Just heard from one of the folks we know who updated his iPhone 3GS to iOS 4 yesterday only to lose his AT&T data connection completely. Today, when he could no longer even place calls, he contacted AT&T who, after several escalations, informed him there are isolated incidents of iPhone 3GS on iOS 4 with unlimited data plans no longer being able to connect to the network.

They told him they're working with Apple to fix it, and put him name on a contact list so they could let him know when it was resolved.

Our question is -- how isolated is isolated? How's your AT&T unlimited data post iOS 4 update?

Princeton University IT, which keeps a close eye on their infrastructure, claims the iPad is not being a good network citizen:

The malfunction we see is that the iPad uses DHCP to obtain a lease, renews the lease zero or more times (as expected), but then continues using the IP address without renewing the lease further. The iPad allows the DHCP lease to expire, but it continues using the IP address after allowing the lease to expire. The incident continues for some time (typically hours); usually it ends when the iPad asks for a new DHCP lease, or the iPad disconnects from the network.

Princeton says they're communicating with Apple to resolve the issue, but in the meantime they -- and other large institutions -- are keeping the misbehaving iPads off their network.

This won't be an issue for a home user or anyone on a small network where IP conflicts are unlikely, but it's something Apple will need to fix in the near future, especially for the education market for which the iPad is so well suited.

Have you noticed any iPad problems on your network? Let us know in the comments!

[Thanks to everyone who sent this in!]

The Wall Street Journal, which kicked up the old iPhone on Verizon rumors (and iPhone HD!) again this week, now says AT&T doing everything they can to fix their network and optimize their service before any customers even think of jumping to Big Red.

First, they've very publicly acknowledged problems in cities like New York and San Francisco and invested in infrastructure to get them up to speed (and to stop the drop). In total, $2 billion will be spent building out the network.

However, dealing with the sheer volume of iPhone users on AT&T is an unparalleled problem, they maintain (and one many suspect would have plagued any single carrier).

For example, AT&T said when iPhone customers started checking their email and surfing the Web from their high-rise offices, AT&T repositioned its cellular antennas to point up, instead of down. Rivals will start the process of making the same changes only after the phones hit their networks, it said.

AT&T also spent a lot of time with Apple trying to fine-tune how the iPhone dealt with their network, and give the Cupertino engineers a crash course on wireless. They returned regularly and, according to the WSJ, helped with new technologies to "lighten the load".

Apple rejiggered how its phones communicate with AT&T's towers. As a result, the phones now put less of a load on the network for such simple tasks as finding the closest tower or checking for available text messages.

It's an interesting, if decidedly pro-AT&T article but well worth the read. If you're on AT&T, are you seeing actual, real world improvements? Would you jump on (or jump to) a Verizon iPhone if one is ever released?

The UK's leading provider of mobile phones, O2, is openly admitting to having iPhone related network issues. It seems AT&T might not be alone after all, as O2 chief executive Ronan Dunne recently let out the secret that the last 6 months the iPhone had overwhelmed its network in London. Of course, O2 is claiming that the issues are all fixed and the network is as healthy as ever.

O2 iPhone owners have suffered through some of the same issues as some AT&T customers have experienced - dropped calls, data inconsistency, etc... So the big question is what has O2 done to solve the issue at hand? They've gone and spent a ton of money on the network to meet it's demand including adding 200 additional mobile towers in the areas of need.

Any of our readers currently on the O2 network care to comment on your current service? Let us know in the comments below!

[Via 9to5Mac]

UPDATE: TiPb asked for AT&T's side of this story, and here it is:

"The AT&T wireless network is designed and engineered to deliver the highest possible levels of capacity and performance. Our standing as the nation's fastest 3G network is validated by multiple third-party testing organizations on the basis of millions of drive tests annually.  

"We believe that recent online speculation regarding AT&T wireless network configuration settings is without foundation. Allegations in these posts regarding packet loss network settings are incorrect."

ORIGINAL: In a post entitled Has AT&T Wireless data congestion been self-inflicted? the blog Communications explores whether the iPhone-on-AT&T problems we keep hearing about are the result of misconfigured buffers in AT&T's mobile core network leading to congestion collapse.

