In defense of xinetd:

I can tell you, that anyone who wants to install a Nagios network
monitoring system will absolutely need to know xinetd.  The "nrpe" daemon
that gets installed on a Linux machine that's being monitored uses xinetd
for both daemon startup and access control.  So, until systemd takes over
the world, I would favor keeping xinetd in the exam objectives.


On Tue, Jul 15, 2014 at 9:05 AM, <[email protected]> wrote:

> Send lpi-examdev mailing list submissions to
>         [email protected]
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
> or, via email, send a message with subject or body 'help' to
>         [email protected]
>
> You can reach the person managing the list at
>         [email protected]
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of lpi-examdev digest..."
>
>
> Today's Topics:
>
>    1. Re:  Status of LPIC-1, LPIC-304 and Linux Essentials
>       Objectives (Marc Baudoin)
>    2. Re:  Status of LPIC-1,    LPIC-304 and Linux Essentials
>       Objectives (Anselm Lingnau)
>    3. Re:  Status of LPIC-1,    LPIC-304 and Linux Essentials
>       Objectives (G. Matthew Rice)
>    4. Re:  Status of LPIC-1,    LPIC-304 and Linux Essentials
>       Objectives (Anselm Lingnau)
>    5. Re:  Status of LPIC-1, LPIC-304 and Linux Essentials
>       Objectives (Alessandro Selli)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 15 Jul 2014 14:48:50 +0200
> From: Marc Baudoin <[email protected]>
> Subject: Re: [lpi-examdev] Status of LPIC-1, LPIC-304 and Linux
>         Essentials Objectives
> To: "This is the lpi-examdev mailing list." <[email protected]>
> Message-ID: <[email protected]>
> Content-Type: text/plain; charset=utf-8
>
> G. Matthew Rice <[email protected]> ?crit :
> > On Tue, Jul 15, 2014 at 5:34 AM, Marc Baudoin <[email protected]>
> wrote:
> > > So I'm going to break the silence.
> >
> > But on the day that we're supposed to go from draft to released?  I'm
> > gonna get you for this. ;)
>
> Was there a calendar? Sorry about that.  Remind me next time we
> meet so I can apologize with the correct amount of beers ;-)
>
> > > I've already said that a few years ago but the order of the
> > > objectives is not logical.
> >
> > Agreed.  But I'll let the other responses cover that.
>
> I give up on that.
>
> > > In 103.1, . and source would make more sense in 105.2.
> >
> > I think that it would better in 105.1 (ie. for sourcing bashrcs and
> > login/logouts).  And revisited in 105.2 (or left as an 'aha' moment).
> >
> > Opinions?  I've moved it to 105.1 until I hear otherwise from people.
>
> 105.1 is OK too (at least better than 103.1).
>
> > > Is exec
> > > really useful (if it is, I would also suggest to move it to
> > > 105.2)?
> >
> > Yes, it's useful.  Why have a shell process kicking around when you
> > really want something else to take over.
> >
> > Plus, I find using exec to redirect file descriptors very useful.
> > This isn't the best article/forum post on the matter but it
> > illustrates my point:
> >
> >
> http://unix.stackexchange.com/questions/18899/when-would-you-use-an-additional-file-descriptor
> >
> > I don't think that we cover this aspect of exec in the LPI exams, though.
> >
> > I think that you're right about it being in 105.2, though.  Doing an
> > exec on the command line isn't usually what someone wants.
>
> It could be a good joke.
>
> > > In 103.2, is the pr command (designed for matrix printers, which
> > > are seldom used today) really useful? I also doubt expand, fmt,
> > > join, paste, split and unexpand are heavily used.  On the other
> > > hand, more and less are useful filters, which are not listed
> > > here.
> >
> > hilarious.  We've missed less on that list for years.  I've added it.
>
> Thanks.
>
> > Personally, I use(d) pr.  Although, I'm more likely to use enscript
> > nowadays but I don't think that we should cover that.
>
> enscript or a2ps.  But I agree they're not relevant here.
>
> > And fmt is pretty handy when you want to make your paragraphs look
> > nice.  I use fmt in vi all the time I should probably make a map for
> > it.
>
> I use Vim and gqap (which does the same thing but without forking
> a process).
>
> > > In 103.3, xz should be added to the list of compression programs.
> >
> > Added.
>
> Thanks.
>
> > > The title for topic 102, "Linux Installation and Package
> > > Management", is not accurate.  Linux installation is not dealt
> > > with.
> >
> > Isn't 102.1 (designing hard disk layouts) and 102.2 (installing a boot
> > manager) part of doing a Linux install?
>
> Of course but 102.1 and 102.2, if I'm not mistaken, focus on the
> manual part of things.
>
> > > I still find it hard to deal with shared libraries with people
> > > with no knowledge of C programming.  Let's face it, most systems
> > > administrators nowadays know nothing about C and the POSIX API,
> > > which also makes their understanding of the operating system much
> > > more superficial.  Does the 102.3 subtopic, with only 1 question
> > > in the exam, needs to be maintained here? It would make much more
> > > sense in 201.
> >
> > I'd vote 'yes' to keeping it.  IMO, it's essential knowledge.  And
> > dropping it to move it to 201 would mean a 4 year window (our next
> > major update to LPIC-2) where it isn't covered at all.
> >
> > What about the use case where someone wants to download something new
> > and try it out in their user account?  Especially, something that
> > hasn't been packaged for them.
>
> So sad people don't use the source code anymore.  Everybody is
> boasting they're doing open source stuff but nobody uses the
> source anymore.
>
> But I see your point.
>
> > > Enough for 101, let's move to 102.
> > >
> > > In 105.1, /etc/bash.bashrc is missing in the awfully long list of
> > > bash configuration files
> >
> > Added.
>
> Thanks.
>
> > > In 106.1, should xhost still be covered given its coarse
> > > security?  Basic knowledge of xauth should replace it, knowing
> > > that ssh -X automatically does everything necessary.
> >
> > I think that most people would argue for less X in general for the
> > exam.  I have a soft spot for it, though.  Porting X11R5, Motif and
> > CDE to a new platform was my second job out of university (a long time
> > ago ;)).
>
> Don't tell.  That was when people used to compile X (see above).
> I also remember downloading X11R6 the day it got released and
> compiling it on a Sun SparcStation.  Sigh...
>
> > > 109.4 should immediately follow 109.2 and so
> > > /etc/resolv.conf, /etc/hosts and /etc/nsswitch.conf should not be
> > > covered in 109.2.
> >
> > hosts + nsswitch.conf (hosts: files ...) != DNS
>
> Well, it's not but it's pretty close.
>
> > > In 110.2, shouldn't inetd and xinetd be removed? I think they
> > > haven't been used by default on any distribution for quite some
> > > time now.  Standalone daemons are the default now (even if I
> > > still launch my UW IMAP from inetd, but that's me).
> >
> > It's still around and important stuff.  Plus, some people use LPIC-1
> > (particularly) as a Unix certification.
>
> BSD systems still come with inetd by default (I'm a huge user of
> NetBSD, I can tell for this one) but it's still rarely used
> because most daemons are standalone.
>
> > How about we wait until the next update.  By then systemd will have
> > subsumed everything
>
> Or not...
>
> http://boycottsystemd.org/
>
> > and we can decide on whether it should replace the (x)inetd
> > coverage? :-D
>
> Why not.
>
> > Thanks, Marc.  These were great comments.
>
> You're welcome.
>
> --
> Marc Baudoin
> Linagora formation - responsable p?dagogique
> http://formation.linagora.com/
>
> La pr?sente transmission contient des informations
> confidentielles appartenant ? Linagora, exclusivement destin?es
> au(x) destinataire(s) identifi?(s) ci-dessus. Si vous n'en faites
> pas partie, toute reproduction, distribution ou divulgation de
> tout ou partie des informations de cette transmission, ou toute
> action effectu?e sur la base de celles-ci vous sont formellement
> interdites. Si vous avez re?u cette transmission par erreur, nous
> vous remercions de nous en avertir et de la d?truire de votre
> syst?me d'information.
>
> The present transmission contains privileged and confidential
> information belonging to Linagora, exclusively intended for the
> recipient(s) thereabove identified. If you are not one of these
> aforementioned recipients, any reproduction, distribution,
> disclosure of said information in whole or in part, as well as
> any action undertaken on the basis of said information are
> strictly prohibited. If you received the present transmission by
> mistake, please inform us and destroy it from your messenging and
> information systems.
