The good and bad on bluetooth

[Ed Hansberry]

A week or so ago I said I would post a semi-retraction on my eternal ranting on bluetooth technologies, but I can't find the post to link to so you'll have to trust me. Duncan probably has it bookmarked, so ask him. <img src="http://www.pocketpcthoughts.com/forums/images/smiles/icon_wink.gif" /><br /><br /><img src="http://www.pocketpcthoughts.com/images/hansberry/2002/20020820-bluetooth.jpg" /><br /><br />So, I got a T68 with bluetooth and infrared to use with my iPAQ and absolutely love it. To the noticeable annoyance of the nice sales lady at the VoiceStream store, I had figured out just enough about the T68 to get it paired with my iPAQ right there in the store as she was explaining my service. Her eyes and my wife's eyes were rolling in unison. I have no clue how to check my voice mail, but hey, my devices are bonded. If you leave me a message, you'll know why I never called back. <img src="http://www.pocketpcthoughts.com/forums/images/smiles/icon_smile.gif" /><br /><br />But the truth of it is, bluetooth still isn't idiot proof, and I guess that is OK as long as those are the expectations. No one expects WiFi to be plug and play by someone that doesn't know TCP/IP from a hole in the ground, yet bluetooth is always advertised as a simple and easy to use technology to replace cables. Hogwash. I still don't understand why I have to configure all of those tabs on my iPAQ's bluetooth manager. Furthermore, if you want to hose your T-68's contact listing, mass beam a bunch of contacts from your iPAQ to the T-68 and hit the X button on the transfer screen instead of "quit." The next time you send data over, it will remember those old contacts and duplicate (or <i>triplicate</i>) them. IR software on the iPAQ and T-68 prevent this error recognizing dupes. So, is that the fault of the programmers, the bluetooth standard, the maturity of OBEX over bluetooth? Who cares?<br /><br />Bottom line, bluetooth has a place in my mobile world for some time to come and <i><b>I LOVE IT!</b></i> But it isn't always a bed of roses, and it certainly isn't for the masses just yet, no more than configuring a spanking new 9,600 modem was in 1994.

[Mobile Bob]

Ed,

Correct me if I am wrong, but I seem to recall that you have, or have used, the Pocket PC Phone Edition (O2 XDA/T-Mobile) device. So my question is: which do you prefer? the PPC PE or the IPAQ/T68 combo?

Bob

[Chris Forsberg]

It's amazing, Ed, I always thought I was a "one body" guy. I simply love the XDA with its all-in-one solution. But, as I'm going to the US tomorrow, I needed tri-band. So today I got my iPAQ 3970 bonded with my T39, and after a firmware update (had to go to the telecom store, no EUUs there), the Running Voice GSM (http://www.pocketpresence.com/) software makes the T39 stay in my pocket where it belongs. All I'm missing now is the bluetooth headset.

Like you, I simply love it! The fact that I can do everything from my iPAQ except the actual talking (now, through my wired handsfree) is simply amazing. And that with no jackets, CF cards, or wires (oh yes, except for the handsfree). I can even do GPRS without stopping Running Voice (new feature in 2.0), and I get the connection stats as a bonus. I can dial, get call logs, make notes during calls, and SMS messages pop up in my iPAQ as they arrive. I can use many of the Pocket PC Phone Edition features that I have grown to love, but the integration is not there. I cannot simply make the phone call from within Contacts, but from the contacts list inside Running Voice GSM, and so on. So, when it comes to software, Pocket PC Phone Edition still is king. What if I could have Phone Edition on my iPAQ, and use it with my phone?

[GregWard]

For what it's worth I have to agree with both your sentiments! BT is great but still has a way to go.

One thing I find annoying (ok I know there are some basic things to fix first!!!) is that there's very little "smarts" between two bt devices (yet). Like you I have connected my T68 and iPaq and I use this as a replacement for my Nokia 9210 when in the US.

