[Novalug] [Opinion] The age of "Guilt-By-Knowledge" and how the anti-system crowd has "won"

Zachary Zebrowski zak.zebrowski@gmail.com
Wed Apr 29 21:13:17 EDT 2015


There are tools & instructions to make windows 8 like windows 7.  See:
http://www.makeuseof.com/tag/make-windows-8-8-1-look-like-windows-7-xp/

That make it at least makes windows ok.  I'm typing this on a windoh's 8.1
machine, because I'm locked in with the laptop I purchased.  It's ok
enough, and I have to use windoh's for the variuos microsoft apps, so why
they are not my favourite things, at least it works.  Of course, all my
servers at work and my servers at home (and toys at home) are linux
variants...  This is an el'cheapo ($300 ish) laptop that works.  Not great,
but works.

Zak

On Wed, Apr 29, 2015 at 9:08 PM, William Sutton via Novalug <
novalug@firemountain.net> wrote:

> And Microsoft does some of these things too.  Like Windows 8 and the tiles
> interface.  Or try logging out of Windows Server 2012.  I'll bet dollars to
> donuts that anyone who hasn't been shown the trick will either be hitting
> google or staring like a moron for the missing Start Menu button in the
> lower left corner of the screen.
>
> William Sutton
>
> On Wed, 29 Apr 2015, John Franklin wrote:
>
>  Shit.  if I'm agreeing with JIM and JIM agrees with Will, something is
>> wrong.
>>
>> - Grub2 is too damned complicated.  There is no earthly reason why it
>> should take a 300+ line file to boot a computer.
>>
>> - SELinux -- I understand this one pretty well, and I still turn it off
>> when it gets in the way.  Usually, I'm trying to get some service to start
>> on a development VM where the benefits of SELinux don't apply.
>>
>> - ip and ifconfig's deprecation I had heard about around the time the
>> iPhone debuted, and I thought I saw a mention of that in the ifconfig man
>> page, but I can't find it now.
>>
>> The objection we have with many of these technologies is that these new
>> and improved technologies break the way the work with our machines.  You
>> could call it "who moved my cheese,"  but the cheese isn't just on another
>> shelf -- it's gone.  Or it's been electrified or put inside a Chinese
>> puzzle box.
>>
>> Example 1: Losing ifconfig is a big deal.  This is a command we have in
>> our muscle memory.  Changing to use ip requires significant retraining,
>> which I'm sure is why ifconfig and ip have co-existed in systems for years
>> now.
>>
>> Example 2: Sometimes the changes are subtle.  Case in point, Apple Mail
>> introduced a change in Lion or Mountain Lion.  Previously, deleting a
>> message would move he cursor to the next message down.  Now, it moves the
>> cursor to the next message in the direction you were previously heading,
>> unless the message in one direction is new and the message in the other
>> direction is not, in which case it'll select the new message and change the
>> direction.  And there is no way to restore the old behavior.
>>
>> When I started using it, I was confused.  I hit delete twice because I
>> could already tell that the next message can be deleted, too -- except it
>> didn't delete the two I thought it would, it deleted a new message.  And
>> now it's deleting up.  WHY IS MY MAIL CLIENT RANDOMLY SELECTING THE NEXT
>> MESSAGE?!
>>
>> (You can bet Apple didn't tout that one on any marketing page.)
>>
>> Example 3: Gnome 3.  Features we use regularly and expect to be there are
>> gone, invisible, or buried.   Such as logging out.  Let's do an
>> experiment.  Get 100 KDE users who have never used Gnome 3 before, tell
>> them we're going to do some usability testing and that time will be a
>> factor.    Step 1: Login.  Step 2: Logout.  I'll bet coffee and donuts for
>> the LUG for a year over half of them take more than 2 minutes to figure it
>> out and at least 25% restart the machine to accomplish the task.
>>
>> For every "upgrade" we put in quotes, there are many we're happy to
>> have.  Bash completion, for example.  The ability to add some scripts so
>> bash can auto-compete command parameters based on context is huge.  "cd
>> b<tab>" won't suggest the file "bar", but will bring up the directory "bas"
>> instead, because cd doesn't work on files.  "git push o<tab>"  returns
>> "origin", even if you have a file or folder called "orangutan" in the
>> current directory, because the new smarter bash-completion understands that
>> git is looking for a remote repo name, not a local filename.
>>
>> The difference between bash completion and Gnome 3?  Bash completion
>> doesn't get in the way.  It doesn't become a puzzle we have to decode in
>> order to complete a task.
>>
>> jf
>>
>> On Apr 29, 2015, at 1:27 PM, "James Ewing Cottrell, III via Novalug" <
