Currently I am following the naming of ODE.jl which follows the naming from
Matlab, which uses the input parameter tspan. Now this is problematic even
in Matlab when you get the sol struct version back as it gives the fieldname
as .x for this tout variable. Looking at codes like Dopri5.f I see that
they use xspan which is a far more meaningful name. The only issue I see
is that then in a sense you would want to document your codes with a call
signature of function dydt(x, y) which feels a bit strange to me. Maybe I
should just go all in on the tspan name and have the solution struct return
.t for the fieldname? Yes. I am going to do this. The ODESolution will
contain a .t field instead of .x.
Currently I am following the naming of
ODE.jlwhich follows the naming from Matlab, which uses the input parametertspan. Now this is problematic even in Matlab when you get thesolstruct version back as it gives the fieldname as.xfor thistoutvariable. Looking at codes likeDopri5.fI see that they usexspanwhich is a far more meaningful name. The only issue I see is that then in a sense you would want to document your codes with a call signature offunction dydt(x, y)which feels a bit strange to me. Maybe I should just go all in on thetspanname and have the solution struct return.tfor the fieldname? Yes. I am going to do this. TheODESolutionwill contain a.tfield instead of.x.