>
>
> ------------------------------
>
> Message: 2
> Date: Tue, 15 Jul 2014 14:54:59 +0200
> From: Anselm Lingnau <[email protected]>
> Subject: Re: [lpi-examdev] Status of LPIC-1,    LPIC-304 and Linux
>         Essentials Objectives
> To: [email protected]
> Message-ID: <2722154.aZKB2N1a5O@ceol>
> Content-Type: text/plain; charset="utf-8"
>
> Matt Rice wrote:
>
> > > In 103.1, . and source would make more sense in 105.2.
> >
> > I think that it would better in 105.1 (ie. for sourcing bashrcs and
> > login/logouts).  And revisited in 105.2 (or left as an 'aha' moment).
> >
> > Opinions?  I've moved it to 105.1 until I hear otherwise from people.
>
> OK ?
>
> > Yes, it's useful.  Why have a shell process kicking around when you
> > really want something else to take over.
>
> It is useful and takes about one minute to explain (two with a
> demonstration).
> Let's keep it in.
>
> > Personally, I use(d) pr.  Although, I'm more likely to use enscript
> > nowadays but I don't think that we should cover that.  I'm okay with
> > dropping it if we can get a 'second' from someone on this nomination.
>
> I usually spend some time (a few minutes) on pr(1), not because it is a
> tool I
> think everybody should use all the time, but because it is a good
> illustration
> for the sort of thing one can accomplish with pipelines. It does come in
> handy
> every so often, though.
>
> > Also, personally, I use expand on other people's source code quite
> > often.  Mixed spaces/tabs are really annoying when people don't use
> > the right tabstops. ;)
>
> Seconded.
>
> > And fmt is pretty handy when you want to make your paragraphs look
> > nice.  I use fmt in vi all the time I should probably make a map for
> > it.
>
> Seconded.
>
> > > The title for topic 102, "Linux Installation and Package
> > > Management", is not accurate.  Linux installation is not dealt
> > > with.
> >
> > Isn't 102.1 (designing hard disk layouts) and 102.2 (installing a boot
> > manager) part of doing a Linux install?
> >
> > Arguable, so is managing shared libraries when you're installing
> > (additional) software.
>
> We haven't covered Linux installation in the strict sense in our classes
> for a
> very long time, because (a) how to do it is distribution-specific, and (b)
> it
> isn't exactly rocket science ? if you pop in the CD (or USB stick) and keep
> hitting RETURN you will usually end up with a reasonable Linux system.
>
> I agree that hard disk layout is something worth covering, simply because
> it
> is easy for people to shoot themselves in the foot. I don't know who still
> does detailed boot manager configuration these days (other than people
> compiling their own kernels, which would suggest that the topic might
> eventually move to LPIC-2).
>
> > [Shared libraries]
> > I'd vote 'yes' to keeping it.  IMO, it's essential knowledge.  And
> > dropping it to move it to 201 would mean a 4 year window (our next
> > major update to LPIC-2) where it isn't covered at all.
> >
> > What about the use case where someone wants to download something new
> > and try it out in their user account?  Especially, something that
> > hasn't been packaged for them.
>
> Yes, because it takes all of a five-line C program to demonstrate most of
> what
> there is to demonstrate about static vs. shared libraries. If people can't
> sort-of grok a five-line C program if it is explained to them, even if they
> haven't seen C before, they have no business doing LPI exams. It's not as
> if
> we're asking them to write configure.in files.
>
> > [bash vs. zsh]
> > Agreed.  At least for command line use.  I still use bash for
> > scripting, though.  Least common denominator and all.  Actually, I
> > just switched from #!/bin/sh to #!/bin/bash a year or two ago for all
> > script writing.  I've had it with having to remember when I started using
> > bashisms in my scripts.
>
> We explain /bin/bash vs. /bin/sh and the caveats involved. We mention the
> existence of other shells but don't go into detail.
>
> > I think that most people would argue for less X in general for the
> > exam.  I have a soft spot for it, though.  Porting X11R5, Motif and
> > CDE to a new platform was my second job out of university (a long time
> > ago ;)).
>
> X11 used to be a huge problem back when one would have to figure out one's
> monitor timings in order not to fry one's display. In previous reviews of
> the
> exam, the level of X knowledge required has been downgraded again and
> again,
> and as far as I am concerned that trend should continue. For example, I'd
> kick
> out all the display manager configuration details in favour of more
> high-level
> knowledge (i.e., what is Wayland and why is it interesting?).
