That makes sense.  I guess we could merge internal + external metadata on a 
regular basis.  

That said, I think it would be good to be able to REQUIRE a URL in the same way 
it’s possible to clone a package from a URL.

thanks,
wade

On Jul 26, 2014, at 4:14 PM, Iain Dunning <[email protected]> wrote:

> Not possible as far as I know, but is a reasonable request. A solution is to 
> make a fork of METADATA for your internal company network, that you 
> periodically sync with upstream. Maybe on a issue?
> 
> On Saturday, July 26, 2014 2:45:54 PM UTC-4, S Wade wrote:
> The problem is that these are internal packages that are not shareable 
> outside our company network (their urls are only internally accessible). 
> 
> thanks, 
> wade 
> 
> On Jul 26, 2014, at 2:21 PM, [email protected] wrote: 
> 
> > You should try to registered them, otherwise who needs the central package 
> > repo if anybody can configure dependencies from various sources. Or, you 
> > can write an install script that clones all dependencies for user. 
> > 
> > On Saturday, July 26, 2014 12:37:17 PM UTC-4, S Wade wrote: 
> > Hi all, 
> > 
> > I have a number of unregistered packages that depend on each other.  I’d 
> > like to be able to specify these dependencies, so I tried to add dependent 
> > URLs to REQUIRE but it seems you can’t just add a URL to the require file. 
> > 
> > I’d like users to be able to run Pkg.clone() to install the package and its 
> > dependencies. Is there some way to accomplish this?  I couldn’t find an 
> > obvious way in the Package documentation.a 
> > 
> > thanks, 
> > wade 
> > 
> 

Reply via email to