[fonejunkie]

I posted this earlier in the .423 OS thread, in response to a comment about the T-Mobile 9700. I am posting it here as a standalone thread because it's an interesting take on the situation. Mods can delete/close/ignore as appropriate.

I saw this article earlier today... regarding the two flavors of 9700 (T-Mobile and everyone else)... talking about mixing the OS between versions that could cause some issues due to the specific tweaks for the different 3G network frequencies... YMMV.

ATTN BlackBerry Bold 9700 Users: All OS Code Not Created Equally... - CIO.com - Business Technology Leadership

[aiharkness]

Wirelessly posted

Thanks for posting the link. Interesting.

[ifonline]

I had read that earlier. Strange. Is this the dawn of a new OS that is not only model specific but carrier specific? Are the days of "it doesn't matter what carrier you are on" over?

What stinks is that if this is the start of carrier specific OS builds, then AT&T customers are going to be miserable. AT&T is terribly slow to release updates with their seal of approval.

[lovescakemix]

OS's are like 130MB in space. they usually contain everything for the different carriers and all the different languages. The OS should read the vendor ID and install the carrier specific files. im on .423 and bell doesnt use our bands yet i have great 3g connection and never had problems. so this will not change anything with regards of people downloading the newest OS.

[Dude]

So what do you think of this quote from the article?

Here's a comment from RIM PR on the subject:

"[T]he two versions of the Bold 9700 are identical other than the band support you mentioned. Because of this difference, the AT&T and TMO devices do use different versions of the BlackBerry OS - an OS created for one carrier's device may not run well on the other."

[JSanders]

What do you expect RIM to say? You won't be reading the RIM PRincesses stating, "Oh, it's ok. Loading an OS from any carrier on your device is perfectly fine", otherwise, they wouldn't allow carriers to do the releasing.

And that's one point the author got flat out wrong.

First and foremost, you should be aware that installing any unofficial, or "beta," BlackBerry OS is a risk; remember, it's unofficial for a reason, and if it was "Ready-for-Primetime," so to speak, RIM would've probably made the code official.

RIM does not make OS releases "official". The carriers do, after certifying, testing, and primarily TRAINING their CSRs to answer questions and hopefully handle the phone calls regarding that OS. That in itself is one reason we're seeing such a long wait for OS5 to reach the 9000, 8900 and 9630, not so much because the code is not good, but because the ATT's of the world have to be ready to support it. The author of the article is a smart guy and well-read. He knows this.

[Doctor A]

I experimented putting T-Mobile .330 and then the latest .423 OS to see effect. On the About screen 3G bands 1,4,8 is what showed on both OS. I do not see why the OS release should make a difference. This was on T-Mobile 9700-I need to update my profile.
Doc

[F0nage]

Originally Posted by ifonline

I had read that earlier. Strange. Is this the dawn of a new OS that is not only model specific but carrier specific? Are the days of "it doesn't matter what carrier you are on" over?

What stinks is that if this is the start of carrier specific OS builds, then AT&T customers are going to be miserable. AT&T is terribly slow to release updates with their seal of approval.

This already happened in the Nokia/Symbian world. Every carrier has the option to brand their phones and then your firmware (OS) updates are locked to that carrier. Some carriers (like mine) produce no updates, we have very few choices of phones to begin with and if the phone comes out with crappy firmware (apparently Nokia policy) and bugs are fixed later, we have no upgrade path. We're stuck on one OS for the life of the phone.

And you have people all over the world with different levels of OS on the same hardware. This makes Nokia's job of troubleshooting harder, they're just too stupid to understand how it makes sense to have one OS for one phone and track updates to it instead of forking their OS all over the place and having a zillion different versions (at least one per carrier, globally) to have to deal with.

That leads us to hack and debrand our phones (no known downside so far except that in the case of any serious problems we have to rebrand and reflash, blah blah blah).

Really stupid, inefficient, wasteful, etc. Nobody wins with this deal. At first the carriers think they can lock people in but then some guy smarter than them comes along and figures out how to debrand, where to get a generic OS and install it, etc.