It appears AT&T Wireless has configured their RNC buffers so there is no packet loss, i.e. with buffers capable of holding more than ten seconds of data. Zero packet loss may sound impressive to a telephone guy, but it causes TCP congestion collapse and thus doesn't work for the mobile Internet!

The article is way over our heads, but give it a read and let us know your take-away. There's obviously something going on that results in iPhone users on AT&T, especially in New York and San Francisco having connection issues, dropped calls, etc. Could this, at least in part, explain it?

UPDATE: AT&T has contacted TiPb to say infrastructure spending is NOT the reason for iPhone connection problems.

ORIGINAL: Saturday Night Live's Weekend Update sideswiped iPhone network problems last night, the latest in a series of mainstream reporting on the issue:

"It was reported this week that Google would soon launch its own cellphone as a challenge to the iPhone. Also a challenge to the iPhone? Making phone calls."

Whether AT&T is to blame, or some combination of how AT&T's network and the iPhone work together, we don't know. While AT&T was content to increase data revenue while decreasing infrastructure investment all in the name of shareholder value, it's perversely harder to ignore bad publicity than it is unhappy customers.

To quote Fake Steve, whose ill-conceived Operation Chokehold did succeed in bringing a lot of the current attention to bear:

AT&T, a huge wireless provider in the United States, cannot reliably connect calls in New York City. How can this be?

SNL video after the break!

[via Engadget]

[YouTube link]

If Apple ends US iPhone exclusivity with AT&T in 2010, could Verizon handle the handset that currently crushes service in data-dense cities like San Francisco and New York? BusinessWeek scored the quote from Verizon Wireless Chief Technology Officer Anthony Melone:

"We have put things in place already. We are prepared to support that traffic."

"It comes down to backing that process with money. We've been more consistent than any carrier in the last 10 years investing year over year."

"We will handle it if we ever get it."

To make sure we're absolutely crystal clear, no one is saying Verizon will be getting the iPhone next year, Verizon is just claiming their network is up to the task if they do.

When reached for comment on that claim, AT&T wouldn't address it directly but offered:

"We think we are leading the way in how people use their wireless phones. We operate a great network."

This comes as AT&T is facing increased media (and satirical) attention over their lack of infrastructure investment despite the increase in data revenue the iPhone has brought them. To their credit, however, they are continuing to publicize the "improved wireless network experience" they have invested in for certain areas of the US.

[Thanks to the Reptile!]

AT&T Mobility CEO Ralph de la Vega has told the Wall Street Journal that despite what he said back in October and earlier in December:

"We have not made any decision to implement tiered pricing."

Which is a weasel-wordy way of saying, contrary to previous statements that suggested they might do just that, AT&T isn't planning on charging customers by the byte -- yet. And they're still going to call their plans "unlimited" even as they try to get customers to use less 3G data via:

• More WiFi hotspots
• Microcells (femtocells) which re-route connections through your own broadband cable or DSL modem, effectively creating a mini cell tower in your home.
• Incentives (though de la Vega declined to comment on what those might be exactly)

Thanks in large part to the iPhone, AT&T data volumes have shot up 5000% and, according to a lot of recent blog-spun controversy, they've seen profits increase as well while simultaneously investing less and less in the network meant to support that usage.

Fake Steve Jobs, the nom de guerre of Newsweek's Dan Lyons, got a ton of attention for his Operation Chokehold campaign to effectively DDoS the AT&T data network, including from AT&T itself and the FCC, not to mention pretty much every commenter on the interwebs who, while they might have applauded the cause, didn't much appreciate the method.

Along with a quick follow up conversation with Fake AT&T CEO Randall Stevenson, where Fake Steve again absolutely excoriates the iPhone's lone US carrier for making billions in profit on iPhone data plans while apparently cutting investments in the very network infrastructure on which the iPhone is supposed to use that data, he first tried to back track a tad, and then just went... a little nuts.

As has happened in the past, Fake Steve is taking the criticisms and spinning it into a farce including hooks into the Tiger woods scandal, historic figures of social conscious, terrorists, former and current heads of state, and now pretty much everything short of a kitchen sink app for iPhone.

In the end, we can't help but think his original point -- that AT&T isn't investing in a network to support the iPhone and future mobile computing platforms even though they have the resources to do so, and are maximizing short-term shareholder profits over long-term share-holder and customer value -- has gotten lost in the gimmick.

