lucenenet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Paul Irwin <pir...@feature23.com>
Subject Re: Branch Status? branch_4x vs master
Date Tue, 25 Nov 2014 21:54:15 GMT
Yep, I found a number of those bugs and I'm squashing them now. Making good
progress on getting a reliable test run (despite lots of failing tests)
that doesn't take hours to complete.

Here's the ICLA: http://www.apache.org/licenses/icla.txt

Print and scan it and send to secretary@apache.org, make sure to note to
notify project Lucene.net.


Paul Irwin
Lead Software Engineer
feature[23]

Email: pirwin@feature23.com
Cell: 863-698-9294

On Tue, Nov 25, 2014 at 4:41 PM, Wyatt Barnett <wyatt.barnett@gmail.com>
wrote:

> Hi Paul -- working on getting stuff cleaned up but you can look at
> https://github.com/wwb/lucene.net to see what I have been doing. Probably
> the easiest global trick would be to add [assembly: Timeout(20000)] in the
> Test project's assemblyinfo.cs file -- it will hard cap every test at 20
> seconds which is the happy medium, there are perhaps a half dozen that get
> clipped by that and part of the cleanup operation was to fix those.
>
> If you are looking for hanging tests look for [Explicit()] attributes -- I
> used those to mark the hanging tests as I went. In cases where I had
> narrowed down some causality I tried to note that in the comment in said
> attribute. Overall one thing I did notice is that there is a pretty big
> problem with the TestHelper's Random() class that blows up many, many
> tests. That would be a really good place to start on fixing things as it is
> used heavily.
>
> No, I haven't got the CLA squared away -- where should I start?
>
>
> On Tue, Nov 25, 2014 at 2:48 PM, Itamar Syn-Hershko <itamar@code972.com>
> wrote:
>
> > Paul, Wyatt's PR is just to skip those tests. If you can get a look at
> them
> > that'll be way better :)
> >
> > --
> >
> > Itamar Syn-Hershko
> > http://code972.com | @synhershko <https://twitter.com/synhershko>
> > Freelance Developer & Consultant
> > Author of RavenDB in Action <http://manning.com/synhershko/>
> >
> > On Tue, Nov 25, 2014 at 9:44 PM, Paul Irwin <pirwin@feature23.com>
> wrote:
> >
> > > I just read back through Wyatt's emails and it does seem to be the same
> > > issue in BaseTokenStreamTestCase. Looking forward to your PR, Wyatt.
> Have
> > > you gotten the CLA squared away yet?
> > >
> > >
> > > Paul Irwin
> > > Lead Software Engineer
> > > feature[23]
> > >
> > > Email: pirwin@feature23.com
> > > Cell: 863-698-9294
> > >
> > > On Tue, Nov 25, 2014 at 2:35 PM, Paul Irwin <pirwin@feature23.com>
> > wrote:
> > >
> > > > Working on running the tests now and the
> > > > TestDoubleMockGraphTokenFilterRandom test runs seemingly indefinitely
> > on
> > > my
> > > > machine. It appears like the test completes as the tear-down method
> > > > executes, but I'm sensing that the threads that are spawned in this
> > test
> > > > are running in the background preventing the test from finishing
> > > according
> > > > to the UI (VS Test Explorer with the NUnit adapter). I keep seeing
> the
> > > same
> > > > repeated test in the output window as well, i.e.:
> > > >
> > > >   pass-through: return pending token
> > > >> LTF.nextToken inputPos=15 outputPos=15 tokenPending=False
> > > >> LTF.peekToken inputPos=15 outputPos=15 tokenPending=False
> > > >> LTF.nextToken inputPos=15 outputPos=15 tokenPending=False
> > > >> LTF.peekToken inputPos=15 outputPos=15 tokenPending=False
> > > >>   input.incrToken() returned True
> > > >>   now inputPos=16
> > > >>   call afterPosition
> > > >>   next position: outputPos=16
> > > >>   pass-through: return pending token
> > > >>   input.incrToken() returned True
> > > >>   now inputPos=16
> > > >>   call afterPosition
> > > >>   next position: outputPos=16
> > > >>   pass-through: return pending token
> > > >> LTF.nextToken inputPos=16 outputPos=16 tokenPending=False
