Re: We didn't touched anything... really...
Paul Tomblin
>I wonder: you have a setup that works, some moron tech changes this-and-that
>and suddendly nothing works anymore. Wouldn't you think that *something*
>got screwed in what he did? No, of course not. The problem must be in *my*
>end of things.
Are you kidding? I got the same shit from a developer at my company who
changed some of his code, something stopped working, and so he immediately
demanded that I drop what I'm doing so I can find what's wrong in my code,
which hasn't changed in two years!
--
Paul Tomblin <ptom...@xcski.com> http://xcski.com/blogs/pt/
Unix is simple, but it takes a genius to understand the simplicity.
-- Dennis Ritchie
J.D. Baldwin
In the previous article, Paul Tomblin <ptomblin...@xcski.com> wrote,
quoting davidey...@onlyforfun.net:
> >I wonder: you have a setup that works, some moron tech changes
> >this-and-that and suddendly nothing works anymore. Wouldn't you think
> >that *something* got screwed in what he did? No, of course not. The
> >problem must be in *my* end of things.
>
> Are you kidding? I got the same shit from a developer at my company
> who changed some of his code, something stopped working, and so he
> immediately demanded that I drop what I'm doing so I can find what's
> wrong in my code, which hasn't changed in two years!
And, then, on the other side of the moron coin, we have the manager
(usually a manager, at least 'round here) who has slavishly memorized
the "What changed?" method of troubleshooting, and Absolutely. Nothing.
Else.
"The Boobookistan site isn't routing to Lesser Angora."
"Well, what changed recently?"
"Uh ... we pushed that new rulebase to the firewall in East Tremolo
... but that doesn't have anything to do with the routing in question,
it's probably a hardw--"
"Roll back the firewall change!"
--
_+_ From the catapult of |If anyone disagrees with any statement I make, I
_|70|___:)=}- J.D. Baldwin |am quite prepared not only to retract it, but also
\ / bal...@panix.com|to deny under oath that I ever made it. -T. Lehrer
***~~~~-----------------------------------------------------------------------
News
> I wonder: you have a setup that works, some moron tech changes this-and-that
> and suddendly nothing works anymore. Wouldn't you think that *something*
> got screwed in what he did? No, of course not. The problem must be in *my*
> end of things.
Luser Rule 1: It's always the fault of the firewall.
Luser Rule 2: Never admit that you fsck'ed with it.
Luser Rule 3: There is no difference between a $600 PC and a $6000 server.
Luser Rule 4: Technical excellence is determined by the installed base, not
by the qualities, or lack thereof, of the product.
Luser Rule 5: A plug will always fit a socket... if you push hard enough.
Luser Rule 6: There is no rule 6.
Luser Rule 7: Any switch port in a packet storm.
--
It's summer in the UK so we just had three very hot days and today the
thunderstorms came. It's cloudy, hot and humid. It's hard to think.
Tech Mav
> Are you kidding? I got the same shit from a developer at my company who
> changed some of his code, something stopped working, and so he immediately
> demanded that I drop what I'm doing so I can find what's wrong in my code,
> which hasn't changed in two years!
>
>
problem, and was convinced it was the scripting. Never mind the fact
that the scripting hasn't changed in at least a year, and that it works
for everyone on the box except for one customer. $SeniorAdmin trying to
explain to him that the script wasn't the issue fell on deaf ears, due
to the fact that $NewAdmin used to be a programmer, and knew more than we
did. No, he wasn't a programmer here, but he still knew more about the
script than we did.
Oddly enough, it turned out the problem was the customer inputting bad
data. Imagine that.
--
TechMav AKA The Guy In The Funny Black Hat
When the FBI/CIA/NSA/FDA/and other three-letter government agencies come
looking, you don't know me, you never saw me, never heard of me. get it?
got it? good!
J.D. Baldwin
In the previous article, News <ne...@buffy.sighup.org.uk> wrote:
> Luser Rule 3: There is no difference between a $600 PC and a $6000 server.
Of course there is. $5,400.
What do I win?
Kenneth Brody
>
> In a previous article, davidey...@onlyforfun.net said:
> >I wonder: you have a setup that works, some moron tech changes this-and-that
> >and suddendly nothing works anymore. Wouldn't you think that *something*
> >got screwed in what he did? No, of course not. The problem must be in *my*
> >end of things.
>
> Are you kidding? I got the same shit from a developer at my company who
> changed some of his code, something stopped working, and so he immediately
> demanded that I drop what I'm doing so I can find what's wrong in my code,
> which hasn't changed in two years!
Then there are the cases of someone getting a new version of our
software, and then complaining about how a "change in behavior" on
our end has broken a program that they wrote. I, being the head of
the development team, know better, and I know that that particular
functionality hasn't changed, and what they claim has been "working
for years" could never have worked.
It's amazing[1] how many times that I've challenged them to demonstrate
their code "working" on any previous release of our software, only to
have them come back with "I know it used to work, but I can't get it
to work on our previous copy".
[1] Well, not to any ASR regular.
