Jump to content

NotHereToPlayGames

Member
  • Posts

    6,714
  • Joined

  • Last visited

  • Days Won

    83
  • Donations

    0.00 USD 
  • Country

    United States

Everything posted by NotHereToPlayGames

  1. It's the only effective solution for this. Been using it for twenty years! Long Live Proxo.
  2. Pleasse count the .dll's for other anti-virus programs and report back your findings on them also. McAfee for one is also very big into this "DLL file invasion and abuse". Taking this thread to such "anti-Kaspersky" extremes serves your viewers a dis-service and not a "service". What @UCyborg posted is of USEFUL IMPORTANCE and is INFORMATIVE to the viewer. We need more of THAT instead of useless "massive security concern" innuendos without actual proof or verification.
  3. Totally agreed. It's like anti-alias sub-pixel fonts. I get migraines from anti-alias sub-pixel fonts and most of my coworkers can't even "see" the difference. I've also noticed those crappy quality images even here at MSFN. But they weren't .webp at the time, they were .png images and it was tied to what they used in order to do the screencap. It doesn't seem that widespread to me. My news and financial web sites don't rely on .webp as they are not that graphic-intensive. And I don't do games, so no frame of reference there. It is extremely surprising that even if I set up an "accept" header that should indicate "don't serve me weppy", the web sites IGNORE the "accept" header and serve them anyway.
  4. It's still version 86 under the hood and the letters "webp" do not exist in the changelog.
  5. Actually, allow me to rephrase that. We (the sort of folks that become members of forums such as MSFN) may not agree on MVPS Hosts versus hardware firewall versus software firewall versus real-time full-time anti-virus versus on-demand malware scans versus Proxomitron versus uBO versus uMatrix versus DoH versus NoScript verus HTTPS Everywhere versus a hundred different things, but what we all do have in common is that we do Practice Safe Hex in our own preferred ways. I'll use uBO as an example. While this "Weppy Scare" does supposedly exist "in the wild", my hunch is that their is a uBO "list" that already safeguards from the "in the wild" web site that technically only exists "in theory". I wouldn't mind knowing EXACTLY where this "in the wild" actually IS. But they never seem to tell you that, it's just the normal "update now!" routine, "You are not safe unless you update now! Update your OS! Update your browser! Update Now!" "Blah blah blah" - https://www.youtube.com/watch?v=mfJhMfOPWdE
  6. This is not going to be "for everyone", some folks believe in every "scare tactic" thrown at them. But "for me", I'm opting to do NOTHING in regards to this Weppy Scare. My computer does NOT protect against Meltdown and Spectre and it will also NOT protect against Weppy. We used to have a saying, "Practice Safe Hex". I've never been hit with a virus or malware and I don't visit the sorts of web sites where one is prone to these "dark shadows". To each their own, of course. But to me, this is just hype and propaganda. Much ado about nothing. Mileage may vary.
  7. To be honest, I have opted to do NOTHING. I'm not in the least bit concerned with all of this recent "hype and propaganda" regarding .webp. To each their own, of course. But to ME, it's nothing more than any other virus or malware out there that has never hit my machines. Reminds me of back in the day when I was a "church goer" and everybody knew I could fix computers so everybody would have me fix them for them. You'd get the same people over and over again. To the point that you FINALLY have to tell them (due to the items discovered on their computers), "You wouldn't get these if you stopped visiting p0rn sites." Sure, there is always the THEORY that .webp could come in via a third-party "advertisement" that your otherwise SAFE web site wouldn't otherwise have, but these 'in the wild" reports DON'T WORK THAT WAY.
  8. Confirmed, sadly. In newer versions of Ungoogled Chromium, there is a flag to set your "accept" header (ie, for those that don't use Proxomitron). The "gallery test page" displays these webp images whether the "accept" header indicates the browser can render webp or not. But I also wonder if this "gallery test page" is even sending true .webp images because IE8 does not support webp and this "gallery test page" shows these .webp images even with an IE8 useragent and client hints disabled. Though I am also on Win10 at the moment so unsure if that is playing a factor.
  9. Agreed. Coincidence ≠ Proof Q.E.D.
  10. So Schrödinger's cat is both alive and dead, got it.
  11. Agreed. A very easy fix using Proxomitron. At least it should be, I haven't tried as of yet.