And that's a shame, because AT&T really needs to invest in their network and give iPhone users the infrastructure they're paying for.

UPDATE: Gizmodo also shows, in nifty, PowerPointy form, how AT&T is making more money, yet spending less on networks, since the original iPhone was released.

[Thanks to everyone who sent in many and wonderful variants of all this!]

Is AT&T to blame for the poor iPhone experience in cities like San Francisco and New York, where calls drop, data fails, and bars depict signal strength with no real connection behind them? And if so, what can they do about it -- build more network infrastructure, create tiered pricing, or maybe just give up on exclusivity?

Dan Lyons, writing under his nom-de-guerre Fake Steve Jobs recently posted a curse-filled parody, describing an entirely fictional, frighteningly plausible conversation between his character and an equally fake AT&T CEO, Randall Stephenson. It's climax:

And now here we are. Right here in your own backyard, an American company creates a brilliant phone, and that company hands it to you, and gives you an exclusive deal to carry it — and all you guys can do is complain about how much people want to use it. You, Randall Stephenson, and your lazy stupid company — you are the problem. You are what’s wrong with this country.

I stopped, then. There was nothing on the line. Silence. I said, Randall? He goes, Yeah, I’m here. I said, Does any of that make sense? He says, Yeah, but we’re still not going to do it. See, when you run the numbers what you find is that we’re actually better off running a shitty network than making the investment to build a good one. It’s just numbers, Steve. You can’t charge enough to get a return on the investment.

AT&T has made billions in profit off of its user base (and off the iPhone!) and many of those users think it would behoove AT&T to take a large portion of those profits and re-invest them in expanding and improving their network. AT&T claims they're doing just that, especially in high iPhone-density cities like San Francisco (now getting the 850Mhz band) and Dallas (upgrade to 7.2Mb HSPA). And as Fake Steve so deliciously skewered, AT&T Mobile CEO, Ralph de la Vaga has unfathomably discussed stopping users from using their devices under the "unlimited" data plans AT&T markets to them.

But is the problem really AT&T?

The New York Times recently ran an article claiming AT&T had a great network despite consumer dissatisfaction... a great network for every other phone other than the iPhone. Of course, few other data-centric phones are as numerous as the iPhone, and none are as easy to use, or have as many users using as many data-centric features. Not to mention other carriers, such as Rogers in Canada and GSM networks across Europe don't seem to report the sheer number of problems AT&T users do. (We also remember with horror what happened when CrackBerry.com's Kevin took his just-release Rogers BlackBerry Bold to New York.) Perhaps it's the unique combination of AT&T's specific network setup and Apple's iPhone radios.

Either way, the perception problem is entirely AT&Ts at the moment and even with new customer-facing strategies like "Mark the Spot", an app that lets iPhone users report problem areas, it's not likely to change any time soon.

So let's say AT&T does invest billions in infrastructure -- more fast 3G HSPA bars in more places. It's the right and logical things to do, and the thing Fake Steve absolutely nails AT&T for being too greedy to go about doing. The end result of that could be higher user satisfaction -- and where do that lead?

Many suffer poor AT&T service just to own an iPhone. If they didn't have to suffer any more, if AT&T's network was considered as vast and solid as Verizon's, how many more might jump on it? Could even a greatly enhanced and expanded AT&T handle 10 million more people getting iPhones and using even more data, requiring billions more to keep up, and who knows how much to actually get ahead of demand?

AT&T's stick to go along with their network expansion carrot is, of course, capped data and tiered pricing. 3% of users "watching video" (or unofficially tethering, perhaps), using 40% of network resources. (And again, AT&T sold their bill of goods as "unlimited" so it's hard to sympathize). But even capping, throttling, and/or tiering those 3%-ers won't stop the millions of other hitting AT&T's towers over and over again like high volume machine-gun fire. It's not tenable. (Unless they're willing to accept their destiny and become "dumb pipes", then we'll talk).

So that leaves moving the iPhone out of AT&T exclusivity and onto other US networks. It's happened in the UK and Canada.

AT&T acknowledges it will happen eventually. The date is unknown to anyone outside the contract-signers, but exclusivity is generally pegged to end in 2010 -- perhaps the end of 2010.

