Additions to published papers:
Addendum to Harbourne constants and conic configurations on the projective plane
At the end of page 892 we discuss (really sketch) positivity of K'. It is quite easy to see that K' can be rewritten as an effective divisor using only the exceptional divisors and the strict transforms of irreducible components of our configuration. Indeed, if n = 2 we can observe that K' is nef, especially for a configuration of 4 curves of bidegree (1, 1) and 12 double points one has (K')^2 = 0 and e(Y) = 12. If now n >= 3, then using standard considerations we can show that K' is big and nef - in the paper we wrote that K' is ample. However, and fortunately, in this particular case this might be viewed as a typo since our idea was to emphasize only positivity of K' and the fact that our surface Y is of general type in order to apply the Bogomolov-Miyaoka-Yau inequality, and it does not interfere on the main result.
Addendum to On conic-line arrangements with nodes, tacnodes, and ordinary triple points
There is a second projectively inequivalent realization of the weak combinatorics
(d,k;n_2,t,n_3)=(3,2;0,5,3).
Consider the two smooth conics
Q_1 : z^2 + xy = 0
and
Q_2 : (y-x-z)^2 + xy =0,
together with the three lines
L_1 : x=0,L_2 : y=0,L_3 : x-y=0.
Hence the corresponding conic-line arrangement is defined by
\mathcal{CL}'\colon f=0,
where f is the product of the above 5 equations.
The arrangement \mathcal{CL}' has exactly five tacnodes and three ordinary triple points, and no nodes. Moreover, it is free with exponents (3,3). In contrast to the previously considered realization, where the three lines form a triangle, the lines L_1, L_2, L_3 are concurrent at (0:0:1). Since concurrency is preserved under projective transformations, these two realizations are not projectively equivalent. This means that there is the second branch in Theorem 5.7(4) and this particular weak combinatorics does not determine the arrangement up to the projective equivalence.This ommision does not touch other results obtained in the paper.