1. JasonM.'s Avatar
    4.7 build when they've compiled into the .2XX's? It seems that if .148 is good that .180 or .190 or .200 would be better no? It seems to me like RIM is just sending VZW a build to fail at this rate... Anyone have a reason that RIM would give VZW .141 or .148 to test vs. the latest compiled 4.7 build?
    05-19-09 03:57 PM
  2. Joel S.'s Avatar
    RIM has a bunch of builds on a publisher site, VZW grabs them as they see fit. They likely have multiple builds in different stages of testing.
    05-19-09 04:06 PM
  3. jdk2's Avatar
    Multiple builds being tested simultaneously in the hopes that one will pass their rigorous testing standards. On the surface, it makes sense. From our viewpoint, it's disconcerting and likely not to end any time soon.
    05-19-09 04:56 PM
  4. yapkuen's Avatar
    Also, the highest number build is not necessarily the best build. As far as I'm concerned, .141 sucked way more than either .122 or .132.
    05-19-09 06:38 PM
  5. JasonM.'s Avatar
    Also, the highest number build is not necessarily the best build. As far as I'm concerned, .141 sucked way more than either .122 or .132.
    I agree with that assessment to a degree. .141 is 19 builds up from .122 and 9 builds up from .132. I can understand the fundamental "fix one thing break another" logic behind that. At .200 though we're talking 59 builds on .141. There must have been some really choice builds out of those somewhere. I'm not sure why we haven't seen one, especially with them just sitting there compiled and untouched now.
    05-19-09 06:53 PM
  6. hookbill's Avatar
    I agree with that assessment to a degree. .141 is 19 builds up from .122 and 9 builds up from .132. I can understand the fundamental "fix one thing break another" logic behind that. At .200 though we're talking 59 builds on .141. There must have been some really choice builds out of those somewhere. I'm not sure why we haven't seen one, especially with them just sitting there compiled and untouched now.
    Ever seen .114? You never hear about that one. That was a bad build.

    Posted from my CrackBerry at wapforums.crackberry.com
    05-19-09 07:32 PM
  7. patches152's Avatar
    but the platform version was different, the version was higher but lower platform, or something like that, i forget which one is which sometimes its been a loooong day so far.
    05-19-09 07:44 PM
  8. markdental's Avatar
    Yeah. .114 had a lower platform than .113.

    Posted from my CrackBerry at wapforums.crackberry.com
    05-19-09 07:47 PM
  9. JasonM.'s Avatar
    Yeah. .114 had a lower platform than .113.

    Posted from my CrackBerry at wapforums.crackberry.com

    Ah yes, that old red-headed stepchild. Busted numbers on the phone app right?
    05-19-09 07:56 PM
  10. Crucial_Xtreme's Avatar
    I agree with that assessment to a degree. .141 is 19 builds up from .122 and 9 builds up from .132. I can understand the fundamental "fix one thing break another" logic behind that. At .200 though we're talking 59 builds on .141. There must have been some really choice builds out of those somewhere. I'm not sure why we haven't seen one, especially with them just sitting there compiled and untouched now.
    No it doesn't really work that way. You can't really say that there's 3 builds between 131 and 134. Reason being is that some builds only increase one increment whereas others are two or three, it just depends. So for the record, there's not 100 builds between .75 and .175.

    As to why Verizon doesn't have the absolute newest. Mainly because it's their choice in how they choose to test revisions. But most carriers do not jump builds just because there's something newer. Good testing practices don't involve jumping 20 builds ahead. Complicated to explain, but it's not good practice in white & gray box testing with RTM builds.
    05-19-09 08:02 PM
  11. patches152's Avatar
    No it doesn't really work that way. You can't really say that there's 3 builds between 131 and 134. Reason being is that some builds only increase one increment whereas others are two or three, it just depends. So for the record, there's not 100 builds between .75 and .175.

