Re[2]: stop that shit
ÐÑÑниÑа, 17 ÑнваÑÑ 2014, 17:58 +05:30 Ð¾Ñ Jay Patel : >Well one thing what foundation can do for future is start selling hardware >with software like routers and all directly from their website. > >don't know how much this is feasible.or something like system76.or do some >consultancy. > >Regards, > >Jay. who needs comedys when one can simply subscribe to misc@openbsd.org. Jay bro, that was awesome! keep trollin' -- Mo Libden
Re[2]: Unbound in base
14 QP5P2QP0P;Q 2012, 12:59 P>Q Gregory Edigarov : > On Tue, 14 Feb 2012 08:09:16 + > Peter van Oord van der Vlies wrote: > > > Hello, > > > > Why replacing bind ? > > Because bind is full of security related bugs and a bloatware. Oh come on! They say about the same thing about sendmail for years (decades already?). Still it is in the base. Are you spreading FUD here, or there are real cases with the version of BIND that is in the base? > Yours C. O.
Re[4]: Unbound in base
14 QP5P2QP0P;Q 2012, 14:31 P>Q Oliver Peter : > On Tue, Feb 14, 2012 at 01:23:01PM +0400, Mo Libden wrote: > > 14 QP5P2QP0P;Q 2012, 12:59 P>Q Gregory Edigarov : > > > On Tue, 14 Feb 2012 08:09:16 + > > > Peter van Oord van der Vlies wrote: > > > > > > > Hello, > > > > > > > > Why replacing bind ? > > > > > > Because bind is full of security related bugs and a bloatware. > > > > Oh come on! > > They say about the same thing about sendmail for years (decades already?). > > Still it is in the base. > > smtpd(8) is underway. Also there is no proper MTA implementation out > there served under the BSD license (i.e. Postfix has IBM license). good thing it is. i see lotsa commits, but after years it still seems to be not ready to replace the sendmail, so it's good to hear it's still gonna be the replacement. like with opencvs and pcc. i see things slow down sometimes. > Unbound (and also nsd) is a good and lightweight alternative to > sendmail using the BSD license. License stuff is more important than > it sounds. "to sendmail"? i guess you meant "BIND" here. and i know the concerns with the licenses, which is what i love in OpenBSD. > IMO the separate development of a resolver (unbound) and an authoritive > nameserver (nsd) is better than having all functionality within one > server (named). this sounds very reasonable. i don't need auth NS myself in most of my installations. add here bits answered earlier (10 goes for python etc.) and now it's clear. it's just that the captain's answer wasn't that obvious. sounded like classical FUD :-)
Re: My OpenBSD 5.0 installation experience (long rant) (even longer)
07 PQ Leonardo Sabino dos Santos : > I start answering the installer's questions. Keyboard layout. Root > password. Configuration of network interfaces. I'm not actually paying > a whole lot of attention to the questions as this is just a test > installation and I figure I can always explore and configure the > system later. you can go paying a little attention when you do it for 101st time. when you're doing installation of a new OS for the first time, having you're sane, you're expected to pay attention. > Next, the disk stuff comes up. A lot of partition information appears > on the screen, followed by the question: > > Use (W)hole disk or (E)dit the MBR? [whole] > > At this point I'm actually trying to remember if there's a way to > scroll back the console, because some information has scrolled of the > screen. I try PageUp, PageDown, Ctrl-UpArrow, Ctrl-DownArrow, but > nothing works, so I press Enter. well, you know what? it is exactly the same key combination as it is in linux: Shift-PgUp / Shift-PgDn. even if it wasn't, you can interrupt the installation any time by pressing obvious ^C and start again, paying attention at the parts you missed earlier. > And my partition table is gone. Poof! Instantly, with no confirmation. > I immediately realized what had happened and rebooted. Too late. I got > a "No OS" message. It seems that the OpenBSD installer actually > overwrites the partition table the instant you press Enter. man, you're miserable! I remember when I installed OpenBSD 2.1 (if memory serves), in 1996 off twenty-something floppy disks. at that time README used to suggest having paper/pencil/calculator handy. at that time you needed to do calculations and know exactly what are those sectors, cylinders, heads. it was even better if you had knowledge about hard disk partitioning and all that stuff. i knew that stuff, i printed and carefully read README.i386 few times, i had all those things handy and yet on my first install I f*cked up my disk having my Windows for workgroups 3.11 disk suffer lost directories and files, along with broken file content. i never complained about it, i did my homework and did proper installation. these days you're relieved of these things. installation procedure is condensed, fdisk is much better, disklabel editing goes without vi, you're asked just few question, but the main thing about OpenBSD remains unchanged: YOU NEED TO KNOW WHAT YOU'RE DOING. if you don't, you're probably best off without going there. But being asked if you want to use the whole disk for OpenBSD, confirming that and expecting it not to overwrite the partition table? what kind of logic is that?
