Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Communications The Almighty Buck Wireless Networking Hardware

Treo Bluetooth Bounty Efforts Unsuccessful 98

Posted by timothy from the born-to-lose dept.
UberGeek28 writes "The development effort pushed by TreoCentral (previously discussed on /. here) seems to have failed. After raising a bounty of $5,812 for the first developer to meet the requirements of a working Bluetooth driver for PalmOS 5.0 with the Treo 600 in mind, no developer has come forward to claim it. The official word has come here. Maybe another effort with wider impact could succeed where this one failed?"
This discussion has been archived. No new comments can be posted.

Treo Bluetooth Bounty Efforts Unsuccessful More

Treo Bluetooth Bounty Efforts Unsuccessful

Comments Filter:

  • Wow, they gave it an entire 3 months! (Score:4, Insightful)

    by Anonymous Coward on Sunday September 12, 2004 @03:28PM (#10229236)
    Learn to have some patience before throwing in the towel.

  • Complicated Treo? (Score:4, Interesting)

    by aklix ( 801048 ) <aklixproNO@SPAMgmail.com> on Sunday September 12, 2004 @03:30PM (#10229243) Homepage Journal
    Since when did an external device become too complicated to program a driver for? Have devices really become that bloated?
    • Since they offered $5000 for it? LOL
      • Since they offered $5000 for it? LOL

        No kidding!
        A top-notch embedded systems programmer can make $50-$100 an hour (salary vs contract).
        That $5k is at best a week or two worth of work.
        Not to mention whomever does it is going to need several thousand dollars worth of various blue-tooth enabled devices to test it with since, (like every other standard that's ever existed), many companies haven't implemented it quite perfectly.

    • Re:Complicated Treo? (Score:5, Informative)

      by Fnkmaster ( 89084 ) on Sunday September 12, 2004 @06:07PM (#10230496)
      See my post here [slashdot.org]. In short, it's not just talking to an arbitrary external device. There is no documentation for any of the several layers of driver components here, at least not without NDAs with PalmOne and others. So the best guide you can find is the Palm OS 4 drivers (at least one PRC file for BT, BT serial, BT/SD), which weren't written for ARM hardware. So take those drivers, disassemble them (like I said, at least 3 PRC files), now go through opcode-by-opcode and figure out what each of the undocumented API calls are doing, so you can write something that's compatible with the built-in Bluetooth API.


      Oh yeah, and there's really no public documentation for writing drivers of any sort for OS 5/ARM - you're talking about stuff operating below the level of the public PalmOS API. So it's not a matter of bloating here, just lack of information because the whole system is relatively closed and undocumented.


      I'm not an expert by any means, just a guy with a modest amount of Palm application development experience who took a crack at this problem only to throw up my hands in frustration after a day or so, realizing it was much harder than it looked at first glance without the proper documentation for anything.

      • I hear ya... apparently viable SDIO development requires a somewhat hefty sum to join the SDIO organisation, along with which comes an NDA preventing disclosure of said specs. I embarked on this project to see what I could come up with, but I hit a brick wall when I could not find the necessary documentation. I'm sure *technically* it can be done, but with no documentation nor a legal avenue to distribute said driver, it's not going to come about from anyone but PalmOne / PalmSource.

        From an economic poi

  • That is some good pay man, and in view of all that whinging on /. sometimes I really wonder why no one took up that offer.

    Anyway why are they still bothering with OS5? I thought OS6 is already coming soon? They might as well save the $ for the new OS.

    • Re:People too picky on jobs these days? (Score:5, Insightful)

      by dioscaido ( 541037 ) on Sunday September 12, 2004 @04:04PM (#10229399)
      > That is some good pay man, and in view of all that whinging on /. sometimes I really wonder why no one took up that offer.

      heh... probably because 95% of the slashdotters that trash Windows on a daily basis couldn't write an OS module if their life depended on it.
    • A USD5.8K "bounty" for three month's work? "Good pay"?

      That's NZD8.8K - yearly eqiv NZD35.2K. If you're able to actually achieve what's required for that bounty, then you should be able to get a job that pays more than that.
      • Ah, its not for 3 months work. They have been trying to find a develepor for 3 months to do the work and they haven't found one so they are calling off the search. Palm are including Bluetooth in the next Treo model so it kind of defeats the purpose of the bounty.

        Cheers
        VikingBrad

      • Half the people here are saying that 3 months is too short. But if they increased it to 6 months, you'd effectively get paid at half the rate - and who wants that? (The other half are saying that the rate is too low already, so the time should be decreased.)

        I'm holding the $5812 sum fixed, not going up or down. The reason is because that's the price that Treo 600 owners have committed for a driver. Whether this is too high or too low is debatable, however, at the end of the day, that's the dollar value.
    • Good Pay?

