Re: Web + email hosting recommendations
On Wed, May 08, 2019 at 08:04:05AM +, Christopher Pinon wrote: > When I ran NetBSD on Vultr over a year ago, the $2.5/m plan included a > dedicated IPv4 address, but alas this is no longer the case. (That plan > was also difficult to get, because they were nearly always "sold out".) I still own one from that lot! Yes those were difficult to get and now no more offered. Besides IPV4 they had 20GB HDD and 1TB BW. Now $2.5 is available with 10GB HDD, 500GB BW and IPV6 only. You get IPV4 for $3.5. Both these are regularly available (unlike older 2.5 plan that was not easy to get.) BTW what is the downside of having IPV6 only? Isn't it the future anyway? Mayuresh
Re: Web + email hosting recommendations
On Wed, May 08, 2019 at 07:23:29AM +, Christopher Pinon wrote: > Nevertheless, depending on the provider and the question, sometimes > they're willing to help with certain questions regarding well-known > Linux distributions. For example, if your question is "How do I set up > IPv6 on Debian?", a lower-cost provider may be willing to help (despite > the fact that the VPS is unmanaged), whereas if your question is "How do > I set up IPv6 on NetBSD?", you're really on your own simply because most > lower-cost VPS providers don't have (much) experience with *BSD. I see your point. But, unless the question is very specific to cloud provider's setup I think most people would ask in forum such as this. But basically, I guess, those who opt for BSD consciously are likely to already have used BSD before for a while and may not face too many questions. [Contrary to newbies opting for Linux because it's more widely used.] Mayuresh
Re: Web + email hosting recommendations
On Wed, May 08, 2019 at 07:19:05AM +0100, Sad Clouds wrote: > UK-SSD-KVM-1024 > 2 CPU Core (Equal Share) > 1024 MB Ram > 10 GB Pure NVMe SSD Disk space > 1000 GB Bandwidth @ 1 gbit (shared) > 1 x IPv4 address > 1 x /64 IPv6 > Full daily backup > > Price is 2.50EUR per month and no setup fees, which sounds rather cheap. I was thinking Vultr to be the cheapest in the low end space. Vultr low end plan is 1core/512MB/10GB/500GB/no IPV4/1xIPV6/no backup for $2.5. So above is lot better. But on the link given I didn't notice custom ISO. Do they support? Mayuresh
Re: Web + email hosting recommendations
On Tue, May 07, 2019 at 09:20:57PM +, Christopher Pinon wrote: > but naturally the downside is that the provider won't be able to help > with NetBSD-specific issues. What could be NetBSD-specific issues that would require provider's help? I have faced none till now. Just curious. Mayuresh
Re: Web + email hosting recommendations
On Tue, May 07, 2019 at 01:47:10PM +0100, Andrew Luke Nesbit wrote: > Also, don't forget SDF. They offer NetBSD VPS's. Slightly OT: There is a poll going on, on sdf to choose between Debian and NetBSD for MetaArray IV. sdfers on the list may please vote with their choice! Mayuresh
Re: Web + email hosting recommendations
On Tue, May 07, 2019 at 01:37:51PM +0100, Sad Clouds wrote: > Hello, do you know what hypervisor they use for your NetBSD VPS? Sorry, no idea. Mayuresh
Re: Web + email hosting recommendations
On Tue, May 07, 2019 at 12:46:54PM +0100, Sad Clouds wrote: > Hello, could anyone recommend web hosting providers for the following cases: I do not know whether it suits all your criteria, but a big advantage of vultr VPS is they allow custom iso based installation. I have a vultr VPS running NetBSD with a (small) website and email for a few users. Mayuresh
code above 'hal' : knowledge of assembler neccessary?
if i wish to write code at the kernel level, but above the 'hal' would it be necessary for me to have a good command over low level assembler? or just a superficial introduction to assembler be good enough? thanks.
Re: A lightweight authentication scheme for website
On Mon, May 06, 2019 at 10:32:42AM +0200, ignat...@cs.uni-bonn.de wrote: > Hello, > > On Sat, May 04, 2019 at 08:37:28PM +0530, Mayuresh wrote: > > > I like the simpicity of htpasswd method. But not being able to logout is > > a little undesirable. > > This is not true. E.g. in Firefox 60esr: To clarify, I meant, server can't log you out. Mayuresh
A lightweight authentication scheme for website
This is probably an OT. Apologies for the same. I am looking for a lightweight authentication scheme for a community website where: - the user count would never grow beyond 300 or so - all user's email ids would be pre-regsitered, no "register me" link needed - `change my password' is nice to have but not strictly necessary (can manage with system generated passwords sent to pre-registered users) - the content on the website is mostly static html, documents etc. no interactive forms etc. - would be preferable if the method works with a lightweight httpd like bozotic, but that's not strictly necessary. I like the simpicity of htpasswd method. But not being able to logout is a little undesirable. If my understanding is right, the web server supported (`standard'?) methods are limited to just htpasswd. Looked for lightweight frameworks where I may get this functionality easily and zeroed on python-flask. Yet to explore in detail, but that's not still looking as light as I'd have imagined. Would be interesting to know simple authentication solutions used in NetBSD community for above scenario. Mayuresh
Re: amd64 SBCs on which NetBSD would run ?
On Sat, May 04, 2019 at 05:49:58PM +0800, Travis Paul wrote: > You mentioned that you were looking for an amd64 board. Have you looked > at the PCEngines APU2 boards[1]? I have not personally tried them but > perhaps they fit your needs. Thanks. Looks interesting though I could not find A. international availability / shipping B. Whether NetBSD would run on it. Surprisingly amd64 SBCs are so rare to find, and further one has to find one with NetBSD compatibility. Mayuresh
Re: amd64 SBCs on which NetBSD would run ?
On Sat, May 04, 2019 at 09:53:32AM +0100, Chavdar Ivanov wrote: > I was thinking of buying https://dlidirect.com/products/atomic-pi , > which should work, but is apparently out of stock. Looks rather > appealing, though. Yes, looks good, though unavailable. But any references of it having worked with NetBSD? Mayuresh
amd64 SBCs on which NetBSD would run ?
I am using RPI2 with NetBSD for a certain requirement. There are some rough edges (wifi support, usb hub not working, media player not working etc.) Besides it's too slow to do any builds of pkgsrc. Was wondering whether there are SBC boards where I can use my amd64 packages compiled on other devices. The board itself need not have a high end configuration (RPI like configuration is good enough) - just that NetBSD should work on it and it should have amd64 arch. Tried searching, but most SBCs seem arm based. Among those that are amd64 based it's hard to figure out whether NetBSD would support it. Please do share recommendations / experiences. Mayuresh
BFD .. invalid string offset .. for section `.strtab'
# uname -a NetBSD pi 8.99.37 NetBSD 8.99.37 (RPI2) #1: Thu Apr 25 16:01:51 UTC 2019 root@pi:/usr/src/sys/arch/evbarm/compile/RPI2 evbarm BFD: /usr/pkg/lib/libpango-1.0.so.0: invalid string offset 12338 >= 11106 for section `.strtab' BFD: /usr/pkg/lib/libfribidi.so.0: invalid string offset 1447 >= 1158 for section `.strtab' BFD: /usr/pkg/lib/libfribidi.so.0: invalid string offset 1652 >= 1158 for section `.strtab' Core was generated by `netsurf-gtk'. Program terminated with signal SIGSEGV, Segmentation fault. #0 0x698315f0 in memcpy () from /usr/lib/libc.so.12 [Current thread is 1 (process 1)] (gdb) where #0 0x698315f0 in memcpy () from /usr/lib/libc.so.12 #1 0x000d9888 in __memcpy_ichk (len=, src=, dst=0x7fe8ef34) at /usr/include/ssp/string.h:82 #2 curl_start_cert_validate (certs=certs@entry=0x7fe906e0, f=, f=) at content/fetchers/curl.c:967 #3 0x000d9f20 in fetch_curl_done (result=CURLE_PEER_FAILED_VERIFICATION, curl_handle=) at content/fetchers/curl.c:1132 #4 fetch_curl_poll (scheme_ignored=) at content/fetchers/curl.c:1223 #5 0x000d6474 in fetch_fdset (read_fd_set=read_fd_set@entry=0x7fe907a8, write_fd_set=write_fd_set@entry=0x7fe907c8, except_fd_set=except_fd_set@entry=0x7fe907e8, maxfd_out=maxfd_out@entry=0x7fe907a4) at content/fetch.c:404 #6 0x00183bc0 in nsgtk_main () at frontends/gtk/gui.c:404 #7 0x002325e0 in main (argc=, argv=) at frontends/gtk/gui.c:1206 Above trace occurred on netsurf core dump. There is a long chain of the "BFD:" errors on various libraries, only 2-3 samples of that are attached above. I think during build of the packages I had seen those errors as well, but not sure. Please advise. Mayuresh
urndis0 works on amd64-8.0 but not on evbarm-current
# uname -a NetBSD pi 8.99.37 NetBSD 8.99.37 (RPI2) #1: Thu Apr 25 16:01:51 UTC 2019 root@pi:/usr/src/sys/arch/evbarm/compile/RPI2 evbarm # dmesg | grep urndis [47.388677] urndis0 at uhub2 port 3 configuration 1 interface 0 [47.388677] urndis0: SAMSUNG (0x4e8) SAMSUNG_Android (0x6863), rev 2.00/4.00, addr 8 [47.388677] urndis0: IOERROR [47.388677] urndis0: unable to get init response [47.388677] urndis0: IOERROR [47.388677] urndis0: unable to get query response [47.388677] autoconfiguration error: urndis0: unable to get hardware address The same mobile phone works fine as urndis0 on NetBSD 8.0 amd64. I do not know whether this is a problem with -current or with evbarm. (I do not have amd64 -current or evbarm 8.0, unfortunately.) What more can I try out? Mayuresh
bozohttpd, ~user + cgi
On NetBSD 8.0_RC1 the man page of httpd says: -c cgibin Enables the CGI/1.1 interface. The cgibin directory is expected to contain the CGI programs to be used. bozohttpd looks for URL's in the form of /cgi-bin/ where is a valid CGI program in the cgibin directory. In other words, all CGI URL's must begin with /cgi-bin/. Note that the CGI/1.1 interface is available with ~user translation using -E switch. I am able to get cgi work without ~user, e.g. with -c /var/cgi But when I want to use ~user/cgi-bin URL, what should be the argument to "-c" above? (It is clear that -u and -E should be specified, I have those in my httpd_flags.) Can someone share example of command line options to enable ~user/cgi-bin? Mayuresh
Questions about cgd
Curious about usability of cgd over network for encrypted cloud storage. The use case is mentioned in the NetBSD user guide here [1]. - Is such use case feasible with sshfs? - Will it be conservative on bandwidth, such as minimizing the blocks exchanged over network for read/write (particularly with sshfs)? - Is the filesystem size growable safely as the space requirement goes up? - One of the drawbacks is, cgd is NetBSD specific, so can't mount on Linux. Can I possibly export it (plain text) as NFS from a NetBSD system and let Linux mount it? - Are there alternatives (userspace acceptable). Particularly encfs comes pretty close to meeting the requirement. But a bit concerned about reported security issues with it[2], supposed to be fixed in 2.0. (Current version is 1.9.5.) - Also read about cryfs[3]. There is a bit old wip/cryfs package. Haven't tried. But cryfs seems in its very early stages still 0.10.1 though it appears to be an active project. [1] https://www.netbsd.org/docs/guide/en/chap-cgd.html#chap-cgd-suggestions-warnings [2] https://en.wikipedia.org/wiki/EncFS#EncFS_1.8_security_concerns [3] https://www.cryfs.org/ Mayuresh
Re: NetBSD current on RPI2 (vs 8.0_RC1)
On Tue, Apr 23, 2019 at 10:00:16PM +0200, Michael van Elst wrote: > You can find most details about config.txt documented here: > > https://www.raspberrypi.org/documentation/configuration/config-txt/ Thanks. Disabling overscan solved it. Mayuresh
Is anyone using NetBSD on N900?
