I think you're trying to a generic rule to automatically create argument sets. Firstly you only need the -o flag (override) if you are requiring different amounts of bits for different addresses. If all addresses require the same bit sizes you just list them all with multiple -r flags eg: hashcash -cdb20 -r 'webmaster@xxxxxxxxxxxxxx' -r '*@suspicious.org' etc. (by default wild card addresses are enabled (-M), also available as optiosn are simple string -S (no wild cards) and -E regular expressions) (in fact with that example the 1st one is redundant as it would be accepted by the 2nd -r anyway, but you get the idea). However if you want different addresses to have different bit requirements and you do this (no -o): hashcash -cdb20 -r 'webmaster@xxxxxxxxxxxxxx' -b 15 -r '*@suspicious.org' then webmaster@xxxxxxxxxxxxxx would be acccepted with 15 bits under the 2nd -r, which is presumably not what you wanted. So to force the webmaster requires 20 bits but other addresses (specified with wild card that also matches webmaster) require only 15 bits you would put a -o switch to say that 'webmaster@' overrides '*@': hashcash -cdb20 -r 'webmaster@xxxxxxxxxxxxxx' -b 15 -or '*@suspicious.org' that means if webmaster fails, it will not consider rule for '*@'. You will see message like this if it happens: hashcash -cvd -b10 -or adam@xxxxxxx -b5 -or '*@foo.com' 1:5:041019:adam@xxxxxxx::/7w1dJqcCfMYeqkp:000000000t checking against resource: adam@xxxxxxx no match: token has insufficient bits checking against resource: *@foo.com no match: overridden resource rejected: no valid tokens found check: failed Next question will -or as the only resource work? Well strictly speaking the -o applies to the preceding rule. However if it is the first or only rule the -o is effectively ignored, and so for the first rule -or is equivalent to -r. But I don't think I changed anything related to -o processing in 1.13. I expect this behavior to have been there for some time. Adam On Tue, Oct 19, 2004 at 02:33:01AM -0400, Atom 'Smasher' wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA256 > > when checking stamps, should i expect everything to work fine when the > first (or only) resource to be checked is specified with "-or"? > > this seems to work in 1.13, but IIRC it didn't work in an earlier version. > i'm not sure if this is a feature or a bug, but i'd like to request that > this behavior be documented and remain intact. if i can count on this, it > will make it *MUCH* easier for me to publish a dot-qmail recipe for > accepting valid stamps... if anyone's interested...