<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Sat, Nov 11, 2017 at 12:37 PM, Satish Balay <span dir="ltr"><<a href="mailto:balay@mcs.anl.gov" target="_blank">balay@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Sat, 11 Nov 2017, Matthew Knepley wrote:<br>
<br>
> > In the long term - Barry wants to get rid of next..<br>
><br>
><br>
> 1) I think next really prevents master from getting screwed up (witness<br>
> next)<br>
><br>
</span><span class="">> 2) I think we are actually finding interaction bugs there.<br>
><br>
> Are those points wrong, or is there another way to do these things?<br>
<br>
</span>Next is an intergration testing mechanism. The prerequisite for it [I<br>
think] is - one should test the branch properly before merging to<br>
next. However we are not doing proper testing before merge to next -<br>
and relying on next to do this part aswell.<br>
<br>
So with current next - it one has to figure out which branches are<br>
breaking the tests [takes time - which most of us are not doing] - and<br>
then hope it gets fixed quickly. Otherwise next stay broken for a long<br>
time [and other branches in next - which could be clean - don't<br>
receive sufficient confidence to graduate to master]<br>
<br>
So Barry's thought wrt getting next is to have a better way of testing<br>
feature branch one wants to test (i.e master+feature). Its not clear<br>
to me how many integration issues we've addressed with our current<br>
next model. [Its mostly been indvidual branch issues]<br>
<br>
Also if feature-1 and feature-2 are feature branches that are tested<br>
in next [wrt integration]. The following should be equivalent to<br>
testing 'master + feature1 + feature2' - aka current next model:<br>
<br>
1. test master+feature1<br>
2. success => merge feature1 to master<br>
3. tets master+ feature2<br>
3. success => merge feature2 to master<br>
<br>
Note: my next-tmp is an attempt to get closer to 'master+feature1'<br>
testing from 'master+feature1+feature2' testing [yeah - its more like<br>
master+2/3 branches in next-tmp vs master+10/15 branches in next.]<br>
<br>
Also I'm restarting next-tmp from a clean master when merging new set<br>
of branches to test. And throwing away branches with problems - and<br>
retest only after it has fixes [This way a broken branch does not keep<br>
next-tmp broken until it gets fixed.]</blockquote><div><br></div><div>I don't think we have the resources to run full tests on every branch one at a time. Do we?</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="HOEnZb"><font color="#888888"><br>
Satish</font></span></blockquote></div>-- <br><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>-- Norbert Wiener</div><div><br></div><div><a href="http://www.caam.rice.edu/~mk51/" target="_blank">https://www.cse.buffalo.edu/~knepley/</a><br></div></div></div></div></div>
</div></div>