While tinkering with my RPI2 kernel I came across a conf file for N900. That's how I realized it might be being supported. I recently did something wrong on my N900 Maemo installation and it has got bricked. Will be quite happy to try out NetBSD on it. Would be great if someone could confirm whether / how well it works and any specific version etc. Mayuresh
Re: SSL_library_init not found in libssl in current
On Fri, Apr 26, 2019 at 10:54:00AM +0530, Mayuresh wrote: > Basically the thread has moved on to recognize 1. ssl version upgrade in > current 2. squid3 being "old" and "not recommended" by upstream and 3. > wip/squid4 being created and compiling fine with either version of ssl. Withdraw the last point about compiling with either version of ssl. May be ssl is not enabled by default. Will start a separate thread for wip/squid4.
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 10:05:12PM -0700, Greg A. Woods wrote: > Not a bad idea to post the actual error messages, and maybe the command > that failed too, and maybe even the code of the test file that failed! :-) Basically the thread has moved on to recognize 1. ssl version upgrade in current 2. squid3 being "old" and "not recommended" by upstream and 3. wip/squid4 being created and compiling fine with either version of ssl. Mayuresh
Re: SSL_library_init not found in libssl in current
On Fri, Apr 26, 2019 at 07:53:56AM +0530, Mayuresh wrote: > [Will have to sort out why my ssh keys aren't working. Wonder whether > those would be deleted if not used for long!] Everything ok. My keys of 2015 still work! I have just initialized wip/squid4 and checked compilability on amd64. Pushed it as initial commit. Will work further on it. If anyone wants to start developing it please go ahead. Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 10:47:15AM -0700, Greg A. Woods wrote: > Checking to see if the symbol mentioned in the message is present in the > library you think will be used for linking is of course a good idea, but > you should also always examine the "config.log" file in the build > directory to see the real error message(s) from the configure test. Yes, that was the first thing I did. Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 11:28:08AM -0700, Greg A. Woods wrote: > Of course as mentioned squid-3 is heading on to old and should soon be > replaced by all users with squid-4, and especially pkgsrc should > deprecate squid-3 and provide the latest squid-4 by default. Just a heads up. I created wip/squid4 (yet to push) and it seems to be building fine on 8.0 amd64. Yet to try on current. [Will have to sort out why my ssh keys aren't working. Wonder whether those would be deleted if not used for long!] Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 04:57:10PM +0200, Martin Husemann wrote: > Note that the issue is the openssl version, not the architecture (or > am I missing something?) I'd think so. Just that mail from JP on this thread says the symbol is present on current amd64. I personally have only evbarm current and 8.0 amd64. Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 09:52:33AM -0500, John D. Baker wrote: > This issue is logged in PR pkg/53409, (and again in pkg/54132, closed > as duplicate) but that's as far as it has gone. The port on which it > was originally observed is sparc, but do feel free to log the issue seen > on other ports to the PR (pkg/53409). Maybe it will get some attention. Incidentally I just checked upstream and looks like squid 3.x is already regarded as too old! 3.5 is under the following comment: "Old Versions: Provided for archival purposes only. Not intended for general use in new installations." Time to create squid4? Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 09:39:57AM +0200, Martin Husemann wrote: > On Thu, Apr 25, 2019 at 01:06:08PM +0530, Mayuresh wrote: > > configure scripts (squid3 for one, but may be there are others) look for > > SSL_library_init. How to deal with that? > > They need to be adjusted for newer openssl versions. > > Martin I guess that might be a change at autoconf level or something? I think quickest I can do for now is to link with pkg openssl. It has the missing symbol. Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 09:28:38AM +0200, Martin Husemann wrote: > On Thu, Apr 25, 2019 at 12:54:46PM +0530, Mayuresh wrote: > > On Thu, Apr 25, 2019 at 12:50:45PM +0530, Mayuresh wrote: > > > # uname -m && nm /usr/lib/libssl.so | grep SSL_library_init > > > amd64 > > > 00022411 T SSL_library_init > > That is an old symbol, the function has been replaced by OPENSSL_init_ssl() > in newer versions. configure scripts (squid3 for one, but may be there are others) look for SSL_library_init. How to deal with that? Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 12:50:45PM +0530, Mayuresh wrote: > # uname -m && nm /usr/lib/libssl.so | grep SSL_library_init > amd64 > 00022411 T SSL_library_init > # > > > # uname -m && nm /usr/lib/libssl.so | grep SSL_library_init > evbarm > # > > Why this difference? +port-arm
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 09:03:47AM +0200, tlaro...@polynum.com wrote: > The problem was that at installation time, the pkg used INSTALL_PROGRAM > instead of INSTALL_DATA for installing libraries, and INSTALL_PROGRAM is > set to install with the "-s" that is the stripped flag. Basically: # uname -m && nm /usr/lib/libssl.so | grep SSL_library_init amd64 00022411 T SSL_library_init # # uname -m && nm /usr/lib/libssl.so | grep SSL_library_init evbarm # Why this difference? Mayuresh
Re: SSL_library_init not found in libssl in current
On Thu, Apr 25, 2019 at 12:47:15AM +, JP wrote: > SSL_library_init is in there on current amd64. > > "library 'ssl' is required" -- is libssl on there at all? configure errors can be misleading. The library is there but nm does not show the specific symbol in it. Mayuresh
SSL_library_init not found in libssl in current
Got this error when trying to build squid: checking for SSL_library_init in -lssl... no configure: error: library 'ssl' is required for OpenSSL on a recently picked current snapshot for RPI2. # uname -a NetBSD pi 8.99.37 NetBSD 8.99.37 (RPI2) #0: Mon Apr 22 04:10:33 UTC 2019 mkre...@mkrepro.netbsd.org:/usr/src/sys/arch/evbarm/compile/RPI2 evbarm The symbol is present in base libssl on amd64 8.0. Wonder whether this is specific to RPI2 or specific to current. Should pkg openssl be used in such case? Mayuresh
Re: NetBSD current on RPI2 (vs 8.0_RC1)
On Tue, Apr 23, 2019 at 03:55:41PM -, Michael van Elst wrote: > The display is handled by the firmware, depending on your HDMI display > you need to configure overscan in /boot/config.txt. I had a feeling that it was /boot/cmdline.txt which shows fb options. For either of them is there a documentation available? Mayuresh
NetBSD current on RPI2 (vs 8.0_RC1)
Just installed this image [thanks Herbert] on RPI2 http://nycdn.netbsd.org/pub/NetBSD-daily/HEAD/201904220430Z/evbarm-earmv7hf/binary/gzimg/armv7.img.gz This is not exactly a problem report, so not writing many details. I do not know on which of the following items work is/was going on and whether they were supposed to work anyway. If such is the case with some of the items please let me know, I'll share more details to describe the problems which may help fix the issues. Most of these were reported with details at 8.0_RC1 stage. - Garbled HDMI display on console (worked fine with x11) appears fixed now. But it still fails to occupy the full width and height of the screen. - run0 interface (a USB wifi adapter) still fails to function. - Tried connecting a couple of mobile phones' with usb tethering. These work fine on amd64 8.0_RC1, but are not recognized on RPI2. In any of these, if there any configuration things that I need to try please let me know. If more details are needed for investigation will be happy to share as well. Mayuresh
Which image for RPI2 NetBSD-current?
Wish to try out current on RPI2 by flashing an image. Is following the right one for this purpose? http://nycdn.netbsd.org/pub/NetBSD-daily/netbsd-8/latest/evbarm-earmv7hf/installation/instkernel/netbsd-RPI2_INSTALL.bin.gz What is the purpose of .bin.gz, .gz, .srec.gz, .symbols.gz etc? Mayuresh
why 2 mails every time?
why do i get 2 emails every time there's a reply to any email by me or to me? earlier i thought it was my mail client (mailx) which was doing something crazy, but it isn't so, a simply reply to netbsd-users goes out and sends me 2 copies of that same mail. like clock-work.
Re: netbsd : internals : bach book : good to start-off?
> From bounces-netbsd-users-owner-mayuresh=sdf@netbsd.org Mon Apr 22 > 04:55:39 2019 > From: Robert Elz > To: Mayuresh Kathe > cc: netbsd-users@netbsd.org > Subject: Re: netbsd : internals : bach book : good to start-off? > > Date:Mon, 22 Apr 2019 04:34:44 GMT > From:Mayuresh Kathe > Message-ID: <201904220434.x3m4yici026...@sdf.org> > > | just nitpicking, isn't bach's book reasonable enough for unix internals? > :) > > You think there is just one "unix" to have internals? Or that they are > all really similar, or something? > > As I recall (it has been a long time since I looked, but I think I > have a copy of that one, or some edition of it anyway somewhere) Bach's > book mostly describes System V. > > Even in the early 90's (in the vintage of McKusick's 4.3BSD book) > System V and BSD had diverged quite a lot internally. > > In the decades since, even moreso. There is (that I know of anyway) > no book that will come really close to describing NetBSD internals, > with all the bus_map and mem management (incl UVM), and locking, and ... > that are more or less unique to NetBSD - and yet are all fundamental > to a true understanding of the internals. > > Even McKusick's FreeBSD book (as similar aas FreeBSD is to NetBSD in > some ways) will contain much that is not relevant to NetBSD (including > soft mounts, and all related to that) and be lacking much, but it > is going to be much closer to NetBSD and so get you further than Bach's > book would. > > But if all you want is a guide to how some arbitrary unix system might > be implemented, or if you really want to know SysV internals, then yes, > that one should be just fine. > > kre > > ps: if you're really looking for a user or programmer's guide, then you > want something quite different. > > i am not looking for a user or programmer's guide. i have no knowledge of any operating system internals, leave alone unix. so, since bach's book is so light (in terms of page count) and affordable i thought it would be a good starting-off point into operating system internals. i know and fully acknowledge that i will have to work hard to understand netbsd internals, and the currently, the only way to do so is by reading the source, over and over again till i get comfortable with it.
Re: netbsd : internals : bach book : good to start-off?
> From bounces-netbsd-users-owner-mayuresh=sdf@netbsd.org Mon Apr 22 > 04:31:24 2019 > From: Robert Elz > To: Mayuresh Kathe > cc: netbsd-users@netbsd.org > Subject: Re: netbsd : internals : bach book : good to start-off? > > Date:Mon, 22 Apr 2019 03:54:20 GMT > From:Mayuresh Kathe > Message-ID: <201904220354.x3m3skvh008...@sdf.org> > > | would "the design of the unix operating system" by maurice bach > | be good as a starting off point to understand netbsd internals? > | of-course, there's no substitute to reading and re-reading the > | source, but just something to act as a spring-board! > > You'd probably be better with one of McKusick's books, on BSD internals > (even the FreeBSD one) than that - though any (reasonable) unix internals > book would at least get you started. > > kre > > just nitpicking, isn't bach's book reasonable enough for unix internals? :)
Re: is netbsd actually a toolkit?
thanks for that jay, looking at the netbsd using lynx prevents one from going to a lot of places on the website. > From jaypatel@gmail.com Mon Apr 22 04:19:08 2019 > From: Jay Patel > Subject: Re: is netbsd actually a toolkit? > To: Mayuresh Kathe > Cc: "netbsd-users @ netbsd. org" > > --c987ac058716c4d3 > Content-Type: text/plain; charset="UTF-8" > > Hi Mayuresh, > > This explains everything about NetBSD http://netbsd.org/about/ hope that's > what you are looking for. > > Regards, > Jay > > On Mon 22 Apr, 2019, 9:46 AM Mayuresh Kathe, wrote: > > > freebsd has an internal focus to become a good server operating system. > > openbsd has an internal focus to become a highly secure operating system. > > what is netbsd's internal focus? can't be just a highly portable operating > > system! is it more to be a really good toolkit for people to build their > > own operating systems for their desired hardware? > > > > --0000c987ac058716c4d3 > Content-Type: text/html; charset="UTF-8" > Content-Transfer-Encoding: quoted-printable > > Hi Mayuresh, >This explains everything about NetBSD=C2=A0 http://netbsd.org/a= > bout/">http://netbsd.org/about/ hope thats what you are looking fo= > r.=C2=A0Regards,=C2=A0<= > /div>Jay ir=3D"ltr" class=3D"gmail_attr">On Mon 22 Apr, 2019, 9:46 AM Mayuresh Kathe= > , mailto:mayur...@sdf.org;>mayur...@sdf.org wrote: r> -left:1px #ccc solid;padding-left:1ex">freebsd has an internal focus to bec= > ome a good server operating system. > openbsd has an internal focus to become a highly secure operating system. r> > what is netbsds internal focus? cant be just a highly portable op= > erating > system! is it more to be a really good toolkit for people to build their > > own operating systems for their desired hardware? > > > --c987ac058716c4d3-- >
is netbsd actually a toolkit?
freebsd has an internal focus to become a good server operating system. openbsd has an internal focus to become a highly secure operating system. what is netbsd's internal focus? can't be just a highly portable operating system! is it more to be a really good toolkit for people to build their own operating systems for their desired hardware?
netbsd : internals : bach book : good to start-off?
would "the design of the unix operating system" by maurice bach be good as a starting off point to understand netbsd internals? of-course, there's no substitute to reading and re-reading the source, but just something to act as a spring-board!
Re: Alternative DVCS to git: hg?
On Tue, Apr 16, 2019 at 09:57:02PM +0530, Mayuresh wrote: > On Tue, Apr 16, 2019 at 11:12:10AM -0500, J. Lewis Muir wrote: > > > I am just intrigued by it being written in python (except may be for the > > > merge algorithm which is in C). Wouldn't most engineers prefer C/C++ for > > > such a low level and key component? > > > > I'm sure some would. But others believe it's a better choice to write > > software in a high-level language (for various reasons which might > > include speed of development, ease of readability, security (e.g., > > built-in protection from certain classes of security vulnerabilities), > > libraries, ease of cross-platform development, etc.). If something > > is known to be or is discovered that is measurably too slow and the > > application spends a significant amount of time there, then the > > developers will spend effort improving the speed there. This may be > > done within the high-level language, or it might be done by writing > > parts of the application in C. > > Came across this: plan to use rust for hg: > https://www.mercurial-scm.org/wiki/OxidationPlan > > Quite interesting. Would be good to see A. hg using a compiled language B. > rust getting another big user. In 2/3 days of hg usage (on backdrop of several years of git usage), no functional problems and I enjoy the ease of command lines. But there indeed is a significant difference in performance. git is way too faster. Would be watching the hg-rust space quite eagerly. Mayuresh
Re: Alternative DVCS to git: hg?
On Tue, Apr 16, 2019 at 12:57:33PM -0400, Greg Troxel wrote: > It would be even cooler if rust built on every system NetBSD ran on with > moderate amounts of CPU time! On my last rebuild on a 2006-vintage i386 > laptop (Core Duo, 4G RAM), it took 7h45m to build. But at least it > built. I think these are relatively early days for rust and even llvm (in terms of optimizations etc). Both look promising and more applications adopting them would make them only better over time. Mayuresh
Re: Alternative DVCS to git: hg?
On Tue, Apr 16, 2019 at 11:12:10AM -0500, J. Lewis Muir wrote: > > I am just intrigued by it being written in python (except may be for the > > merge algorithm which is in C). Wouldn't most engineers prefer C/C++ for > > such a low level and key component? > > I'm sure some would. But others believe it's a better choice to write > software in a high-level language (for various reasons which might > include speed of development, ease of readability, security (e.g., > built-in protection from certain classes of security vulnerabilities), > libraries, ease of cross-platform development, etc.). If something > is known to be or is discovered that is measurably too slow and the > application spends a significant amount of time there, then the > developers will spend effort improving the speed there. This may be > done within the high-level language, or it might be done by writing > parts of the application in C. Came across this: plan to use rust for hg: https://www.mercurial-scm.org/wiki/OxidationPlan Quite interesting. Would be good to see A. hg using a compiled language B. rust getting another big user. Mayuresh
Re: Alternative DVCS to git: hg?