>> novalug@firemountain.net> wrote:
>>
>>  You know, William Sutton and I don't often agree, but I think he is Spot
>>> On here. I remember finally buckling down to learn Upstart one day, only to
>>> find out a week later that Systemd was now King. Gee, what was wrong with
>>> the System V file Structure in the First Place?!?
>>>
>>> Same with IP vs ifconfig. Of course, Route Assignment was something that
>>> also seemed to change and I had to Look Up all too often.
>>>
>>> SELinux? First off, it's NOT even Universal. There is Another thing
>>> called AppArmor which does the same thing. And I'll stick with Simple Root
>>> Privilege. I really don't fancy ACLs either, but I can Deal with them.
>>>
>>> Udev? WHY? And I could NEVER get anyone to explain, even Peter or Bryan,
>>> how to do the One Most Useful Thing I'd like. Simply RESCAN all the
>>> Devices, without me having to Reboot the Computer.
>>>
>>> Firewalld? Turn it off and use iptables natively.
>>>
>>> LVM? Well, that one is Quite Useful....but it's it's also Relatively
>>> Simple.
>>>
>>> Grub2? OMFG. Actually, I am finally wading thru the docs for that, but
>>> Why Is It So Ugly?!?
>>>
>>> And WTF is up with all the Thousand Options we have to set on the Kernel
>>> Command Lines these days?!?
>>>
>>> Actually, I have somewhat made my piece with systemd. One of the nice
>>> things it does is subsume chkconfig.
>>>
>>> But it all comes down to Future Shock. People are simply Throwing Change
>>> at us Too Fast and often Gratuitously.
>>>
>>> But Bryan, please don't confuse Yourself with the Technologies people
>>> complain about. Their is no Guilt By Knowledge; it's somewhat like Guilt by
>>> Association, but in your case it's Guilt By Identification. You don't have
>>> to feel Guilty just because you have Mastered new ideas, whether they are
>>> Good or Bad. This stuff isn't YOU.
>>>
>>> JIM
>>>
>>> P.S. Welcome back
>>>
>>> ----- Original Message -----
>>> From: "William Sutton via Novalug" <novalug@firemountain.net>
>>> To: "Ed James" <edward.james@gmail.com>
>>> Cc: "NoVaLUG@firemountain.net" <novalug@firemountain.net>
>>> Sent: Wednesday, April 29, 2015 8:39:15 AM
>>> Subject: Re: [Novalug] [Opinion] The age of "Guilt-By-Knowledge" and how
>>> the anti-system crowd has "won"
>>>
>>> I think sometimes there is a perception that the existing tools are "good
>>> enough", so why should we move to something else?  Sometimes that
>>> perception is right, and sometimes it is wrong.
>>>
>>> For example, when the ip utils came out, my perception was that someone
>>> didn't like the existing disorderliness and wanted to rearrange things
>>> for
>>> the sake of "elegance".  That opinion, held for several years, changed
>>> the
>>> first time I had to reassign the default route on a machine and found
>>> that
>>> the old tools were frankly horrible compared to the ip utils.
>>>
>>> On the other hand, I've long held a similar opinion about udev, and none
>>> of my experiences with udev have altered it.  In particular, when I
>>> upgraded my Gentoo system some years ago and ran into the UDEV-200
>>> issues,
>>> I was beside myself [1].
>>>
>>> Red Hat, which is the driving force behind systemd, has a reputation for
>>> pushing things at the expense of the community.  See, for example, the
>>> KDE/Gnome wars [2].  No less a luminary than Linus himself has written,
>>> at
>>> considerable length, about the things that are wrong with systemd.
>>>
>>> I don't know if systemd is "good" or "bad".  It doesn't particularly
>>> matter at the end of the day because, despite the martyred rant that
>>> started this thread, systemd /is/ being incorporated into the major
>>> distributions.  But if Linus thinks there are issues with it, maybe those
>>> should have been addressed before systemd was pushed out.
>>>
>>> Moreover, pushing it out despite the various concerns that were being
>>> raised tends to reinforce the idea that the people behind systemd care
>>> more about creating their own closed dependencies, rather than helping
>>> the
>>> community.  In a community that consists of very independent thinkers
>>> with
>>> a fierce independent streak, pushing something from the corporate top
>>> down to the masses is a surefire way to arouse resentment, distrust, and
>>> a
>>> knee-jerk rejection.
>>>
>>> Just my $0.02.
>>>
>>> William Sutton
>>>
>>> [1] UDEV-200 broke existing software RAID and network device
>>> configurations.  On the RAID side, it would assign md volumes in the
>>> order
>>> it detected them, so on one startup, it might assign a particular set of
>>> devices to boot, and the next, it might assign them to swap.  Depending
>>> on
>>> what they were configured for, the system might not boot.  Only by
