[concurrency-interest] dealing with people that question visibilityproblems

Shaffer, Darron Darron_Shaffer at stercomm.com
Wed Feb 21 11:37:26 EST 2007


You can always use FUD in a good cause.

Explain that these problems will likely pop up at a customer site, in
production, at 2 AM on a national holiday during critical end-of-quarter
processing.  And take a minimum of 5 days to figure out. 

-----Original Message-----
From: concurrency-interest-bounces at cs.oswego.edu
[mailto:concurrency-interest-bounces at cs.oswego.edu] On Behalf Of Peter
Veentjer
Sent: Wednesday, February 21, 2007 9:28 AM
To: concurrency-interest at cs.oswego.edu
Subject: [concurrency-interest] dealing with people that question
visibilityproblems

I have a question about dealing with other developers that question
the usefulness of preventing visibility problems.

At the moment I'm almost finished with writing a blogpost about
potential visibility problems in Spring. But one of the questions I'm
going to receive (and already have received) is : why should we go for
all the extra trouble? It works at the moment, we haven't seen any
visibility problems yet, so why should we fix what isn't broken.

There are a lot of theoretical reasons I can think of why writing
threadsafe code is a good thing. But if it doesn't go wrong in
practice, it is hard to convince other developers.

My guess is that with the introduction of multi core systems (with an
increasing number of cpu's), memory coherence is going to be weaker,
and the chance of visibility problems is going to increase. But if you
don't get errors today, why should someone bother?

Does someone have experience with dealing with this situation?
_______________________________________________
Concurrency-interest mailing list
Concurrency-interest at altair.cs.oswego.edu
http://altair.cs.oswego.edu/mailman/listinfo/concurrency-interest



More information about the Concurrency-interest mailing list