On Mon, Apr 15, 2019 at 01:52:26PM -0500, J. Lewis Muir wrote: > Yes, it's a good alternative. I use it for most of my projects. It's > also used by a number of large projects such as Firefox, nginx, and > OpenJDK, and I gather it's on a short list of VCSes being evaluated by > NetBSD as its next VCS (which would replace CVS). I also read somewhere facebook picking it, and also heavily contributing to it. I am just intrigued by it being written in python (except may be for the merge algorithm which is in C). Wouldn't most engineers prefer C/C++ for such a low level and key component? Regarding NetBSD, pkgsrc-wip was moved to git and pkgsrc has a git view of its CVS repo. So I thought git was more likely candidate in NetBSD. It will help if list members with experience of both git and hg could share some pros and cons of both. (No flame war meant!) Mayuresh
Alternative DVCS to git: hg?
As shared in some other threads, I am trying to get an encrypted revision control repository using encfs. git seems to be pushing the limits of a fuse file system one way or the other. libperfuse catered to some of them, but quite peculiar use of filesystem by git (and I wonder why it had to be coded that way) is making it impossible to use git on encfs file system (on NetBSD). Probably later versions of libperfuse might deal with all quirks of git, but at present I am looking for a less quirky distributed revision control system - similar to git. I have tried out hg and it worked without any problems on an encfs mount. But I have used hg very little. Is that a good alternative to git or are there better options? Mayuresh
Re: Propose to link fuse-encfs against pkg libperfuse [Was Re: mmap errors related to using encfs]
On Sun, Apr 14, 2019 at 12:30:24PM -0400, Greg Troxel wrote: > Opinions differ about whether libfuse/perfuse and librefuse should be > the standard approach. For a very long time, librefuse has been the > standard approach. So obviously just changing to use libfuse is out of > the question. > > Adding some kind of preference variable so that people who wish to have > all packages that need the FUSE high-level API use libfuse/perfuse > instead (leaving the default) seems fine; that's similar to how we let > alternative implementations of other things be selected. I suppose you meant perfuse is for low-level APIs, or did I miss something? So if I understand things, the options are these: fuse-* to by default link with librefuse. [I do not know technical reasons behind this preference, but that apart. Users have to be aware that they won't get some features of the filesystem e.g. mmap, instead of them discovering this the hard way]. fuse-* to provide an option to link with pkg fuse, for those who want low level APIs. Most fuse-* patches would need to become conditional. Going by encfs as an example, most patches may not be needed if you choose pkg fuse. [I think since pkg fuse layer exists, where Maya has got all patch upstreaming done already, probably it's not productive to chase upstream for fuse-* patches that are nb specific. That's just my 2 cents.] pkg fuse need not link with librefuse as fuse-* can link with it directly. [This I am not 100% sure about. Just seeking views.] pkg fuse may have an option to link with pkg libperfuse or base libperfuse. Or may be instead of user choosing this, it can be decided by the following logic: As confirmed by Emmanuel, up to NetBSD 8.0 it needs to link with pkg libperfuse only. For NetBSD>8.0 base libperfuse should be usable. Please let me know if above analysis is correct. Mayuresh
fuse-encfs write permission for non root
As described in a couple of other threads, filesystems/fuse-encfs is nearly working well on NetBSD 8.0. I am able to mount as non root user as well. But I can't still write to the mounted directory as a non root user. The encrypted directory, mount point, /dev/putter all are owned by the non root user. I have also set "vfs.generic.usermount=1" (without which a non root user was not able to mount). If write permission issue is solved, that will complete the functionality of encfs on NetBSD. Please do share suggestions. BTW this is a thread of 2010, with more or less the same state... https://mail-index.netbsd.org/pkgsrc-users/2010/02/06/msg011657.html Mayuresh
Re: error writing fsbn, too soon on a new flash drive
On Sun, Apr 14, 2019 at 12:49:25PM +0530, Mayuresh wrote: > Just in case it matters: It was NetBSD 8.0_RC1 amd64 on the HDD while it was NetBSD 8.0 amd64 on the flash drive mentioned in my last post. Mayuresh
error writing fsbn, too soon on a new flash drive
My 3 year old laptop internal hard drive had started showing "error reading fsbn". The drive is still working, but I suppose that's a sign I should prepare to replace it. As a quick measure I installed NetBSD on an unused USB 2.0 64GB Sandisk flash drive (aka stick/pen drive). It worked fine for a week or so. I haven't done any major disk write exercise such as any builds etc. on this drive. But within a week of usage it started showing "error writing fsbn". It hung at "checking filesystem" and after a long wait I had to reboot it. But it booted fine this time around and as yet hasn't shown any sign of fsbn error. Following questions come to mind now: - Is it a workable idea to use a usb flash drive as the main disk i.e. would it last for 2/3 years with regular daily usage, which is not disk intensive (particularly no major and frequent builds on this drive). No major use of swap. If someone is using / has used such setup, please do share your experiences. - Since I faced fsbn error on an almost new and unused flash drive, which rebooted fine on power cycle with no sign of the same error now, what are the chances that this error was a false alarm and I can continue to use this drive for some time now? Mayuresh
Re: Propose to link fuse-encfs against pkg libperfuse [Was Re: mmap errors related to using encfs]
On Sun, Apr 14, 2019 at 10:48:51AM +0530, Mayuresh wrote: > 3. We might as well take this opportunity to encfs-1.9.5. I have already > confirmed the results on the list. Missing words: upgrade to. Mayuresh
Propose to link fuse-encfs against pkg libperfuse [Was Re: mmap errors related to using encfs]
On Sat, Apr 13, 2019 at 12:31:26AM +0530, Mayuresh wrote: > Is it possible that the issue / limitation is with librefuse? (And in that > case shall we be trying using filesystems/fuse? I pursued above point further. I took offline help from Emmanuel Dreyfus. Thanks Emmanuel for your inputs. For reference, this old post gives the genesis of libperfuse: https://mail-index.netbsd.org/tech-userlevel/2010/08/22/msg003843.html Indeed the low level interface mentioned in the post is needed by some filesystems. E.g. git wasn't working on encfs mounted dir due to unsupported mmap. But if we link with libperfuse it is found working now. Would also mention that linking to base libperfuse isn't helping as (at least as of 8.0) it is not up to date. pkg's libperfuse works fine though. (current should work fine, though I think better to link with pkg). I propose the following: 1. in mk/fuse.buildlink3.mk we should include filesystems/fuse/buildlink3.mk for NetBSD as well. Alternatively provide a choice to do so. [I have changed the code in an ad hoc manner right now. If the proposal is ok I'll try to do it neater and submit a patch. But I am not so conversant with changing mk/ files, do's and dont's of doing so etc.] 2. Two of the patches in fuse-encfs (1 added last week and 1 older) seem to be no more required if we link with filesystems/fuse: patch-encfs_FileUtils.cpp patch-encfs_main.cpp Not sure about the third one: patches/patch-encfs_DirNode.cpp. (Will test without it and confirm.) 3. We might as well take this opportunity to encfs-1.9.5. I have already confirmed the results on the list. Request your comments. Mayuresh
Re: mmap errors related to using encfs
On Sat, Apr 13, 2019 at 01:29:26PM -, Michael van Elst wrote: > Yes. You need to increase kern.sbmax. Thanks. Yes, that solved this issue. But there is io error occurring repeatedly. E.g. if I do ls in mounted dir, it works once, but second time around just shows io error. /var/log/messages shows following kinds of errors repeatedly when io error mentioned above occurs: encfs[1434]: ERROR withFileNode: error caught in fgetattr: fh=139135124959488 not found in fuseFhMap Mayuresh
Re: Scanner experiences
On Fri, Apr 12, 2019 at 07:58:40PM -0400, Bob Bernstein wrote: > At this point the wiki page gives instructions for configuring and > re-making the kernel. Two concerns give me pause: I have an epson scanner, too, though not same model number. Currently it's set under Linux but I recollect it was working on NetBSD as ugen0 itself. So I think you might not require reconfiguring kernel since yours is shown as ugen0. You might just have to tinker with the device option (-d) of scanimage. [I do not have that setup now, otherwise I'd have shared my exact device string.] Mayuresh
Re: mmap errors related to using encfs
On Sat, Apr 13, 2019 at 09:24:17AM +0530, Mayuresh wrote: > encfs: perfuse_open: setsockopt SO_SNDBUF to 2162688 failed: No buffer space > available Env var, something like, PERFUSE_BUFSIZE=131072 solves above error. (Larger than that doesn't work. May be some sysctl setting would make it work.) However now I get input/output error on the mounted encfs. E.g. # ls ls: .: Input/output error Usually first time after mounting, ls works, 2nd command onwards it stops working. Mayuresh
Re: mmap errors related to using encfs
On Sat, Apr 13, 2019 at 12:31:26AM +0530, Mayuresh wrote: > Is it possible that the issue / limitation is with librefuse? (And in that > case shall we be trying using filesystems/fuse? Just giving it a try. Was able to build filesystems/fuse-encfs against filesystems/fuse. # ldd /usr/pkg/bin/encfs /usr/pkg/bin/encfs: -lfuse.2 => /usr/pkg/lib/libfuse.so.2 -lperfuse.0 => /usr/lib/libperfuse.so.0 -lpuffs.2 => /usr/lib/libpuffs.so.2 -lc.12 => /usr/lib/libc.so.12 -lpthread.1 => /usr/lib/libpthread.so.1 -lssl.12 => /usr/lib/libssl.so.12 -lcrypto.12 => /usr/lib/libcrypto.so.12 -lcrypt.1 => /lib/libcrypt.so.1 -lintl.1 => /usr/lib/libintl.so.1 -lstdc++.8 => /usr/lib/libstdc++.so.8 -lm.0 => /usr/lib/libm.so.0 -lgcc_s.1 => /usr/lib/libgcc_s.so.1 When I run it I get these errors: encfs: perfuse_open: setsockopt SO_SNDBUF to 2162688 failed: No buffer space available encfs: perfuse_open: setsockopt SO_RCVBUF to 2162688 failed: No buffer space available encfs: perfuse_open: setsockopt SO_SNDBUF to 2162688 failed: No buffer space available encfs: perfuse_open: setsockopt SO_RCVBUF to 2162688 failed: No buffer space available encfs: perfuse_open: setsockopt SO_SNDBUF to 2162688 failed: No buffer space available encfs: perfuse_open: setsockopt SO_RCVBUF to 2162688 failed: No buffer space available Will investigate by looking at code, but I have no other background. Sharing so that I get help/inputs as well. Mayuresh
Re: mmap errors related to using encfs
On Fri, Apr 12, 2019 at 12:14:08PM -0400, Greg Troxel wrote: > The following hints may seem obvious, but because you didn't report on > your investigations into the source code of encfs and git, I'll send > them anyway :-) Thanks. These are useful. In encfs I did not yet find any clue, but it might be alright for it to not support mmap, as you suggest. I'll try it out on Linux to see what happens. In git code I have not found it backtracking if mmap fails, but I do not know whether that's a conscious choice or not. I'll log a bug upstream. BTW it limits utility of encfs for me, as the issue is not just git. I was planning to use encfs for various kinds of files, applications of some of which might run into mmap issue. Mayuresh
mmap errors related to using encfs
Trying out filesystems/fuse-encfs on NetBSD 8.0 amd64. # mount | grep puff /dev/puffs on /mnt/foo type puffs|refuse:encfs (nodev, nosuid) # cd /mnt/foo/ # git init --bare error: unable to mmap '/mnt/foo/config': Operation not supported fatal: could not set 'core.filemode' to 'true' This problem occurs with encfs mounted directory only. From the error message it does not look like a git problem, rather some interaction between mmap and encfs. Would be interesting to know what's happening and is it possible to deal with this error. Mayuresh
concurrencykit in the kernel!
