[Tinymux] Features that strike my fancy
Mark hassman
mark at moosh.net
Thu Mar 24 22:45:50 EST 2005
Things I see that are likely not high on anyone's priority list either:
Local Extensibilty: similar to Penn's but tailored to MUX. I'm working
this and hope to have something useful to submit in a few weeks while
still in alpha period. The idea is to have very few entry points.
Primary will to be register callback functions keyed off specific events
(load db, save db, create obj, dest obj, ...). Those callbacks will
then be called with appropriate arguments on that event happening.
Second portion is to allow extending the command and function tables.
When finished, it should be capable of decoupling the various add-ins
(BT, WOD_REALMS, and HSpace) and make it available to various other systems.
The other aspect I was asked to look at was fiddling @mail. It appears
to suffer from the "many hands" syndrome but I've not really delved into
it. I'd gut and replace it, personally. Features requested are
differentiation on the To:, CC:, and BCC: fronts; separation of messages
between the note and the original when doing forwards; tracking of
senders by dbref to deal with senders swapping names in between;
capability of handling anonymous mail senders (configurable); ability to
re-send a @mail message if you forgot a recipient via recalling the
message from another recipient into the send buffer. I'm sure there
are numerous other ideas on that front.
On the longer term front,
Replace the configuration parameter handling with a system capable of
writing the current configuration to a file (either the current config
file or an alternate). The historical load only setup is rather
antiquated and should be replaced, IMO. Being able to dump the
configuration makes it more flexible for admins who are shell impaired
and easier for those admins who want to make a change on the fly and
forget it. The temporary behavior can easily be retained by not
writing the conf file.
Beyond that, since MUX is in the C++ world, I'd start to migrate away
from the entire predefined buffer lengths. The std::string class
provides enough to eliminate most of that but may have portability
problems with a few platforms. That would be a major shift but should
be better in the end.
Lastly, just toss up a list of "addressable items" that those of us not
in the pimary development loop can take a look at. You can always
reject the patches if you feel it doesn't meet the requirements of what
you want.
My $0.13.
-Mark
Stephen Dennis wrote:
>Perhaps it isn't a high-priority on anyone's list, but I'd like to
>enabling Chinese-language MUSHes by supporting UTF-8.
>
>Also, I'd like to change the way ANSI is handled and replace the LBUFs
>with an ANSI-aware string object. TinyMUX is using C++, so I could
>leverage that. Maintaining speed is the still the trick though, but
>with this, color is pervasively understood. Like so:
>
> 'add(%xr123,%xg123)' ---> '246'
>
>There are at least a handfull of things I'd like to tackle in MUX at some point.
>
>Brazil
>_______________________________________________
>Tinymux mailing list
>Tinymux at tinymux.org
>http://www.tinymux.org/mailman/listinfo/tinymux
>
>
More information about the Tinymux
mailing list