Slashdot Mirror
Android 4.0 Upgrade For Sony Xperia Smartphones Opens a Pandora Box
First time accepted submitter ctrl-alt-canc writes "The udpdate to Android ICS offered for free by Sony to the Xperia smarphone users has caused plenty of troubles. Not only the decision by Sony of not updating Xperia Play phones to ICS caused rage among customers, but those who were lucky to get an upgrade for their smartphones discovered that WiFi connection did not work anymore. Up to now, the only suggestion proposed by Sony to fix the problem is to turn off the encryption, and reboot the smartphone and the access point."
If it opened a pandoras box, then all hell on earth would break out.. all manner of terrible things would be unleashed upon the world.
Having a few problems with your phone is not a pandoras box.. at best, you could say it opened a can of worms.
If you need web hosting, you could do worse than here
Didn't we learn our lesson from the last N rounds of user-hostile actions taken by Sony?
Who in their right mind buys Sony gear any more, after them suing their customers, removing advertised features from products after you bought them, root-kitting people's PCs, leaking your private data to the whole world, etc etc?
Seriously? People are STILL buying from Sony? Why on earth would anyone want to encourage their behaviour?
Yes. Sometimes they're also miserable at hardware.
USE HOT GRITS WITH STATUE OF NATALIE PORTMAN (NAKED AND PETRIFIED)
Sony sucks at proper software upgrades.
The real problem is that Sony totally blew at testing the WPA stack before release. This isn't altogether hard to fix, but it's time that could be taken away from their next great phone they want you to buy. Android vendors are hit and miss on this upgrade thing. Some vendors are really amazing at providing updates and some just blow. Being able to root one's phone is the only real salvation. So I don't think this deserves the hype to author is putting into this, Sony sucks, is there anyone of us surprised by this? Will they fix it? Who knows, maybe, however this is yet another reminder that this mobile stuff is still pretty new shit. Tread with caution.
Android 4.0 Upgrade For Sony Xperia Smarphones Opens a Pandora Box
It's a smartphone, ffs, and the box is Pandora's.
I'm beyond even thinking about asking Slashdot to edit; I know that's too much to ask for. Could you at least run a fucking spell check?
"We just did it to make you open your wi-fi to the world, so everyone can have access using your internet connection! :P
That way people can pirate our movie and music works on your opwn Wi-Fi. Then we'll send threatening letters to you about this infringement. You'll settle the case without going to court, and we'll profit!"
Sadly, despite them kicking us in the junk all the time, they do make a lot of superior products.
My most recent purchase was a PS3, not for the games, for the bluray player. The first one I bought was complete garbage. And was pleasantly surprised to find I could stream movies from my extensive video library on my computer too.
But that's probably why they're abusive and still around, there are enough people that tolerate the abuse because they otherwise have the best product.
I work for the Department of Redundancy Department.
a baggie of thunder-monkeys
rewriting history since 2109
Opening a can of worms would mean that there is an open can containing worms. I don't believe that is the situation.
I've got an unlocked Arc (LT15) myself and Sony pushed out the ICS update for me two weeks ago. Wi-fi works perfectly. Maybe part of the reason they decided to hold back the update for the Play was because they knew it might bork the wi-fi... any time you flash unofficial firmware you're taking your chances without a safety net.
Meanwhile my WP7 phone (like all phones running Windows Phone) is up-to-date running silky smooth.
And yet slashdot continues to make posts about how terrible WP is and how great Android is.
Only one guy has managed to get his Xperia connected to the 'Net so far.
Have gnu, will travel.
Maybe I'm just spoiled by Linux, but it really irritated me that Telus (in Canada) (aka one-third of the oligopoly that controls all cel phones in the country) took months to upgrade my Google branded Nexus S to ICS. Short of rooting the damned thing there wasn't a thing I could do about it.
We've reached a point where phones are becoming computing appliances, and end users shouldn't be held hostage by this sort of nonsense. If a major upgrade is available, I should have the option of installing it now, not when some bean-counter in Toronto decides it can no longer be avoided.
Three Squirrels
This _is_ Sony, right?... Let's see... rootkits, Playstation network security problems... Nope, not surprised.
Its not Sony's fault. (Well, its no ONLY Sony's fault).
This is a long standing problem with Android and certain wifi chip sets and ICS. It first reared its head in the Nook first edition
a long time ago, then it disappeared for quite a while only to come back with a vengeance with ICS.
Simply google the words: connecting authenticating obtaining ip loop
and you will see this is common to a LOT of Android handsets and tablets.
Google search click here.
The entire flagship HTC One line has similar problems. Samsung has the same problem, as does ZTE, Achros, Huawei and several others.