  12. I do this via Proxomitron. But for the non-Proxo 360Chrome user, I can follow @Dixel's suggestion and upload a revision within the next day or two. I'd likely only upload a new rev for build 13.5.2036 as it is the only version I still use. Unknown which versions we still have MSFN Members using.
  13. Agreed. That's my take also. Until when-and-only-when we see a Google Bug Report specifically cite MSFN, then NO, they don't base their development from our "rants". Same goes for REDDIT. I promise you, and any search engine is "the proof is in the pudding", for every ONE "rant" regarding Google here at MSFN, there are TWO HUNDRED over at Reddit. Google isn't "listening" to Reddit as far as altering development as far as trying to "prevent bad PR". I kind of think we "see ourselves" as much more "influential" than we really are. "Nobody is listening."
  14. So I guess the question is technically a "yes or no" question. Would a mathematical mind that follows the Scientific Method claim a gigantic software company makes software decisions based on posts here at MSFN that THEORETICALLY could be "data-mined" by AI as part of Google's marketing department agenda to "avoid bad PR"?
  15. A fellow MSFN Member is claiming that the reason that Chrome version 119 works in Windows 10 1507 is because the creators of Chrome read MSFN posts and was afraid of bad PR if they didn't make v119 work in 1507. The claim is that it is just too much of a coincidence for this to have happened without Google reading MSFN posts and making software decisions based on those posts. No taking sides!
  16. I will accept @AstroSkipper's perspective on this - but he cannot "agree with you just because it is you", WAY too much of that going on here at MSFN. Him being a mathematician will surely not accept "theoretical speculative heresay" over mathematic logic and reasoning. That is, if he's even inclined to comment. Totally up to him, of course. This "theoretical speculative heresay" would never hold up "in court of law", as the saying goes. Under a paradigm of theoretical speculative arguments, one must admit that one can claim ANYTHING that they want, no matter how far-fetched, and place it under an umbrella of "theoretical cause and effect" versus a true Scientific Approach. I'm being serious, just ponder for a moment the MOUNTAINS of arguments one could EASILY pass off as "reality" if speculation and "the help of AI" were our Litmus Test. *I* could claim that *I* invented anti-lock brakes because I slid off the road in the Winter of 1970 and posted something at MSFN regarding the incident and that post was THEORETICALLY "data mined" before the term AI ever became mainstream. We like to claim Google reads MSFN and creates google-isms based on that reading - sorry, it's not happening, it's just our minds creating our own reality. The difference between reality, illusion, and imagination is not that big of a psychological concept - the difference exists solely within one's mind. My mind. Your mind. Insert-name-here's mind. I work in Engineering. If there isn't DATA and NUMBERS to "back it up", then it doesn't exist.
  17. I kind of doubt it. I am not positive, but I think that some web browsers (may only be macOS) do convert-in-place conversions and display .webp images even when .jpg was issued. You would have to alter the HTTP "Accept" Request Header in order to prevent this.
  18. Win10+ users also have the option of removing the wepb (pronounced "weppy", if I recall correctly) -- Get-AppxPackage -allusers Microsoft.WebpImageExtension | Remove-AppxPackage –allusers Those familiar with PowerShell will know how to use that line.
  19. Yes, but also just a theory. Without the company's web site saying right there in black and white for all to read, "Due to discussions at MSFN, we have altered our policy so as to avoid any possible bad PR", then this theory is nothing but SPECULATION. And let's face it, if I posted such SPECULATION here at MSFN, we would have a small "gang" of three or four MSFN Members make it their mission to "run me out of town". West Side Story. AI is not a mature technology yet. AI can be "wrong" just as easily as it can be "right". No mature company is going to rewrite corporate policies based on some AI feed giving them "misinformation". But anyway, lol. I'll leave it for the "gang" to settle.
  20. No it isn't. More like proof that you "see what you want to see". And obviously this post is "just as" OT as yours. West Side Story. And no, it's not a "western". Good day, mate.
  21. I'm in the "not concerned" camp.
  22. On your Win10 1511, do you have a "System and compressed memory" process always running at 15% to 30% for CPU when sitting IDLE? Win10 1607 on the same exact laptop idles at 1% to 3%.
  23. Yes. Roughly 5 to 7 seconds in Win 10 1511 (inline with XP). Roughly 14 to 17 seconds in Win 10 1607. First launch and first launch only coming out of hibernate or full shutdown.
×
×
  • Create New...