heard about freebsd guys using the concurrency kit (concurrencykit.org) in their kernel. anyone has any idea about how it could be useful from a netbsd perspective too? i was under the impression that the kit would be useful only for userland projects, wonder what role it would play at the kernel level.
Re: how to use netbsd with ubuntu?
running a netbsd vm on a google compute engine is a nice idea, but that makes me dependent on the network, something which cannot be trusted to not fail. > From bsieg...@gmail.com Fri Apr 12 09:05:51 2019 > From: Benny Siegert > Subject: Re: how to use netbsd with ubuntu? > To: Mayuresh Kathe > Cc: NetBSD Users > > Another option would be to run Ubuntu as a Xen Dom0 and run NetBSD in > a DomU, which is well-supported. Myself, I use the lazy solution, > which is a NetBSD VM on Google Compute Engine that I spin up when > needed. > > On Fri, Apr 12, 2019 at 9:33 AM Mayuresh Kathe wrote: > > > > as in a previous email, i need to use ubuntu as my primary desktop. > > i researched various options via googling around as well as asking > > this mailing list and it looks like the best thing would be to run > > netbsd in a virtual environment. for the same, would "qemu" be > > considered good enough? if it is, should i upgrade my system memory > > from 4gib to 8gib? i have a 1tib hard disk, would running with a > > 256gib 'ssd' instead pose any unforeseen problems? thanks. > > > > -- > Benny >
Re: how to use netbsd with ubuntu?
> From er.abhinav.upadh...@gmail.com Fri Apr 12 07:45:24 2019 > From: Abhinav Upadhyay > Subject: Re: how to use netbsd with ubuntu? > To: Mayuresh Kathe > Cc: NetBSD Users Mailing List > > On Fri, Apr 12, 2019 at 1:03 PM Mayuresh Kathe wrote: > > > > as in a previous email, i need to use ubuntu as my primary desktop. > > i researched various options via googling around as well as asking > > this mailing list and it looks like the best thing would be to run > > netbsd in a virtual environment. for the same, would "qemu" be > > considered good enough? if it is, should i upgrade my system memory > > from 4gib to 8gib? i have a 1tib hard disk, would running with a > > 256gib 'ssd' instead pose any unforeseen problems? thanks. > > What do you mean by running with Ubuntu, Is dual boot an option? If > virtualization is the only option, I prefer VirtualBox over qemu when > running NetBSD inside a VM. > > - > Abhinav > no, dual boot is not an option, i need ubuntu for chrome to surf the web, especially for a few websites which prefer chrome over chromium. may i know your rationale for leaning towards virtualbox over qemu?
how to use netbsd with ubuntu?
as in a previous email, i need to use ubuntu as my primary desktop. i researched various options via googling around as well as asking this mailing list and it looks like the best thing would be to run netbsd in a virtual environment. for the same, would "qemu" be considered good enough? if it is, should i upgrade my system memory from 4gib to 8gib? i have a 1tib hard disk, would running with a 256gib 'ssd' instead pose any unforeseen problems? thanks.
bozohttpd : how to enable TLS
The man page of bozohttpd in the base (NetBSD 8.0) says TLS 1.1 and 1.2 is supported. But I couldn't find much documentation about how to enable it. Are there any tutorials on how to use TLS with bozohttpd? Mayuresh
Re: filesystems/fuse-encfs compilation error; upgrade
On Tue, Apr 09, 2019 at 09:46:31PM +0530, Mayuresh wrote:
> BTW I notice that pkgsrc's filesystems/fuse is version 2.9. Thus it may
> not require compat22 patch.
>
> But I am not sure how to make fuse-encfs pick pkgsrc's fuse rather than
> from base. I installed filesystems/fuse and did a clean build (by not
> applying above patch), but it results in same error implying, fuse-encfs
> is using base fuse.h.
Pasting this from mk/fuse.buildlink3.mk
It appears that fuse/buildlink3.mk is included only for certain platforms.
May be on NetBSD pkgsrc's fuse is not recommended? Please confirm.
. elif ${OPSYS} == "Linux"
.include "../../filesystems/fuse/buildlink3.mk"
. elif !empty(MACHINE_PLATFORM:MSunOS-5.11-*)
.if !exists(/usr/include/fuse/fuse.h)
PKG_FAIL_REASON+= "Couldn't find fuse headers, please install
libfuse."
.endif
.include "../../filesystems/fuse/buildlink3.mk"
Mayuresh
Re: filesystems/fuse-encfs compilation error; upgrade
On Tue, Apr 09, 2019 at 08:58:16PM +0530, Mayuresh wrote: > +#ifdef __NetBSD__ > + fuse_unmount_compat22(mountPoint); > +#else >fuse_unmount(mountPoint, nullptr); > +#endif BTW I notice that pkgsrc's filesystems/fuse is version 2.9. Thus it may not require compat22 patch. But I am not sure how to make fuse-encfs pick pkgsrc's fuse rather than from base. I installed filesystems/fuse and did a clean build (by not applying above patch), but it results in same error implying, fuse-encfs is using base fuse.h. Please advise. Mayuresh
Re: filesystems/fuse-encfs compilation error; upgrade
On Tue, Apr 09, 2019 at 08:58:16PM +0530, Mayuresh wrote: > It builds fine but I am yet to test it. This is to confirm that in quick tests this is working fine for me. Knowing the stability would take some usage over time. I'd feel hugely relieved if it works stably on NetBSD. It would solve a lot of problems for me. Please do comment on my patch (prev mail). Mayuresh
Re: filesystems/fuse-encfs compilation error; upgrade
On Tue, Apr 09, 2019 at 10:34:16AM -0400, Greg Troxel wrote:
> Please feel free to create an updated package in pkgsrc-wip.
I just did the following to get it to compile:
- Upgraded to latest release 1.9.5
- Reconciled existing 2 patches
- Created following new patch
$NetBSD$
--- encfs/FileUtils.cpp.orig2018-04-27 08:52:22.0 +
+++ encfs/FileUtils.cpp
@@ -1734,7 +1734,11 @@ RootPtr initFS(EncFS_Context *ctx, const
void unmountFS(const char *mountPoint) {
// fuse_unmount returns void, is assumed to succeed
+#ifdef __NetBSD__
+ fuse_unmount_compat22(mountPoint);
+#else
fuse_unmount(mountPoint, nullptr);
+#endif
#ifdef __APPLE__
// fuse_unmount does not work on Mac OS, see #428
// However it makes encfs to hang, so we must unmount
It builds fine but I am yet to test it.
I just found fuse_unmount_compat22 with a somewhat compatible signature in
fuse.h but I do not know whether it can be used in place of fuse_unmount.
Would appreciate if someone could comment on this.
ccing netbsd-users also.
