Re: [DNG] I'm not part of the Debian project
Quoting Steve Litt (sl...@troubleshooters.com): > I'm the originator and 2 year maintainer of the VimOutliner project. > The Debian package substituted a double backslash \\ for the double > comma ,, command prefix. When I said that the double comma was > selected for speed purposes and that double backslash was a huge > slowdown, the Debian maintainer gave me this song and dance about > customary defaults in Debian and Vim. Meanwhile, as "upstreams", we > were forever asked to debug Debian installations because they didn't > work. After a few attempts, we just told them Uninstall your Debian > VimOutliner and install from our tarball. That was the quickest fix. As our Aussie friends say, 'Good on ya.' Thanks for VimOutliner. During that period, it would have been a fine and good thing (if you felt like it) for you to produce a .deb and publish it at an apt-compatible repo URL. Nobody has any right to demand this, and people should be grateful for your work -- but IMO in general installing from upstream tarballs _if there is a reasonable alternative_ is a very bad idea and should be strongly discouraged for new Linux users, for many individually compelling reasons. During the long period when I was one of the main editors for _Linux Gazette_, I wrote about this several times (often as an editorial footnote to well-intended articles) to make sure our readers were not mislead by bad advice, such as here: http://linuxmafia.com/~rick/weatherwax.html#1 As you know, Steve, I am an avid and appreciative reader of your technical articles -- but one of the things you frequently urge on readers that I feel is misguided is (IMO) urging immediate resort to upstream tarballs in cases where this is _not_ necessary and not the best choice. (To phrase my specific point about VimOutliner a different way, you do the free-software world a notable favour by maintaining VimOutliner, and the correct comment first and foremost is 'Thank you', so: Thank you. If you felt like helping people even more, helping them better, and helping more people, taking that small extra step using debhelper to create a package would also be cool.) OTOH, there certainly are cases where dispensing with the extra work of making a package is perfectly reasonable because there isn't enough advantage to justify the extra work. An init system and/or supervisor such as runit would probably be a good example of that. (I noted these exceptions in the last two paragraphs of my footnote.) ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] I'm not part of the Debian project
> > > I have no interest in half-way fixing Debian's broken packages, but > I'll be glad to help create an excellent DeVUan package that respects > what the software's author was trying to accomplish (but without any > new directories off the root). If the new Devuan runit package work out > as well as I think it will, I can help do the same thing for s6, and I > talk to s6's developer on a regular basis and I think he'd be helpful. If you feel like doing that, I say its a great idea. Runit-init would be a good addition to the repo. > > SteveT > > Steve Litt > October 2017 featured book: Rapid Learning for the 21st Century > http://www.troubleshooters.com/rl21 > ___ > Dng mailing list > Dng@lists.dyne.org > https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] I'm not part of the Debian project
Hi all, A couple days ago I volunteered to be one of a two person team to bring a good, high quality runit package to DeVUan. The package would be capable of using runit as a simple process supervisor, or assuming all init responsibilities. I would consult the creator of runit so our package would reflect his vision of how runit is supposed to work. And now here comes a troll I /dev/nulled in 2015, apparently suggesting I fix DeBIan bug number such and such to achieve a DeVUan runit. U, no. Once upon a time I used Debian, and of course installed their daemontools and runit packages. I'll leave space here for the troll to tell others (I don't see his messages, only his vestiges in others' quoted context) that I used the wrong package names... The Debian runit and daemontools didn't work without a lot of initial debugging. There was no Debian documentation to match the (head scratchingly arbitrary) Debian changes. I gave up in disgust and installed the real stuff, from the real tarball, issued by the real developers. That wasn't my first rodeo with Debian. I'm the originator and 2 year maintainer of the VimOutliner project. The Debian package substituted a double backslash \\ for the double comma ,, command prefix. When I said that the double comma was selected for speed purposes and that double backslash was a huge slowdown, the Debian maintainer gave me this song and dance about customary defaults in Debian and Vim. Meanwhile, as "upstreams", we were forever asked to debug Debian installations because they didn't work. After a few attempts, we just told them Uninstall your Debian VimOutliner and install from our tarball. That was the quickest fix. I have no interest in half-way fixing Debian's broken packages, but I'll be glad to help create an excellent DeVUan package that respects what the software's author was trying to accomplish (but without any new directories off the root). If the new Devuan runit package work out as well as I think it will, I can help do the same thing for s6, and I talk to s6's developer on a regular basis and I think he'd be helpful. SteveT Steve Litt October 2017 featured book: Rapid Learning for the 21st Century http://www.troubleshooters.com/rl21 ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Debian testing drop redis
