[lsb-discuss] LSB 3.1 Update 1 status (and minutes for 2/20 and 2/27)

Stew Benedict stewb at aysenterprises.com
Fri Mar 2 02:48:40 PST 2007


On Fri, 2 Mar 2007, Ian Murdock wrote:

> On 3/1/07, Stew Benedict <stewb at aysenterprises.com> wrote:
> >
> > On Thu, 1 Mar 2007, Ian Murdock wrote:
> >
> > > Stew reports that, at the command line level, the switch is "-T core",
> > > which is mildly confusing (since it's actually Core and C++) and "-T
> > > desktop" (or no option at all, since "everything" is the default here
> > > too). Since most users will be using the DTK Manager, I'm not too worried
> > > about -T core, but I suppose if it's easy to fix, we could change it.
> > >
> >
> > Could probably be changed easily for libchk,cmdchk. What text do you want
> > to use for the -T argument(s)?
> 
> Are those the only tools for which it applies? Is runtime test core only?
> In other words, to do the "Core and C++" profile, you'd run libchk and
> cmdchk with the appropriate -T option and just not run the desktop tests?
> 

Yes, the rest X tests are in vsw4, xts5 for the distribution testing.

> If it can be done consistently, how about changing the possibilities
> to "-T core,c++" and "-T core,c++,desktop" with the latter being the
> default if -T is not specified?
> 

OK

> > I was still seeing a problem with appchk, in that it was testing for and
> > passing X symbols when I specified "core", possibly because in the past,
> > the X libs were part of core. I don't know if Jeff got any further with this
> > bit. Pkgchk would probably inherit the same problem.
> 
> Let's make sure we look at this. We definitely want X to be part of the
> desktop profile.
> 

Yes. on the TODO, just not solved yet.

-- 
Stew Benedict




More information about the lsb-discuss mailing list