It won't be an easy transition -- T-Mobile uses a different frequency for their 3G bands and Verzion and Sprint use an entirely different radio technology. (Yes, even if Apple sold the iPhone 3GS unlocked, for use on any carrier, the only US carrier that whose 3G network is compatible right now is AT&T). That means, even with Verizon being interested, Apple would have to add T-Mobile's bands, perhaps switch to an entirely new, GSM/CDMA hybrid radio so that it can reach America's three other networks. A non-trivial solution to say the least, but perhaps a necessary one now.

If volumes keeps growing, even Verizon couldn't handle the iPhone by itself either. Just like new highways ease traffic congestion, letting the iPhone speed along several carriers might just make it better for everyone involved -- including AT&T.

If anyone can do it, Apple can. If not, Google might just be waiting in the wings...

AT&T has released a new iPhone app called AT&T Mark the Spot [Free - iTunes link] intended to allow customers to send in real-time, location-specific feedback about dropped calls, coverage gaps, or other network problems, have occurred.

AT&T is committed to providing its customers with the best network experience possible.

This application will help contribute towards this goal and its utilization is greatly appreciated.

AT&T is often criticized for poor signal quality and network availability when it comes to the iPhone, is this a positive sign that they're trying to make things right?

[Thanks Keith and Gregg for the tip!]

According to a Gizmodo reader who took his iPhone to the Apple Store Genius Bar due to issues with dropped calls, he was told a 30% failure rate in New York City is normal.

Now, we all know AT&T's network crumbles beneath the weight of the iPhone (and suspect any other single network might as well), but it's not often we get numbers to go with it.

AT&T claims to be improving their network and adding more frequency, but with a fail rate that high, New Yorkers will believe it when their calls stop dropping.

Boy Genius is reporting that AT&T MMS seems to be down for some users in some states, and what's more:

a quick call to AT&T’s customer care line revealed that there is a known latency issue with MMS in all states with no estimated time of repair.

So how about it? Are you experiencing any delays in sending or receiving MMS? Any outages? Either way, let us know where you are, and how long you've been having the problem.

Has AT&T been delaying the US launch of iPhone 3.0's MMS and tethering services due to concerns about their network being able to handle it? Um, yeah, that would have been our guess... The New York Times, however, states it as fact:

[AT&T] has also delayed bandwidth-heavy features like multimedia messaging, or text messages containing pictures, audio or video. It is also postponing �tethering,� which allows the iPhone to share its Internet connection with a computer, a standard feature on many rival smartphones. AT&T says it has no intention of capping how much data iPhone owners use.

How big is the concern? AT&T claims they've diverted $18 billion to upgrade and expand their 3G network to handle the load, but that getting local approval to build towers takes time, as does upgrading existing infrastructure.

Analysts quoted seem to agree with TiPb readers that AT&T may just have been hit first and hardest by the iPhone, but other networks will face the same problem if/when they start to see iPhone class devices hitting their towers in the same numbers.

So, is AT&T doing the right thing delaying MMS and tethering until, you know, they'll actually work, or do you just want your features and want them now? Or do you just not buy this whole "data usage conspiracy" at all?

Techcrunch asks the impertinent question: can AT&T handle the iPhone. The pertinent answer thus far is: no. The iPhone is a consumer success the likes of which no smartphone has experienced before. There are more users using more features that consumer more bandwidth that likely even Apple or AT&T ever estimated, and it's put an extreme hurt not only on existing infrastructure, but a hurt that's growing faster than infrastructure expansion can handle.

The answer to many is simply to have the iPhone on Verizon, which is believed to offer a better network. While obviously splitting the iPhone between AT&T and Verizon would lesson the individual demand on both -- load balancing the user pool, so to speak -- we're curious as to whether or not Verizon could have, or could still, handle the iPhone all by itself.

It's largely reported that Verizon was the first US carrier offered the iPhone after all. If they'd said yes, and if the iPhone grew on Verizon as fast (or faster, given their reach) than AT&T, would Verizon have suffered the same problems -- and bad reputation -- AT&T is suffering now? CDMA towers, while serving more with less, still have their limits, after all. (TiPb's heard that some feel BlackBerry hits CDMA networks hard -- they ain't seen anything like the freight-train of hurt the iPhone is bringing.)