Of course, the T68 has it's own e-mail capabilities - limited but alright for checking/receiving/sending urgent emails when you are (for example) out for a few drinks in the evening. However, if you then connect the iPaq to the T68 it has zero knowledge of any previous activity via the T68 directly - so it brings down the emails all over again! Even though they're already in the phones memory. Ok - I guess I could probably set my server to delete messages the T68 had received - but this isn't what I want - the T68 might only have the first few lines and no attachments. In an ideal world the iPaq would talk to the phone and say "what have you received since we last talked?" only after this would it poll the server and offer to bring down any extra data. The way it works I have to either a. not use one device or b. accept paying for two downloads.

So my "school report" for bt - shows good promise but must work harder!

[Will T Smith]

A week or so ago I said I would post a semi-retraction on my eternal ranting on bluetooth technologies, but I can't find the post to link to so you'll have to trust me. Duncan probably has it bookmarked, so ask him. http://www.pocketpcthoughts.com/forums/images/smiles/icon_wink.gif

http://www.pocketpcthoughts.com/images/hansberry/2002/20020820-bluetooth.jpg

So, is that the fault of the programmers, the bluetooth standard, the maturity of OBEX over bluetooth? Who cares?



As a professional programmer I will try to nail down the culprit ... It's either

1) That guy who writes most of his code with Cut/Paste. He produces something that looks like it works, satisfies the VP, and is looked upon highly for it. It's shipped and someone else ends up debugging his sloppy "cookie cutter code" after customers complain.

2) The Manager/VP/Marketer(the worst) breathing down a responsible implementer to get it done NOW despite all seaming reason and practicality (who ironically also has a "zero-tolerance" bug policy). Ulitmately, It's takes LONGER than if he would have been patient in the first place dealing with poor planning, lack of robustness, inconsistent, ad-hoc in-house testing (there's NEVER time to test until just before the product is released).

3) A contracter who did just enough to get paid so he could move on to his next job.

----------------------

I've read the Bluetooth spec. It's very well defined. They have a good program in place to certify interoperability. Though, they don't check robustness, nor do they delve into human factors.

The stupid interfaces are NOT defined by the Bluetooth committee. The reason it's difficult is that the product introduces the user to details that he shouldn't HAVE to worry about. There is a difference between a test environment for implementers/testers and a product for end users. A lot of production side people either don't realize this difference ... or don't care.

Seriously, when you've been hacking on something for a week straight working 14 hour days your consideration for human factors tends to lessen and the "it works, I'm finally done" impulse takes hold of the most responsible of implementers.

I've worked with dozens of individuals who had the attitude that because they can do it easily (after implementing everything theselves for THEIR personal purposes, knowing all the factoids, special cases, Stupid quirks, program level process flow) believe that everyone else SHOULD be able to. It's actually a big ego grease for these folks to design Rube Goldberg software and believe themselves superior since ONLY THEY can understand something so "clever" and "well-designed" that it is so simple for them, as if they were Einstein describing general relativity in 1924.

Want to fix these issues. Write your congressman and voice support for software quality and liability laws. It was once written that if buildings were produced the same way as software that our cities would crumble from the first errant landing of an unexpected butterfly.

As long as software companies are shielded from liability by those stupid little stickers and "End User License Agreements", software QUALITY will only get lip service. To market time is EVERYTHING in the software biz, there is no effective counterbalance because Microsoft will never be liable for bringing your Business or Home to a screaching, expensive halt from the newest "Butterfly Virus".

[Ed Hansberry]

Correct me if I am wrong, but I seem to recall that you have, or have used, the Pocket PC Phone Edition (O2 XDA/T-Mobile) device. So my question is: which do you prefer? the PPC PE or the IPAQ/T68 combo?

I would prefer the PPC PE concept, but not quite as implemented by the T-Moblie. Here is why.

