Fix issues with Aki-Richards convention

[Esteban82]

Done with Claude 5.0

This PR fixes the issues reported in #6840 for the Aki-Richards convention.

I use the following input data:

112 32 25  30  90   0  4.5  112 33 Strike-slip
115 34 15  30  60  90  5    115 32 Reverse
118 32 45  30  60 -90  5.4  118 34 Normal

The fix was tested using the following scripts:

Issue: -S+l may not work as expected

The script doesn't plot any beachballs with +l.

gmt begin 6840 png
    gmt basemap -R110/120/30/35 -JM10c -Baf -B+t"Don't plot any beachball"
    gmt meca meca.dat -Sa1c+l -T0
gmt end

Output with this fix:

Issue: Sometimes the nodal plane lines are not drawn correctly:

gmt begin beachball5_T0 png
    gmt basemap -R110/120/30/35 -JM10c -Baf -B+t"Nodal planes lines not drawn (T0)"
    gmt meca meca.dat -Sa1c -T0
gmt end

Output with this fix:

Issue: -T0/red doesn't work

gmt begin beachball5_T0red png
    gmt basemap -R110/120/30/35 -JM10c -Baf -B+t"-T0/red doesn't work but"
    gmt meca meca.dat -Sa1c -T0/red
gmt end

Output with this fix:

Issue: -Lblue doesn't work

gmt begin beachball5_Lblue png
    gmt basemap -R110/120/30/35 -JM10c -Baf -B+t"-Lblue doesn't work"
    gmt meca meca.dat -Sa1c -Lblue
gmt end

Output with this fix:

Issue: -A+pred also doesn't work,

Output with this fix:

gmt begin beachball5_A+p png
    gmt basemap -R110/120/30/35 -JM10c -Baf -B+t"-A+pred also doesn't work"
    gmt meca meca.dat -Sa1c -A+pred
gmt end

Output with this fix:

Fix #6840

[joa-quim]

But we must stop submitting new PRs before 6.7.0 is created

[Esteban82]

But we must stop submitting new PRs before 6.7.0 is created

Do I merge it? or should I wait after 6.7.0?

[joa-quim]

Go ahead and merge.
True that we could make branches that hold until after release but we know how that works.