Re[2]: My OpenBSD 5.0 installation experience (long rant)
08 PQ "Dmitrij D. Czarkoff" : > On Wed, 2012-03-07 at 16:23 +, Dennis den Brok wrote: > > > > "Use (W)hole disk (writes to disk immediately) or (E)dit the MBR? [whole]" > > > > While the FAQ is indeed clear, the installer's simplicity appears > > at that point a little deceptive, in that one (I know I was) is > > tempted to think that such a user-friendly installer would not harm > > one so easily... > > Don't you think it all gets too far? One should generally expect that > choosing "use the whole disk" means that all the data on disk will be > lost. If the user doesn't pay attention to installer, this wording won't > help. Furthermore, the more chatty installer is, the less amount of > newcomers would be reading the messages. my sentiments exactly. if they don't think about what's written, will it make it better to write some more? besides, what does exactly "writes to disk immediately" mean? ok, it writes, so what? will it change MBR? will it change something else? or will it just read sector and write it back (i.e. no actual change)? may be the install script may be changed so that it does dd if=/dev/sdXc of=/tmp/sdXc.mbr count=1 so that after chosing the "whole disk" option and breaking the install script, you still have an option to get your partition table back. it needs to be documented of course...
Re[2]: My OpenBSD 5.0 installation experience (long rant)
12 PQ Fredrik Staxeng : > Once upon a time, mkfs used to make a 10-second pause before starting. > That's the way you do it. oh, this is B.C.! then some dork will appear on the list saying the 10 sec was not long enough for him. when you install an OS (in all other cases too by the way, but let's stay with the point) you need to think. that head of yours are not just there to put hat on it. program: i am going to use the whole disk there user: ok program: ...using... user: WTF?! you're writing to the disk now?!
Re[2]: posix_spawn(3)
10 PQ Ted Unangst : > On Fri, Mar 09, 2012, Frank Denis wrote: > > Hi Matthew, > > > > Good catch. I'm going to fix that. > > > > On Mar 9, 2012, at 10:54 AM, Matthew Dempsky wrote: > >> I briefly looked over this code, and the diff looks good to me except > >> for one subtle FreeBSDism: memory writes done in a vfork(2)'d child > >> process will not affect the parent process. This is critical for how > >> FreeBSD's posix_spawn() propagates errors from the spawned child > >> process. > > I believe you may be able to just use rfork. If the parent waiting > behavior of vfork is abslutely ocritical, we can add it to rfork very > easily. I think that makes more sense than making userland jump > through hoops. now this is intriguing. AFAIK, classical vfork was invented in earlier BSD to avoid expensive duplication of a parent process in case all the child does is launch of other executable. SysV solved it with CoW, BSD came up with vfork. Now, how come OpenBSD has vfork which does not work as the classic vfork, instead we can use rfork, which does what vfork is supposed to do?
how to get userland pthreads working?
Hi! I wrote a two string program to check rthreads, it looks really interesting! I am interested how do I get pure userland library back in case I need? Just setting kern.rthreads=0 seems to be not enough: pthread_create() fails this case as not supported. Do I need to link the program against another library? Anything else? Thank you in advance! m/ P.S. sorry if I am asking something obvious ;-)
Re[2]: how to get userland pthreads working?
