1 |
hasufell <hasufell <at> gentoo.org> writes: |
2 |
|
3 |
> |
4 |
> Ciaran McCreesh: |
5 |
> > On Thu, 05 Feb 2015 11:21:37 -0500 |
6 |
> > Michael Orlitzky <mjo <at> gentoo.org> wrote: |
7 |
> >> I'll start with the silliest: |
8 |
> >> |
9 |
> >> Explain briefly the purpose of the following tools: grep, cut, sed, |
10 |
> >> cat, wc, awk |
11 |
> > |
12 |
> > This is in no way silly when viewed in its proper historical context. It |
13 |
> > was added because before the quiz existed, someone recruited a couple of |
14 |
> > developers for Gentoo/Alt who didn't know what grep was, and they broke |
15 |
> > the tree in a horrific way. |
16 |
> > |
17 |
> |
18 |
> That just shows that our workflow is so broken that we have to ask such |
19 |
> things in a quiz. |
20 |
> |
21 |
> "Let's track him for one month and revert any major breakage quickly" is |
22 |
> our answer to what the rest of the world does: review. |
23 |
> |
24 |
> |
25 |
|
26 |
|
27 |
Hello. |
28 |
|
29 |
I just wanted to bring up a quick point which hasn't been made yet. At |
30 |
several points, there has been discussion of removing certain questions, |
31 |
like the aforementioned "Explain briefly the purpose of the following |
32 |
tools." |
33 |
|
34 |
Rather than removing such questions, and working off of a previously |
35 |
mentioned web interface for the quizzes, I thought it worth mentioning that |
36 |
an interactive test module, akin to Computer Based Testing (CBT). The idea |
37 |
is simple: there are two kinds of questions, those that are are text based, |
38 |
like policy and documentation questions, and those that are task oriented, |
39 |
like the above mentioned UNIX tools, or ebuild editing. |
40 |
|
41 |
Since we are all familiar with the first kind of questions, let me just |
42 |
describe what I envision for testing the second kind of question. For |
43 |
example, with the UNIX tools, the quiz taker is presented, via a web |
44 |
interface, some input, and the desired output. Then they are told that |
45 |
using cmd foo (substitute whatever UNIX tool) they are expected to achieve |
46 |
the desired output. This could be implemented any number of ways, so I |
47 |
won't bother going into that now. Ebuild editing could follow a similar |
48 |
process, whereby syntax is edited by the user and the output checked, via |
49 |
repoman, or via some other method (depending on the scope of the problem). |
50 |
|
51 |
This new system could be offered as an alternative to the current system, |
52 |
however, based on my perception of the consensus in this thread (that the |
53 |
current quiz is inadequate), my opinion is that, once functional, the new |
54 |
quiz format should be the only method offered. |
55 |
|
56 |
So, in summary, my suggestion is to consider transforming the quiz process |
57 |
into two separate sections, the first is a standard text based test like the |
58 |
current system, and the second is an (semi-)interactive system where the |
59 |
user has to problem solve, giving a more practical approach to quiz |
60 |
questions. |