> > > >> LTF.peekToken inputPos=16 outputPos=16 tokenPending=False
> > > >> LTF.nextToken inputPos=16 outputPos=16 tokenPending=False
> > > >> LTF.peekToken inputPos=16 outputPos=16 tokenPending=False
> > > >>   input.incrToken() returned True
> > > >>   now inputPos=17
> > > >>   call afterPosition
> > > >
> > > >
> > > > I'm assuming this is similar to what Wyatt was running into? Anyone
> > else
> > > > seen this or have any ideas?
> > > >
> > > >
> > > > Paul Irwin
> > > > Lead Software Engineer
> > > > feature[23]
> > > >
> > > > Email: pirwin@feature23.com
> > > > Cell: 863-698-9294
> > > >
> > > > On Tue, Nov 25, 2014 at 11:32 AM, Itamar Syn-Hershko <
> > itamar@code972.com
> > > >
> > > > wrote:
> > > >
> > > >> Hi Paul,
> > > >>
> > > >> Use the master branch. The branch_4x one is with your work, however
> > the
> > > >> port that was made for master was done from scratch so we are going
> to
> > > >> discard that branch...
> > > >>
> > > >> For your R&D time, I think your best shot would be at looking
at the
> > > >> failing tests in the core and taking it from there.
> > > >>
> > > >> @Wyatt -- any ETA for your PR with marking the faulty tests?
> > > >>
> > > >> Thanks,
> > > >>
> > > >> --
> > > >>
> > > >> Itamar Syn-Hershko
> > > >> http://code972.com | @synhershko <https://twitter.com/synhershko>
> > > >> Freelance Developer & Consultant
> > > >> Author of RavenDB in Action <http://manning.com/synhershko/>
> > > >>
> > > >> On Tue, Nov 25, 2014 at 6:26 PM, Paul Irwin <pirwin@feature23.com>
> > > wrote:
> > > >>
> > > >> > Awesome, thanks Prescott and Wyatt. I'll get started there.
> > > >> >
> > > >> > Anyone have any insight into why master and branch_4x (where
4.x
> > > porting
> > > >> > work was happening previously) diverged so much? I assume they're
> > > pretty
> > > >> > much unmergeable at this point and branch_4x will be discarded.
> > > >> >
> > > >> >
> > > >> > Paul Irwin
> > > >> > Lead Software Engineer
> > > >> > feature[23]
> > > >> >
> > > >> > Email: pirwin@feature23.com
> > > >> > Cell: 863-698-9294
> > > >> >
> > > >> > On Tue, Nov 25, 2014 at 11:19 AM, Prescott Nasser <
> > > >> geobmx540@hotmail.com>
> > > >> > wrote:
> > > >> >
> > > >> > > Master is where we're at for the moment
> > > >> > > ________________________________
> > > >> > > From: Wyatt Barnett<mailto:wyatt.barnett@gmail.com>
> > > >> > > Sent: 11/25/2014 7:14 AM
> > > >> > > To: dev@lucenenet.apache.org<mailto:dev@lucenenet.apache.org>
> > > >> > > Subject: Re: Branch Status? branch_4x vs master
> > > >> > >
> > > >> > > I think one should be working off the new master branch
-- or at
> > > least
> > > >> > that
> > > >> > > is where the CI efforts are going, see
> > > >> > > https://github.com/apache/lucene.net
> > > >> > >
> > > >> > > On Tue, Nov 25, 2014 at 10:06 AM, Paul Irwin <
> > pirwin@feature23.com>
> > > >> > wrote:
> > > >> > >
> > > >> > > > Hello all,
> > > >> > > >
> > > >> > > > I have some available R&D time so I'd like to jump
back in.
> I'm
> > a
> > > >> > little
> > > >> > > > confused about the status of branch_4x vs master, it
looks
> like
> > > >> they've
> > > >> > > > diverged significantly:
> > > >> > > >
> > > >> > > > "This branch is 462 commits ahead, 131 commits behind
master"
> > > >> > > >
> > > >> > > > Can someone fill me in on what's going on here? It
looks like
> I
> > > >> should
> > > >> > > jump
> > > >> > > > in to master, but it appears to not match the commit
history
> of
> > > >> > branch_4x
> > > >> > > > work that happened a while back.
> > > >> > > >
> > > >> > > > Thanks!
> > > >> > > >
> > > >> > > > Paul Irwin
> > > >> > > > Lead Software Engineer
> > > >> > > > feature[23]
> > > >> > > >
> > > >> > > > Email: pirwin@feature23.com
> > > >> > > > Cell: 863-698-9294
> > > >> > > >
> > > >> > >
> > > >> >
> > > >>
> > > >
> > > >
> > >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message