Tue, 10 Apr 2012 11:10:09 -0700 P>Q Philip Guenther : > On Tue, Apr 10, 2012 at 10:41 AM, Mo Libden wrote: > > I wrote a two string program to check rthreads, it looks really interesting! > > > > I am interested how do I get pure userland library back in case I need? > > Just setting kern.rthreads=0 seems to be not enough: pthread_create() > > fails this case as not supported. > > > > Do I need to link the program against another library? Anything else? > > Running -current with userland threads is not supported. If you need > userland threads, you need to run 5.1 or earlier. That was what I really needed. Thank you! It is not stated clearly in the pthreads man page. The man page still says it is a userland library implementation, which obviously is not true anymore. That's why I asked. Now, for the following, what's up with the immediate disclaimers? Only asked a question, sheesh. Nobody asked for support or anything similar. Being with OpenBSD since 1996, I kinda know the versioning and the terms for the support. > As for how to do > that, well, to quote section 5.1 of the FAQ: > -- > One should also understand that the upgrade process is supported in > only one direction: from older to newer, and from -stable to -current. > You can not run 5.0-current (or a snapshot), then decide you are > living too dangerously, and step back to 5.0-stable. You are on your > own if you choose any path other than the supported option of > reloading your system from scratch, do not expect assistance from the > OpenBSD development team. > -- > > > Philip Guenther
Re: how to get userland pthreads working?
Wed, 11 Apr 2012 02:00:28 -0700 P>Q Philip Guenther : > On Wed, Apr 11, 2012 at 12:28 AM, Mo Libden wrote: > > Tue, 10 Apr 2012 11:10:09 -0700 P>Q Philip Guenther : > ... > >> Running -current with userland threads is not supported. If you need > >> userland threads, you need to run 5.1 or earlier. > > > > That was what I really needed. Thank you! > > > > It is not stated clearly in the pthreads man page. > > The man page still says it is a userland library implementation, > > which obviously is not true anymore. That's why I asked. > > Perhaps if we hit a blocker that would keep rthreads from being ready > for the 5.2 release we'll be reverting the switch. Some parts of the > docs might not be updated until we are confident that won't occur. Thank you, that's clear. I hope rthreads will make it, they're hot :-) > BTW, any particular reason for wanting to switch back? Something not > working right or a regression in performance or behavior? No, nothing in particular. I just used to think ahead of other possible scenarios like "what if this something goes wrong, how do i..." Since rthreads seem to be controlled with sysctl, I had impression that setting kern.rthreads=0 would turn old libpthread back. Like libpthread has some kind of check for kern.rthreads and chooses what functions to use according to the value. Now, thanks to you I know how the things currently are. > > Now, for the following, what's up with the immediate disclaimers? > > Only asked a question, sheesh. Nobody asked for support or > > anything similar. Being with OpenBSD since 1996, I kinda know the > > versioning and the terms for the support. > > Ah misc@, where half the people need the FAQ read to them and the > other half are insulted when someone tries to answer the obvious next > question. No, in my head it's not "obvious next question". If I had come with some problem like "i compile package X and it fails with rthreads", that might be the case. However, my e-mail started with "I wrote a two string program to check rthreads, it looks really interesting!" So it is obvious I was just playing around with this, looking how the rthreads are processed by the OS (checked OS tools like ps, top etc.). Anyway, we can drop this part :-) > Philip Guenther Thanks!
rfork(2) considered harmful? [Re: CVS: cvs.openbsd.org: src]
Thu, 12 Apr 2012 06:33:04 -0600 (MDT) P>Q Theo de Raadt : > CVSROOT: /cvs > Module name: src > Changes by: dera...@cvs.openbsd.org 2012/04/12 06:33:04 > > Modified files: > sys/kern : kern_fork.c syscalls.master > sys/sys: param.h proc.h vmmeter.h > include: unistd.h > lib/libc/sys : Makefile.inc > lib/librthread : __tfork_thread.3 > sbin/sysctl: sysctl.c > usr.bin/kdump : kdump.c kdump_subr.h mksubr > usr.bin/vmstat : vmstat.8 vmstat.c > Removed files: > lib/libc/arch/alpha/sys: rfork.S > lib/libc/arch/amd64/sys: rfork.S > lib/libc/arch/arm/sys: rfork.S > lib/libc/arch/hppa/sys: rfork.S > lib/libc/arch/hppa64/sys: rfork.S > lib/libc/arch/i386/sys: rfork.S > lib/libc/arch/ia64/sys: rfork.S > lib/libc/arch/m68k/sys: rfork.S > lib/libc/arch/m88k/sys: rfork.S > lib/libc/arch/mips64/sys: rfork.S > lib/libc/arch/powerpc/sys: rfork.S > lib/libc/arch/sh/sys: rfork.S > lib/libc/arch/sparc/sys: rfork.S > lib/libc/arch/sparc64/sys: rfork.S > lib/libc/arch/vax/sys: rfork.S > lib/libc/sys : rfork.2 > > Log message: > remove rfork(); ok guenther miod Wow. If memory serves, rfork() availability was a feature. Now it is gone... Any reasons to share please? It allowed creation of interesting types of processes, awesome flexibility regarding share of memory space and/or file handle tables.
Re: rfork(2) considered harmful? [Re: CVS: cvs.openbsd.org: src]
Thu, 12 Apr 2012 11:32:54 -0400 P>Q Ted Unangst : > On Thu, Apr 12, 2012, Mo Libden wrote: > > > Wow. If memory serves, rfork() availability was a feature. > > Now it is gone... Any reasons to share please? > > > > It allowed creation of interesting types of processes, > > awesome flexibility regarding share of memory space > > and/or file handle tables. > > rthreads should be awesome enough for anybody. :) "640K ought to be enough for anybody" eh?
Re: OT: SSH not secure?
Wed, 9 May 2012 09:20:44 -0600 P>Q Alvaro Mantilla Gimenez : > According these guys connect trough SSH to a remote server is not secure... > > http://www.wziss.com/ > > Look in "Case Studies" > What a disgusting way of promoting one's product! Content of "Case Studies" is just ridiculous. "If somebody has keys from your apartment, they can enter it! Locks are not secure!" You can make it as secure as you want, then there is also the wrench solution: http://xkcd.com/538/ :-)
Re[2]: OpenBSD forked
Sat, 16 Jun 2012 15:15:05 -0600 (MDT) от Theo de Raadt : > They started the fork because they got kicked out because one > developer (Marco) hired 5 other developers for his startup company, > and attempted to hire around 10 other developers in a sneaky and > underhanded way. They were told, oh i forget they were "asked", to > not tell anyone else in OpenBSD that this was happening, probably > because people "including Theo" would be upset. hah, it's like little kiddies: "don't tell mom we've eaten all the candies" WTF people, couldn't it be handled without playing these games? it's sad of course, dividing forces isn't nice, but then again what the hell, it's life. bright thing is - Theo keeps the development of the OS on the course of evolutionary changes, so all is well in the world and god is in his heaven
Re[2]: Free domains
Sat, 30 Jun 2012 12:44:20 +0200 от Henning Brauer : > * Tomasz Marszal [2012-06-20 09:46]: > > Could somebody give me an advice where i can register free domain for non > > commercial use ( cv script howtos etc ) > > can anyone give me advice where I can get free bananas? LOL, i already started to forget that i am one of the bunch of masturbating monkeys. thanks Henning for the reminder :-)
Re[2]: problem in fstab
Tue, 24 Jul 2012 18:46:55 + от Alexander Polakov : > * HvN [120724 17:17]: > > I booted into single user mode, mounted / and /usr according to FAQ 8. > > However, when I try to use vi to change fstab, it says "unknown terminal > > type". Any suggestions ? > > export TERM=vt220 may be vi can be modified to ask for terminal type if it is not defined, something like the following: vi: unknown terminal type, please enter the type of the terminal [vt220]: _
Re[4]: problem in fstab
Thu, 26 Jul 2012 07:06:02 -0430 от Andres Perera : > the problem with this logic is that there are numerous curses > programs: less, top, systat, vi; just to name the ones i recall from > base. since you top-posted, I will too. what's the problem with top? will it bomb out if there is no TERM? i think rather than hurrying to answer, you'd better do you homework. yes, you lose the interactiveness, but all your mentioned programs do work with dumb terminal no problem! $ echo $TERM xterm $ unset TERM $ top load averages: 0.44, 0.12, 0.04v1.my.domain 05:46:41 18 processes: 17 idle, 1 on processor CPU states: 0.5% user, 0.0% nice, 7.4% system, 0.4% interrupt, 91.7% idle Memory: Real: 7944K/35M act/tot Free: 206M Cache: 16M