On Wed, Sep 17, 2025 at 02:36:55PM +0000, Colin Percival wrote: ! On 9/17/25 06:51, Peter 'PMc' Much wrote: ! > It is about a deliberate decision to change system defaults in a ! > way as certain commercial powers want it, no matter whether that ! > implies beating up some ordinary users like me. ! ! Release engineer here. I'm not paid by any nefarious "commercial powers" ! (although I wouldn't object to getting paid something for the hundreds of ! hours I'm spending on managing this release), and I approved the change.
Great! Thanks a lot for responding! ! When our CPU is pretending to be a CPU from the 1980s there isn't enough ! memory to support everything we'd like to do, and some tradeoffs were ! needed. Warner consulted within the project and I think we picked the ! right tradeoffs. Okay. I have no problem with the decision itself (even if I dont really understand the pros and cons), as long as I get to know the fact that I can roll my own. ! The long-term solution here is to just stop using BIOS boot mode. There ! are very few places (mostly obscure virtualization environments) where we ! can't boot EFI at this point. Just FYI: the dedicated servers which I rent from my hoster can only boot MBR. Not even GPT. (But then also, I never got the idea to try and run graphics over IDRAC. Might actually be possible...) But also then, if there were a good statement somewhere, along the line of "booting into hi-res will no longer be supported for non UEFI-Installations. If you need to use that, either boot via UEFI or compile your own loader.> That then would be something in the good old style that catches attention, telling us thet something is *NOT* working anymore. BTW, I don't know why the release-notes now read like advertisement sheets - not even IBM ever did that (with them everything sounded like either advertisement or lawyers task, except the release-notes: these carried the valuable information about what one can and cannot do). cheerio, PMc