      The last device driver I wrote under contract paid US$88,000. Plus, I had documentation and access to the hardware developers to ask questions, so it was a fairly easy gig. For $5k this would only be worth it if I wanted the driver for myself too, and that's assuming it's even possible. Without docs it would be a considerable amount of work to even figure out if it could be done.
  • Although I am really siked for the new Treo's:
    Link for Treo Ace [engadget.com]

  • Only 1135 signed up so far (Score:4, Informative)

    by Anonymous Coward on Sunday September 12, 2004 @03:32PM (#10229256)
    http://www.savebetamax.org/ [savebetamax.org] Let's Slashdot the Senators. Please sign-up and follow through. If the INDUCE act passes, we all lose.

  • Unrealistic Expectations (Score:5, Insightful)

    by Noksagt ( 69097 ) on Sunday September 12, 2004 @03:34PM (#10229274) Homepage
    I know they announced the dealine in advance, but their 3 month project was horribly unrealistic for such a small bounty. The initial slashdot article had comments to this effect (out of the relatively few comments it had--this is a BAD sign: no one really cares). I can see MAYBE an undergrad who didn't have a summer job working on it, but no one else would put everything else on hold for a pathetic bounty for an already ambitious timeline.
    • The whole "open funding" philosophy seems to be crap compared to old fashioned "just hire a guy." Look at this laundry list of stuff they wanted:

      So Treo developers, the challenge is up to you. A working Bluetooth driver is a driver that enables the use of a common Bluetooth SD card (Toshiba's, PalmOne's, or Socket's) in the Treo. However, Bluetooth would be useless without drivers written for certain applications. To win the bounty, this driver must enable the Treo to at minimum wirelessly sync to a deskt

  • Impossible (Score:5, Interesting)

    by comwiz56 ( 447651 ) <comwizNO@SPAMgmail.com> on Sunday September 12, 2004 @03:35PM (#10229278) Homepage
    From the article:
    "As we got more publicity, more people who knew Bluetooth started to get involved in the discussion. Sadly, these people only had bad news to share. Developers started to tell us that what we were asking was impossible, because it was physically impossible for the Treo to access the voice stream from the radio. This meant that at best, a driver would only be capable of doing data over Bluetooth. But, as our conditions stated that it must support the headset profile, a driver that only did data would not have won the bounty." ... and they're suprised nobody could do something impossible?
    • ... and they're suprised nobody could do something impossible?

      I guess they're surprised that it turned out to be impossible, as per the Konigsberg Bridge Problem.

      However, if it really is impossible they should have sucked it up and awarded the bounty to the first person who proved it, as per Euler and the Konigsberg Bridge Problem.

      Knowing for a fact that something can't be done is itself valuable information.

      KFG

    • Re:Impossible (Score:5, Interesting)

      by Fnkmaster ( 89084 ) on Sunday September 12, 2004 @05:56PM (#10230394)
      I can also assure you even writing a driver for the data capabilities of the BT card was nearly impossible without extensive reverse engineering of the existing drivers - there are several layers of drivers, all of which are undocumented in the public PalmOne hardware/Palm OS API documentation. The SD card has no documentation to speak of. There were, IIRC, three different driver files you would have to rewrite and get compiling and working under OS 5/ARM. Oh, and did I mention, you can't use any of the usual Palm development tools since they don't generate ARM native code, you need to use a totally different toolchain (GNU with ARM target, I guess).


      There was an easier way to get BT working, which is to build an add-on device that connected to the external serial line and ran the signal through a UART to one of the many complete BT chip systems out there. I made some progress toward doing this, but I lost my motivation when they announced the Treo Ace (and I got busy with other stuff). My takeaway from this experiment was that the internals of Palm OS software and hardware are sadly extremely closed these days, and even figuring out the general things (like how to write drivers for OS 5) is nearly impossible. Things didn't used to suck so much in Palm land (before the PalmOne/PalmSource split I guess?).


      And a hardware add-on solution would have made voice possible too (by plugging into the headset connector) - as this article said, there was no way to access the voice audio stream from software to redirect it to the BT card, even if it had the capability to do so (which is also doubtful - apparently the relevant voice pins on the BT chip in these cards may not have even been connected).


      So no, nobody in their right mind would have done THAT much work for 5 grand. I've done plenty of Palm programming in the past, and had this been a simple Palm app, I would've whipped something up in a minute.

  • No wonder (Score:5, Insightful)

    by janoc ( 699997 ) on Sunday September 12, 2004 @03:52PM (#10229343)
    Three month of development for OS 5.x, which is a messy hybrid of old Dragonball code and new ARM stuff with mediocre development tools (e.g. POSE does not emulate OS 5.x machines, AFAIK) ? Are they nuts ? Or did they really think, that somebody is going to plop down money for the dev. kit, sign few NDAs to get the docs and implement it for $5k which probably does not even cover the price of the dev. kit.

