On 2018-03-06, at 7:22 AM, G Alexander wrote:

> I like the auto calculate hashes
> For grabbing something from url sounds like having a lot of switches, 
> for archive type and depedencies

I was kinda hoping that the archive type might be figured out from the tail of 
the URL. All the use_bzip2 yes business if plenty confusing for people. Even 
the whole idea of master_sites and distnames takes getting used to.

dependencies would have to be manually specified. A basic template could be 
laid down for them, however, in the generated Portfile.

> what if link is a redirect to cdn, or something like Dropbox or symbolic

some things would have to be manually done, unless AI can sort this out in ways 
I don't know how to do...


Reply via email to