On 10/28/2017 01:00 AM, Rick Moen wrote: > Quoting zap (calmst...@posteo.de): > >> I still have a mixture of irritation and amazement, that systemd is >> considered libre software yet causes so many issues and >> security/stability flaws. amazing yet terrifying... > Nothing at all prevents libre software from being absolutely wretched. > The term merely denotes right to fork code and use it for any > purpose without fee, as detailed in your choice of 'free software' > definitional document (FSF's or Debian's being the usual two, either > being for practical purposes same as OSD, modulo marketing focus). You are most definitely right, I was naive before and thought libre software couldn't be that screwed up Well... I was wrong. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Debian testing drop redis
On 10/28/2017 03:28 PM, goli...@dyne.org wrote: > On 2017-10-27 21:34, Steve Litt wrote: >> >> Everyone: John Hughes has history here. See the following: >> >> https://lists.dyne.org/lurker/mindex/d...@20151218.114434.e5e44af4.en.html >> >> >> And here's what I wrote after he posted his fiftieth prop-systemd or >> ambi-systemd on DNG in a period of a couple days. >> >> https://lists.dyne.org/lurker/message/20151219.183351.831454aa.en.html >> >> SteveT >> >> > > And remember when he reappeared here, he assumed he was banned! > > https://lists.dyne.org/lurker/message/20171020.144827.42a007e5.en.html > > So he remembered those encounters and imagined the result . . . > > golinux > He must be very paranoid. Sometimes I wonder if trolls arguing is a good reason for people to get popcorn and sit and watch... ;) > > > > ___ > Dng mailing list > Dng@lists.dyne.org > https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] Getting to know Chris Lamb
I was just looking at some old threads over at DUF and ran across this: https://www.debian.org/vote/2017/platforms/lamby Haven't given it a thorough reading but what I did see I found interesting . . . golinux ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Clear communication (Was: Debian testing drop redis)
On 10/28/2017 02:06 AM, John Hughes wrote: While keeping your eyes peeled is obviously a good thing please remember the downsides of crying wolf when the wolf isn't there. Clear communication is also a good thing. Perhaps the words "[D]rops the Debian-specific support for ... in favour of using systemd" were not the best choice of words to summarize something that was "not a sysvinit-specific change" or to communicate that "this change is completely initsystem agnostic". In fact, the change itself was actually fine, but a summary like "in favour of using systemd" will provoke a strong negative reaction on DNG because those words don't mean "initsystem agnostic" to most DNG readers. Patrick ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Debian testing drop redis
On 2017-10-27 21:34, Steve Litt wrote: Everyone: John Hughes has history here. See the following: https://lists.dyne.org/lurker/mindex/d...@20151218.114434.e5e44af4.en.html And here's what I wrote after he posted his fiftieth prop-systemd or ambi-systemd on DNG in a period of a couple days. https://lists.dyne.org/lurker/message/20151219.183351.831454aa.en.html SteveT And remember when he reappeared here, he assumed he was banned! https://lists.dyne.org/lurker/message/20171020.144827.42a007e5.en.html So he remembered those encounters and imagined the result . . . golinux ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] xrandr on Ascii for two displays
I have two displays, one built-in to a laptop and another external, and am not able to get any signal to the external one. The same arrangement works fine with other distros. I've tried with the display managers for both XFCE4 and LXQt, as well as xrandr. For example, I have tried the following, but the external screen stays black: $ xrandr --output eDP-1 --auto \ --output HDMI-1 --auto \ --same-as eDP-1 $ uname -sr; lsb_release -rd Linux 4.9.0-4-amd64 Description:Devuan GNU/Linux 2.0 (ascii) Release:2.0 What should I be looking for with xrandr to cause a usable signal? /Lars ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] ..duh, nvidia killed nouveau, was: ..console & X setup these days...
