[LSST|dm-users 555] Re: Problem with "Testing the Installation"

[Jim Bosch]

On Tue, Feb 10, 2015 at 11:03 AM, Will Dawson <will at dawsonresearch.com>
wrote:

> Thanks to many of your helpful comments I was able to build the stack.
> However when I ran the steps in Testing the Installation
> <https://confluence.lsstcorp.org/display/LSWUG/Testing+the+Installation> the
> resulting detected-sources.txt catalog is vastly different from the
> detected-sources.txt.expected catalog. I have put snippets of these files
> on:
>
> https://gist.github.com/292447e1c70fd494a765
>
> As you can see many of the outputs are the same (e.g. coord_ra, coord_dec)
> but many are different (e.g. id apparently offset by 20) but most notably
> the parameter columns in the files are very different.
>
> Note that this seems to be different from
> https://jira.lsstcorp.org/browse/DM-1086.
>
>
This is a known issue, and it's just that you're pulling the master version
of the demo package, and it's ahead of the versions of the rest of the
packages you're using.  We need to improve the versioning and release
system for the demo package so they're consistent (it's relatively rare
that they're out of sync, because changes to the demo package are rare, but
that's certainly the case right now).

Finally, shouldn’t I be submitting this stuff as an issue on github or
> bitbucket rather than emailing this list?
>


Perhaps; I think this is working okay for now.  We'd ask you to escalate to
a JIRA issue (https://jira.lsstcorp.org/secure/Dashboard.jspa) if it was
something we didn't already understand as a known issue.

I don’t seem to have access to git.lsstcorp.org though (perhaps I just have
> to wait for my gitolite account).
>
>
We're actually in the process of moving to GitHub in the very, very near
future, so you probably won't ever get one.  In the meantime, you should
use the ready-only anonymous git access described here:
https://confluence.lsstcorp.org/display/LDMDG/LSST+Code+Repositories

Once we're on GitHub, you'll be able to create PRs against our code in the
usual way there.

Jim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.lsst.org/pipermail/dm-users/attachments/20150210/2f279708/attachment-0001.html>