So, what do you think?

What's causing all the commotion down in Austin, Texas right now? Turns out it's not only iPhone news at South by Southwest 2009 (SXSW), but it's everyone using iPhone 3Gs as well! It's SMASHING AT&T's network.

We've joked before that AT&T had basically gone out and strapped rabbit ears on their EDGE antennas and called it 3G, well if that's the case, they're rapidly strapping even more ears on round Austin way right now, according to Business Insider:

AT&T (T) tells us it's adding wireless capacity in downtown Austin to deal with "unprecedented" demand (that's resulted in service that's flaky at best). This, of course, a result of thousands (tens of thousands?) of South by Southwest attendees bringing 3G iPhones into this part of Austin for the first time. The telco expects service to improve tonight and throughout the rest of the show.

Good news to most techies, however: while AT&T was down, at least Twitter was mostly up...

[via Daring Fireball]

Know how some people are complaining that they have trouble connecting to AT&T's 3G network? How they drop calls? How they blame Apple? (Despite the phone working pretty dang well in other countries on other carriers). Remember the theory that there were so many iPhone 3Gs hitting the market that AT&T couldn't handle the load? (That their network was basically rabbit ears tied to old antennas? -- okay, we made that last one up!)

Now imagine that each and every one of those iPhones, especially in high-density areas like San Francisco, Los Angeles, and New York -- where each tower is already slicing bandwidth to razor thin margins -- suddenly found itself barraged by laptops tethering themselves on for the ride.

That might just be what's keeping AT&T from allowing iPhone tethering: fear it will crush their already strained 3G network. MacBlogz claims to have a source saying just that (via Gizmodo):

“Regardless of how many billions of dollars AT&T pours into their 3G network, it hasn’t been stable enough to handle all you iPhone users.”

Roughly Drafted is claiming a source close to AT&T has spilled the beans on what's really going on with the iPhone and its 3G connection problems, and what 2.0.2 did to fix it.

In a nutshell? An iPhone 3G running 2.0 or 2.0.1 tries to pull too much power from the network, so when multiple iPhones connect, a tower can actually run out of juice and start dropping calls and losing data.

Why hasn't upgrading to 2.0.2 already fixed the problem? Simple: some people haven't yet upgraded, so their 2.0 and 2.0.1 iPhones keep pulling too much power, causing the same problem even for people who have upgraded but are stuck on the same tower (or same high-density city like San Fran or NY). Only when most users have patched to 2.0.2 will people stuck on high-demand towers see improvements.

Earlier reports and theories have lain blame on everything from the 3G radio and antenna, to the Infineon chipset and Apple firmware, to the carriers themselves. We here at TiPb have long been saying the problems were likely a combination of factors, and firmware that pulls too hard on networks that aren't that hardy seems a far better explanation than any one previously offered. It also goes a long way to explaining why Bluetest didn't find any hardware issues, and why both Apple (via their website) and AT&T (via SMS) have really stepped up the push for this update.

So, do we finally have our answer, or is this just the next "shot in the dark"? Are you still having 3G problems? Is your neighbor still on 2.0 or 2.0.1? Tell them to upgrade now and then let us know if it helps!

When the Swedish engineers over at Bluetest revealed that, when measured at their facilities, the iPhone 3G radio performed roughly the same as 3G handsets made by Sony Ericsson and Nokia, some (including a few of TiPb's own, very astute, commenters!) cried foul. Not ones to be dissuaded by a little doubt, however, the Swedes brought in some of the people who complained about 3G reception problems, and put their iPhones to the test.

The results? According to Apple Insider, pretty much the same as before:

Wieselgren reported that the lab found that all these iPhones to "have no problems with the 3G communication in the test chamber. They send and receive signals in a fully normal manner. They do not disconnect earlier than the others we have tested when the signal becomes weaker." The iPhone using updated 2.0.2 software reported slightly better numbers, but Bluetest indicated there was no statistical significance, as a difference of up to 1dB in the results "can occur due to measurement uncertainty and random fluctuations."

Does this mean all the problems we keep having and hearing about are the exclusive fault of the carriers and their networks?