• No compact flash. I have too much money invested in CF memory to ditch all that for SD and I need the storage for music when I travel.
• No ability to use a 56K CF modem, CAT5 Ethernet or WiFi. I download Audible content when on the road - about 1.8-2MB per download. Doing that via GSM at 9600 is painful. Doing that via GPRS is costly. So I need a basic landline or ethernet/VPN solution.
• Less important, but it is just one more thing - I need that external stowaway. I cannot and will not ever use a thumbboard. I can FITALY faster than that.

So, PPC PE is great, I just need a device wrapped around it that allows CF.

[mgd]

Welcome aboard, Ed.

[st63z]

MS and the BT industry seem to be moving to simplify both the development and end user experience by being selective, in addition to some future proofing? There was a recent mini-thread rehashing some of this:

http://www.pocketpcpassion.com/forum/showthread.php?s=&postid=81965#post81965

Do ya'll agree with this direction? Also, I wonder if our T68s and iPAQs will still require 3rd-party add-on stack support even after the MS stack is released this year?

[Jeff Rutledge]

• No ability to use a 56K CF modem, CAT5 Ethernet or WiFi. I download Audible content when on the road - about 1.8-2MB per download. Doing that via GSM at 9600 is painful. Doing that via GPRS is costly. So I need a basic landline or ethernet/VPN solution.
• Less important, but it is just one more thing - I need that external stowaway. I cannot and will not ever use a thumbboard. I can FITALY faster than that.


I agree that these two items are crucial. Does anyone know of any current or planned device/device configuration that would accomodate this? Maybe the iPAQ 5000 (at least, what's rumoured so far...)

[Mobile Bob]

Correct me if I am wrong, but I seem to recall that you have, or have used, the Pocket PC Phone Edition (O2 XDA/T-Mobile) device. So my question is: which do you prefer? the PPC PE or the IPAQ/T68 combo?

• Less important, but it is just one more thing - I need that external stowaway. I cannot and will not ever use a thumbboard. I can FITALY faster than that.



I've considered getting a stowaway keyboard, but I'd like to use one with both my Toshiba e740 and my T-Mobile PPC PE, and I don't want to have to buy two stowaway keyboards (as far as I know, neither device has a compatible stowaway available as of now, anyway). Are you using FITALY with your PPC PE device? Maybe I should consider FITALY for my two PPCs (?).

Back to the topic of the thread, I think I'll be investing in a T68i and a CF Bluetooth card for my e740 at year-end. Which CF Bluetooth card are you using? Or, which CF Bluetooth card would you recommend?

Thanks for the info Ed. :D

p.s. That was smart to establish a bond between your IPAQ and the T68 in the Voicestream store before plunking down your dollars. Especially considering your trials with Bluetooth in the recent past. :wink:

[Ed Hansberry]

I've considered getting a stowaway keyboard, but I'd like to use one with both my Toshiba e740 and my T-Mobile PPC PE, and I don't want to have to buy two stowaway keyboards (as far as I know, neither device has a compatible stowaway available as of now, anyway). Are you using FITALY with your PPC PE device? Maybe I should consider FITALY for my two PPCs (?).

Back to the topic of the thread, I think I'll be investing in a T68i and a CF Bluetooth card for my e740 at year-end. Which CF Bluetooth card are you using? Or, which CF Bluetooth card would you recommend?
I use FITALY on all PPCs I use. I have an iPAQ 3870, so bluetooth is built in. On a card, I hear good things about the Socket BT card from www.socketcom.com.

[Pony99CA]

A week or so ago I said I would post a semi-retraction on my eternal ranting on bluetooth technologies, but I can't find the post to link to so you'll have to trust me.

So, is that the fault of the programmers, the bluetooth standard, the maturity of OBEX over bluetooth? Who cares?


As a professional programmer I will try to nail down the culprit ... It's either

1) That guy who writes most of his code with Cut/Paste. He produces something that looks like it works, satisfies the VP, and is looked upon highly for it. It's shipped and someone else ends up debugging his sloppy "cookie cutter code" after customers complain.

