[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 2 Apr 2020 15:39:33 +1100
From: Ulisses Albuquerque <ulisses.montenegro@...il.com>
To: oss-security@...ts.openwall.com, Jeffrey Walton <noloader@...il.com>
Subject: Re: Deficient engineering processes
In my experience, specific advice and recommendations trump generic advice
every single time. If your teams are failing to meet quality standards,
identify those individually, document what failed and how it could be
addressed, and *only then* show how a more structured approach would solve
the issues. This can also be an interesting exercise for you, as going
through specific problems and correlating those to the proposed
"engineering best practices" might make you realise that those would not
actually solve the problems.
More than anything, though, be empathetic. Approach teams with a "can you
talk me through your process" rather than with a "you are doing it wrong"
angle. People are far more likely to accept input when you hear them first,
and they might even be upfront about known shortcomings of their current
approach.
Regards
On Thu, 2 Apr 2020 at 13:47, Seth Arnold <seth.arnold@...onical.com> wrote:
> On Wed, Apr 01, 2020 at 07:42:38PM -0400, Jeffrey Walton wrote:
> > My question is, how to convince someone that following standard
> > project management procedures is a good thing? How do we get them
>
> I've heard variations on the phrase "we don't have time to fix these bugs
> before release" or "this new feature is our top priority" from dozens of
> projects over the years.
>
> The impression is that fixing bugs won't win new customers, or finding
> bugs proactively means you might spend time fixing bugs your users might
> not encounter in practice (thus that time is wasted).
>
> But we have all seen software that's too buggy to be enjoyable, or even so
> buggy it is not fit for use. We've all got horror stories of a known, but
> ignored, bug, that cost thousands or millions of dollars. (I imagine a
> handful of people even know of billion-dollar errors. The usual example is
> https://en.wikipedia.org/wiki/Tony_Hoare#Apologies_and_retractions
> but this is probably far from the only case.)
>
> The costs of unknown or unfixed bugs is largely hidden from view, until
> the cost is large and impossible to ignore.
>
> We all also have examples of bugs that we're very glad to have caught
> before release: the bugs that would have cost thousands, or millions, of
> dollars to repair after release, if it's possible at all. These are much
> less known.
>
> Perhaps we need to talk more about our successes, too? Not just the cases
> where we went wrong, but also the cases where we went right, and thus
> saved a fortune?
>
> Thanks
>
--
“If debugging is the process of removing software bugs, then programming
must be the process of putting them in.” - *Edsger Dijkstra*
Powered by blists - more mailing lists
Please check out the
Open Source Software Security Wiki, which is counterpart to this
mailing list.
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