Well, no. We go back to our original theory that it's a confluence of conditions at work (which is why Apple says they can address some of the problems via another firmware update). Even if the antenna is fine, combine some dodgy networks with software that may be a little too sensitive to fluctuation, or too conservative in its reporting, and there are all many of problems that can arise.

2.1 may fix things on Apple's end, while public outcry (especially in France, where Orange has just been caught... er... red handed throttling down 3G traffic) could speed up the notoriously slow and stingy carriers to invest in their networks, and our future.

Make any sense? (Provided you can connect to the network long enough to read it...)

Last week Apple and Infineon were getting all the heat for shoddy 3G performance. Now AT&T is getting its share of the blame with a dizzying array of combinations. First it was Wired's fairly damning survey and the Swedish antenna tests that pointed further fingers at the network, and now Gizmodo head-honcho Brian Lam has had the chance to chat with AT&T CTO John Donovan:

I asked Donovan if caution was the overriding strategy behind waiting to match Sprint's initial 3G rollout, he replied, "I'd like to say we're deliberate. " He added that initially meeting the voice quality and data rates of Sprint's 3G network would have been both technically and financially impossible, despite the customer benefit. (One only needs to look at Sprint's financial weakness now to appreciate the wisdom of his point.) He also pointed out that by waiting, they got to leapfrog the limitations of Sprint's EVDO networks, referring to the extended data rates their network will eventually run at, at a better value. "The most astute thing you can do is be as late as possible and as fast as possible. Because it's going to cost you more if you do it too early, and if you do it too late, you don't get the features you want."

Well bully for AT&T, but where exactly does that leave frustrated customers with dodgy 3G reception? According to Donovan, they have a multipart plan to make sure AT&T really, truly, eventually delivers on the "more bars in more places" promise.

Lam likes having them on the record, so they can be held accountable. We think customers would prefer having them simply get the job done, so that dead zones, dropped calls, downgraded connections, and basically everything else that's currently broken about AT&T's 3G network is fixed and fast.

Remember the Wired.com Global iPhone 3G Study Casey posted about a week or so back? Well, the results are in, and Wired's conclusion is interesting to say the least:

In our view, this data is a strong indicator that performance of the mobile carrier's network is affecting the iPhone 3G more than the handset itself.

Wired further cites the recent Swedish engineering tests, run by Bluetest, which showed that the iPhone 3G's eponymous radio and antenna performed roughly the same as those of the other handsets they tested.

So where does this leave the last, 2.0.2 firmware tweaks from Apple, and hopes that 2.1 would further fix 3G connection problems? Chipset and firmware finger-pointing may have been misplaced, or at least been only part of a larger overall 3G connectivity knot. Either way, it looks like the tennis ball has just been rocketed back into AT&T (and the other carriers') court for now.

From an end-user's perspective, however, especially if you're one of the unfortunate 2%, other than providing some entertaining distraction, does the blame game really help us? Or do we just want it fixed, like, yesterday?

(Thanks Bad Ash and yc for the tips! Thanks to the TiPb faithful who went over there to help out with the study, and thanks to Wired for recognizing them!)

More Blackberry Bold coverage? What are we, CrackBerry.com? Fair point. But come on, RIM's latest/greatest did just launch in North America yesterday, so it's only polite that we pay it some attention. How does RIM repay us, however? It's not enough they iClone our glossy black and silver edged form factor, now they've got to iClone our 3G network connectivity issues as well?

I mean, we've gone on, and on, and on (or should it be off?) about the reports of the iPhone suffering from a string of perplexing 3G connection problems. Blame has been laid with the carriers and their networks, with chip manufacturer Infineon and their hardware and drivers, and with Apple and their 2.0h-no firmware. But thus far, while 2.0.2 supposedly addresses some issues for some users, no widespread solution has yet appeared.

However, reports are also now surfacing of the brand spanking new BlackBerry Bold suffering from very similar sounding problems. Says Electronista:

[Citigroup investment analyst Jim Suva] also notes that the BlackBerry suffers from the same problems of the iPhone, including 3G connection problems; the device will frequently drop 3G in favor of a slower, 2G EDGE link when downtown. Suva speculates that the flaw may likewise stem from rough software and that AT&T may have delayed its launch primarily to stabilize 3G performance.

