[vtk-developers] [VTK 0014170]: Segfault in vtkSmartVolumeMapper

[David E DeMarle]

About the UI - agreed. No one likes it or finds it intuitive.

Try the search field in the top right: (and no I don't know what the entire
set of keys is either)
for example:
entity:topic status:open project:VTK
entity:change reviewer:"John Biddiscombe" status:open
entity:topic owner:"John Biddiscombe" status:merged

About long topics/patchsets - also agreed. The list decided that best
practice is to limit topics to a handful of commits in length. It makes it
easier for others to understand what you did and review it, and makes it
easier to rebase or merge the topic into an ever changing master. What a
handful means exactly is up for interpretation.

Another point - everyone should understand the distinction between topics
and patchsets.

A topic is a just of series of git commits. Kitware's gerrit is unique in
that we've modified it to know about topics. Upstream gerrit only knows
about commits. Topics are a good thing, but they introduce even more
wrinkles. We only allow merges at the topic level and regression tests are
kicked off from the commit level for example.

A change set is a series of attempts at the same thing. This can be at the
commit or topic level. Take the commit level for simplicity. If I submit
something, and Bill rejects it in review. I will fix whatever dumb thing I
did and resubmit. I'll keep the same Change Id my commit message (git
--amend) to tell gerrit that I am working on the same thing. Gerrit will
show this as a new patchset on the same topic. Bill will review that and
eventually I'll either merge it, or abandon it. When I merge it, git master
only sees the more recent patch set and the wrong paths I took are
forgotten.

In any case, to make it easier to test someone's elses long set of changes,
use the download button to copy the "git fetch..." string from the most
recent patchset on the topic level so that you get the most recent patchset
of the entire topic all at once.

David E DeMarle
Kitware, Inc.
R&D Engineer
21 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-881-4909


On Fri, Jul 19, 2013 at 3:48 AM, Biddiscombe, John A. <biddisco at cscs.ch>wrote:

>  Dave****
>
> Thanks for the explanation. Since I’m not a member of the “core” group, I
> guess this doesn’t apply to me and I have to wait in line like everyone
> else.****
>
> ** **
>
> I’ve been playing with gerrit a bit and reviewing people’s patches (I’m
> warming to the system). It’s hard to navigate and find ones that are
> ‘doable’ - many have huge patch sets and a proper review would require a
> full checkout of the branch with some local testing. I would like to get up
> to speed with working with the system as my intern this summer will
> hopefully tart up the paraview volume render interface and I want this to
> go into the official distribution with testing etc.****
>
> ** **
>
> I’ve added VTK/ParaView to my watched projects but there doesn’t seem to
> be a simple way of saying “show me open topics in these two”, ****
>
> If I use “My/Changes/Watched” I see VTK/ParaView, but they nearly all date
> back to 2012, if I try “All/topics/open”, then I get more recent ones, but
> I also see all the other projects that I’m not involved with.****
>
> ** **
>
> If there’s a simple way of seeing all VTK/ParaView topics that are in need
> of review, please tell.****
>
> ** **
>
> Ta****
>
> ** **
>
> JB****
>
> ** **
>
> *From:* David E DeMarle [mailto:dave.demarle at kitware.com]
> *Sent:* 18 July 2013 16:41
> *To:* Biddiscombe, John A.
> *Cc:* David Cole; Bill Lorensen; vtk-developers at vtk.org
>
> *Subject:* Re: [vtk-developers] [VTK 0014170]: Segfault in
> vtkSmartVolumeMapper****
>
>  ** **
>
> Here is how it should work:****
>
> ** **
>
> Anyone in the core:
> http://review.source.kitware.com/#/admin/groups/11,members has a build
> and ctest spawned on mac, linux and windows immediately once they git
> gerrit-push.****
>
> ** **
>
> Anyone gets test spawned only when someone in the core requests the tests
> to run (for security)****
>
> ** **
>
> In either case click on the topic in gerrit and click on the filtered
> CDash results to get to them.****
>
> ** **
>
> Anyone in the core can then merge the change (assuming the reviewers were
> not grumpy) into master through gerrit. After this it is just like in the
> old days, the continuous testers immediately test it a bit more thoroughly,
> and then at 9pm the nightlies test it even more thoroughly.****
>
> ** **
>
>
> ****
>
> David E DeMarle
> Kitware, Inc.
> R&D Engineer
> 21 Corporate Drive
> Clifton Park, NY 12065-8662
> Phone: 518-881-4909****
>
> ** **
>
> On Wed, Jul 17, 2013 at 4:10 AM, Biddiscombe, John A. <biddisco at cscs.ch>
> wrote:****
>
> Not very impressed with this gerrit stuff. In the old days I’d commit a
> change and the next morning the dashboard would show the files updated and
> the tests run (this code wasn’t tested, so I’m not expecting problems).***
> *
>
>  ****
>
> How will I know when to monitor the dashboard when the commits sit inside
> gerrit until one day something presumably happens?****
>
>  ****
>
> JB****
>
>  ****
>
> *From:* Biddiscombe, John A.
> *Sent:* 15 July 2013 17:17
> *To:* Biddiscombe, John A.; David Cole; Bill Lorensen
> *Cc:* vtk-developers at vtk.org
> *Subject:* RE: [vtk-developers] [VTK 0014170]: Segfault in
> vtkSmartVolumeMapper****
>
>  ****
>
> On another machine the change-ID is generated correctly. So I’ve pushed
> from there.****
>
>  ****
>
>  ****
>
>  ****
>
> *From:* vtk-developers-bounces at vtk.org [
> mailto:vtk-developers-bounces at vtk.org <vtk-developers-bounces at vtk.org>] *On
> Behalf Of *Biddiscombe, John A.
> *Sent:* 15 July 2013 16:03
> *To:* David Cole; Bill Lorensen
> *Cc:* vtk-developers at vtk.org
> *Subject:* Re: [vtk-developers] [VTK 0014170]: Segfault in
> vtkSmartVolumeMapper****
>
>  ****
>
> I did the setup stuff first and tried many times to redo the commit, but
> it won’t generate one. Unfortunalty, I’m trying to work on another branch
> and if I switch back to the topic branch the entire paraview will rebuild
> again L so I’m reluctant to experiment.****
>
>  ****
>
> JB****
>
>  ****
>
>  ****
>
> *From:* David Cole [mailto:dlrdave at aol.com <dlrdave at aol.com>]
> *Sent:* 15 July 2013 16:00
> *To:* Bill Lorensen; Biddiscombe, John A.
> *Cc:* vtk-developers at vtk.org
> *Subject:* Re: [vtk-developers] [VTK 0014170]: Segfault in
> vtkSmartVolumeMapper****
>
>  ****
>
> > I can’t do a gerrit push because it keeps asking for an Id and I don’t
> want to make one up!****
>
> > ****
>
> > Where do I get it from?****
>
>  ****
>
>  ****
>
> It comes from the stuff set up by the setup for development script.****
>
>  ****
>
> If you made the commit before running the script, it won’t have one.****
>
>  ****
>
> If you just do “git commit --amend” now on that commit, one should be
> added for you.****
>
>  ****
>
> Then you should be able to push.****
>
>  ****
>
>  ****
>
> HTH,****
>
> David C.****
>
>  ****
>
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/listinfo/vtk-developers
>
> ****
>
> ** **
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtk-developers/attachments/20130719/5494ce07/attachment.html>