised that the former AT&T manager got $30 million for reducing the value of his company so that the cost to the buyer, SBC, would be less.

The AT&T 3G network isn't widely deployed in the U.S. Since the iPhone supports WiFi, high speed access to networks will be available to people on the iPhone in WiFi hotspots. EDGE will work fine for SMS text chats and email. When the 3G HSDPAnetwork gets rolled out more widely, we will undoubtedly see 3G capable iPhones to use it. Meanwhile, EDGE is deployed in 13,000 cities, 3G in only a few dozen.

See the AT&T Wireless coverage chart (and see link below the chart for a list of 3G cities).

Colorado is not even covered at all yet. Around 100 major metro areas does not a country of metro coverage make.

American mobile-phone technology is five, maybe seven years behind Europe and Asia.

This is due in large part to geographic size, and the nature of the marketplace as mobile telephone services were rolled out in the United States.

Large metropolitan areas have coverage more or less on par

Features which are acceptable in the USA (e.g., EDGE,

It's not tht EDGE is "acceptable"; it's that it's what most of the coverage on AT&T's network actually is. And a large part of it, believe it or not, is economies of pure geographic size. AT&T does have a lot of 3G coverage in metro areas, and CDMA carriers such as Sprint and Verizon have even broader 3G coverage.

simlocks,

The US isn't the only place that does carrier SIM locking, though I'll admit it's definitely standard practice for US carriers. However, manufacturers do sell unlocked phones in the US (such as Palm), and carriers (inlcuding AT&T) will unlock phones for customers (permanently) for things like international travel and the use of prepaid SIMs.

contract-locked Wi-fi

There is no reason I can think of that you'd have to have a "contract" to use WiFi on iPhone. The stories going around implying this is the case are assuming because someone said you needed a contract to get an iPhone (and therefore use the iPhone's WiFi) that you must need a contract to use WiFi from a technical standpoint. That's garbage. What if the phone is out of AT&T (or any) coverage? The phone can't "know" whether it's on contract; do you actually believe the WiFi simply won't work? That's ridiculous. WiFi isn't "locked" to anything.

Since you need a two-year AT&T contract anyway, this is a moot point. But if you let the contract expire or pay a termination fee, WiFi won't just stop working.

etc) are so archaic as to provoke spontaneous laughter when described to non-US mobile users. Just look at the terminology -- fully half* the phone users outside the USA would have no idea what a "cellular" phone is.

This actually proves my point. This kind of technology (mobile telephones) was prevalent in the US long before it was anywhere else in the world.

It's a mobile phone. Mobile across networks, user SIMS, and national borders.

So is iPhone.

The simple fact that the parent post asks rhetorically "would it be cool if the first-gen Iphone had 3G?" amazes me. Jesus, is it still 2002 in the USA or something? If Apple takes that attitude to Europe it'll get laughed at. And it is.

To reiterate something I said in another post:

Large numbers of people in major metro areas will still purchase iPhone, and WiFi will also mitigate the need for 3G for a lot of people. Those who really need 3G in a handheld device don't have to get an iPhone.

Obviously, future generations of iPhone will have added functionality. Apple has already said 3G is coming in the future. Further iPhone is a half-year out for Europe, and it's not a foregone conclusion that today's iPhone specs MUST mean that 3G won't be there when iPhone ships for Europe, and even if it isn't we already know it's coming soon, because Apple specifically said so at the iPhone introduction in January. So if you NEED 3G, don't get the first gen iPhone. How hard is that?

Its not about a data plan, the story is that in order to use WIFI on your phone(a non AT&T network feature), you will be forced to sign up for an expensive data plan.

But what if they said instead that, "In order to buy the phone, you must buy the data plan. WiFi comes with the iPhone, but the iPhone isn't available without the data plan."

Personally, given how vague and anonymous the claim is that "One anonymous AT&T store manager said users will get their WiFi...when they sign a contract locking them into a data plan and EDGE," I'm not sure how to take that. It might even mean that at&t is going to provide WiFi service as part of a package deal with EDGE service.

In other words, until you know what this claim means and whether it's true, there's nothing to complain about.

