Re: Goodbye uk.rec.cycling; Helllo U.R.C.moderated

"Peter Grange" <peter@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:kvq5e59eaq7qgcq83ok6qpuadheehagiba@xxxxxxxxxx
On Fri, 23 Oct 2009 22:53:42 +0100, "Clive George"
<clive@xxxxxxxxxxxxxxxxxx> wrote:

"Peter Grange" <peter@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:du14e5d2243jm489l2bbmef6dj9p5gfu6m@xxxxxxxxxx
On Fri, 23 Oct 2009 18:18:16 +0100, "Clive George"
<clive@xxxxxxxxxxxxxxxxxx> wrote:

"Peter Grange" <peter@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:g9n3e59gdn4q99k281fhqkd3s9t6lpjqp9@xxxxxxxxxx

Anyway, it looks as if the problem has been recognised in posts in
urcm, so at least people are aware of what's happening. If someone
posts from Highwinds I can tell by the headers if the problem on that
server is fixed. If no-one posts I may need to from time to time, not
to demonstrate it's still bust but to find out if it's fixed. There's
no point in posting anything of cycling interest as most of the people
won't see it, so it will need to be a test message.

Can we leave it at that?

Since you appear to have acknowledged that you don't need test messages
approving, yes.

At the moment I don't, as the only problems I know about I can test
for, but I don't know what's coming up tomorrow, do you?

I can't forecast with 100% certainty, no, but I can make a darned good
guess. From what I know about the system, which is very little but
sufficient, I can get a test strategy for any problems external to chiark
which doesn't require allowing test posts. Internal problems would be
shown
by the regular activity - posts appearing in the logs but not making it
into
the group would be noticed quite quickly.

My job is making software like this work - not had to do NNTP yet, but the
underlying concepts are will be similar. And I'm pretty good at it,
especially at diagnosing and solving problems. I realise this is an appeal
to authority, and hence an invalid debating tactic, but I feel it's worth
mentioning nonetheless.

The problem with an appeal to authority is that there is a danger that
it may lead to a "my willie's bigger than your willie" contest.

Anyway, here's my willie:
I've been making hardware and software work, including some harware I
designed myself, and some software I wrote myself, since 1967 (until I
retired last year anyway). This includes designing and fixing hardware
with an oscilloscope in the days when it wasn't all on one chip (or
even any chips!), writing and debugging hardware diagnostic software,
writing enhancements to and supporting mainframe operating systems,
designing the hardware for and writing the software for microprocessor
driven interfaces to large mainframe (Multics) systems, designing and
developing software for a file transfer and email system to interface
to Janet (Colour Books), supporting Unix networking software and, for
the last ten years, specifying, developing and implementing helpdesk
and order management systems.

I guess I must have been pretty good at most of it as people kept
asking me to keep on doing it, so I do know a bit about testing
things.

In all that time I can't remember a single instance where I have tried
to discourage a user from testing anything he wanted. I may have asked
said user to make it easier for me by making his testing obvious,
perhaps by prefixing his transactions with "Test". In the case we are
discussing I might suggest a single test thread for example.

Well, I don't let our users test anything they want. There's many cases
where that's a really bad idea. Depending on the problem, I might test it
using the test system, rather than live, or diagnose with eg the transaction
they're trying to process. But generally speaking our users don't even ask
to test stuff - they ask for stuff to work, and testing is our problem.

The trouble with your mini-CV is that you've now demonstrated that you
really ought to understand about testing, and ought to understand why the
system as provided at the moment is sufficient.

I said to Phil Lee that if he could come up with a suitable counterexample,
I'd consider revising my position. You apparently are in a good position to
come up with a suitable counterexample, so how about it? Use the skills
you've just told me you've got.


.

Relevant Pages

  • Re: Goodbye uk.rec.cycling; Helllo U.R.C.moderated
    ... If no-one posts I may need to from time to time, ... I've been making hardware and software work, ... This includes designing and fixing hardware ... even any chips!), writing and debugging hardware diagnostic software, ...
    (uk.rec.cycling)
  • Re: Goodbye uk.rec.cycling; Helllo U.R.C.moderated
    ... If no-one posts I may need to from time to time, ... I've been making hardware and software work, ... This includes designing and fixing hardware ... An important part of delivering a service is user confidence. ...
    (uk.rec.cycling)
  • Re: Lies I Have Told In Usenet
    ... because I am not one of those who hates writing but enjoys having written ... your forgiveness as I am a great fan of your important usenet posts wherein ... it was the writer Elliot Gould who said "good artists steal and good artist ...
    (misc.writing)
  • Re: Lets Get Something Straight!
    ... Walk In The Woods" for days now. ... but at the top of each of her posts she puts - by Janey ... Her writing style sucks, if you ask me. ... I have agreed to leave her posts alone completely because I need to ...
    (alt.med.fibromyalgia)
  • Re: Why physicists conceal the evidence on God, eternal soul, another world, etc. (PDE JP)
    ... physics, by claiming all these preposterous things, as that there is ... And you actually do this in your writing above - i.e. ... I flood them with information against their beliefs. ... my original posts, you shall shut up and rot. ...
    (sci.physics)