    As to why Verizon doesn't have the absolute newest. Mainly because it's their choice in how they choose to test revisions. But most carriers do not jump builds just because there's something newer. Good testing practices don't involve jumping 20 builds ahead. Complicated to explain, but it's not good practice in white & gray box testing with RTM builds.
    at any given time, how many builds are available to take on the partner site? and not every build gets posted on the site, correct? RIM does internal testing to ensure its worth the time of the testers, right?
    05-19-09 08:04 PM
  12. Crucial_Xtreme's Avatar
    at any given time, how many builds are available to take on the partner site? and not every build gets posted on the site, correct? RIM does internal testing to ensure its worth the time of the testers, right?
    haha, well in the past, before the Storm there would be maybe 2-6 bundles available depending on the device. Whereas now, it's pretty much the same but the Storm is one of a kind and there's access to more. There is also the TKC which has pertinent information about all devices, Resource Center, Solutions Guide, and a few other things which provide information about the device itself and revisions. It's the CrackBerry Addicts internet page heaven really. lol. Yes RIM does test builds internally and sometimes something may not be put up, but once you get to a certain level of stability in the code, all revisions are available.
    05-19-09 08:15 PM
  13. patches152's Avatar
    haha, well in the past, before the Storm there would be maybe 2-6 bundles available depending on the device. Whereas now, it's pretty much the same but the Storm is one of a kind and there's access to more. There is also the TKC which has pertinent information about all devices, Resource Center, Solutions Guide, and a few other things which provide information about the device itself and revisions. It's the CrackBerry Addicts internet page heaven really. lol. Yes RIM does test builds internally and sometimes something may not be put up, but once you get to a certain level of stability in the code, all revisions are available.
    makes sense, like we've discussed before, i only see one aspect of the testing process, so any info from RIM's end always help piece the puzzle together.

    just curious, but why is RIM still posting 4.7 stuff after they started 5.0 development? isn't it kind of expensive for carriers to release two OS builds? wouldn't it be in everyone's best interest to start on 5.0 stuff and avoid the hassle?
    05-19-09 08:19 PM
  14. CrackBerry Storm meh's Avatar
    makes sense, like we've discussed before, i only see one aspect of the testing process, so any info from RIM's end always help piece the puzzle together.

    just curious, but why is RIM still posting 4.7 stuff after they started 5.0 development? isn't it kind of expensive for carriers to release two OS builds? wouldn't it be in everyone's best interest to start on 5.0 stuff and avoid the hassle?
    They are taking care of the little matters before taking on a new OS.
    05-19-09 08:23 PM
  15. JasonM.'s Avatar
    haha, well in the past, before the Storm there would be maybe 2-6 bundles available depending on the device. Whereas now, it's pretty much the same but the Storm is one of a kind and there's access to more. There is also the TKC which has pertinent information about all devices, Resource Center, Solutions Guide, and a few other things which provide information about the device itself and revisions. It's the CrackBerry Addicts internet page heaven really. lol. Yes RIM does test builds internally and sometimes something may not be put up, but once you get to a certain level of stability in the code, all revisions are available.
    The percieved difficulty in making a stable, well performing OS for this device is mind boggling. I understand it's RIM's first touch screen and first time playing with an accelerometer but is it really this difficult? We're talking over a year of software development at this point... It might be time to hire some more people from Apple.
    05-19-09 08:23 PM
  16. Crucial_Xtreme's Avatar
    makes sense, like we've discussed before, i only see one aspect of the testing process, so any info from RIM's end always help piece the puzzle together.

    just curious, but why is RIM still posting 4.7 stuff after they started 5.0 development? isn't it kind of expensive for carriers to release two OS builds? wouldn't it be in everyone's best interest to start on 5.0 stuff and avoid the hassle?
    Oh it's definitely very expensive, hence why Verizon is taking it's sweet time, because they do not want to release a 4.7 update and have to turn around and release 5.0 shortly after. The 4.7 Dev is done, so why not post it, is more or less the attitude taken. But 5.0 builds are accessible as well.
    05-19-09 08:25 PM
  17. blackmannx's Avatar
    The cost in rolling out updates is often overlooked.. As I have stated before, an update places loads on the network, call centers, and data hubs.. If Vzw released all the "beta" builds that CB has floating around the network would work as well as AT&T's... No bars in most places...
    05-19-09 08:54 PM
  18. Dave12308's Avatar
    Yeah. .114 had a lower platform than .113.