Mayuresh
Re: 8.0 amd64 poweroff reboots
On Tue, Apr 02, 2019 at 05:19:20PM +0200, Martin Husemann wrote: > > ?() at 0 > > usb_transfer_complete() at netbsd:usb_transfer_complete+0x... > > xhci_softintr() ... > > usb_soft_intr() ... > > softint_dispatch() ... > > OK, which kernel is that exactly and can you give the +0x at the end of > the first two lines above in details? # uname -a NetBSD asusnetbsd 8.0 NetBSD 8.0 (GENERIC) #0: Tue Jul 17 14:59:51 UTC 2018 mkre...@mkrepro.netbsd.org:/usr/src/sys/arch/amd64/compile/GENERIC amd64 Was downloaded yesterday from: https://cdn.netbsd.org/pub/NetBSD/NetBSD-8.0/images/NetBSD-8.0-amd64-install.img.gz ?() at 0 usb_transfer_complete() at netbsd:usb_transfer_complete+0x148 xhci_softintr() ...+0x1df usb_soft_intr() ...+0x38 softint_dispatch() ...+0x91 Mayuresh
Re: 8.0 amd64 poweroff reboots
On Tue, Apr 02, 2019 at 04:50:19PM +0200, Martin Husemann wrote: > > sd0(umass0:0:0:0) generic HBA error > > sd0: cache synchronization failed > > fatal page fault in supervisor mode > > trap type 6 code 0x10 rip 0 cs 0x8 rcflags 0x10282 cr2 0 ilevel 0x5 rsp... > > curlwp 0x pid 0.6 lowest kstack 0x > > kernel: page fault trap, code=0 > > Stopped in pid 0.6(system) at 0: invalid address > > Does the keyboard work here? Can you do a "bt" and show the output? Yes. ?() at 0 usb_transfer_complete() at netbsd:usb_transfer_complete+0x... xhci_softintr() ... usb_soft_intr() ... softint_dispatch() ... DDB lost frame for netbsd:Xsoftintr+0x4f, trying 0x --- interrupt --- bb6 Mayuresh
Re: 8.0 amd64 poweroff reboots
On Tue, Apr 02, 2019 at 04:02:10PM +0200, Martin Husemann wrote: > This likely is a crash late in the shutdown sequence, maybe try Yes. The trace shows somewhat like: sd0(umass0:0:0:0) generic HBA error sd0: cache synchronization failed fatal page fault in supervisor mode trap type 6 code 0x10 rip 0 cs 0x8 rcflags 0x10282 cr2 0 ilevel 0x5 rsp... curlwp 0x pid 0.6 lowest kstack 0x kernel: page fault trap, code=0 Stopped in pid 0.6(system) at 0: invalid address # disklabel /dev/sd0 # /dev/sd0: type: unknown disk: Cruzer Blade label: flags: bytes/sector: 512 sectors/track: 63 tracks/cylinder: 16 sectors/cylinder: 1008 cylinders: 65535 total sectors: 123174912 rpm: 3600 interleave: 1 trackskew: 0 cylinderskew: 0 headswitch: 0 # microseconds track-to-track seek: 0 # microseconds drivedata: 0 16 partitions: #sizeoffset fstype [fsize bsize cpg/sgs] a: 11921112032 4.2BSD 2048 16384 0 # (Cyl. 0*- 118265*) b: 3963760 119211152 swap # (Cyl. 118265*- 122197*) c: 12317488032 unused 0 0# (Cyl. 0*- 122197*) d: 123174912 0 unused 0 0# (Cyl. 0 - 122197*) Mayuresh
8.0 amd64 poweroff reboots
I have been using poweroff command (without any arguments) to power off a laptop on NetBSD 8.0_RC1 amd64. Recently I installed 8.0 on a usb stick and notice that poweroff now reboots the laptop instead of powering off. The /etc/rc.shutdown is not altered by me and neither it is different between 8.0_RC1 and 8.0. Do not know whether it matters - 8.0 boots from a usb stick and 8.0_RC1 from native hard drive. An aside: The laptop hard drive had started showing "error reading fsbn". I presume it means disk will need a replacement soon. Hence as a precaution I switched to a usb stick based installation. The drive is barely 3 years old and also hosts Linux which has not shown any warnings. Am I right in interpretation of above error or is there something I can try with the hard drive to extend its useful life? Mayuresh
Re: netbsd development on a raspberry pi
> From bsieg...@gmail.com Tue Apr 2 09:30:51 2019 > From: Benny Siegert > Subject: Re: netbsd development on a raspberry pi > To: Mayuresh Kathe > Cc: NetBSD Users > > On Tue, Apr 2, 2019 at 10:55 AM Mayuresh Kathe wrote: > > the question that i don't have an answer to is whether it is possible > > to do development on an 'rpi' running netbsd? development would not > > be restricted to just userland, but would also spill over into > > kernel zone too. > > Sure! NetBSD runs well on the RPi with the evbarm port. You'll want to > run -current for best results. See http://invisible.ca/arm/ for images > that you can just dd onto a memory card. > > Personally, I have an "OrangePi Plus 2E" that I bought based on > jmcneill's recommendation. It is a bit faster than the RPi > (particularly for storage and network), has been running super stable > doing continuous builds and costs about the same as an RPi. > > -- > Benny > what could be an ideal storage size for the micro-sd card?
netbsd development on a raspberry pi
hello, returning to netbsd after a long time, feels good to be back. i am in a peculiar situation for which i can't figure out a solution. i have my primary desktop, an 'hp-aio' based on a dual-core celeron. i have to run ubuntu on it since i need to run "google chrome". i tried "chromium" but that doesn't fit my needs as well, as chrome. i intend to do netbsd specific development in the long term but due to lack of desk space can't afford to setup another 'pc'. i was musing about setting up a "raspberry pi 3 b+" to run netbsd. i would be working with netbsd on the 'rpi' via a 'ssh' session. the question that i don't have an answer to is whether it is possible to do development on an 'rpi' running netbsd? development would not be restricted to just userland, but would also spill over into kernel zone too.
Re: Qemu virtual machine
On Sat, Mar 16, 2019 at 11:48:27AM +0100, Pedro Pinho wrote: > I'm planning to create a qemu virtual machine on my NetBSD laptop, so I can > set-up a minimal linux system running Firefox and watch Netflix on NetBSD. > I'm wondering though, which one is better for this user case, HAXM or NVMM? > Any thoughts? I am also interested in widevinecdm based contents on NetBSD. It's a pain to leave your default OS and reboot to Linux every time to watch a video. Haven't tried qemu of late. But it used to be too slow for this purpose. Linux emulation layer can be an alternative. Flash plugin, adobe reader, libreoffice-bin etc have been working pretty fine on NetBSD through Linux emulation. Not sure what it would take to make firefox-bin available on NetBSD. Xen is yet another alternative. But some limitations like NetBSD Dom0 can use only 1 cpu make it less attractive for this purpose. You'd rather want a Linux DomU to use 1 cpu and let NetBSD Dom0 use the rest. Looking forward to a wider discussion on this. Mayuresh
Re: syslogd exits with fatal error [Was Re: newsyslog does not find...]
On Wed, Mar 13, 2019 at 07:58:46AM +0100, Martin Husemann wrote: > > > This is on 8.0_RC1 amd64. > > Have you considered updating? There are 3854 lines in doc/CHANGES* > since that version. It is on my TODO list, though there is some inertia. Will try to do it soon. Mayuresh.
syslogd exits with fatal error [Was Re: newsyslog does not find...]
