Announcement

Collapse
No announcement yet.

Intel Clears Up Microcode Licensing Controversy - Simpler License, Allows Benchmarking

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

  • chilinux
    replied
    Be warned that the tweet from Imad Sousou last year was extremely misleading. The Intel VP Chief Architect of spin for Intel "open" source implied that they did more than just change the microcode license and that they fixed the Intel license in general.

    However, the Intel license including an anti-benchmark clause still continues to be used. It can be found here:
    https://software.intel.com/en-us/lic...=20555&elqat=2

    You can find references to it still being used with such products as the Intel OpenVino Toolkit as can be found here:
    https://software.seek.intel.com/open...olkit?os=linux

    So, not only do conditions Intel acknowledged are problematic to the open source community still exist in Intel licensing, they are referenced by products Intel state as being "open." That issue wasn't completely cleared up a year ago and is still going on today.

    Best part is that if the courts determine benchmarks are a standardized way to review a computer company's product, then the clause might be illegal under the USA federal Consumer Review Fairness Act along with laws in several other states.

    I believe 3mdeb that they heard Intel claim they would do changes to licensing that sounds like it would be helpful. However, I also believe Intel is willing to use provisions that may be illegal, use the term "open" in a misleading way and just are all around extremely dishonest. If Intel actually follows through on being helpful, I will be pleasantly surprised. But I still have learned to still expect Imad Sousou style of over promoting "changes" made by Intel while actually under delivering on the results.

    Leave a comment:


  • juanrga
    replied
    Originally posted by Candy View Post
    Intel is going to die a horrible death, once everything is moved on with ARM (no matter for their license to fuck around with ARM cpus)
    ARM and RISC-V. Don't forget RISC-V!

    Moreover, there is a rumor about Intel working in a hybrid architecture that can execute both x86 and ARM.

    Leave a comment:


  • Michael
    replied
    Originally posted by GrayShade View Post
    Michael will you run some benchmarks on the new version?
    It's my goal for this weekend.

    Leave a comment:


  • chilinux
    replied
    "but knew would be corrected"

    How did you know that? Imad Sousou's (the guy that tweets as "#IAmIntel" because he is a VP/GM) statements to The Register on Tuesday seems to make clear he had no intention of correcting it.

    "didn't feel like reporting the hype train"

    "Following all of the public attention and pressure, Intel was quick to clear up the situation. "

    If it takes reporting on the situation to get Intel to clear up the problem, is it really fair to reduce the reporting to being a "hype train"?

    Anyone that has talked with Randal Schwartz (one of the top experts in Perl) about Intel knows that when Intel releases "some overzealous Intel legal folks" on someone that REALLY BAD things happen. Taking care of this sooner instead of later was kind of critical given what was implied by the license was highly restrictive (the open ended reference to "benchmark" may have also included prohibiting code profiling) and was holding up critical security patches to Debian users. Once Bruce Perens posts about an issue to his blog, it usually is safe to say that other options to communicate with Intel has been exhausted.

    I really highly recommend doing an interview with Randal Schwartz. If you do, I would then ask that you please reconsider your position to what degree the community/press should be cautious about Intel using overly broad legal language and the degree Intel needs to be called out when they do. They are powerful enough that even little "mistakes" have a huge impact.

    Leave a comment:


  • GrayShade
    replied
    Michael will you run some benchmarks on the new version?

    Leave a comment:


  • Guest's Avatar
    Guest replied
    Man, the kinds of things people obsess about... If it says "GNU" before "Linux", if it's GPL2 or GPL3, if the proprietary software is in the repo or outside or if there's an alert when you install the OS, etc etc. Can you imagine if all this energy was put into making distros betters, writing newer better GUI toolkits and better apps using said toolkits?

    Leave a comment:


  • Michael
    replied
    Originally posted by sunweb View Post
    Where is the original article about Benchmark's Bans?
    I knew it was going to be rectified so had to wait until it actually was to report on it rather than just spreading news that wasn't entirely accurate.

    Leave a comment:


  • sunweb
    replied
    Where is the original article about Benchmark's Bans? Its similar to Nvidia's situation. When they *ucked up you didn't write anything, like it never happened. Now when Intel fixed it you wrote about it. Journalism ain't the same anymore...

    Leave a comment:


  • ids1024
    replied
    Originally posted by Xicronic View Post

    Shouldn't be legal to write stuff like this
    It's standard language in closed source software. It isn't surprising to see in microcode. Reverse engineering for "interoperability" is protected, but it's not clear what counts. I would hope there's some protection for security research, but I'm not aware that there is.

    The Atari v. Nintendo court case is an interesting one relating to reverse engineering.

    Leave a comment:


  • hgoldfish
    replied
    Is the link of new license broken?

    Leave a comment:

Working...
X