I agree that EDGE isn't exactly exciting, but it does have much better coverage than the newer 3G Networks in the U.S. It would be nice if the AT&T data plan for the iPhone bundled their AT&T WiFi Hotpot aka LaptopConnect service with EDGE.

The future of the AT&T wireless network shows signs of being much brighter. Their HSDPA is being improved and it appears that they are upgrading the existing sites to 14.4 Mbit/sec service. If they roll that technology out nationwide I for one would welcome our new high speed wireless overlords.

I agree that EDGE isn't exactly exciting, but it does have much better coverage than the newer 3G Networks in the U.S. It would be nice if the AT&T data plan for the iPhone bundled their AT&T WiFi Hotpot aka LaptopConnect service with EDGE.

The future of the AT&T wireless network shows signs of being much brighter. Their HSDPA is being improved and it appears that they are upgrading the existing sites to 14.4 Mbit/sec service. If they roll that technology out nationwide I for one would welcome our new high speed wireless overlords.

AT&T's smartphone unlimited data plan is 19.99/month. 9.99 for the 5MB/month plan.
Whether this applies to the iPhone is still anybody's guess, however.

AT&T is not AT&T now, because the name was sold to an abusive west coast telephone company named SBC.

My understanding is that everything else of value in the original AT&T was sold piece-by-piece, and SBC bought mostly just the name. My understanding is that the SBC trademark was worse than useless because the company is so abusive. So, the managers bought another name.

Apparently, for $16 Billion SBC got AT&T's VOIP customers, and the AT&T name.

AT&T's VOIP customers were Sheila and Gerald Funk, who have since moved to Elbonia. Wait... That last sentence my contain an error.

So, what we are seeing is SBC mismanagement under a new name. Soon just saying the name AT&T will cause people to become upset.

>> "was and is a key goal when C++ was developed."
> Oh really? Well thats the first time I've heard of it. Provide a link.

http://www.research.att.com/~bs/hopl-almost-final. pdf

especially section 2, anything on generic programming (like 7.2.1), ...

> Well thats helpful when you're a maintenance programmer faced with 10,000 lines of code that could be
> doing anything and you have no idea what because the guy who wrote it thought it would be cool to
> overload every operator he could think off and to hell
> with readability and comprehension 2 years down the line.

we were talking about boost here, where the interfaces are reviewed to be both correct and sensible.
your vectorAdd is a great example. The natural expression (as any mathematician will tell you) is v1 + v2.
Yes, if you use templates and operator overloading you cannot "count cycles" any more, i.e. you have no idea if, when or how many functions are called. There might be expression templates which optimize the operation, or even offload them to the GPU.

Of course tracing and debugging is not as easy at first, but that is mainly a matter of the experience and the approach you take. Traditional debugging by "extensive code gazing" will not work so well, but proper test cases will.

>> "This helps to really rise the abstraction level of the language to solve problems of the problem
>> domain, in a very efficient way."

> Bullshit. Its no more efficient than function calling , its just a way for coders show off.

Expression Templates are a great optimization, and operator overloading gives you a reasonable syntax
i.e. do you really consider M = matVecMul(A, v).vecAdd(b); more readable than M = A * v + b;
try more involved examples which include precedence, like M = A * (v1 + v2) + v3;

I love this comment, from Bjarne Stroustrup's home page (href=http://www.research.att.com/~bs/bs_faq2.html #memory-leaks)

Q) How do I deal with memory leaks?

A) By writing code that doesn't have any. (goes on to advocate vector & string)

And also: C++ Is my favorite garbage collected language because it generates so little garbage (http://www.research.att.com/~bs/bs_faq.html#reall y-say-that)

Over the past 6 months or so, I've really made an effort to better my usage of C++ (using Effective C++, Effective STL and C++ Coding Standards). With a combination of STL, references, RAII, std::string and boost::shared_ptr, all of my memory, ownership & null-pointer problems just went away. I hardly ever actually write 'new' any more. The Java model of just leaking objects and hoping they'll get collected sooner or later seems horrible.

But I'm not maintaining old code, so this is completely -1 Offtopic.

I love this comment, from Bjarne Stroustrup's home page (href=http://www.research.att.com/~bs/bs_faq2.html #memory-leaks)

Q) How do I deal with memory leaks?

A) By writing code that doesn't have any. (goes on to advocate vector & string)

And also: C++ Is my favorite garbage collected language because it generates so little garbage (http://www.research.att.com/~bs/bs_faq.html#reall y-say-that)

Over the past 6 months or so, I've really made an effort to better my usage of C++ (using Effective C++, Effective STL and C++ Coding Standards). With a combination of STL, references, RAII, std::string and boost::shared_ptr, all of my memory, ownership & null-pointer problems just went away. I hardly ever actually write 'new' any more. The Java model of just leaking objects and hoping they'll get collected sooner or later seems horrible.

But I'm not maintaining old code, so this is completely -1 Offtopic.

what would a US citizen be doing with a GSM phone besides travelling? Despit the Global name, they don't work in the US now do they?

http://www.wireless.att.com/learn/why/technology/g sm-gprs.jsp

You're welcome.

will be available in parts of Southern California communities initially.

It is also available in San Antonio, TX, Dallas, TX, Austin, TX, Milwaukee, WI, and Indianapolis, IN. Source

UWIN is great, much higher performance than Cygwin, and higher level of standards compliance. Open source and well supported.

http://www.research.att.com/sw/tools/uwin/

It does seem that a large part of the T-1 price is due to the upload speed. If we compare it to symmetric DSL, it is priced similarly. Here are the prices that I just got from http://www.att.com/dsl:

AT&T Business Class DSL
Download/upload speeds (users/monthly price)
  Multi-IP IDSL 144-144Kbps ($142.45)
  SDSL 192-192Kbps ($142.45)
  SDSL 384-384Kbps ($179.95)
  SDSL 768-768Kbps ($188.96)
  SDSL 1.1-1.1Mbps ($244.96)
  SDSL 1.5-1.5Mbps ($279.96)

All these posts are completely missing the point and not understanding at all how T1 (voice or data) pricing is determined. This is not a technical or economic issue at all (that would be too logical). It is simply tariffs. The PUC decides what LECs and CLECs can charge. CLECs can charge less (to improve competition), but your LEC has a fixed price that it *must* charge for T1 in your area. There is *no* getting around this. It really comes down to room full of old white men deciding the price.. Here is an example for at&t in California.

http://serviceguide.att.com/tariff/business/ext/fi les/CAPL11PLM1.pdf#page=4

These are the prices *period* (few exceptions for special state contracts such as CALNET)

Movielink's website may not support Linux but their service is available on Linux based set-top-boxes. If someone wanted to make an application (or possibly an extension) to interface with the Movielink service for Linux, they would probably be happy to cooperate as it would increase business.

Indeed it was.

I prefer tools like Ggobi http://www.ggobi.org/, its predecessor XGobi http://www.research.att.com/areas/stat/xgobi/ or commercial products like SAS' insight or jmp. I'm sure there are others. They allow you to tour and manipulate the data through linked plots and displays and selectively turn on and off elements.

First, keep in mind that it's possible to get unlimited service through smaller mobile providers in the area of $30-$70/month. This article is just about one national provider's offerings.

I assume that Finland (like most countries in Europe) charges more to call a mobile phone from a landline. AT&T (landline) tells me that there is a $.12/min surcharge to dial a mobile phone in Finland versus a landline.

These charges don't exist in North America. Mobile phones are numbered exactly the same as landlines.

So even if you have an "unlimited" plan in Finland, presumably your friends dialing from a landline are paying around 5 euro/hour to talk to you. An unlimited plan here is completely unmetered for both incoming and outgoing calls. Don't forget to figure the cost of both sides when making a comparison!

You could start here: http://www.research.att.com/~njas/sequences/A00079 6

AV Rule 25 (MISRA Rule 127)
The time handling functions of library <time.h> shall not be used.

Saw Hugh at the RSA conference, where he told this story and talked about a new internet talk show that he's hosting. Don't know that it's for everyone, but the slashdot crowd should dig it.

Is anyone (here) surprised by this? It seems painfully obvious to me, that most such services obviously wouldn't work.

My home phone service is through AT&T CallVantage VoIP. AT&T has a FAQ on its CallVantage information pages specifically about this issue. And I quote:

No, this service does not support home alarm or security systems.

What's more, I seem to remember I was shown this information during the sign-up process and had to acknowledge on a terms-of-service agreement form that I understood that this service was not to be used for home security systems.

Elsewhere on the site AT&T discusses the fact that a power outage will knock out your phone service -- in fact, it's in bold type.

I agree with the others who say it's probably the ILECs behind this kind of FUD(*). It's not that the problems don't exist ... it's that you shouldn't be complaining about them after you've gone ahead and signed up for the service, because nobody is concealing this information.

(* Ironically, however, my VoIP provider is now technically the same company as my ILEC ... though the two services are offered by very different divisions.)

Is anyone (here) surprised by this? It seems painfully obvious to me, that most such services obviously wouldn't work.

My home phone service is through AT&T CallVantage VoIP. AT&T has a FAQ on its CallVantage information pages specifically about this issue. And I quote:

No, this service does not support home alarm or security systems.

What's more, I seem to remember I was shown this information during the sign-up process and had to acknowledge on a terms-of-service agreement form that I understood that this service was not to be used for home security systems.

Elsewhere on the site AT&T discusses the fact that a power outage will knock out your phone service -- in fact, it's in bold type.

I agree with the others who say it's probably the ILECs behind this kind of FUD(*). It's not that the problems don't exist ... it's that you shouldn't be complaining about them after you've gone ahead and signed up for the service, because nobody is concealing this information.

(* Ironically, however, my VoIP provider is now technically the same company as my ILEC ... though the two services are offered by very different divisions.)

http://www.research.att.com/viewProject.cfm?prjID= 69

Rebuild the Death Star? FCC says "Cool! Watch out for snub fighters!"

Merge two niche radio markets? Nuh uh. Not gonna have any of that. No interoperability for you!

The FCC should go back to policing wardrobe malfunctions.

See also a table of the first 1,000 greater of prime twins, and more references and stuff at:

"A006512 Greater of twin primes."
  at The On-Line Encyclopedia of Integer Sequences
http://www.research.att.com/~njas/sequences/A00651 2

5, 7, 13, 19, 31, 43, 61, 73, 103, 109, 139, 151, 181, 193, 199, 229, 241, 271, 283, 313, 349, 421, 433, 463, 523, 571, 601, 619, 643, 661, 811, 823, 829, 859, 883, 1021, 1033, 1051, 1063, 1093, 1153, 1231, 1279, 1291, 1303, 1321, 1429, 1453, 1483, 1489, 1609, ...

-- Prof. Jonathan Vos Post

Likewise for memory leaks: http://www.research.att.com/~bs/bs_faq2.html#memor y-leaks

Beyond pandering to end-users' collective inferiority complex, what's Platt's central thesis? Is he really saying that high IQ causes bad software? That programmers are control-freaks who create functionality rather than usable software? O.K.

So what's he suggest? VB? Should we just assign stupider people to UI? Maybe he's of the school of development theory that assails Bjarne Stroustrup regularly with questions like, "Is C++ too hard for most programmers?" [1] (Bjarne's answer, btw, is always "No, C++ is exactly as hard as it is, although it's true some people shouldn't write C++").

Frankly, it's not smart programmers who write crappy error dialogs, but rushed, sloppy or dumb programmers and tech writers. I'd sooner blame poor quality on a misguided belief that the general dumbing down of programming tasks makes life less complicated and makes programs more robust. It isn't, and it doesn't [2].

Software development has always been exruciatingly dependent upon individual talent, and increasingly people don't want to pay the freight for polished, elegant work.

[1] http://www.research.att.com/~bs/MIT-TR-original.pd f
[2] see Mythical Man-Month, or Code Complete.

Have you seen the new AT&T deathstar? It's like the softer side of the Sith!