However, in this particular case from the OP, I think what we're talking about *is* somewhat acceptable. RIM produced two different models of the same handset which they *had* to do under the circumstances because of band allocations in different regions. They very well may be running two different chipsets for radios to operate in these bands, which in turn requires different drivers. Therefore it's reasonable (but not exactly genius) to have two versions of the OS. You would have thought that they would be smart enough to write code that is generic enough or to use plugins but apparently they are not smart enough, or because we're talking about a relatively constrained device they chose to economize and not give you drivers you don't need. That would be the most acceptable scenario, but who knows.

[xeroxide]

but when you install a new OS, it does not unbrand your phone by default.
it's weird cause i thought most gsm phones were able to select multiple bands, auto-selecting the right one for your sim.

[aiharkness]

Wirelessly posted

Originally Posted by ifonline

I had read that earlier. Strange. Is this the dawn of a new OS that is not only model specific but carrier specific? Are the days of "it doesn't matter what carrier you are on" over?

What stinks is that if this is the start of carrier specific OS builds, then AT&T customers are going to be miserable. AT&T is terribly slow to release updates with their seal of approval.

Yeah. I was thinking sort of the same thing. The 3g issue I can grasp. The network tweeks is what caught my eye.



I really don't have experience with running OS's released by other carriers because I've never been motivated, but mostly because I'm lazy. So it makes me leary about continuing to pass along the conventional wisdom in the future.

EDIT I should have finished the thread before posting. There's some good points, especially about what has to be said publicly versus what is true, and about the obvious error in the article.

[ifonline]

Yep. Not sure how to take it at this point. It will be interesting to see where this goes... if it goes anywhere at all.

[cmink79]

so where does .423 stand t this point, released by bell if i may ask?

[ifonline]

Meaning what? I have it on my AT&T phone and it works great. So from my perspective, install it regardless of carrier. It works fine.

[HaTaX]

The software for each model blackberry is the same, only difference to this is the CDMA phones that have switched from ESN ids to MEID (8830 & 8830m as an example).

If the model numbers are the same, you can pretty much load any carriers OS for your phone.

That being said, since each of the releases are targeted towards a particular carrier (330 for T-Mo and 296 for AT&T at the release of the 9700), there IS the possibility that the radio file COULD be tweaked for that particular carrier's network. This is about the only downfall I can think of with using releases not designed for your carrier.

Maybe they put updates in .330 for the AWS 3G bands that T-Mobile uses and .296 that was for AT&T had tweaks for their 3G network. Could happen, but I doubt they build unique radio files all together for different carriers. Nobody knows unless they can compare the source code to find differences for particular bands / frequencies.

[fonejunkie]

The article is talking about the frequency differences between the T-MO and non-T-MO versions of the 9700... because T-MO dorked around and missed out on the 3G bandwidth auction, they had to kludge together 3G at 1700 mHz... and then get phone manufacturers to accommodate them. The article seems to indicate that there are functional hardware differences that can cause non-T-MO software to not run right on a T-MO 9700... and cause T-MO OS to puke on a non-T-MO (read: everyone else's) 9700. Fact? Fiction? Hard to say... but it's something to keep in mind, and as always, YMMV.

BTW, this is NOT the he same issue as CDMA ESN/MEID issue... And I think RIM screwed the pooch on this one by making 2 (different) hardware devices with the same model number. Perhaps it should have been 9700 for everyone but T-MO and 9750 for T-MO... because the hardware IS different.

[HaTaX]

Hate to be a bugger on this, but the hardware IS the same. The 9700 as it exists hardware wise is only one model, however there is different firmware for the T-Mo and AT&T versions that tell the 3G chip which bands to operate on. This isn't like the firmware you load onto your phone with these OS releases, this block of firmware can only be modified by RIM engineers.

There's been discussion on this topic before here:
http://www.blackberryforums.com/gene...bb-9700-a.html

Also, the FCC only approved one ID for the 9700, if they had produced two different hardware models that use different frequencies, there would be two IDs. L6ARCN70UW is the ID if you're curious and want to do a OET Exhibit lookup.

[aiharkness]

The UK folks may take issue with your vocabulary, HaTaX, but the discussion is interesting.

[HaTaX]

Heh, whoops! They would probably take issue with ME for using that vocabulary!

I honestly didn't even think about it until I read your comment, that's the joy of the PC world we live in. Did give me a laugh though..