Re: [Savannah-hackers-public] Don't multiplex mirror URLs that use nonfree JavaScript ?

[Thérèse Godefroy]

Le 04/02/2024 à 19:11, Ineiev a écrit :
> Hello, Thérèse;
>
> On Fri, Feb 02, 2024 at 04:44:58PM +0100, Thérèse Godefroy wrote:
> ...
> > I made a test list [1] with underscores as markers:
> >    '_' for URLs that use nonfree JavaScript
> >    '__' for mirrors that use Cloudflare services.
> ...
> > It probably wouldn't take much to adapt the Mirmon scripts so they
> > select all URLs for the "allgnu" list, and unmarked URLs for the "gnu"
> > list [4].
>
> Probably no, it wouldn't.
>
> > WDYT?
>
> I wonder if we could implement a more flexible approach
> to mark mirrors, something along these lines,
>
> NR - https://wicked-js.biz/gnu [*]
> SM - http://walled.mil/mirrors [**]
> TD - https://gnu-mirror.td [*] [**]
>
> [*] The mirror uses unethical JavaScript.
> [**] The mirror is part of a centralized surveillance network.

So basically you want 2 markers instead of one, after the URL rather 
than before? This will make the regex a little more complicated, but 
that's OK. I'll try to fix /prep/FTP in staging.


> > I have another question: how do these scripts select against alpha
> > mirrors? Do they look for "alpha" in the URL? Are the parentheses
> > necessary?
>
> The script that extracts mirror list from /prep/FTP skips anything
> written in parentheses, so alpha mirrors are ignored.

Thanks!