On Tue, Feb 19, 2019 at 08:08:42AM +0530, Mayuresh wrote: > This is on 8.0_RC1 amd64. > > newsyslog: /var/run/syslogd.pid: No such file or directory > syslogd is sometimes exiting, not sure why, there is no disk space issue, nothing with heavy use of memory/swap is running. Last log says: localhost syslogd[293]: Fatal error, exiting I could enable syslogd verbose option as a last resort; but not sure about when it would trigger next, hence would preferably like to avoid keeping verbose enabled. Are there any other known issues why syslogd would exit? Mayuresh
newsyslog sometimes does not find syslog.pid
This is on 8.0_RC1 amd64. Occasionally (not regularly) I get the following email alert: newsyslog: /var/run/syslogd.pid: No such file or directory What does this indicate and what can I check in this regard. Mayuresh
Re: postfix alternatives on NetBSD / pkgsrc
On Sat, Jan 19, 2019 at 05:56:31PM +0530, Mayuresh wrote: > In rejectall > /./ REJECT 550 5.1.1 > > Now gmail does not complain. However I still don't know why it still shows > 554 5.7.1 first and then 550 5.1.1 Dropped the word REJECT and now it works fine. Mayuresh
Re: postfix alternatives on NetBSD / pkgsrc
On Fri, Jan 18, 2019 at 08:34:21AM -0600, Edgar Pettijohn wrote: > The only way I know is through an access(5) map. But I'm not sure if it > can be done with this specific use case. I replaced `reject' with a regexp in class definition: insiders_only = check_sender_access hash:/etc/postfix/insiders, check_sender_access regexp:/etc/postfix/rejectall #was just reject here (Well why doesn't posfix let me write the code right after reject instead of having to create another regexp? I think it believes in making itself a black art of sorts.) In rejectall /./ REJECT 550 5.1.1 Now gmail does not complain. However I still don't know why it still shows 554 5.7.1 first and then 550 5.1.1 Jan 19 17:45:24 localhost postfix/smtpd[8783]: NOQUEUE: reject: RCPT from mail-it1-f176.google.com[209.85.166.176]: 554 5.7.1 : Sender address rejected: 550 5.1.1; gmail says: 554 5.7.1 : Sender address rejected: 550 5.1.1 Mayuresh
Re: postfix alternatives on NetBSD / pkgsrc
On Fri, Jan 18, 2019 at 08:03:41AM -0600, Edgar Pettijohn wrote: > > 554 5.7.1 > > Seems like 550 would be a better error code for this situation. I was trying to set that (as I noticed gmail didn't complain for a mail that was bounced "normally" - such as non existent id). But struggling to find out an example of how to do it - how do I relate my reject point with a certain reject code? Mayuresh
Re: postfix alternatives on NetBSD / pkgsrc
On Fri, Jan 18, 2019 at 02:49:15PM +0100, Tobias Ulmer wrote: > On Fri, Jan 18, 2019 at 07:50:52AM +0100, Niels Dettenbach (Syndicat IT & > Internet) wrote: > > The security footprint is very good. > > https://www.cvedetails.com/vulnerability-list/vendor_id-10919/product_id-19563/Exim-Exim.html I am not an expert in comparing these and I am not taking any side. But let's put both on the table to make a fair comparison: https://www.cvedetails.com/vulnerability-list/vendor_id-8450/product_id-14794/Postfix-Postfix.html Mayuresh
Re: postfix alternatives on NetBSD / pkgsrc
On Fri, Jan 18, 2019 at 06:45:06AM -0600, Edgar Pettijohn wrote: > I think you should post the logs from your postfix test with Gmail > issue. I bet someone here knows an option to correct it. Not much I can see. I think it has more to do with the error code interpretation by gmail. For other rejects such as mails directed to non existent users gmail doesn't call the server as misconfigured. Jan 18 09:21:15 localhost postfix/smtpd[28050]: connect from mail-lj1-f177.google.com[209.85.208.177] Jan 18 09:21:15 localhost postfix/smtpd[28050]: NOQUEUE: reject: RCPT from mail-lj1-f177.google.com[209.85.208.177]: 554 5.7.1 : Recipient address rejected: Access denied; from= to= proto=ESMTP helo= Jan 18 09:21:16 localhost postfix/smtpd[28050]: disconnect from mail-lj1-f177.google.com[209.85.208.177] ehlo=1 mail=1 rcpt=0/1 data=0/1 quit=1 commands=3/5 Gmail bounced to y...@gmail.com says: Message not delivered Your message couldn't be delivered to x...@myhost.com because the remote server is misconfigured. See technical details below for more information. The response from the remote server was: 554 5.7.1 : Recipient address rejected: Access denied I have also posted my postfix conf in previous mail. Mayuresh
Re: postfix alternatives on NetBSD / pkgsrc
On Fri, Jan 18, 2019 at 07:50:52AM +0100, Niels Dettenbach (Syndicat IT & Internet) wrote: > We use EXIM since decades now from small satellite mailer setups to very > large ISP setups after migrated from sendmail and postfix as they brought our > hardware down in performance with heavy mail loads. > > EXIM is very (!) efficient - especially when build from sources the > "official" way (what is provided by pkgsrc by build options). This means you > just compile fucntionality / code into the binary what you really need. Thanks a lot - a first hand account really helps. In general searches on comparison between the two, most often claim postfix to have better performance than exim (some qualify the statement saying "for large queues" - which does not bother me for my use case, but in your case you have seen it scaling well as well). > The security footprint is very good. > > The config is very flexible but of consistent syntax (developed my a > mathematican - Phillip Hazel) - for me much more transparent then on postfix. > There are many of good examples and howtos out there which provide single > config files you could easily adapt and use. But you can split config files > too if you prefer that. By profession I am a in programming languages researchers and have created many DSLs in my career. I can say in light of whatever little experience of inventing notations I have, postfix notation does not really sound intuitive, particularly when the problem domain does not require it to be that complex. I'll definitely give exim a try on this aspect. Mayuresh
postfix alternatives on NetBSD / pkgsrc
Short story: A quick search shows exim as the main alternative. I am looking for efficiency and if possible clearer semantics (than postfix!) of the configuration files. Please do suggest alternatives. Long story: There is a separate mail thread in which I am sharing my experience of setting up a mailing list (on an experimental basis right now). I'd prefer using an MTA first for this, the list being mostly static. Would go to MLM software only if I fail to get it right with MTA. I think I have got nearly everything right with postfix (my current and default MTA), except that when I `reject' an email sent by a non member to the list, the sending servers (such as gmail) report that my mail server is not configured properly. The status code returned is (554 5.7.1) actually fine, but I am not sure whether there is indeed any issue with my configuration that draw this remark. Bouncing of an email for protecting senders to an id should not be such an unusual scenario for the sending server. Further I am not sure whether to just ignore the error that senders would get. This is because there is some probability of sending server blacklisting the domain on recurrence of such bounce. So thought of giving an alternative server a try to see if for a similar situation sending servers complain or not. Mayuresh
Re: Mailing list manager on NetBSD
On Thu, Jan 17, 2019 at 11:50:44AM -0200, Silas wrote: > On Wed, Jan 16, 2019 at 07:09:56AM +0530, Mayuresh wrote: > > Since your friend runs multiple lists, may be he has been able to put such > > restriction. Please do share if you can. > > I talked to him. He let it be clear that it is a [ugly?] workaround. > Actually, he wanted to use a mailing list manager, although he agreed that > it might not free the server from the load that happens when someone (like > the university administration) sends an email to 1+ students. Thanks, and please convey my thanks to your friend for sharing this! I was tinkering with this idea and came up with my version of ugliness. While I have used postfix to set up simple mail server, I do not have much experience of setting up the kind of things I illustrate below. The documentation is really intimidating and semantics not easy to grasp. This is adopted from /usr/share/examples/postfix/RESTRICTION_CLASS_README main.cf: # `permit' is really my trial and error outcome, most examples show # "reject" there, but that leads to rejection of mails sent to non-list # ids smtpd_recipient_restrictions = ... other things... check_recipient_access hash:/etc/postfix/protected_destinations permit insiders_only = check_sender_access hash:/etc/postfix/insiders smtpd_restriction_classes = insiders_only protected_destinations: insiders_only insiders (generatable from the alias list): OK OK ... This is kind of working for me, except that when a non-member's email gets rejected, e.g. from gmail, gmail shows "the remote server is misconfigured". The status code is 554 5.7.1, which as per [1] looks ok for the scenario. Firstly why should gmail call it misconfigured, but the bigger worry is it might blacklist the server if this happens too many times! Looking for help on this. Mayuresh [1] https://tools.ietf.org/html/rfc3463
Re: Mailing list manager on NetBSD
On Tue, Jan 15, 2019 at 03:38:35PM -0200, Silas wrote: > There are some caveats, like not managing bouncing (there are cases when the > student is not registered anymore and his e-mail account is cancelled, but > it is still on the list -- it is just an alias --, they doesn't seem to have > a fully automated solution!), but I can't remember of other big problems > they had with this approach. The is indeed an interesting solution. BTW postfix does have a way to redirect bounces from an alias to a specific id specified as `owner-aliasname' (man 5 postconf). > Mailing list managers insert headers and contents to make e-mails match with > (kind of) recent e-mail rules for mailing lists. This is specially > important if you are going to deliver to other SMTP servers. I am testing things on a really small list right now to discover these things. As yet the mails have not bounced. But may be I can study headers of mailing lists and do the same in mine. That is very easy with header_checks. > But since it is an internal solution, you see any other problems with the > approach of just using Postfix aliases? I am yet to solve one issue : how to restrict sending mails to a list to the members of the list only. postfix documentation [1] does give this scenario, but I find the documentation to be cryptic and things aren't yet working. Since your friend runs multiple lists, may be he has been able to put such restriction. Please do share if you can. Mayuresh [1] /usr/share/examples/postfix/RESTRICTION_CLASS_README
Re: Mailing list manager on NetBSD
On Wed, Jan 09, 2019 at 09:34:16PM -0600, ed...@pettijohn-web.com wrote: > man aliases > > list-name:include:/path/to/file/with/aliases > > look into allow_mail_to_commands and allow_mail_to_files for postfix > so that you can also pipe it to a script that saves it in some db. For a nearly static mailing list using postfix aliases is looking quite economical, presumably it would have the smallest footprint as MTA itself is doing the core job. For my second requirement of web archive I was looking for a good efficient (preferably C based) archiver preferably a separate component rather than a part of some mlm. These seem surprisingly few in number. I came across blists[1]. Integratable via procmail (or equivalent) as well as can be run as a cron job. It uses cgi to generate pages on the fly. It might make it conservative on space. But I think cgi would add to security hassles. I'd have preferred static htmls instead. Sample archives are here[2]. I am not particularly liking lack of threaded view (though thread links are present once you click on an item). But anyway, not finding too many alternatives that are efficient. Would appreciate comments and suggestions on postfix-blists combination for my requirements (i.e. mailing list with archives for a closed group of around 300, with web browsable archives and small resource footprint). I am considering adding blists it to pkgsrc-wip, after knowing the comments from the list. Mayuresh [1] https://github.com/aabc/blists [2] https://lists.openwall.net/linux-kernel/
Re: Mailing list manager on NetBSD
On Thu, Jan 10, 2019 at 02:21:53PM +0100, Julian H. Stacey wrote: > If there's a possibility a mail list server might have to later support > non tech users, avoid server software that don't support MIME. In order of importance I think a mailing list does the following: 1. bounce / forward the mails to a list of registered addresses. 2. Maintain web archives: MIME may be possibly an issue here. 3. Administration - such as subscribe / unsubscribe etc. MIME may matter here. If: For #1 I use postfix/procmail/equivalent (MTA/MDA) instead of an MLM. For #2 use a MIME aware archiver (not sure which, HyperKitty to name one that mailman 3 uses) While #3 does not matter much to me (will maintain list manually for a more or less static group) Then: Am I missing any MIME related issues? Mayuresh
Re: Mailing list manager on NetBSD
On Wed, Jan 09, 2019 at 09:34:16PM -0600, ed...@pettijohn-web.com wrote: > man aliases > > list-name:include:/path/to/file/with/aliases > > look into allow_mail_to_commands and allow_mail_to_files for postfix > so that you can also pipe it to a script that saves it in some db. Thanks. It seems more economical to do it postfix level. Just that procmail spec can be written without postfix/root privileges giving some comfort on that aspect. Would there be some downsides of doing it with procmail - besides cpu cycles? Mayuresh
Re: Mailing list manager on NetBSD
On Tue, Jan 08, 2019 at 09:45:16PM +0530, Mayuresh wrote:
> I am looking to set up a mailing list manager on a NetBSD server.
