mvhub-dev team mailing list archive

[Dan MacNeil <dan@xxxxxxxxxx>]

All this is from the top of my head, w/o double checking the code, your 
mileage may vary.

> After this step, I think we need to edit the BAD_* variables in
> nsp.conf to match those in conf/mvh.conf.

As far as I know, the important vars get set by mv_update_* , the ones 
that aren't set are cosmetic. Matching the two conf files probably 
shouldn't be a priority.

> Also, why is nsp.conf not
> created the same time mvh.conf is?

It is used as the source of various config variables BASE.dir for example.

>[email tests]

yeah, actually sending email isn't a very good test right now as it 
mostly tests the library we use to send the mail, which has it's own 
tests and doesn't return false on failure anyway.

changing them so they don't require a working email server, (which btw, 
isn't that hard) would be a good idea, t

If people are working on the email code, then the email tests should run.

All this is good and valid (thank you ) but outside the scope of your 
current project.

I have made a note to

On 02/06/2012 11:11 PM, Emaad Ahmed Manzoor wrote:
> Hi Dan,
>
> I thought I'll note down some issues I had while installing MVHub on
> my Debian Squeeze VM:
>
> 1. sudo cp link-to-live-code/app-mvhub/project-tools/templates/template.conf
> conf/nsp.conf
>
> After this step, I think we need to edit the BAD_* variables in
> nsp.conf to match those in conf/mvh.conf. Also, why is nsp.conf not
> created the same time mvh.conf is?
>
> 2. *.$USER.testing123.net
>
> Sites with names of this format didn't work for me for some reason,
> even after I added entries in /etc/hosts. I fixed that by changing the
> site_name in the conf to *.$USER.localhost instead.
>
> Also, there are some tests
> (lib-mvhub/t/Utils/Setup/get_dev_home_files.t,
> send_emails_contained_in.t) that make sense when run on the
> development server but probably not on a home development machine.
> Requiring a contributor to pass all these tests will raise the
> contribution barrier. Should we separate these tests at some point?
> Something like, tests that the contributor runs on his machine, and
> tests we run on bold.
>
> -- Emaad Ahmed Manzoor


--
dan@xxxxxxxxxx   ===> Official Community Software Lab, not private
dan@xxxxxxxxxxxx ===> Personal, private, side jobs.
cell: 978-455-1885