>
> > > In 107.1, vipw is dearly missing...
> >
> > I believe that vipw (and vigr) were in the objectives but dropped a
> > few years ago to keep vi from creeping into the 102 exam.  Although,
> > the EDITOR environment variable will fix that problem for the
> > vi-challenged.
> >
> > Isn't this where usermod/groupmod come in, though?
>
> Imagine you're rearranging the /home layout for your 200-user file server.
> Personally I'd prefer using ?:s///? in vipw to 200 calls to usermod (even
> in a
> shell loop). The latter is still an improvement if you'd otherwise have to
> use
> a GUI, but even so ?
>
> (vipw/vigr have been part of our training materials from the very
> beginning.)
>
> > We're getting into distro specifics here (albeit, prevalent ones).
> > I'm not against adding them at an "awareness" level but I don't think
> > that we need to cover the innards.
> >
> > Opinions anyone?
>
> In our training materials we go over the network configuration files for
> the
> most important Linux distributions. I'm not sure detailed knowledge of
> their
> content belongs in the exam, though.
>
> > [inetd/xinetd]
> > How about we wait until the next update.  By then systemd will have
> > subsumed everything and we can decide on whether it should replace the
> > (x)inetd coverage? :-D
>
> Agreed.
>
> Anselm
> --
> Anselm Lingnau ... Linup Front GmbH ... Linux-, Open-Source- &
> Netz-Schulungen
> [email protected], +49(0)6151-9067-103, Fax -299,
> www.linupfront.de
> Linup Front GmbH, Postfach 100121, 64201 Darmstadt, Germany
> Sitz: Weiterstadt (AG Darmstadt, HRB7705), Gesch?ftsf?hrer: Oliver Michel
>
>
> ------------------------------
>
> Message: 3
> Date: Tue, 15 Jul 2014 09:01:01 -0400
> From: "G. Matthew Rice" <[email protected]>
> Subject: Re: [lpi-examdev] Status of LPIC-1,    LPIC-304 and Linux
>         Essentials Objectives
> To: "This is the lpi-examdev mailing list." <[email protected]>
> Message-ID:
>         <CAFKQHHPSbnEJ3jZT7Oiybgi3PjLoS7UGQUq=
> [email protected]>
> Content-Type: text/plain; charset=UTF-8
>
> On Tue, Jul 15, 2014 at 8:48 AM, Marc Baudoin <[email protected]>
> wrote:
> >> Isn't 102.1 (designing hard disk layouts) and 102.2 (installing a boot
> >> manager) part of doing a Linux install?
> >
> > Of course but 102.1 and 102.2, if I'm not mistaken, focus on the
> > manual part of things.
>
> Sure but the concepts are important for doing installs.  And full
> blown installs are beyond the scope (and too distro specific) to
> cover.  So, we cover some important parts of doing an install instead.
>
>
> >> How about we wait until the next update.  By then systemd will have
> >> subsumed everything
> >
> > Or not...
> >
> > http://boycottsystemd.org/
>
> I don't have a strong opinion either way but this sure seems to be
> swimming against the tide.
>
> Regards,
> --matt
>
> --
> G. Matthew Rice <[email protected]>                         gpg id: EF9AAD20
>
>
> ------------------------------
>
> Message: 4
> Date: Tue, 15 Jul 2014 15:13:30 +0200
> From: Anselm Lingnau <[email protected]>
> Subject: Re: [lpi-examdev] Status of LPIC-1,    LPIC-304 and Linux
>         Essentials Objectives
> To: [email protected]
> Message-ID: <8080906.X9WByDppmf@ceol>
> Content-Type: text/plain; charset="utf-8"
>
> Matt Rice wrote:
>
> > > http://boycottsystemd.org/
> >
> > I don't have a strong opinion either way but this sure seems to be
> > swimming against the tide.
>
> These people are circling the wagons but the caravan has long since moved
> on.
>
> Most of their objections are easily debunked, anyway, simply by studying
> the
> systemd docs. Three years from now nobody will remember any of this.
> Besides,
> these objections (overengineering, gratuitous new interfaces, blasphemy
> against the Unix philosophy, ?) are exactly the ones that people brought
> forward when SysV init was new in the 1980s ? and now SysV init is
> supposed to
> be the best thing since sliced bread?