>>> assigning a super-minor number and configuring the device UUID id
>>> mdadm.conf could you fix the problem.  On the network device side, they
>>> went ahead and renamed /existing/ eth* devices to ens*, with a different
>>> device sequence.  That didn't affect me, but there were plenty of
>>> anecdotes of production systems being used for firewall traffic that
>>> suddenly became unusable.  In the most extreme incidents, 8 different
>>> addresses were reassigned completely random ens* device names, with no
>>> clue left as to their original assignments.  You can imagine what that
>>> does to thousands of iptables rules....
>>>
>>> [2] For those not old enough to remember, back in the day we had fvwm,
>>> olwm, and the like, as window managers (not desktop environments).  Then
>>> Enlightenment came along.  It was slick and pretty, and a resource hog.
>>> Then the first version of KDE came into being.  It wasn't cool by today's
>>> standards, but it was pretty impressive at the time, and it was a
>>> community project.  But some people objected to its underlying toolkit
>>> being closed.  So Red Hat started and funded Gnome to be the anti-KDE.
>>> For a long time thereafter, it was a war to see which would win.  In
>>> general now, they co-exist--at the price of having /both/ sets of
>>> libraries installed on your system to run much of anything.  So now we
>>> all
>>> lose by having multiple sets of functionally similar toolkit dependencies
>>> hogging disk space.
>>>
>>> On Wed, 29 Apr 2015, Ed James via Novalug wrote:
>>>
>>>  Bryan,
>>>>
>>>> Methinks yer giving up too easy.  I understand where you're coming from,
>>>> but I've learned that "the loudest voice in the room" isn't always the
>>>> rightest - just the one that get noticed the most.
>>>>
>>>> My pet peeve is often the voice that tells me to dump my old gear and
>>>> buy
>>>> new stuff, when I'm asking a question on how to get something to work.
>>>> Quite often it takes a while, but eventually a voice will actually
>>>> answer
>>>> the original question, rather than offering an annoying and
>>>> counter-productive opinion.
>>>>
>>>> A similar response is "switch to the new software like all the kewl kids
>>>> are doing".  But the old stuff is rock-solid and just needs some sort of
>>>> minor mod.  Eventually I find the mod and move on. Sometimes the new
>>>> stuff
>>>> really is better, but (like all magic) comes at a price, and at a price
>>>> that I don't care to pay (time/money/aggravation/loss of other
>>>> functionality/wotever).  Other times, the new way works out better, and
>>>> while I may kick and scream for a bit, I'll spend the time learning the
>>>> why
>>>> and how of "the new way".
>>>>
>>>> I'm starting to ramble, but the bottom line is...
>>>>
>>>> I hate seeing the useful voices go silent.
>>>>
>>>> Ed James
>>>>
>>>>
>>>> On Wed, Apr 29, 2015 at 12:36 AM, Bryan J Smith via Novalug <
>>>> novalug@firemountain.net> wrote:
>>>>
>>>>  It may not last forever, but for the next few years, I want to
>>>>> congratulate the anti-systemd crowd.  You have "won."  You really
>>>>> have.  I had the epiphany this weekend after trying to help two (2)
>>>>> people, and that was that.
>>>>>
>>>>> And it goes to the greater point.  We're in a "new age" I will call
>>>>> "Guilt-by-Knowledge."
>>>>> ​...
>>>>> ​...
>>>>>
>>>>>
>>>>
>>>>  complaints about systemd.  People like me are the problem.
>>>>>
>>>>> Well done anti-systemd crowd.  Very well done.  You have "won."  It
>>>>> may not last, but it will for at least a few more years.  Very well
>>>>> done.
>>>>>
>>>>>
>>>>>  **********************************************************************
>>>> The Novalug mailing list is hosted by firemountain.net.
>>>>
>>>> To unsubscribe or change delivery options:
>>>> http://www.firemountain.net/mailman/listinfo/novalug
>>>>
>>> **********************************************************************
>>> The Novalug mailing list is hosted by firemountain.net.
>>>
>>> To unsubscribe or change delivery options:
>>> http://www.firemountain.net/mailman/listinfo/novalug
>>>
>>> **********************************************************************
>>> The Novalug mailing list is hosted by firemountain.net.
>>>
>>> To unsubscribe or change delivery options:
>>> http://www.firemountain.net/mailman/listinfo/novalug
>>>
>>
>> --
>> John Franklin
>> franklin@elfie.org
>>
>>
>>
>>
>>  **********************************************************************
> The Novalug mailing list is hosted by firemountain.net.
>
> To unsubscribe or change delivery options:
> http://www.firemountain.net/mailman/listinfo/novalug



More information about the Novalug mailing list