Several others have responded. I've tried to not overlap. | From: o1bigtenor via talk <talk@gtalug.org> | On Wed, Feb 17, 2021 at 12:53 PM D. Hugh Redelmeier via talk | <talk@gtalug.org> wrote: | The kernel report has things at 5.4.99 already. What does that mean? Generally speaking, you can have many versions of the kernel installed and at boot time choose which to boot. Simple heuristic: boot the latest one that works. Bonus: you can "bisect" your problem. That means: test a sequence of kernels to find out when the problem showed up. As the word "bisect" suggests, if the sequence is long, you can simply do a binary search in that sequence so that instead on expecting to do N/2 test boots, you can find the point using about log-base-2(N) boots. debian probably has a repo with a large sequence of kernels, many of which you have never tried. Bisecting in this set would narrow the focus even further. You can do an even finer job by bisecting on the kernel.org git tree. git even has a "bisect" command to help. You should then be able to tell exactly which kernel patch introduced your problem. The above assumes - the bug is a single one - the bug is in the kernel - the bug came from upstream In your investigation, you might discover that one or more of these assumptions is wrong. Does this sound like a lot of work? Maybe. Who is best to do this: you. After all, you have the system that manifests the problem. But there is a shortcut you should try first: has someone else hit this already? If so, have they solved it? (Very often it is hard to tell if two sets of symptoms are manifestations of the bug.) | > Are you using the proprietary nvidia driver? If so, good luck getting | > support. With my distro, bug reports are only dealt with if no | > proprietary drivers are loaded. Ditto for upstream (the kernel | > bugzilla). | > | > Nouveau is not proprietary so it is OK. But it is pathetic. | | Hmmmmmm - - - - that really isn't good news. | | Oh well sounds like same old same old! To get help, you really have to answer questions. You didn't. What do you mean by "same old same old" here? I do know what the phrase normally means. | From: o1bigtenor via talk <talk@gtalug.org> | 1. You're still using google for a search engine - - - - - wow - - - - can | I introduce you to Duckduckgo? So far, at least, they're not not harvesting | your data and all your searches. I am using "google" as a generic verb. I use Duckduckgo. I'm not confident that it will save me. | 2. A multi-gpu and multi-monitor system on LInux is considered very extreme | fringe. Almost all the upper level coders seem to find that more than one | gpu is no fun so they run maybe 2 monitors. Its not the same thing when you | get to multi-gpu AND multi-monitor. Those with serious technical expertise | just don't seem to run this kind of stuff. I'm a programmer. I find multiple monitors no longer very useful I just use UltraHD monitors and they fill my field of view. On my latest toy, I can drive three UltraHD monitors. But I don't. I do have three UltraHD monitors around me, visible from my chair (but I have to rotate). Ergonomics are very personal. But that doesn't mean that each of us knows what's best for our self. I often report to this list what works for me. | what used to be called the power user - - - - in (and on) LInux is considered | not worth wasting time on - - - so finding expertise - - - - - | incredibly difficult. I don't think that this is true. | From: o1bigtenor via talk <talk@gtalug.org> | There are a lot of these 'official' volunteers that are working day jobs that | strangely are very very like their volunteer interests. This sounds like a veiled accusation. Please spell it out or withdraw it. | When I was beating my head against the wall for over 2 months when I | first set up this system the amount of information available or offered to | me was close enough to zero so as to make it a moot point for open source | support. 1. you need to improve your problem reports 2. you need realistic expectations. 3. useful information may not be organized the way you understand the problem Of course open source helps: it lets anybody solve a problem. | In the years since I've found that most coder types really don't get into | screen real estate - - - - in fact some serious programmers were still using | 19" monitors (within the last year) where I had moved to a 1600x1200 crt | some 20 years ago. I don't like the word coder, at least when applied to me. I'm a programmer. (Other's liked the term "Systems Analyst", but I don't.) When I started out, we didn't even have screens. | Interesting - - - - the one thing that you don't mention and that I think | might be the case here is that the actual bug may be a specific confluence | of things. So even with a complete listing of all of what you have - - - - the | actual problem still isn't visible. I spent a few hours reading through files | in /var/log/ and haven't been able to find anything. Which is why I was | asking for assistance in where to ask. (The assumption seems to be | that I don't have any idea - - - - and maybe I don't . . . . .) If you had an idea, I assume that you'd have solved the problem. "Have you checked if the computer is plugged in?" When helping someone, you really need to double-check all the basics. | From: o1bigtenor via talk <talk@gtalug.org> | On Wed, Feb 17, 2021 at 5:14 PM Dhaval Giani <dhaval.giani@gmail.com> wrote: | > If it is worth your time, maybe you can develop the expertise. We | > welcome you to contribute to the community. | > | Haven't found any interest - - - - in fact mostly the opposite. debian project is very welcoming of help. Little pockets might have pathologies -- I don't know. But I doubt that's what you've hit. It is easy to jump in and alienate a group. Your postings here have done just that. You should have noticed this. Don't do that if you expect help or sympathy. It's not constructive. It is possible to recover. | From: o1bigtenor via talk <talk@gtalug.org> | Any tool that doesn't have clearly defined limits usually doesn't get installed | here. My trust factor in software isn't very high - - - - if it even | exists in all | honesty - - - that's called experience. I understand that. That policy is unworkable with modern software and hardware. If you are adamant, you must go back to perhaps the 8-bit era. - all modern processors have secret software running all the time that is invisible to the owner. For example, it was discovered a few years ago that every Intel x86 processor has a Minix system embedded inside running on a supervisory processor that is not in the architecture. - all modern OSes are too big to comprehend, even if you have the source. - all modern software has more contributors than you can vet. - all modern hardware surely has bugs that allow security violations. Every year a bunch are reported -- how many are unreported? We are standing on the shoulders of giants (and a lot of ordinary folks). If we don't trust them, we need to go out naked into the wild. A simple trust / don't trust model won't work. | - - - hard to get work done when you | need about 1/2 hour to set up things the way you want them and then | within a few hours you need to do it all over again. Computers are good at automating tasks. Surely you could automate "setting up the things the way you want them". | So what do I do - - - - no clear path at this point yet except perhaps to | install kernel 5.4.99 (last time I looked at the kernel info) and then run that, | maybe I can keep that kernel and just update everything else - - - dunno. Learn to work constructively with the debian community. It's a marvel. I say that as (mostly) an outsider.