Swap: 0K/280M PID USERNAME PRI NICE SIZE RES STATE WAIT TIMECPU COMMAND 1 root 100 452K 352K idle wait 0:01 0.00% init 17516 root 20 3344K 2580K sleep poll 0:00 0.00% sshd 31665 _syslogd 20 412K 712K sleep poll 0:00 0.00% syslogd 15719 user 20 3312K 2112K sleep select0:00 0.00% sshd 665 root 20 640K 476K idle netio 0:00 0.00% pflogd 597 root 20 1608K 2052K idle netcon2 0:00 0.00% sendmail 15280 user 180 740K 520K sleep pause 0:00 0.00% ksh 28645 root 20 396K 752K idle netio 0:00 0.00% syslogd 4803 root 20 1452K 1632K sleep select0:00 0.00% sendmail 1977 root 30 456K 792K idle ttyin 0:00 0.00% getty 10310 root 30 492K 800K idle ttyin 0:00 0.00% getty 44 root 20 304K 796K idle select0:00 0.00% inetd 13251 root 20 868K 1216K sleep select0:00 0.00% sshd 18652 root 20 308K 656K sleep kqread0:00 0.00% apmd 9421 root 20 692K 880K idle select0:00 0.00% cron 3016 _pflogd40 704K 312K sleep bpf 0:00 0.00% pflogd 9936 user 280 348K 768K onproc- 0:00 0.00% top 24414 _sndio 2 -20 488K 428K idle poll 0:00 0.00% sndiod $ less /etc/motd WARNING: terminal is not fully functional /etc/motd (press RETURN) ^OpenBSD 5.2 (GENERIC) #0: Sat Jul 14 11:58:37 EEST 2012 Welcome to OpenBSD: The proactively secure Unix-like operating system. Please use the sendbug(1) utility to report bugs in the system. Before reporting a bug, please try to reproduce it with the latest version of the code. With bug reports, please try to ensure that enough information to reproduce the problem is enclosed, and if a known fix for it exists, include that as well. ~ ~ $ $ systat 1 usersLoad 0.38 0.13 0.05 Fri Jul 20 05:47:04 2012 $ > surely retrofitting them with prompts isn't an option, specially when > having TERM unset isn't the norm > > On Thu, Jul 26, 2012 at 4:33 AM, Mo Libden wrote: > > Tue, 24 Jul 2012 18:46:55 + от Alexander Polakov : > >> * HvN [120724 17:17]: > >> > I booted into single user mode, mounted / and /usr according to FAQ 8. > >> > However, when I try to use vi to change fstab, it says "unknown terminal > >> > type". Any suggestions ? > >> > >> export TERM=vt220 > > > > may be vi can be modified to ask for terminal type if it is not defined, > > something like the following: > > > > vi: unknown terminal type, please enter the type of the terminal [vt220]: _
Re[2]: Suspend stuff on TOSHIBA laptop.
Tue, 04 Sep 2012 03:36:50 -0400 от Ted Unangst : > On Tue, Sep 04, 2012 at 16:23, David Walker wrote: > > I've set the options in wsconsctl.conf to blank the screen which also > > works but this thing has I think what's called "a backlight" which > > means the screen constantly glows. I'm planning to go set this thing > > up, let the screen blank and close the lid. > > I'd like to remove the backlight and the eerie glow. > > I'm unfamiliar with laptops but I've tried zzz and apm -S both kill > > the backlight which is great but network functions cease, yes I did > > not know that. > > I also can't seem to bring it back up form either state short of a > > power cycle but that's moot. > > > > Is there a way to turn off the backlight? > > Is there anything else I can do to sedate this machine? > > I've never seen a laptop that kept the light on when the lid was > closed. Is it really still on? Well I had this for YEARS on my Compaq NC6000. The only cure I found is to disable acpivideo. If acpivideo was loaded during boot time, the screen is lit even when the lid is closed. I can even press the switch (it is mechanical), and the screen backlight is not turned off if acpivideo was enabled.