Stories
Slash Boxes
Comments
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

use Perl Log In

Log In

[ Create a new account ]

barbie (2653)

barbie
  reversethis-{ku. ... m} {ta} {eibrab}
http://barbie.missbarbell.co.uk/

Leader of Birmingham.pm [pm.org] and a CPAN author [cpan.org]. Co-organised YAPC::Europe in 2006 and the 2009 QA Hackathon, responsible for the YAPC Conference Surveys [yapc-surveys.org] and the QA Hackathon [qa-hackathon.org] websites. Also the current caretaker for the CPAN Testers websites and data stores.

If you really want to find out more, buy me a Guinness ;)

Links:
Memoirs of a Roadie [missbarbell.co.uk]
[pm.org]
CPAN Testers Reports [cpantesters.org]
YAPC Conference Surveys [yapc-surveys.org]
QA Hackathon [qa-hackathon.org]

Journal of barbie (2653)

Monday February 02, 2004
11:28 AM

Linux Non-Certification

[ #17150 ]
Last week my time was spent on a Linux Administrators training course, learning about the kernel, processes and services, neworking and other good stuff. I ended up getting alot out of the actual course, if for nothing else but to understand about compiling a kernel and adding to the boot list. For my wireless card I need to patch the kernel, so it was useful to be taught how to do it right.

However, on Friday we had an "exam". It lasted 90 minutes and dispite all the cloak and dagger aspects of a professional exam, it was a limp wristed effort. To qualify, my own performance was not up to scratch and there were some simple questions that I just had brain-fade with. However, of the 8 people who sat the exam, only 2 passed and both just barely.

So was it a hard exam? Most definitely not. It was however, full of ambiguous, woolly and downright confusing questions. Bare in mind that this is an Linux Administrators exam, so what does knowing who writes the Linux HOWTOs enable you to run a cron job? The worst possible aspect of the "exam" was on several questions they wanted a definitive answer to a question that could result in multiple answers, such as in the minutes field of a crontab how do you write 'every 2 minutes'? I wrote both answers and judging from the scorecard I got marked down! Another was asking for a specific result of a theoretical logging application (I assumed it referred to syslog, but who's to say it wasn't something else). Having to second guess examiners should not be part of any exam. What is the "normal exit value of a process"? Without qualifying that, any documented success or error code can be deemed normal.

The other examinees all thought I was doing another exam, seeing as I was writing so much, until they realised I filled in the 'Comments' box for virtually every question. Afterwards I was asked whether I would take the exam again. If the questions are going to be as flakey as this exam, I think not.

I've been using Unix for nearly 18 years, it's not like I'm a novice or anything! While I didn't expect to walk it or anything, I was expecting something closer to the rehearsal test (77%).

With the talk of Perl Certification surfacing again last year, I REALLY hope they can write a decent exam.

The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More | Login | Reply
Loading... please wait.
  • > What is the "normal exit value of a process"? Without qualifying that, any documented success or error code can be deemed normal.

    Wrong.
    • My guess was 0, though EXIT_SUCCESS would also be correct, right?

      I'm assuming you saw "process" and thought "function", which can indeed vary wildly.

      • > My guess was 0, though EXIT_SUCCESS would also be correct, right?

        Correct. There are many possible failures but only one possible success: EXIT_SUCCESS from , and in Unix/Linux EXIT_SUCCESS is defined to be zero. Easily testable in shell by:

        ls /none/such; echo $?
        ls /; echo $?

        • Except it didn't ask about a SUCCESS exit status, it asks for a NORMAL exit status. All documented failure and success return codes can be deemed normal. In this particular question even if you said EXIT_SUCCESS you would have been wrong.
        • What is the normal exit status of false?

          See, it's not so simple...
      • > I'm assuming you saw "process" and thought "function"

        Err... nope. Processes and functions are vastly different. A function can have any return value it feels like, whereas a process should have a known set of exit conditions and values.

        I've been writing both for quite sometime ;)

    • Why is that wrong? If a process fails with a 1 in a set of known conditions, why is that abnormal? The only abnormal exit values are ones returned from unknown exit conditions, such as a core dump.
  • Note that I'm not trying to defend the (obviously faulty) Linux test.

    All I'm saying I see absolutely no uncertainty in the question "normal exit status of a process". It's zero. You can do all the mincing of the words of "normal is different from success" or of "but what if this particular application has defined something else as 'normal'" but that does not change the agreed-upon, documented, and standardized semantics of a zero exit status, not one iota.
    • I agree it is mincing words, but that was the point I was making. The word normal is open to interpretation, and the question itself was too open ended. Any exam question should be clear cut, with no fancy words that can lead to confusion, unless of course its in an interview situation and you're looking for someone who sees the bigger picture. For the record I answered 0.

      I took this up with the director of the organisation, who happened to be there, and he did seem genuinely interested in improving the q

  • What course did you take and what was the exam that went with it?

    I've done the RHCT course and exam, I found the course interesting, the exam was less exciting. Red Hat's exmas are pure practicals, here is a box, make it do the following things - just like it says on their web page. It doesn't test your ability to remember a lot of things, rather your ability to approach problems from the right direction.

    I'm considering taking the LPI [lpi.org] exam, it's all multi-guess based, but they claim to have spent a lot

    --
    -- "It's not magic, it's work..."
    • The course was LPI 102. The actual course was good and the trainer was very thorough. It was just the actual exam I was disappointed with. However, the exam is optional, I only took it because the company was paying for it.

      After the experience, I am very dubious of any automated exam. The selection box type questions are easy to mark, but when it comes to entering a string of characters, it needs a human to verify whether it's correct. Unless you can run it (if its a command) and check the output. However

      • I've been reading a LPIC-I book, and found the content quite comprehensive, but I've found some of the example questions a bit awkward. I like the idea of a vendor neutral exam, and I think a written exam like the LPI's complements the more practical approach of Red Hat.

        I must confess to being less than confident of an automated exam, but the book did come with an example exam on CD, and I suppose it's useful to practice first. I've done the example exam a few times, and been surprised with my mistakes, b

        --
        -- "It's not magic, it's work..."