Of course, our friends over at Crackberry HQ would likely point out that, as usual, truck-sized grain of salt should ship standard with every analyst report, and that AT&T due to the size and perhaps complexity of their network, does a whole heaping lot of testing.

But misery -- and lack of connectivity -- does love company, doesn't it?

For my part, other than Gmail on the iPhone continuing to make my life miserable, 3G is a little slow in connecting at first, but then seems to be working fine.

(Thanks Bad Ash for the tip!)

Ed Baig over at USA Today (via Daring Fireball) is reporting that:

Apple (AAPL) acknowledged Tuesday that a software update for the iPhone partly fixes the connection snags that have caused a global firestorm for the new iPhone 3G. Though mum on details, Apple spokeswoman Jennifer Bowcock said on Tuesday, "The software update improves communication with 3G networks."

Really?

As I mentioned previously, yesterday in Downtown Montreal, it looked like I was unable to connect to the 3G network at all. Last night in the suburbs of Montreal, however, I was able to connect (though it took a while). Today, downtown again... nadda. Or so I thought. I switched back to WiFi and still couldn't connect... to Gmail.

Yup. While I'd tested Gmail, MobileMe, ActiveSync, and MobileSafari yesterday, sometime since then I'd made the mistake of just hitting Gmail in MobileMail.app to see if a connection would pop up. Turns out that was really shoddy testing on my part. See, Gmail on iPhone says I haven't had any messages since 6pm last night. Gmail on the desktop however, while continuously giving me "Server error: too many simultaneous connections (Failure)", shows 50+ more, right up to this very minute. Now, I've Twittered nearly constantly about problems with Gmail IMAP lately, from invalid certificate errors, to server connection problems, to the mail outage they had a week or so back (not coincidentally the same time MobileMe was out... again).

So what's going on? Are their network connection problems or is Gmail IMAP that really buggy (according to Twitter again, it's buggy enough to make some iPhone developers abandon it entirely)? And has this been adding to, or merely confusing my 3G network connection problems?

My guess is the former. Intermittent 3G network connection errors, and Gmail IMAP still really isn't ready for prime time. (And why that doesn't get the blog-focus MobileMe gets, aside from the admittedly free nature of the beast, is a bit perplexing).

I plan to run more (and better) tests today, and hopefully get something of a less obscure picture.

So, this weekend I had a lot of problems connecting to the 3G network. Bars showed full. 3G icon was lit up. But email and web browsing -- any type of network activity really -- either took forever to resolve or timed out completely. Today was even worse. Couldn't get on for most of the day. Zip. Zero. Zilch. And this was AFTER installing yesterday's hot new 2.0.2 firmware (once I got it to download...). So what's going on?

Are there carrier issues resulting from less mature 3G networks? Is there an Infineon 3G chipset hardware problem? Is Infineon dragging their heels about writing better drivers? Is something in Apple's iPhone 3G software stack that's just not connecting well, or timing out too quickly? Or is it a horrible confluence of all of the above, making it an especially tough -- and frustrating -- bug to squash?

Given the lack of any apparent, or at least successful, fix in 2.0.2, Engadget says Apple is "shooting in the dark" trying to resolve the 3G issues. I don't think so. I think, as one of our commenters mentioned, 2.0.2 was scheduled to add support for the addition 20+ countries and carriers coming on line this week, and crammed in whatever minor improvements Apple had ready. Rewriting the 3G drivers, especially if Infineon isn't moving at Apple-required speed, isn't likely to happen before the rumored September 2.1 release (which, as mentioned in the post on turn-by-turn GPS, has already jettisoned Push Notification Server support, hopefully because Apple is laser-focused on delivering an actual, gosh-darn real stable release in 2.1).

I don't know about you, but at this point, that's the priority I want them to have moving forward. Do one thing at a time, do it very, very well, and then move on...

UPDATE: Former Apple employee Chuq Von Rospach has just blogged about a meeting with an unnamed current Apple insider who let slip that:

90% of the disconnects are initiated inside the phone, which would exonerate AT&T. Most of the disconnects are being generated by crashes in the driver code for the 3G chip, which comes from the chip vendor, not something Apple written and outside of Apple's direct control. Complicating this -- even though Apple is handing over "here is the bug, here is the fix, update the driver", the turnaround from the vendor on driver updates is on the order of 2-3 months. Said, um, lack of urgency not exactly making people inside the projects happy.

