05-11-20 01:45 PM
731 ... 1213141516 ...
tools
  1. Thud Hardsmack's Avatar
    It takes time and money to change a product.
    Roll credits!
    02-20-20 08:03 PM
  2. Chuck Finley69's Avatar
    Past Performance is not indicative of future results...

    Even Facebook is claiming to be making changes to address customer (more likely governmental) privacy concerns as is Google.

    The goal is to provide a product that is acceptable to sufficient numbers of individuals so they would buy-in; to vote for development with dollars without the stigma of a crowd funded campaign. So, I would vote with my dollars if I could purchase a third party (i.e. Samsung Knox) slab handset with hardware root of trust that provides the Flow UX, modernized browser using the existing Browser UI, a new Android Player supporting API 28 with DTek and support for both Amazon's App Store and BlackBerry World.

    I hope my choices would minimize new development costs (I may even consider Android API 19) and time to market. Hopefully, if there is enough buy-in it would convince BlackBerry/QNX to further invest to make the platform more attractive to more people and to create a controlled IP sharing model applicable to both businesses and individuals. This is not beyond BlackBerry's purview as it has purchased WatchDox which is now being sold as Workspaces. Workspaces is, however, targeted to business.

    What I chose may not be enough for most people. Which is why I would like to have a survey of what would get people to buy-in.
    What you want doesn’t matter. Just ask BlackBerry what they want and go from there. If BlackBerry doesn’t answer you then you have your answer. Blackberry knows how much was blown on the BB10 epic fail. BlackBerry also knows what could or can’t be salvaged. Quit over complicating this. Either BlackBerry answers you or not. The End.
    Last edited by Chuck Finley69; 02-21-20 at 05:45 AM.
    02-20-20 09:57 PM
  3. app_Developer's Avatar
    I’m going to write a business book based on this thread.

    Working title is “How not to lose more billions after losing most of your billions”

    Manuscript:

    Ch 1: “Don’t do the same things, especially after your market position and assets have deteriorated even more than the last time.”

    cover: [picture of Heins holding Z10]

    That’s it. I’m done. Publish on Amazon you think? $50?
    02-20-20 10:10 PM
  4. conite's Avatar
    I’m going to write a business book based on this thread.

    Working title is “How not to lose more billions after losing most of your billions”

    Manuscript:

    Ch 1: “Don’t do the same things, especially after your market position and assets have deteriorated even more than the last time.”

    cover: [picture of Heins holding Z10]

    That’s it. I’m done. Publish on Amazon you think? $50?
    Make it $5000 per book. They will sell themselves.
    02-20-20 11:35 PM
  5. DonHB's Avatar
    Just curious. How many times do you think you need to ask the same question until suddenly you get a different answer? I would say we're up to about 300 now.
    Theres were no examples how drivers for the X12E used in cars are different from drivers for the X12E used in handsets. Most of what was answered were generalizations misdirection and bordering on smoke and mirrors.

    You could choose another piece of hardware it doesn't have to be the above modem.
    02-21-20 05:43 AM
  6. app_Developer's Avatar
    Theres were no examples how drivers for the X12E used in cars are different from drivers for the X12E used in handsets. Most of what was answered were generalizations misdirection and bordering on smoke and mirrors.

    You could choose another piece of hardware it doesn't have to be the above modem.
    If you want to use the Samsung phones, then you have to support the modems in the phones that you intend to support. Is there any part of that sentence that you would consider misdirection?

    Also if by some luck, you happen to find that it is the same modem then you have the issue of the 32/64 (armv7/8) thunking that you have to go do. That is non-trivial work and is very likely to cause your Samsung to become a hand warmer until the battery dies.

    And if you manage to do all that, you have to go update the browser. That is a huge amount of work.

    None of that is smoke or mirrors. It is simple reality. Buzzwords and bargaining are no substitute for the deep business, technical, marketing, and mobile industry experience that the 5-6 people in this thread have been trying to share.

    Your idea is based on an extremely shallow understanding of the technologies involved, the business strategy of BB, the tech industry, the mobile industry and most of all ARM processors and SoCs that some of us have actually worked with in our careers. So if you cannot or will not listen to any feedback on your idea and you still think you can make this work, then BB should be your next call.
    Last edited by app_Developer; 02-21-20 at 06:50 AM.
    JeepBB likes this.
    02-21-20 06:08 AM
  7. Dunt Dunt Dunt's Avatar
    Theres were no examples how drivers for the X12E used in cars are different from drivers for the X12E used in handsets. Most of what was answered were generalizations misdirection and bordering on smoke and mirrors.

