Is there a reason to not just add any required symbols on to the end of the password, or remove illegal characters (perhaps generating a slightly longer password to account for this)? There's a negligible effect on password entropy and skips all of this bikeshedding around replacing pwgen.
Password generation isn't something done often, and if it is, it's trivial to write a script to generate passwords according to your own needs, and make pass use it by setting EDITOR. I suspect most pass users do not need this feature. On Sun, Nov 13, 2016 at 2:10 PM Michael Hoff <m...@michael-hoff.net> wrote: > On 13.11.2016 21:36, Jason A. Donenfeld wrote: > > On Sun, Nov 13, 2016 at 2:57 PM, Michael Hoff <m...@michael-hoff.net> > wrote: > >> I don't think that `pwgen` is the suitable application for this. > > Why? It already supports lots of different options regarding symbols > > and numbers and whatnot. This seems like just another little flag for > > it. > > My point is that `apg` already provides all required features and - to > my understanding - `apg` is also available in all linux distributions. > Why invest time and effort in `pwgen` instead? > > Of course, it is possible to implement this in `pwgen`, also. > One has to change the central generator function interface `(char > *inbuf, int size, int pw_flags)` to also contain a character set for > symbols (which is consequently no trivial flag) and implement the > corresponding functionality in the random and the phonem generator. > > -- > S/MIME and PGP (0xA79D31C0) supported > https://michael-hoff.net > > > _______________________________________________ > Password-Store mailing list > Password-Store@lists.zx2c4.com > http://lists.zx2c4.com/mailman/listinfo/password-store >
_______________________________________________ Password-Store mailing list Password-Store@lists.zx2c4.com http://lists.zx2c4.com/mailman/listinfo/password-store
