migrated from Trac, where originally posted by clange on 24-Aug-2010 3:13pm
Definitional FMPs are not enough in practice. When they define a symbol implicitly, they are not usable for computation. Therefore, Michael suggested to introduce, in addition to definitional FMPs, computational FMPs. These must have the form
<OMA>
<OMS cd="relation1" name="eq"/>
<OMA>
<OMS cd="the" name="symbol"/>
arg1
...
argn
</OMA>
f(arg1, ..., argn)
</OMA>
(Michael, is there any other possible form?)
See http://arxiv.org/abs/1006.4057, last page of section 2, and the remark on FMPs in section 3 for a practical use case.
migrated from Trac, where originally posted by clange on 24-Aug-2010 3:13pm
Definitional FMPs are not enough in practice. When they define a symbol implicitly, they are not usable for computation. Therefore, Michael suggested to introduce, in addition to definitional FMPs, computational FMPs. These must have the form
(Michael, is there any other possible form?)
See http://arxiv.org/abs/1006.4057, last page of section 2, and the remark on FMPs in section 3 for a practical use case.