2) The Manager/VP/Marketer(the worst) breathing down a responsible implementer to get it done NOW despite all seaming reason and practicality (who ironically also has a "zero-tolerance" bug policy). Ulitmately, It's takes LONGER than if he would have been patient in the first place dealing with poor planning, lack of robustness, inconsistent, ad-hoc in-house testing (there's NEVER time to test until just before the product is released).

3) A contracter who did just enough to get paid so he could move on to his next job.

Now that's funny. As another professional programmer, I can appreciate that.


Want to fix these issues. Write your congressman and voice support for software quality and liability laws. It was once written that if buildings were produced the same way as software that our cities would crumble from the first errant landing of an unexpected butterfly.

As long as software companies are shielded from liability by those stupid little stickers and "End User License Agreements", software QUALITY will only get lip service. To market time is EVERYTHING in the software biz, there is no effective counterbalance because Microsoft will never be liable for bringing your Business or Home to a screaching, expensive halt from the newest "Butterfly Virus".
Well, as a programmer, I'm not sure I want my company being sued every time some user can't figure something out or has his crazy configuration cause an incompatibility with my progam.

With the economy like it is now, this would probably just cause more layoffs. At least let me find my next job first. :lol:

Steve

[denivan]

I first had an Armadillo BT card but wasn't satisfied with it. Performance wasn't what it should have been and I didn't like the form factor. The Socket BT card is they only way to go. It's software is super easy; I had my iPaq and T68i paired in one minute. But what I like most is the formfactor, it doesn't stick out the CF slot, so if you have an expensive vaja case for your iPAQ + CF sleeve, now you can close it. If you choose another BT card, you can't use your case anymore.

[heliod]

Ed Hansberry wrote:

But the truth of it is, bluetooth still isn't idiot proof, and I guess that is OK as long as those are the expectations. No one expects WiFi to be plug and play by someone that doesn't know TCP/IP from a hole in the ground, yet bluetooth is always advertised as a simple and easy to use technology to replace cables. Hogwash. I still don't understand why I have to configure all of those tabs on my iPAQ's bluetooth manager. Furthermore, if you want to hose your T-68's contact listing, mass beam a bunch of contacts from your iPAQ to the T-68 and hit the X button on the transfer screen instead of "quit." The next time you send data over, it will remember those old contacts and duplicate (or triplicate) them. IR software on the iPAQ and T-68 prevent this error recognizing dupes. So, is that the fault of the programmers, the bluetooth standard, the maturity of OBEX over bluetooth? Who cares?

Bottom line, bluetooth has a place in my mobile world for some time to come and I LOVE IT! But it isn't always a bed of roses, and it certainly isn't for the masses just yet, no more than configuring a spanking new 9,600 modem was in 1994.

I couldn't agree more. But we have to take in account that it is mainly a question of implementation rather than definition. I really worship the work done by the guys at SOCKET in the UI of their BT stack for the Pocket PC. Whoever sees it falls in love with it.

One or two days ago there was a small thread in the microsoft newsgroups regarding how beautiful life will be in the day that devices using technologies like BT begin to recognize each other automatically and carry and transfer unique digital signatures. One example really caught my mind: you will enter the hotel lobby and immediately your reservation will be confirmed in the desk's computer, the guest card will be emmitted and the magnetic key recorded for you to take. Doesn't it sound nice?

[Mobile Bob]

I first had an Armadillo BT card but wasn't satisfied with it. Performance wasn't what it should have been and I didn't like the form factor. The Socket BT card is they only way to go. It's software is super easy; I had my iPaq and T68i paired in one minute. But what I like most is the formfactor, it doesn't stick out the CF slot, so if you have an expensive vaja case for your iPAQ + CF sleeve, now you can close it. If you choose another BT card, you can't use your case anymore.