    You could choose another piece of hardware it doesn't have to be the above modem.
    Don.... no one here really knows the absolute answers to all your different technical questions.

    But again... what does it matter. BlackBerry had people that understood these things, BlackBerry had people that knew all the other question you have. BlackBerry looked for ways to monetize BB10...

    Do you really think that they didn't consider what your proposing?
    02-21-20 08:31 AM
  8. DonHB's Avatar
    I guess my primary problem with all these arguments is the question of how reuse impacts cost. Engineers working on the next generation of SoC reuse hardware designs and software code to support it. Do drivers for the next generation of GPU in a next generation SoC have just a small fraction of code in common with its previous generation GPU? It seems that from what has been written here that moving a platform to a new generation of a currently supported SoC requires about the same work as supporting a wholly new architecture. Also, reasons for abstractions in the OS with regard to code reuse seems, from what has been written here, to have minimal impact in reducing this cost. It isn't true that as a platform matures QA and testing becomes a larger percentage of the overall development cost?

    I would like to understand why the above are true in more detail than have been provided here so far. Do you have a link that lays this out? It would be great if you can povide it.
    02-21-20 08:38 AM
  9. Thud Hardsmack's Avatar
    Theres were no examples how drivers for the X12E used in cars are different from drivers for the X12E used in handsets. Most of what was answered were generalizations misdirection and bordering on smoke and mirrors.

    You could choose another piece of hardware it doesn't have to be the above modem.
    Unless the hardware is an exact match, new drivers are required.
    02-21-20 08:41 AM
  10. Dunt Dunt Dunt's Avatar
    I guess my primary problem with all these arguments is the question of how reuse impacts cost. Engineers working on the next generation of SoC reuse hardware designs and software code to support it. Do drivers for the next generation of GPU in a next generation SoC have just a small fraction of code in common with its previous generation GPU? It seems that from what has been written here that moving a platform to a new generation of a currently supported SoC requires about the same work as supporting a wholly new architecture. Also, reasons for abstractions in the OS with regard to code reuse seems, from what has been written here, to have minimal impact in reducing this cost. It isn't true that as a platform matures QA and testing becomes a larger percentage of the overall development cost?

    I would like to understand why the above are true in more detail than have been provided here so far. Do you have a link that lays this out? It would be great if you can povide it.
    Sorry no link....

    But going from 32Bit to 64Bit wasn't an easy thing back when it happened for most OEM years ago - most all did it in stages. And that's very different from having an active development team working on YoY generational improvements in SoCs.

    Besides BB10 development ended with the SD601 that about seven generations ago...
    02-21-20 08:52 AM
  11. conite's Avatar

    I would like to understand why the above are true in more detail than have been provided here so far. Do you have a link that lays this out? It would be great if you can povide it.
    Without having a QNX/BB10 developer symposium at BlackBerry, you aren't going to get the granular details you are requesting.

    None of this matters of course, since the things we DO know make this economically undoable anyway.

    You're simply trying to get a good deal on deck chairs for the Titanic.
    02-21-20 09:00 AM
  12. Chuck Finley69's Avatar
    I guess my primary problem with all these arguments is the question of how reuse impacts cost. Engineers working on the next generation of SoC reuse hardware designs and software code to support it. Do drivers for the next generation of GPU in a next generation SoC have just a small fraction of code in common with its previous generation GPU? It seems that from what has been written here that moving a platform to a new generation of a currently supported SoC requires about the same work as supporting a wholly new architecture. Also, reasons for abstractions in the OS with regard to code reuse seems, from what has been written here, to have minimal impact in reducing this cost. It isn't true that as a platform matures QA and testing becomes a larger percentage of the overall development cost?

    I would like to understand why the above are true in more detail than have been provided here so far. Do you have a link that lays this out? It would be great if you can povide it.
    Again Don, to echo others. What does any of this matter since it’s not possible without BlackBerry Limited involvement. Contact them directly as you feel our replies are speculation. BlackBerry Limited is the only party that can provide the factual answers to your level of satisfaction. Why aren’t you asking their people these questions?
    02-21-20 09:01 AM
  13. bb10adopter111's Avatar
    I guess my primary problem with all these arguments is the question of how reuse impacts cost. Engineers working on the next generation of SoC reuse hardware designs and software code to support it. Do drivers for the next generation of GPU in a next generation SoC have just a small fraction of code in common with its previous generation GPU? It seems that from what has been written here that moving a platform to a new generation of a currently supported SoC requires about the same work as supporting a wholly new architecture. Also, reasons for abstractions in the OS with regard to code reuse seems, from what has been written here, to have minimal impact in reducing this cost. It isn't true that as a platform matures QA and testing becomes a larger percentage of the overall development cost?

    I would like to understand why the above are true in more detail than have been provided here so far. Do you have a link that lays this out? It would be great if you can povide it.
    Drivers are constantly rewritten in response to changes and optimizations in hardware design. They are the most dynamic code sets in the entire stack. Most hardware companies have huge teams of engineers working constantly on these drivers so that the drivers are delivered with the hardware, with the price built in.

    This means that any small OEM not ordering millions of units really needs to use the drivers provided. Custom driver development is prohibitively expensive. When BB10 launched, BlackBerry had to make huge purchase commitments in order for their hardware providers to supply new drivers.

    Any new BB10 development would have to either interface with the existing drivers (meaning the entire code base would need to be redeveloped) or guarantee many billions of dollars in revenue to convince the hardware providers to develop custom drivers.

    The logic and structure of the existing code is likely reusable, but the implementation, optimization, testing, etc. would essentially require a completely new effort.

    I write this with the assumption that you'll argue with me, and I'm sorry that I don't have time to do all the research to expand on this short answer.

    I will give you an example, however. For storage devices, each OEM has a team of dozens of software engineers optimizing drivers for every generation of product and for every generation of interface. That means that every time Samsung releases a new SSD, the drivers are re-written and optimized for each existing platform and each time a new platform is released (such as a new Intel or AMD chipset), the drivers are reevaluated for each supported Samsung storage product. Since the volumes for these products are so large, the costs of this development effort are small on a per device basis. But, if someone invented a new platform and wanted to use Samsung storage products, they would either have to tailor their code to the products, or guarantee the purchase of tens of millions of units to Samsung to compensate them for their efforts.

    Z10 = BB10 + VKB > iOS + Android
    app_Developer likes this.
    02-21-20 09:25 AM
  14. DonHB's Avatar
    Don.... no one here really knows the absolute answers to all your different technical questions.

    But again... what does it matter. BlackBerry had people that understood these things, BlackBerry had people that knew all the other question you have. BlackBerry looked for ways to monetize BB10...

    Do you really think that they didn't consider what your proposing?
    I do believe they have.

    Since the beginning, a survey of what features would get buy-in was the limit of what I had hoped to do. When ever I have tried, the most "vocal" of people here have detoured the discussion to financial, business and technical questions which only BlackBerry is fully qualified to answer.

    Because BlackBerry is unlikely to create such a survey due to all the arguments against such propositions that have been made here over the years, it could be useful information that would otherwise be unavailable. At best the survey results may have caused BlackBerry to reevaluate previous decisions on the subject. At worst (or for some here even better) it could have allowed the naysayers the means to avoid further discussions by directing people to answer the survey.
    02-21-20 09:25 AM
  15. DonHB's Avatar
    Drivers are constantly rewritten in response to changes and optimizations in hardware design. They are the most dynamic code sets in the entire stack. Most hardware companies have huge teams of engineers working constantly on these drivers so that the drivers are delivered with the hardware, with the price built in.

    This means that any small OEM not ordering millions of units really needs to use the drivers provided. Custom driver development is prohibitively expensive. When BB10 launched, BlackBerry had to make huge purchase commitments in order for their hardware providers to supply new drivers.

    Any new BB10 development would have to either interface with the existing drivers (meaning the entire code base would need to be redeveloped) or guarantee many billions of dollars in revenue to convince the hardware providers to develop custom drivers.

    The logic and structure of the existing code is likely reusable, but the implementation, optimization, testing, etc. would essentially require a completely new effort.

    I write this with the assumption that you'll argue with me, and I'm sorry that I don't have time to do all the research to expand on this short answer.

    I will give you an example, however. For storage devices, each OEM has a team of dozens of software engineers optimizing drivers for every generation of product and for every generation of interface. That means that every time Samsung releases a new SSD, the drivers are re-written and optimized for each existing platform and each time a new platform is released (such as a new Intel or AMD chipset), the drivers are reevaluated for each supported Samsung storage product. Since the volumes for these products are so large, the costs of this development effort are small on a per device basis. But, if someone invented a new platform and wanted to use Samsung storage products, they would either have to tailor their code to the products, or guarantee the purchase of tens of millions of units to Samsung to compensate them for their efforts.