Some of these vendors (HTC) have promised fixes (and all have failed to deliver as best as I can tell).
If you fiddle with it long enough it will connect, eventually. Often rebooting your router will work, but you can't always do that.
Often connecting to your neighbor's "guest" network will work, and then subsequently reconnecting to your own network will also work fine.
(especially if said guest account is an open network with no encryption).
Its never a problem of a bad password. Its not something you can fix with a static IP.
The problem is in the actual authentication layer of the wifi connection, before it even gets around to asking for an IP Address.
Seems to affect 802.11G routers more than others. Its not specific to certain router brands.
I've alogcated my self into a stupor and dug thru some of the opensource code.
The only thing I can see is that it appears some token is supposed to be incremented by the handset with each authentication attempt, and it is not
being incremented, so the router disconnects the client. But so much of wpa_suplicant is running in binary blobs that the end user is at a
serious disadvantage trying to dig through this stuff.
I can induce this error at will on my HTC One X, and I can recover from it by simply connecting to an unprotected wifi "guest" account
near by, then wait 30 seconds, and re-connect to my wpa2 secured router. I also solved it by running a spare router with no security
and leaving the router unconnected to anything. I use this for connection, wait 30 seconds, then reconnect to my home router.
Its a major pain. But its not SONY's fault, I suspect its Google's fault or the wifi chipset manufacturer's fault.
Sig Battery depleted. Reverting to safe mode.
Oh, the token I mentioned was the EAPOL replay counter. (What ever the hell that is).
See this thread http://forums.fedoraforum.org/showthread.php?t=177798 where the resolution (near the bottom) was as I mentioned, connecting to any unsecured router than connecting back again.
Sig Battery depleted. Reverting to safe mode.
I compared Sony to a beautiful but abusive ex a few years ago. You swear you'll never talk to her again but a year goes by and you run into her at the mall and she's looking hot. Dinner seems safe enough. Public place and all. One thing leads to another...
Someone pointed out that /.ers don't know what it's like to have a hot ex and I should stick to car analogies.
I've sworn off Sony a couple of times but then they put out a product with the perfect mix of features and price so I make an exception "just this one time".
This is very good information, but... isn't it Sony's responsibility to test their products before they ship?
Reminds me of all those 1st-gen SSDs powered by JMicron controllers. They studdered and froze the machine constantly while trying to flush the buffer, rendering the whole PC virtually unusable. When I bought an OCZ Apex and had nothing but problems with it, several people yelled at me and told me that I should expect to have problems with such an early, immature product. For long-term reliability, sure, but for extremely obvious problems at launch? Fuck that. I sold that OCZ drive to a Linux junkie who was willing to tweak it until it worked, and got myself a Corsair P128, which is still working flawlessly.
I didn't blame JMicron. Supplier issues are not my problem. I squarely blamed OCZ for not testing their product properly and deciding to ship such a buggy piece of junk.
Strange how smart phones and tablets are far more closed and proprietary than PCs, and manufacturers are still having the same interoperability issues. No, wait... it isn't.
My most recent purchase was a PS3, not for the games, for the bluray player.
Which has a nasty habit of ramping up its power supply fan to take-off speeds about 30 minutes into a movie and staying that way until powered off, sounding much like a vacuum cleaner with a wad a paper stuck in the nozzle. I don't know about you, but that's a killer flaw from where I sit.
When all you have is a hammer, every problem starts to look like a thumb.
Absolutely the manufacturer's fault.
Considering that every ICS handset manufacturer seems to see this problem, why is anyone releasing any ICS build without testing for this specific issue? Are they so internally focused that they are unaware that this is going on in the market as a whole?
Sig Battery depleted. Reverting to safe mode.
For the two people who haven't figured it out, MAC address spoofing is trivial, and finding a valid MAC address is as easy as listening in on ONE packet from a connected device.
It may keep the average user out, but it'll barely slow down even the lamest of script kiddies.
My sig can beat up your sig.
No, but it's a decent indicator that the fault doesn't lie in the OS's network stack (and thus not Google's doing), but rather in the specific hardware (including drivers).
That being said, Google isn't the ones who've pushed out a defective OS image to devices in the field. As far as Xperia phones go, this is ENTIRELY Sony's fault for skipping on proper QA. Proper testing would've discovered this problem before it hit end users, ESPECIALLY if they are smart enough to look at known issues for OTHER ICS devices as part of their QA process.
My sig can beat up your sig.
Nobody has the advantage.
Yeah, our product is defective, but so is everyone else's, so you might as well buy ours.
Computer marketing in a nutshell. These same people sit around wondering how Apple makes so much money.