>
> Anselm
> --
> Anselm Lingnau ... Linup Front GmbH ... Linux-, Open-Source- &
> Netz-Schulungen
> [email protected], +49(0)6151-9067-103, Fax -299,
> www.linupfront.de
> Linup Front GmbH, Postfach 100121, 64201 Darmstadt, Germany
> Sitz: Weiterstadt (AG Darmstadt, HRB7705), Gesch?ftsf?hrer: Oliver Michel
>
>
> ------------------------------
>
> Message: 5
> Date: Tue, 15 Jul 2014 15:12:42 +0200
> From: Alessandro Selli <[email protected]>
> Subject: Re: [lpi-examdev] Status of LPIC-1, LPIC-304 and Linux
>         Essentials Objectives
> To: "This is the lpi-examdev mailing list." <[email protected]>
> Message-ID: <[email protected]>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> G. Matthew Rice wrote:
> > On Tue, Jul 15, 2014 at 5:34 AM, Marc Baudoin<[email protected]>
>  wrote:
> [...]
>
>
> >> In 103.1, . and source would make more sense in 105.2.
> > I think that it would better in 105.1 (ie. for sourcing bashrcs and
> > login/logouts).  And revisited in 105.2 (or left as an 'aha' moment).
> >
> > Opinions?  I've moved it to 105.1 until I hear otherwise from people.
>
>    I agree.
>
> >> Is exec
> >> really useful (if it is, I would also suggest to move it to
> >> 105.2)?
> > Yes, it's useful.  Why have a shell process kicking around when you
> > really want something else to take over.
> >
> > Plus, I find using exec to redirect file descriptors very useful.
> > This isn't the best article/forum post on the matter but it
> > illustrates my point:
> >
> >
> http://unix.stackexchange.com/questions/18899/when-would-you-use-an-additional-file-descriptor
> >
> > I don't think that we cover this aspect of exec in the LPI exams, though.
> >
> > I think that you're right about it being in 105.2, though.  Doing an
> > exec on the command line isn't usually what someone wants.
>
>    I usually use it when I run su on remote terminals, to make sure I
> won't leave around an idle network login when I exit just once.
>
> [...]
>
> >> I still find it hard to deal with shared libraries with people
> >> with no knowledge of C programming.  Let's face it, most systems
> >> administrators nowadays know nothing about C and the POSIX API,
> >> which also makes their understanding of the operating system much
> >> more superficial.  Does the 102.3 subtopic, with only 1 question
> >> in the exam, needs to be maintained here? It would make much more
> >> sense in 201.
> > I'd vote 'yes' to keeping it.  IMO, it's essential knowledge.
>
>    I think the same way.
>
> [...]
>
> >> In 109.2, configuration files (/etc/network/interfaces,
> >> /etc/sysconfig/network and
> >> /etc/sysconfig/network-scripts/ifcfg-*) should be covered (they
> >> are used by ifup, which is in the objectives).
> > We're getting into distro specifics here (albeit, prevalent ones).
> > I'm not against adding them at an "awareness" level but I don't think
> > that we need to cover the innards.
> >
> > Opinions anyone?
>
>    To me it looks like the same situation as per the package install
> commands: rpm and yum vs. dpkg* and apt*.  On one side, they are
> distribution specific, but not to a single distribution, rather to a
> family of like, widespread distributions.  If we do cover
> /etc/sysconfig/network* RH stuff, we should also cover /etc/network/* DB
> stuff (I actually think covering just /etc/network/interfaces suffices).
>
> >> 109.4 should immediately follow 109.2 and so
> >> /etc/resolv.conf, /etc/hosts and /etc/nsswitch.conf should not be
> >> covered in 109.2.
> > hosts + nsswitch.conf (hosts: files ...) != DNS
> >
> > I did drop resolv.conf from 109.2, though.    It is DNS.
>
> 109.1 "Fundamentals of internet protocols" lists port 53 (domain) in the
> "Knowledge about common TCP and UDP ports" section, though.
>
> [...]
>
> > If anyone made it this far, let me know if you have a problem with any
> > of the changes.
>
>    No problem.  The proposed changes sound reasonable.
>
>
>    Alessandro
>
>
>
> ------------------------------
>
> _______________________________________________
> lpi-examdev mailing list
> [email protected]
> http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
>
> End of lpi-examdev Digest, Vol 81, Issue 5
> ******************************************
>
_______________________________________________
lpi-examdev mailing list
[email protected]
http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev

Reply via email to