    Not to mention which Bluetooth card did they have in mind ? The ultra-proprietary Palm one ? As if there was anything else.

    Unrealistic expectations doomed this project from the start, IMHO.

  • Give me a break (Score:4, Insightful)

    by Anonymous Coward on Sunday September 12, 2004 @03:57PM (#10229370)
    Why not just hire someone as a real employee to do this. I'm sure there are plenty of qualified people who can do this, but they want to be cheap and make it into a contest.

    Nobody is going to get started on a project like this when they don't know how many other people are working on it and how far they have gotten. Why waste a lot of time working on this when at any moment someone else can come out of nowhere and make all your work for nothing.

    Reality check. The types of people who can write a bluetooth driver are not slave labor who will grovel around in the mud for your amusement.

  • Why there was a time limit. (Score:5, Informative)

    by miradu2000 ( 196048 ) on Sunday September 12, 2004 @04:18PM (#10229457) Homepage
    As Senior Editor of TreoCentral.com, the author of the linked article, and overall the man in charge of the bounty, I think it is fair for me to answer the question of why we had a time limit. I do not believe that I addressed this well in the article, so I'll do it here.

    First and foremost, I have other responsibilities come September, and I would have been unable to dedicate the time needed keep this bounty running, answer questions, test possible results, etc. The bounty had an end time because it needed closure at some point - we chose to do a post-collection method, and as each month went by more of the credit cards used for the pledges expired.

    Secondly, when I started the bounty I had full knowledge of the next generation Treo, currently rumored to be released at the end of October/early November. At the time, specific accurate info on the next generation Treo was publicly unknown, so "want" for bluetooth on the current Treo was high. However, I was aware that this info would leak sometime over the summer, and it did - through an article I wrote a few weeks later. Subsequently, after photos were leaked in August, support from users who may have wanted this solution waned. I'm not saying that all support disappeared - we have many users who really want bluetooth on their current Treo, but many others are now resolved to simply upgrade to the next generation device. At the time, I thought the next gen Treo was going to be released early september, so the goal was to have the end of the bluetooth bounty somewhat neatly coincide with the introduction of the bluetooth compatible Treo.

    It's debatable whether 3 months would have been enough time to complete a driver. I have seen some very complex Palm OS applications developed in a much shorter time period. Even so, if in the last weeks of the bounty a developer said "I'm making progress", I would have asked our users to let us extend it. But no developer did, and as explained in the article, the prospects of having a working driver to our specifications were grim.

    -Michael Ducker
    • The problem with your contest is that a developer who writes the code for $5800 is probably going to need at least two, if not all three months. That's between $440-$660/week before taxes for what would undoutably be many late nights, and even then it's just a CHANCE to be the winner... too risky for the reward. You're better off writing some open source software that WILL get used and will end up being a better resume item than "I wrote a contest entry that lost."
    • I actually signed up with Palm's dev site to work on this. Problem: horrid documentation and nothing to test it on. While I have no Palm experience I am comfortable finding my way around code and docs.

      Oh well. Back to googling for bounties.

  • Rather odd sum .. (Score:3, Funny)

    by Anonymous Coward on Sunday September 12, 2004 @04:39PM (#10229672)
    After raising a bounty of $5,812 ...

    Well .. for $5,813 I would have done it .. but as it stands ..

  • Nah, TreoCentral didn't attract any serious developers because of two main factors. They never published a precise spec for claiming the bounty, and the underlying specs for the HW and OS requiring the driver were unpublished. It's hard enough to succeed, and get paid, in the opposite scenario: clear specs from a single customer, and open source/specs for the platform. I'm as impressed with the Treo developers who didn't take the foolhardy risk, as I am with the community for trying to get them to.
    • Obvioulsly you'd call the Manhattan Project too risky not for its nuclear potential, but how the funding was drummed up.
      • What the hell are you talking about? The Manhattan Project was funded from the US Treasury during a life or death total war with Japan (and Germany, at first). It followed the demonstration of atomic fission at the University of Chicago, and was staffed with the most capable scientists of the day, backed by the unsurpassed industrial capacity of the USA, and targetted a simple, clear, though difficult spec: make a single bomb that can destroy a Japanese city after being dropped from a plane, and make two if
  • no developer has come forward to claim it

    Say, ya think maybe that measly 4-digit bounty might have sumthinta do with it?
  • Um, how about a $5,812 bounty for a Treo 600 OS that can go 24 hours without crashing??!?

Slashdot Top Deals

Single tasking: Just Say No.

Close