--
+-------------------------+--------------------+-----------------------+
| Kenneth J. Brody | www.hvcomputer.com | #include |
| kenbrody/at\spamcop.net | www.fptech.com | <std_disclaimer.h> |
+-------------------------+--------------------+-----------------------+
Don't e-mail me at: <mailto:ThisIsA...@gmail.com>
David Cameron Staples
> You need to manage your manager better, filter his requests through a
> sanity filter (to retain yours, not his) so that the conversation gets
> parsed like this:
>
> "The Boobookistan site isn't routing to Lesser Angora."
>
> "Well, what changed recently?"
>
> it's probably a hardware fault."
>
> "Ok find what's broken and get it fixed then!"
>
> Yes, I know it would never work in the real world, but we can dream can't
> we.
If *you* asked the question "what changed" in that context, and got the
answer "Uh ... nothing," would *you* believe it?
Besides, it's just as likely that that firewall in East Tremolo was
actually passing some magic packets back and forth to some third system,
and the upgrade turned off that port/protocol, and now all three systems
have thrown a hissy-fit, and...
--
David Cameron Staples | staples AT cs DOT mu DOT oz DOT au
Melbourne University | Computer Science | Technical Services
Is someone here not having a good time?
J.D. Baldwin
In the previous article, Cameron Biggart <news...@yaffa.com.au>
wrote, quoting me:
> > And, then, on the other side of the moron coin, we have the manager
> > (usually a manager, at least 'round here) who has slavishly memorized
> > the "What changed?" method of troubleshooting, and Absolutely. Nothing.
> > Else.
>
> You need to manage your manager better,
*My* manager isn't the problem. *My* manager lives in California and
has a more or less proper attitude toward work, its role in a larger
life, and a good sense of both his gifts and limitations.
No, it's those jackasses -- strike that, an actual jackass has *some*
kind of functioning brain -- morons in New Jersey and New York who
conspire to make my life hell. Not only are they incompetent in ways
that would make the knotted and combined locks of even this august
group's members to part, and each particular hair to stand on end like
quills upon the fretful porcupine ... they all, to a one, *think* they
are God's gift to the Science and Art of Information Processing.
J.D. Baldwin
In the previous article, David Cameron Staples
<sta...@cs.mu.OZ.AU.spam> wrote:
> If *you* asked the question "what changed" in that context, and got
> the answer "Uh ... nothing," would *you* believe it?
>
> Besides, it's just as likely that that firewall in East Tremolo was
> actually passing some magic packets back and forth to some third
> system, and the upgrade turned off that port/protocol, and now all
> three systems have thrown a hissy-fit, and...
For the record -- and I'm sure this is already clear to 99% of you,
including DCS -- I have no objection to considering and discussing all
possibilities, even those that appear at first glance to be absurdly
unlikely. I object to focusing on "the most recent change" as a
substitute for thinking. Unfortunately, nothing sells better in this
culture than a substitute for thinking.
Joe Zeff
INVALID...@example.com.invalid (J.D. Baldwin) wrote:
>I object to focusing on "the most recent change" as a
>substitute for thinking.
"It was working before, it's not working now, therefore the most
recent change is the most likely cause." It's good logic, and true
often enough to be worth using. Of course, it's not *always* true,
and insisting that the most recent change *must* be the cause is
luserish.
--
Joe Zeff
The Guy With the Sideburns
The only problem with troubleshooting is that
trouble sometimes shoots back.
http://www.lasfs.org http://home.earthlink.net/~sidebrnz
Menno WIllemse
<davidey...@onlyforfun.net> wrote:
>
> So it turned out that the 'isp tech' is really the boyfriend of the
> daughter of a friend of the cousin of someone that work (every now and then)
> for $boss...
On a completely unrelated matter: We seem to be having a spate of
those ads recently that say: "So your job sucks? How about you come on
one of our four-week courses, and we guarantee you an exciting and
fulfilling job in Eye Tee afterwards!"
Outlook not so good.
Cheers,
Menno
Kenneth Brody
>
> On Tue, 13 Jun 2006 14:32:44 +0000 (UTC),
> INVALID...@example.com.invalid (J.D. Baldwin) wrote:
>
> >I object to focusing on "the most recent change" as a
> >substitute for thinking.
>
> "It was working before, it's not working now, therefore the most
> recent change is the most likely cause." It's good logic, and true
> often enough to be worth using. Of course, it's not *always* true,
> and insisting that the most recent change *must* be the cause is
> luserish.
Quite often, the most recent change merely exposed a problem that
was caused by some change long ago. So, technically, undoing the
most recent change will "fix" the problem in the sense that it
will go back to being hidden.
Which, of course, is impossible to explain to a PHB after the latest
change is undone and it starts "working" again.
J.D. Baldwin
In the previous article, Joe Zeff
<the.guy.with....@lasfs.info> wrote:
> >I object to focusing on "the most recent change" as a
> >substitute for thinking.
>
> "It was working before, it's not working now, therefore the most
> recent change is the most likely cause." It's good logic, and true
> often enough to be worth using.
True enough if the change is reasonably coincident in time with the
failure. Consideration of the time domain apart from "most recent" is
not usually a feature of the "reasoning" I encounter in these cases.
Chris Suslowicz
INVALID...@example.com.invalid (J.D. Baldwin) wrote:
>In the previous article, Joe Zeff
><the.guy.with....@lasfs.info> wrote:
>> >I object to focusing on "the most recent change" as a
>> >substitute for thinking.
>>
>> "It was working before, it's not working now, therefore the most
>> recent change is the most likely cause." It's good logic, and true
>> often enough to be worth using.
>
>True enough if the change is reasonably coincident in time with the
>failure. Consideration of the time domain apart from "most recent" is
>not usually a feature of the "reasoning" I encounter in these cases.
Luser failing to test something after the software upgrade $Mumble
years ago, then asking why KBNS doesn't work....
Gah!
Chris. (For values of $Mumble >3 years, of course.)
--
Cheese will not solve *all* the world's problems...
-- Matt Martin
David Cameron Staples
> Joe Zeff wrote:
>>
>> On Tue, 13 Jun 2006 14:32:44 +0000 (UTC),
>> INVALID...@example.com.invalid (J.D. Baldwin) wrote:
>>
>> >I object to focusing on "the most recent change" as a substitute for
>> >thinking.
>>
>> "It was working before, it's not working now, therefore the most recent
>> change is the most likely cause." It's good logic, and true often
>> enough to be worth using. Of course, it's not *always* true, and
>> insisting that the most recent change *must* be the cause is luserish.
>
> Quite often, the most recent change merely exposed a problem that was
> caused by some change long ago. So, technically, undoing the most recent
> change will "fix" the problem in the sense that it will go back to being
> hidden.
>
> Which, of course, is impossible to explain to a PHB after the latest
> change is undone and it starts "working" again.
Which is, if you think about it, what my hypothetical third system story
was about.
--
David Cameron Staples | staples AT cs DOT mu DOT oz DOT au
Melbourne University | Computer Science | Technical Services
guy just walks in and asks me " i bought windows xp 64 and it only shows
16bit and 32bit color but no 64bit ..." -- bash.org/?593954
Paul Colquhoun
| On Tue, 13 Jun 2006 14:32:44 +0000 (UTC),
| INVALID...@example.com.invalid (J.D. Baldwin) wrote:
|
|>I object to focusing on "the most recent change" as a
|>substitute for thinking.
|
| "It was working before, it's not working now, therefore the most
| recent change is the most likely cause." It's good logic, and true
| often enough to be worth using. Of course, it's not *always* true,
| and insisting that the most recent change *must* be the cause is
| luserish.
But "most recent change" *is* always the cause, if you include equipment
failure, act of god, etc, as a "change".
Always blaming the most recent *planned*, *deliberate* change is the
problem.
(Not saying a planned and deliberate change can't be the cause...)
--
Reverend Paul Colquhoun, ULC. http://andor.dropbear.id.au/~paulcol
Asking for technical help in newsgroups? Read this first:
http://catb.org/~esr/faqs/smart-questions.html#intro
Dave Brown
> I object to focusing on "the most recent change" as a substitute
> for thinking. Unfortunately, nothing sells better in this
> culture than a substitute for thinking.
I once had a project manager like that. Whenever anything went
wrong, he would inevitably attribute it to the most recent change,
and certainly not because of the astounding feculence of the
project itself or anything like that.
After a spate of problems, all in his mind caused by recent
changes, he demanded to be notified of every change that took
place, as soon as it happened.
By the end of his week, he was *begging* to be taken off the CVS
notification mailing list (his exact words, which I cherish to
this day: "Dave, please! Take me off that God-damned spam list!").
--Dave
--
<wankel> every unix on the planet has the same route(1M) syntax
<wankel> but noooo, linux has to be different
<wankel> hell, *NT* has normal route syntax
Kenneth Brody
[...]
> But "most recent change" *is* always the cause, if you include equipment
> failure, act of god, etc, as a "change".
Hmm...
Most recent change: yanked power cord out of back of unit.
Second-most recent change: magic smoke released from unit.
Scott Packard
>
> In the previous article, Cameron Biggart <news...@yaffa.com.au>
> wrote, quoting me:
>> > And, then, on the other side of the moron coin, we have the manager
>> > (usually a manager, at least 'round here) who has slavishly memorized
>> > the "What changed?" method of troubleshooting, and Absolutely. Nothing.
>> > Else.
>> [...]
>>
>> You need to manage your manager better,
>
> *My* manager isn't the problem. *My* manager lives in California and
> has a more or less proper attitude toward work, its role in a larger
> life, and a good sense of both his gifts and limitations.
>
> No, it's those jackasses -- strike that, an actual jackass has *some*
> kind of functioning brain -- morons in New Jersey and New York who
> conspire to make my life hell. Not only are they incompetent in ways
> that would make the knotted and combined locks of even this august
> group's members to part, and each particular hair to stand on end like
> quills upon the fretful porcupine ... they all, to a one, *think* they
> are God's gift to the Science and Art of Information Processing.
First rule of work in those states is Dress The Part.[1]
I bet they look like real IT experts.
[1] Watched rerun of Boiler Room recently.