Subject: Re: [Boost-build] Bjam patch to use maximum value of timeout option
From: K. Noel Belcourt (kbelco_at_[hidden])
Date: 2010-11-23 15:41:36
On Nov 22, 2010, at 12:35 PM, David Abrahams wrote:
> At Mon, 22 Nov 2010 12:07:13 -0700,
> K. Noel Belcourt wrote:
>> I discovered that if bjam is presented with more than one occurrence
>> of the timeout parameter (-l) on the command line, bjam will always
>> use the first occurrence. Often during regression testing, more than
>> one -l option occurs on the command line. This patch changes the
>> logic to use the largest specified timeout value on the command line,
>> rather than an order specific selection (i.e. always first, always
>> last, etc...).
>> Tested on Darwin and Linux, okay to commit?
> Needs a doc patch too, no?
Yes, I think so. Working on that.
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk