Announcement

Collapse
No announcement yet.

Intel Makes ControlFlag Open-Source For Helping To Detect Bugs In Code

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Intel Makes ControlFlag Open-Source For Helping To Detect Bugs In Code

    Phoronix: Intel Makes ControlFlag Open-Source For Helping To Detect Bugs In Code

    Last year Intel announced ControlFlag as a machine learning tool for helping to uncover bugs within code. ControlFlag promised impressive results after being trained on more than one billion lines of code and at the end of 2020 was already being used internally on Intel's code-bases from firmware to software applications. We hadn't heard anything more about ControlFlag this year... Until today. Intel has now made ControlFlag open-source for helping to autonomously detect more programming bugs...

    https://www.phoronix.com/scan.php?pa...ag-Open-Source

  • #2
    Originally posted by phoronix View Post
    Intel has made training data available that they generated from 6,000 open-source GitHub repositories.
    Hmm... did they make sure the repositories were under open-source licenses? Because there are many that don't have a license or have restrictive (source-available) licenses, and I remember the Github Copilot controversy...

    Comment


    • #3
      Another brute force tool to facilitate and promote laziness.

      Comment


      • #4
        Originally posted by tildearrow

        Are you sure? Do you mean productivity? ;p
        My observation is the more people rely on computers and "smart" tech to do the thinking for them, the dumber they get. That does not bode well for productivity.

        Comment


        • #5
          Originally posted by ddriver View Post
          Another brute force tool to facilitate and promote laziness.
          That's why technologies exist.

          Comment


          • #6
            Originally posted by ddriver View Post

            My observation is the more people rely on computers and "smart" tech to do the thinking for them, the dumber they get. That does not bode well for productivity.
            I used to hear that about calculators.

            Comment


            • #7
              Originally posted by RedEyed View Post

              That's why technologies exist.
              For you maybe. I'd rather stick to usage models that increase rather than reduce my capabilities.

              You can write write good software, or you can write bad software and rely on more bad software to fix it.

              Originally posted by sinepgib View Post

              I used to hear that about calculators.
              There is nothing wrong with using a calculator, after you know and understand math. There is a lot of merit to understanding and being able to perform math that you will never acquire if you always and only use a calculator. Chances are, you will be quite limited in what you can do with that calculator.
              Last edited by ddriver; 20 October 2021, 03:35 PM.

              Comment


              • #8
                Intel, please use this tool on your own icx/icpx compiler (and while at it, at clang/clang++, too). Thanks to a serious compiler bug I bricked my Arch install today.

                Comment


                • #9
                  Originally posted by ddriver View Post
                  There is nothing wrong with using a calculator, after you know and understand math. There is a lot of merit to understanding and being able to perform math that you will never acquire if you always you as a calculator. Chances are, you will be quite limited in what you can do with that calculator.
                  So we agree then. You can use tools if you understand what they do and just take a part of the cognitive burden from mechanical tasks. On a side note, but keep in mind I don't think this is generally true, I used static analysis many times to learn. I'd run these over existing projects (mostly open source) and tried to understand whether the warnings were false positives and, if not, what they meant and why they were happening. It's all about the mindset and how you use the tools, not about whether you use them or not.

                  Comment


                  • #10
                    Originally posted by sinepgib View Post

                    So we agree then. You can use tools if you understand what they do and just take a part of the cognitive burden from mechanical tasks. On a side note, but keep in mind I don't think this is generally true, I used static analysis many times to learn. I'd run these over existing projects (mostly open source) and tried to understand whether the warnings were false positives and, if not, what they meant and why they were happening. It's all about the mindset and how you use the tools, not about whether you use them or not.
                    If you make such trivial errors that can be caught by a static analyzer, you clearly don't "already know math", and you are not using the "calculator" as a time saver, but as a mitigation of the absence of skill.

                    Could you learn from issues, detected by a static analyzer? Theoretically yes, just like you could and should have learned from actual language learning materials. Does this happen in practice? From my experience mostly no - people will use just about anything to indulge their laziness, in the process reinforcing it further.
                    Last edited by ddriver; 20 October 2021, 03:28 PM.

                    Comment

                    Working...
                    X