>
> The member count is more or less fixed between 300 to 350 and isn't going
> to grow beyond.
>
> The email archive should be browsable and searchable through a web
> interface. ("searchable" is less critical of the two requirements as even
> google search can be used to search through the archive.)
>
> I need the email storage to be in text format so as to be able to write
> tools of my own, on the server, to analyze the emails (say to grep
> patterns or even to do NLP).
After going through the nuances of some available solutions, I wonder:
- Do I really need a specialized mail manager software or can I just use
.forward (or procmail) to bounce the mails to registered members?
- I do not need automated subscribe / unsubscribe, this being for a closed
group of size not exceeding 300/350. Manual registration, with very
occasional changes is fine.
- I need "member-only" restriction for posting to the list email id, which
I think procmail can manage.
- I do need a web archiver with thread view etc. (and ability to write
text pattern searches of my own on the mail texts), for which there
might be alternatives that do just that - archiving. (E.g. HyperKitty
which mailman uses, which can be used standalone also.)
Would appreciate views on whether I am missing something, if I do not use
a proper mailing list software for above requirement.
Mayuresh
Re: Mailing list manager on NetBSD
On Wed, Jan 09, 2019 at 05:54:49PM +0100, Steffen Nurpmeso wrote: > when i used it. (There should be posts on their ML on that, > a couple of years back.) If your users use MIME you have to hook > in scripts, and then it becomes more expensive... Having said > that, AlpineLinux seems to use it for their MLs, and it seems to > work. But there all people use 7-bit clean text mails only. Plain text restriction is suitable (in fact better from storage point of view) for my purpose but can't "fix" everybody's mail client. Most people won't do that. So, yes, if I have to process (such as throw away MIME and retain only text) it will add up. BTW I am not too sure whether mlmmj's mailing list is active. 2018 is conspicuously absent in the archives[1]. (At least archives are not being produced, but how can it remain in that state.) I enquired about this on their list and hardly drew any response - except from 1 user who echoed similar concern. I have to assume their ML to be deeply dormant if not dead. Mayuresh [1] http://mlmmj.org/archive/mlmmj/
Re: Mailing list manager on NetBSD
On Wed, Jan 09, 2019 at 12:18:11PM -0500, Amitai Schleier wrote: > ezmlm is still for qmail, but with the current state of pkgsrc that oughtn't > be a huge constraint. I'd suggest mail/ezmlm-idx over mail/ezmlm to get more > features that are typically useful, and mail/qmail-run to integrate easily > into /etc/rc.conf. > > If I couldn't run qmail I'd look at mlmmj. In any case, it's probably time > that we bring it from wip into pkgsrc proper. ezmlm and idx look dormant with last release in 1997 and 2014 respectively. (As I said before that might just be psychological factor.) Besides I am all set on postfix and would have inertia to change that for the sake of a mailing list. wip/mlmmj has a strange TODO line: CVE-2009-4896 I wrote to the maintainer, but haven't got any response. If someone could clarify that TODO, I can request to move it to pkgsrc on the list or file a PR. Mayuresh
Re: Mailing list manager on NetBSD
On Wed, Jan 09, 2019 at 12:03:46PM +0100, Julian H. Stacey wrote: > But I gradually ran more public lists for non techs, including some > self admitted completely clueless & some other immeasurably lazy > users, many of whom cant think or refuse to think, love to argue, > & freak at command line etc, so the support load on unpaid volunteer > admin time became intolerable, & I was depserate for a list manager > with graphical clickey support to seperate myself from user support. > (Though mailman can be CLI driven too I recall) Thanks for a comprehensive reply. I am currently tending towards mlmmj due to the claims of smaller footprint as I'll be using a VPS to host this. Above para in your reply is something that really forewarns me about what I might almost certainly run into in a year or two ... I'd probably stick to a techy list for now and leave the rest to the future. Mayuresh
Re: Mailing list manager on NetBSD
On Wed, Jan 09, 2019 at 07:19:41AM +0530, Mayuresh wrote: > > I'm not sure if it is in pkgsrc but mlmmj is nice. > > Thanks. Seems to be a successor of ezmlm. But looks like it is not > constrained to work with qmail like ezmlm. > > Seems to be in pkgsrc-wip. For a production use I'd tend to have a little > bias towards pkgsrc. > > Still, are there things with mlmmj that you can mention as specific likes? > I'd even look at slimness and compact resource footprint as pluses. Some more search shows that simplicity and resource footprint are indeed the things people like about mlmmj. Given the liking of this community for such attributes, I wonder how mlmmj managed to lurk around in wip so long! IMO it should get an entry into pkgsrc. I noticed TODOs in wip and the release looks a bit old. Will try to work with the maintainers. But one of the major limitations, as far as my requirement, is its lack of web archive. There is no builtin feature though there are different projects (mentioned in mlmmj documentation[1]) that produce a web archive. Guess, that would need another wip project. Mayuresh [1] http://mlmmj.org/docs/readme-archives/
Re: Mailing list manager on NetBSD
On Tue, Jan 08, 2019 at 11:21:30AM -0600, Edgar Pettijohn wrote: > > Lastly it needs to be available in pkgsrc. > > > > I'm not sure if it is in pkgsrc but mlmmj is nice. Thanks. Seems to be a successor of ezmlm. But looks like it is not constrained to work with qmail like ezmlm. Seems to be in pkgsrc-wip. For a production use I'd tend to have a little bias towards pkgsrc. Still, are there things with mlmmj that you can mention as specific likes? I'd even look at slimness and compact resource footprint as pluses. Mayuresh
Re: Mailing list manager on NetBSD
On Tue, Jan 08, 2019 at 05:54:29PM -0500, Greg Troxel wrote: > Besides those, you may want to look at > https://en.wikipedia.org/wiki/Ezmlm > but that appears to have been released in 1997. Had come across Ezmlm, but it said it is "for qmail". Thought that might become a constraint. 1997 looks scary as well - I know that's not a particularly objective remark, may be just psychological. For same reason I am not considering majordomo, though as a user of several majordomo lists I never faced any problem. Mayuresh
Mailing list manager on NetBSD
I am looking to set up a mailing list manager on a NetBSD server.
The member count is more or less fixed between 300 to 350 and isn't going
to grow beyond.
The email archive should be browsable and searchable through a web
interface. ("searchable" is less critical of the two requirements as even
google search can be used to search through the archive.)
I need the email storage to be in text format so as to be able to write
tools of my own, on the server, to analyze the emails (say to grep
patterns or even to do NLP).
Lastly it needs to be available in pkgsrc.
I first looked up majordomo as that's the one NetBSD mailing lists use.
But looks like it is not an active project, with last release being in
2000.
Saw mailman and sympa to be talked about more and there is a nice
comparison here[1].
Would appreciate inputs on this.
An OT question: Does NetBSD mailing list prefer majordomo or it's a legacy
with no specific reason to change (or are there thoughts about changing
it?)
Mayuresh
[1] https://www.sympa.org/documentation/mailmanvssympa.html
Re: rcvar for locate.updatedb
On Sun, Nov 25, 2018 at 01:10:15AM +0700, Robert Elz wrote: > Provided it doesn't go on forever, and isn't just a rehash of > an earlier discussion, then discussing anything is never a > bad idea > > And it (kind of) seems to have already started. Wasn't aware of prior discussion on this. If a PR helps take a step towards logical conclusion (either way), I have filed one (yet to receive a number). Mayuresh
Re: rcvar for locate.updatedb
On Sat, Nov 24, 2018 at 12:54:35PM +0700, Robert Elz wrote: > Aside from the inertia criteria, which is probably really what > it is, I'd have thought a better test than "critical" would be > "probably useful to the majority of users - particularly the > less knowledgable". I'd think sshd to be useful to many times more of the users - knowledgable or otherwise - than updatedb! But sshd is required to be enabled. I think enabling explicitly is more transparent. There can be exceptions to this, but wonder whether updatedb is one! As you said rightly it's more about "that's how it has always been". But may not be a bad idea to discuss anyway. Mayuresh
Re: rcvar for locate.updatedb
On Sat, Nov 24, 2018 at 10:47:57AM +0700, Robert Elz wrote: > Whether it is important enough to run every week depends > upon your needs. That's why it can be configured. Thanks. The question is about the defaults. Except for critical services / jobs I thought a general guideline is to explicitly enable things. Hence the question, whether it is felt to be critical. Mayuresh
