POV-Ray : Newsgroups : povray.off-topic : More poor planning : Re: More poor planning Server Time
11 Oct 2024 13:15:52 EDT (-0400)
  Re: More poor planning  
From: Stephen
Date: 20 Nov 2007 10:40:01
Message: <web.4742ff4b7ad347ee726bd13c0@news.povray.org>
Invisible <voi### [at] devnull> wrote:
> Stephen wrote:
> > Invisible <voi### [at] devnull> wrote:
> >> (Today I'm writing an unecessarily long and complex plan document
> >> describing in absurd detail how we're going to test a trivial 2-page
> >> Haskell program. I really wish to God I was working somewhere else right
> >> now!)
> >
> > You can generally shorten test scripts by using a spreadsheet or a table with a
> > layout that details:
> > Test Step |Test action & Instruction(s) | Field/Variable data / input | Expected
> > Results | Output | Passed/Failed
> >
> > Of course you modify the script to suit the tests.
>
> Well, let's put it this way: The plan is 17 pages long, and I haven't
> even said what the tests are yet! o_O
>
> Basically there's a standard template for writing test plans. It's
> designed for huge complex systems, and the thing I'm testing is tiny. So
> I've already spend ages deleting all the stuff that's not applicable.
> And it's *still* 17 pages of whaffle.
>
> (This is who we are, this is what the software is, this is were it's
> going to be used, this is who is going to use it, this is who will test
> it, this is what the software is for, this is what we do currently, this
> is why software testing is necessary, this is what the tests attempt to
> demonstrate, this is all the things the testing won't cover, this is
> what we'll do if any tests fail, this is what we'll do if we ever alter
> the software, this is the list of documents that will be produced during
> the test process, etc.)
>
> Believe it or not, that bundle I just wrote there? It's incomplete. (!)

Right! As if anyone ever reads that. I know the sort of template you mean.
Here we have to write design documents about what we are going to do before we
do it and get it authorised by QA. But if there is even a minor change then the
doc has to be re-evaluated and all work on it stops until it goes to the next
status. That sounds fine until you realise that most configure-ers are allowed
to prototype before submitting the design to QA.
Hint, put in a deliberate mistake that you can take out to keep QA happy :)



Stephen


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.