    Posted from my CrackBerry at wapforums.crackberry.com
    Funny thing is, when that build leaked; and I first noticed that, I had a sinking feeling that the build would be a disaster. No real reason why, I just had a hunch.
    05-20-09 07:43 PM
  19. Dave12308's Avatar
    No it doesn't really work that way. You can't really say that there's 3 builds between 131 and 134. Reason being is that some builds only increase one increment whereas others are two or three, it just depends. So for the record, there's not 100 builds between .75 and .175.
    Why the gaps in builds? Is it because sometimes changes are made to code that badly break things; so a particular "attempt" at a build never gets packaged into one, and that build number is skipped? In other words, that particular build never makes it out of the "pre-Alpha" stage in computer software terms?
    05-20-09 07:49 PM
  20. Dave12308's Avatar
    Good testing practices don't involve jumping 20 builds ahead. Complicated to explain, but it's not good practice in white & gray box testing with RTM builds.
    Wouldn't the easiest way to explain it be "newer isn't always better"?

    In other words, they might miss something "golden" if they skipped alot of builds, and hit on something with new bugs introduced?
    05-20-09 07:51 PM
  21. patches152's Avatar
    well not all builds leave RIM, and when the carriers take the builds, not all pass all of the stages of testing, so for example with VZW they've had these builds go to friendly user testing:

    .65 (passed)
    .75 (passed)
    .113 (failed)
    .141 (failed)
    .148 (pending)

    so not all builds get past lab to get to network, and not all builds get past network to get to friendly user, the last stage.
    05-20-09 07:52 PM
  22. radioguinea's Avatar
    I get the feeling they will pass .148
    05-20-09 08:01 PM
  23. markdental's Avatar
    I get the feeling they will pass .148
    Me too. Now let's take bets on when it becomes official.

    Posted from my CrackBerry at wapforums.crackberry.com
    05-20-09 08:08 PM
  24. Ronindan's Avatar
    The percieved difficulty in making a stable, well performing OS for this device is mind boggling. I understand it's RIM's first touch screen and first time playing with an accelerometer but is it really this difficult? We're talking over a year of software development at this point... It might be time to hire some more people from Apple.
    Jason to develop a stable OS is very difficult, painful and long process.

    You can see it unfold right now with RIM/Storm.
    In creating an OS you have to account for all of the phone functions and process that happens all the time.

    Lets look at the touch screen functions, for example when you are in the message box and you have a message open and you swipe on the screen to see the next message. As a developer you have to take to account not just the hardware's capabilities but also the setting that are on the phone and how the user swipe on the screen. Not everybody interacts with their phone on the same manner.

    In some ways you can see here strength of Iphone developers. Apple has been working on touch related GUI for some time now. Their Ipods are proof of that, they are Apple's first generation touch devices. They basically ported their experience on coding those kinds of functions to the Iphone.

    And this also where Apple is weak at, sure they can create a very good GIU but their OS is still has some difficulty working with the machine part of the phone and messaging system.

    Back to RIM you can see the same thing, the BB devices that uses the trackball and full qwerty is very stable and very good. Simply because it is RIM forte. They have been working on these kinds of devices for a long time. I will not be surprised when RIM released the Onyx with the touchpad that it will be very good at the get go because trackball and trackpad are just the evolution of one technology to the other.

    Hope this helps.
    Last edited by Ronindan; 05-20-09 at 08:37 PM.
    05-20-09 08:35 PM
LINK TO POST COPIED TO CLIPBOARD