Thank you Ed and denivan. The Socket CF Bluetooth card is on my list. Hopefully, Voicestream will offer the T68i by the end of this year.

[Ravenswing]

One or two days ago there was a small thread in the microsoft newsgroups regarding how beautiful life will be in the day that devices using technologies like BT begin to recognize each other automatically and carry and transfer unique digital signatures. One example really caught my mind: you will enter the hotel lobby and immediately your reservation will be confirmed in the desk's computer, the guest card will be emmitted and the magnetic key recorded for you to take. Doesn't it sound nice?

Can you say "Big Brother"?

There's an article (actually a couple of articles) in The Economist about the current trend in the use of positional technologies and how it's affecting our privacy. They point out the, rather than the state smothering us with cameras and such to control our lives, we are actually choosing to give away our right to privacy in order to make our lives easier.

Apparently the London Underground are starting a scheme where people will have cards which allow automatic charging of their account for access to the Underground system. The point is, that cards are wireless and work inside a wallet or bag. Machines at every station can detect the cards and notify the system of where each customer is, charginging them accordingly. People are happy to do this, because it speeds their use of the system, but they are giving away their movements within the network (and effectively throught the city) whenever they use the card.

Privacy given away for convenience. So how much of your privacy are you willing to give away?

8) Geez, but it's too hot for philosophical discusions. Forget I mentioned it. :lol:

[Ed Hansberry]

Ed Hansberry wrote:
One or two days ago there was a small thread in the microsoft newsgroups regarding how beautiful life will be in the day that devices using technologies like BT begin to recognize each other automatically and carry and transfer unique digital signatures. One example really caught my mind: you will enter the hotel lobby and immediately your reservation will be confirmed in the desk's computer, the guest card will be emmitted and the magnetic key recorded for you to take. Doesn't it sound nice?

Can you say "Big Brother"?
I sort of agree - and please get your attributions right. I didn't say that. :)

[Ravenswing]

I sort of agree - and please get your attributions right. I didn't say that. :)

Beg pardon. As is often the case, a slip of the quotation edits. I have editied to remove the mistake.

[Jonathon Watkins]

Good post Will. As a professional SW/HW Tester I recognise the situations you are describing very well. :?

dealing with poor planning, lack of robustness, inconsistent, ad-hoc in-house testing (there's NEVER time to test until just before the product is released).
But surely that just is bad planning – testing should never be ad-hoc – there are a variety of methodologies and methods that can be used to ensure that testing is comprehensive and thorough. But I suppose that that is exactly the point you were making..... :)

Seriously, when you've been hacking on something for a week straight working 14 hour days your consideration for human factors tends to lessen and the "it works, I'm finally done" impulse takes hold of the most responsible of implementers.
Yup been there – tested that. But again – a good tester should call the programmer on that and challenge them. A tester needs to have a thick skin to be able to challenge the programmer’s assumptions like this. On one project I logged 350 defect personally – against a general testing average of 150 defects per project! The difference? I logged UI and usability problems as well – which the programmer had told other testers to ignore at it was not in their testing scope. If I am the final layer of testing before a product is shipped I like to put myself in the user’s shoes and think about how they will take to using the product for the first time. It’s also the software tester’s responsibility to ensure that concerns are raised and logged – so any problem is on the Manager’s heads and they take full responsibility for any problems when they sign off the release.

There is a difference between a test environment for implementers/testers and a product for end users. A lot of production side people either don't realize this difference ... or don't care.
Agreed – though surely a good User Acceptance Tester would pick up on these points. Now – as to whether the testers are listened to is another matter.......... :cry: When I worked for HP I had the semi-final sign-off on a product. If I was not happy – I would not sign it off – and the release could get delayed. Of course there was a lot of pressure to get things signed off. However, at many other places I have worked the Testers aren’t in the release loop in this way. Most places pay lip service to quality and ship regardless of how many Critical/Level 1 defects are outstanding. :evil: