recurring "standup"

Mark Atwood fallenpegasus at
Mon Feb 1 23:46:52 UTC 2016

Thank you Eric.

For my own piece of mind, if you are not doing so already, please keep that private development branch of 
the testframe work regularly pushed to a remote on gitlab.  It’s ok if it’s in a private branch under your personal account there.


On February 1, 2016 at 3:29:32 PM, Eric S. Raymond (esr at wrote:

Mark Atwood <fallenpegasus at>:  
> Hello everyone (and hello from AU, I am in Geelong this week for #lca2016).  
> What have you all worked on in the past couple of weeks, what successes and  
> frustrations have you had, and what do you see yourself working on over  
> this next week?  


1. I've gotten back the ability to do Coverity scans and am keeping us  
clean on that front. This is a bit more of a statement than it might  
otherwise be because Coverity upgunned its auditing capability recently;  
it's checking more invariants and throwing more warnings than it used to.  

2. I'm also on top of our issue queue. Our incidence of known bugs not  
inherited from Classic is nil. The dog that isn't barking is more  
bug reports - we're getting peripheral little stuff about port issues  
and documentation but no functional bugs. I like that kind of boredom.  

3. Following the recent polishing pass by Sanjeev, I'm pretty happy  
with the state of our documentation. It has now been eyeball-checked  
against the code's behavioral reality by *three* different people; the  
quality and freshness is way ahead of Classic's.  

4. We appear to be shipping solid code. Our small but active pool of  
test users absorbed 0.9.1 without a ripple. I think there are grounds  
for believing that our next several point releases will be uneventful  
to the point of being boring. Lovely, lovely boredom...  

Ongoing work:  

1. I'm still slogging away at capture/replay. It's hard work with enough  
reverses that I'm now doing it on a private branch. I have concluded that  
the previous strategy of intercepting only the protocol-machine calls  
won't work - doing that there's no way to keep a huge amount of peer state  
up to date. So now I'm trying to mock at the packet-I/O level, in effect  
fully capturing and resimulating ntpd's network environment.  

2. In the near future (later this week and early next) I mean to  
budget some time to cross-port some of the recent (non-security) bug  
fixes from Classic; I'm leaving the security stuff to Daniel. I don't  
believe any of these are actually very important, but it's good PR to  
plug everything Classic has plugged.  


I'm not currently blocked on any resources, nor on actions by others.  
Hal Murray has not yet responded to my suggestion that we divvy up the  
fix cross-ports from Classic, but I can start on those before he does.  


Our next natural breakpoint for a 0.9.2 release will be when either (1)  
Amar finishes the unit tests, or (2) Hal and I finish cross-porting the  
classic fixes.  

General note:  

I've been quiet lately because I've let myself slow down to normal  
hours and more variety of work from the intense tempo and  
near-monomaniacal concentration I was maintaining for most of last  
year. That was effort well spent - it bought us the relatively smooth  
sailing we're having now - but not really sustainable.  

I'm taking it easier now so I can speed up when we need me to. I expect  
that'll be in the immediate run-up to 1.0.  
<a href="">Eric S. Raymond</a>  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the devel mailing list