    Z10 = BB10 + VKB > iOS + Android
    Can old drivers (or their code) be used on new hardware and if so how much (generally) would performance be compromised? Further, how much work would be required to assure a reasonable level of reliability and how does this compare to the cost of new development? Will doing what these questions imply increase technical debt?
    02-21-20 09:52 AM
  16. Chuck Finley69's Avatar
    I do believe they have.

    Since the beginning, a survey of what features would get buy-in was the limit of what I had hoped to do. When ever I have tried, the most "vocal" of people here have detoured the discussion to financial, business and technical questions which only BlackBerry is fully qualified to answer.

    Because BlackBerry is unlikely to create such a survey due to all the arguments against such propositions that have been made here over the years, it could be useful information that would otherwise be unavailable. At best the survey results may have caused BlackBerry to reevaluate previous decisions on the subject. At worst (or for some here even better) it could have allowed the naysayers the means to avoid further discussions by directing people to answer the survey.
    So you think BlackBerry would look to CrackBerry forums to make important business case decisions instead of their own proprietary comprehensive data, that’s verifiable, quantifiable and far more statistically relevant?!?

    I’ll give John Chen a F’n call right now and tell him to answer your questions so you’ll put all to rest....

    Sup JC?!? It’s Chuck, yep, that Chuck... everyone great, thanks for asking...

    Seriously...... We’re only legends in your mind but I appreciate the compliment
    JeepBB likes this.
    02-21-20 09:52 AM
  17. DonHB's Avatar
    So you think BlackBerry would look to CrackBerry forums to make important business case decisions instead of their own proprietary comprehensive data, that’s verifiable, quantifiable and far more statistically relevant?!?

    I’ll give John Chen a F’n call right now and tell him to answer your questions so you’ll put all to rest....

    Sup JC?!? It’s Chuck, yep, that Chuck... everyone great, thanks for asking...

    Seriously...... We’re only legends in your mind but I appreciate the compliment
    Seriously? Enough to merit further investigation, perhaps?
    02-21-20 10:08 AM
  18. Chuck Finley69's Avatar
    Seriously? Enough to merit further investigation, perhaps?
    Their own internal proprietary data tells BlackBerry all it needs to know. CrackBerry is a fan site with too much bias in all directions for serious analytical sampling. It’s the vocal minority sampling used to predict Hillary Clinton’s electoral victory. Same basic concepts, politics excluded
    02-21-20 10:15 AM
  19. Eumaeus's Avatar
    Cunard should raise the Titanic from its resting place, refurbish it, and offer transatlantic trips on it again. It is only four miles down, after all. And there was a lot of work that went into its construction. Just raise it, a little welding, mopping, and window-washing, and it should be good to go. This would have to be cheaper than building a new ship from scratch. I know that Cunard currently has no plans to do this, but perhaps we could take a survey to see how many folks would like to buy tickets. There has to be a niche market for slow ocean crossings on a tiny (by modern standards), poorly-engineered coal-fired steamship from 1912, lacking all modern safety equipment. Even if they only did one or two trans-Atlantic trips, it would surely be worth the money.
    02-21-20 10:16 AM
  20. bb10adopter111's Avatar
    Can old drivers (or their code) be used on new hardware and if so how much (generally) would performance be compromised? Further, how much work would be required to assure a reasonable level of reliability and how does this compare to the cost of new development? Will doing what these questions imply increase technical debt?
    It depends. For Intel enterprise hardware, which is designed to last a decade or more, hardware OEMs will create stable drivers designed for long term compatibility. These prioritize stability and sacrifice performance. They still, however, need to be revised for major new architectures or OS versions.

    In mobile, where the pace of change in OS design and hardware architecture is so much faster, especially over the past decade, I don't know if there is any analogue. The problem.with drivers is that, if they are not 100% perfect, they can easily crash entire applications, subsystems or the OS itself, and there is nothing the user can do about it. Imagine if every time your camera app tries to write a photo to a file in memory, the app crashes and you lose your photo, or your phone reboots, or your battery drains in an hour when using WiFi.

    These kinds of problems were apparent at the BB10 launch precisely because it was all new, slightly buggy code, and was only about 98% optimized, and even doing that required a multi-year, multi billion dollar effort.

    This is why most of us believe you to be underestimating the technical risk and level of effort required to reuse BB10 code.

    Given a choice between a) rebuilding the BB10 application layer (user interface and apps) on an Android fork to fully replicate the BB10 user experience with any desired modifications to privacy and security policies and b) re-engineering the drivers and subsystems to reuse the BB10 application layer and apps, option a would be less risky and less resource intensive, assuming the goal includes using current and future hardware.

    Z10 = BB10 + VKB > iOS + Android
    Last edited by bb10adopter111; 02-21-20 at 11:17 AM.
    02-21-20 10:18 AM
  21. conite's Avatar
    Seriously? Enough to merit further investigation, perhaps?
    Get signatures from 10,000 people that they'd be willing to spend $10,000 for an autoloader, and you might just have a starting point.
    02-21-20 10:27 AM
  22. app_Developer's Avatar

    Since the beginning, a survey of what features would get buy-in was the limit of what I had hoped to do. When ever I have tried, the most "vocal" of people here have detoured the discussion to financial, business and technical questions which only BlackBerry is fully qualified to answer.
    So make the poll. Add the disclaimer that you don't care about technical or financial feasibility about any of it. Respondents should understand that they are choosing between unicorns and rainbows and pixie dust and they can tell you what they would like.

    You can do that right now. Or post the list of features and I'll do it myself.
    02-21-20 11:06 AM
  23. app_Developer's Avatar
    Given a choice between a) rebuilding the BB10 application layer (user interface and apps) on an Android fork to fully replicate the BB10 user experience with any desired modifications to privacy and security policies and b) re-engineering the drivers and subsystems to reuse the BB10 application layer and apps, option b would be less risky and less resource intensive, assuming the goal includes using current and future hardware.
    ^^^ 1000% this. In fact (a) is an interesting, realistic and fun conversation. I'd even contribute code to that effort if someone could get that off the ground and lead it.
    02-21-20 11:08 AM
  24. Dunt Dunt Dunt's Avatar
    So make the poll. Add the disclaimer that you don't care about technical or financial feasibility about any of it. Respondents should understand that they are choosing between unicorns and rainbows and pixie dust and they can tell you what they would like.

    You can do that right now. Or post the list of features and I'll do it myself.
    Reality is even a Poll won't accomplish much....

    Heck most threads on CrackBerry these days don't hit over 10K views unless they go on for a LONG time. And I'm not sure how CrackBerry even counts views... some of us might be counted multiple times as we use different devices connected to different networks. I assume only registered users could vote, so that should give him a picture of just how little interest there is... at least here.

    To get BlackBerry "interested".... next phase would be crowdfunding where the vote is your credit card payment.

    Donation of $50 get's you a BB11 hat.
    $100 get you a BB11 T-Shirt.
    $500 a BB11 Jacket.
    $1000 a signed copy of John Chen's memoirs
    $2,0000 a link to download a future copy of something called BB11
    $3,000 all of the above.

    NOTE: no refund and a minimum goal of $300 million required. Once funding is complete, then they determine just what BB11 will be.
    Troy Tiscareno likes this.
    02-21-20 11:46 AM
  25. Vlad Dl's Avatar
    These kinds of problems were apparent at the BB10 launch precisely because it was all new, slightly buggy code, and was only about 98% optimized, and even doing that required a multi-year, multi billion dollar effort.
    Why on Earth would they ditch this for Android and start all over again with that? They should have expected new system to pick up at last not in a year or so. Makes me wonder how dumb Mr Chen has to be to jump from platform to platform. How much money they have spent securing Android to their liking? Is it really secure? It shows freaking Google search on home screen.
    02-21-20 01:15 PM
731 ... 1213141516 ...

Similar Threads

  1. Exchange online after October 13, 2020?
    By richdb in forum BlackBerry 10 OS
    Replies: 39
    Last Post: 05-08-20, 05:25 PM
  2. Replies: 13
    Last Post: 03-06-20, 03:16 PM
  3. Why do I suddenly have a character limit in my emails?
    By iozier in forum Ask a Question
    Replies: 2
    Last Post: 03-03-20, 03:35 PM
  4. BlackBerry Bold 9900 in Feb 2020?
    By jlscott2 in forum Ask a Question
    Replies: 4
    Last Post: 03-02-20, 07:39 AM
LINK TO POST COPIED TO CLIPBOARD