Understated much? If he's not, as he says, being lied to, Chuq thinks this lack of responsiveness may be why Apple went ahead and bought PA Semi a few months back: to bring the chipset in house and more fully under their own control.

Remember that 3G network connection glitch we mentioned a few days back? The one that might be a hardware problem with a software fix? (Apple Insider weighs in today that this could, in fact, be likely). MacRumors is reporting that Steve Jobs, as he or someone acting on his behalf is sometimes want to do, has responded to an email inquiry about it, and provided the following:

We are working on some bugs which affect around 2% of the iPhones shipped, and hope to have a software update soon.

If this is authentic, and 2% is a solid number, given that the iPhone 3G sold 1 million units its first weekend, and may be over 3 million units now, that's a staggering 60,000+ users potentially affected, -- never mind 20 more countries set to launch later this week.

Unless Apple pushes out a 2.0.2 hotfix, firmware 2.1 has been in beta for a while already, and in general rumored to be heading towards a September-ish release (probably to coincide or follow up Apple's tradition fall iPod and Mac product Special Event, which last year introduced both the iPod Touch and the WiFi Music Store).

Casey recently posted that the current 2.1 beta removed support for the Push Notification services (Apple's conceptual replacement for multitasking functionality) to allow for more internal development, but perhaps also to fast-track the 3G fix?

That could leave the 2% in the slow lane for a while still...

Update: Daring Fireball points out that: "The 3G networking glitches may well be real, but it’s worth pointing out that Richard Windsor is the same jackass who issued a report a year ago about the supposedly faulty “film” on the iPhone touchscreen, when in fact there was no such film."

So add that to the "grain of salt" heap...

Yesterday we asked you if you were having any iPhone 3G network connection problems, and while some of you were fine, many of you were suffering. Well, MacRumors has jumped on the story, providing an interesting perspective (via MSNBC -- and yes, the MS stands for Microsoft) on what might be going wrong:

The report said the most likely cause of the 3G problems is defective adjustments between the antenna and an amplifier that captures very weak signals from the antenna.

Hardware would be bad news for Apple and for chipset supplier Infineon whose 3G chipset is now getting a real-world pounding beyond anything they could have given it in the lab. It's also bad news, of course, for users who'll be considerably more inconvenienced even if some type of fix is eventually offered. However, Business Week has others sources sticking with the software angle for now:

Apple programmed the Infineon chip to demand a more powerful 3G signal than the iPhone really requires. So if too many people try to make a call or go on the Internet in a given area, some of the devices will decide there's insufficient power and switch to the slower network.

They go on to say Apple and Infineon are already testing a firmware fix that should be rolled up into a larger update sometime in September (sounds like 2.1 to us). But here's the question, can 2.1 patches fix flaky chipsets? Can good software overcome bad hardware?

The day after I bought my iPhone 3G I went back to the local Rogers store to see how things were going, and a customer was there complaining that he couldn't get on the 3G network. The customer service rep tried fiddling with his iPhone, but the settings all looked right. Yet there I was, less than 5 feet away, with full, fast 3G speed downloading TiPb's homepage at that very moment. Later that night, I saw some chatter that others thought Rogers was down because they couldn't connect either.

Things had quieted down some for a while, but now more and more reports are spreading of transient 3G connection errors. MacRumors quotes Mark Siegel of AT&T as saying there haven't been an unusual amount of complaints about the iPhone 3G in specific:

How a device performs in individual situations depends on circumstances like where you are in the 3G coverage, how close you are to a cell site. Things like terrain and buildings all come into play. I'm not denying that people are having a less than satisfactory experience, but overall, the phone is doing great.

GigaOm, however, picks up some analyst rumors about potentially flakey 3G chipsets, while iLounge steps it up with T-Mobile and Vodafone laying blame on the same, with the Syndney Morning Herald claiming an unnamed source revealed that Apple only provided 3G test units to carriers the day before launch. Ouch. Any chip experts out there that could help identify what problem would cause reception problems for a fraction of users?

Mine, like I said, is rock-solid so far, only dropping to EDGE in areas where Rogers' coverage is weak to begin with. How about you? Any 3G connection problems?