On Sat, 28 Oct 2017 15:33:49 +0200, Arnt wrote in message <20171028153349.7c49a0ff@d44>: > On Sat, 28 Oct 2017 11:05:10 +, Arnt wrote in message > <20171028110510.3e58558b@d44>: > > > Hi, > > > > ..lost power last night while asleep, laptop batteries ran down, OS > > stability proved good enough for my abuse, I only logged in once > > during the last 130 day and 20-ish hours, according to my last run > > of uptime, too much post-Groklaw litigation meant no reboot even on > > kernel updates. > > > > ..booting up this morning, starts nicely with grub @ > > 1920x1200x32b@60Hz and winds up with no X wailing "no KMS" and vga > > ttys, which sort of defeats the whole point of having > > 1920x1200x32b@60Hz capable hardware. > > > > ..no X prompted a quick update of > > linux-image-4.9.0-0.bpo.3-rt-amd64 to 4.9.30-2+deb9u5~bpo8+1 > > (2017-09-28) and still no X. > > > > ..so, what changes did I miss during my 130 day and 20-ish hour kde > > session? > > > > ..I use vdev-0.1.2, and nouveau, I'm not aware of any news on these. > Did we drop KMS? > ..I somehow managed to get nvidia kill nouveau 'n X this summer. ;oD -- ..med vennlig hilsen = with Kind Regards from Arnt Karlsen ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] ..console & X setup these days...
On Sat, 28 Oct 2017 11:05:10 +, Arnt wrote in message <20171028110510.3e58558b@d44>: > Hi, > > ..lost power last night while asleep, laptop batteries ran down, OS > stability proved good enough for my abuse, I only logged in once > during the last 130 day and 20-ish hours, according to my last run of > uptime, too much post-Groklaw litigation meant no reboot even on > kernel updates. > > ..booting up this morning, starts nicely with grub @ > 1920x1200x32b@60Hz and winds up with no X wailing "no KMS" and vga > ttys, which sort of defeats the whole point of having > 1920x1200x32b@60Hz capable hardware. > > ..no X prompted a quick update of linux-image-4.9.0-0.bpo.3-rt-amd64 > to 4.9.30-2+deb9u5~bpo8+1 (2017-09-28) and still no X. > > ..so, what changes did I miss during my 130 day and 20-ish hour kde > session? > ..I use vdev-0.1.2, and nouveau, I'm not aware of any news on these. Did we drop KMS? -- ..med vennlig hilsen = with Kind Regards from Arnt Karlsen ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] ID Quantique "Quantum" PCI-e RNG's - does anyone have more info?
taii...@gmx.com writes: I can't imagine it being equivalent to a (non-intel/amd) hardware source of entropy when it comes to quality of entropy - have there been any quality analysis performed? Yes. But it misses the point. http://www.irisa.fr/caps/projects/hipsor/publications/havege-tomacs.pdf and http://www.irisa.fr/caps/projects/hipsor/publications/havege-rr.pdf state it clearly: Havege is not a entropy gatherer like the Simtec device or the new Intel CPUs, both of which rely on quantum physics. Neither is it a PRNG. Instead it is a third class, namely an observer of unreproducible state changes. Havege says: Modern computers can be measured more accurately than they can be modelled, and the additional accuracy provides a stream of numbers that are practically like entropy. They use the word "practically". After reading the source and both papers, I accept that there is such a stream. I just don't accept that the stream is as fast-flowing as they think. Now, can you analyse the speed of the stream in a way that's really different from what's in their source code (and mentioned in their papers)? AFAICT you cannot. Their own analysis goes as far as reasonably possible. What remains is the philoshophical point: Do you classify an observer of unreproducible state as a PRNG or as like entropy? It is a shame IDQ is the only vendor with a PCI-e device, and also the only vendor it seems that offers something quality and obtainable (I can't find an open source hardware entropy device that is really for sale right away And IDQ won't take you seriously? Long ago I shared office with an extremely capable salesman for a while. He was a bit of a cynic and very, very skilful. The company's top salesman in terms of money. One of the things he told me was about prospective customers who choose not to buy the product because it doesn't X. He said: "Those never buy. If we X they just find a missing Y." The prospects worth taking seriously were (according to him) are the ones who bought from a competitor and said they'd switch if X. Their present outlay proved their sincerity in his eyes. Arnt ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] ..console & X setup these days...
