Doc in `doc/configuration.md` (has to be started sometime, no?).
Configuration function has to be lamely-named set_algo(), because
set_algorithm() is already declared in algorithm.h (prevent namespace
conflict).
algorithm has to be added as global variable due to the way the
callback is done (by CCAN/opt, which in itself is nice).
This can be cleaned up significantly by (at least) introducing a
global configuration struct, but there is no reason to do it now
just for this - better a wholesale manana.
The config parser has a bug, and can only set boolean options to
non-default state. If a default state is given to an OPT_WITHOUT_ARG
option (as defined in the config table), it is set to the non-default state
anyway, because a CCAN function opt_set_[inv]bool() is used, which is not
meant for this.
Therefore, using such options from the command line is fine, but specifying
them in the .conf file results in above-described behaviour.
E.g., a user tests
"api-listen" : true
in the config, but changes it later to
"api-listen" : false
The option will be set true anyway.
A previous commit (8dd1c2d4af) has allowed
this, whereas previously only setting to true was allowed.
Currently boolead options that default to true and can be set false by
the user are explicitly defined as true, e.g.:
bool opt_restart = true;
There is no simple or right way around this. This is an issue with the
design.
Current help contains conflict statements regarding intensity
priorities.
Per commit d11df698d141988491494aa1f29c7d3595e9712b:
- rawintensity overrides both xintensity and intensity
- xintensity overrides intensity
Could be issue 88 (no reconnect to pools on network connectivity going
down for a while and then back up) is a regression introduced in pool
handling after cgminer 3.7.2. Pull in other changes to pool handling
just in case.
Tests show that this may be an issue with glibc's getaddrinfo(), and that
the last changes in cgminer might mitigate the issue somewhat.