> > On top of that I'll only buy hardware that can be free software down > > to the bios. It should be nothing to them what runs in the bios, so > > I'll wait and see if they can be more flexible there (for however long > > it takes). > > I've been trying that for a while now, but my server is starting to > fail. > > It may not be an option unless I want to get out of computing > altogether.
If one peers closely enough at the various speculative execution bugs alluded to in this thread, and the horrible modern web-browser in the thread next door, one realises that they are just different facets of the same problem. If browsers were simple and didn't contain javascript interpreters, then the CPU bugs wouldn't be that serious on a single user machine as it would be much more difficult to get hostile code to run on the CPU. In the same way, the biggest memory and CPU hog on a modern computer is the web browser - it is not uncommon to have browsers consume multiple gigs of RAM to display data which contains a few hundred bits of actual information (train time-table, bank-balance, tomorrow's weather forecast). So if were not for the browser a PC from the last century would do, and that would mean that a simple, single core CPU without wild speculative execution modes, insane pipelines, signed microcode or management engines would suffice - the kind that enthusiasts occasionally build in verilog, and the role that maybe RISCV will fill more comprehensively. So maybe don't give up on computing, nor give up on the internet, but give up on the web. Maybe we should all host things gopher, ftp or bittorrent and related protocols... talk via irc or smtp, not web forums. Read mail with an actual mail client, not a web frontend. regards marc _______________________________________________ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