Hi, ..lost power last night while asleep, laptop batteries ran down, OS stability proved good enough for my abuse, I only logged in once during the last 130 day and 20-ish hours, according to my last run of uptime, too much post-Groklaw litigation meant no reboot even on kernel updates. ..booting up this morning, starts nicely with grub @ 1920x1200x32b@60Hz and winds up with no X wailing "no KMS" and vga ttys, which sort of defeats the whole point of having 1920x1200x32b@60Hz capable hardware. ..no X prompted a quick update of linux-image-4.9.0-0.bpo.3-rt-amd64 to 4.9.30-2+deb9u5~bpo8+1 (2017-09-28) and still no X. ..so, what changes did I miss during my 130 day and 20-ish hour kde session? -- ..med vennlig hilsen = with Kind Regards from Arnt Karlsen ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Chris Lamb = Good Maintainer (Was: Debian testing drop redis)
On Sat, 28 Oct 2017, John Hughes wrote: > While keeping your eyes peeled is obviously a good thing please > remember the downsides of crying wolf when the wolf isn't there. crying wolf?! we are the wolves!!! while since the move to systemd Debian is just an elephant walking on stills. ciao ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Runit for Devuan: was Debian testing drop redis
Quoting John Hughes (j...@atlantech.com): Just a note to say you strike me as saying useful and constructive things, for which my thanks. I wish some other good and valued folks, bless them one and all, were a bit less quick to pick a fight. There are much more interesting things to discuss. [snippity snip of your useful s6 comments] > You don't have to make a runit package for Debian/Devuan -- it > exists already. > > What is missing is the runit-init package, i.e. runit as pid 1, > which was removed as nobody had the time to fix bug 861536. This is something that's had previous discussion here, FWIW. > I guess Devuan could just live with the bug, or declare it fixed by > documenting it. > > Or someone who was interested could fix it and get runit-init back > into Debian (and, therefore) Devuan. A third (additional) alternative has been mentioned in prior discussions: A variety of other inits can, instead, fill the PID1 role. Even SysVInit's PID1 process doesn't have (IMO) a lot wrong with it. Its deficiencies primarily lie in the init system. I'll also note in passing that Debian bug 861536 in no way prevents use of runit's codebase as PID 1. (See comment #89 on that bug.) It merely prevents doing so solely using package operations using the runit-init _package_. One might say: That is more than nothing, but far less than everything. Which is to say: For people of even modest sysadmin abilities, the problem is small beer indeed. (But sure, fixing the runit-init package would be a nice-to-have.) (I do not speak for Devuan, Debian, or any other distribution in the foregoing. On a good day, I speak for my grumpy orange cat, but not when she's in a particular mood.) ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Chris Lamb = Good Maintainer (Was: Debian testing drop redis)
On 27/10/17 21:56, Jaromil wrote: my training in hermeneutics and epistemology rings a bell about this being the wrong general attitude about changes and regressions, but for now I just rest on the fact redis "just works" in Devuan ASCII (using sysvinit) and that, as I stated at the beginning, this should be an issue addressed by upstream rather than us. Upstream has been warned. Lets all keep our eyes peeled for this and other packages. But what is the issue to be addressed by upstream? What do they have to be warned about? Chris Lamb removed a misfeature that he put in himself. That, undocumented, misfeature never existed in the upstream code. While keeping your eyes peeled is obviously a good thing please remember the downsides of crying wolf when the wolf isn't there. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Runit for Devuan: was Debian testing drop redis
On 28/10/17 03:45, Steve Litt wrote: I'm the original poster of the thread renamed "Runit for Devuan", and I don't understand this email at all. What does a debian bug about s6 have to do with my offer to be one of a two person team to bring a runit package to Devuan? You said > s6's maintainer is on the job every day improving s6, and once again, > someone who knows how to build Devuan package plus me plus a little bit > of guidance from s6' upstream would produce a Devuan package. The Debian bug I quoted was a RFP (request for pakage) for s6, it was the closest I could find to a s6 package. Since then I've found: https://github.com/lwf/s6-packaging Which is the stuff necessary for a s6 .deb Note that, as far as I know, this doesn't provide enough stuff for using s6 as pid 1. By the way, anyone here good at Devuan packaging and also would like a runit package? I could probably put together a Devuan/runit Vagrant file, from which we could make a package. You don't have to make a runit package for Debian/Devuan -- it exists already. What is missing is the runit-init package, i.e. runit as pid 1, which was removed as nobody had the time to fix bug 861536. I guess Devuan could just live with the bug, or declare it fixed by documenting it. Or someone who was interested could fix it and get runit-init back into Debian (and, therefore) Devuan. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Debian testing drop redis
On 28/10/17 01:39, zap wrote: On 10/27/2017 04:51 AM, John Hughes wrote: On 27/10/17 07:21, Steve Litt wrote: No Debian fan, Chris Lamb or otherwise, Debian fan? He's one of the people who build the distribution Devuan is based on. He's not a "fan". That is very impressive indeed. I would have never guessed that he was that skilled at development. I wonder if he uses sid... or ceres probably... I assume he uses one of those. :) Given that Chris Lamb is currently Debian Project Leader and a Debian developer since 2008 I suspect he uses sid. https://chris-lamb.co.uk/about ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng