Unfortunately there seems to be an embarassingly large number of 
mistakes in my papers, though all pretty harmless:

-----------------------------------------------------
Bounded delay packet scheduling in a bounded buffer,
Operations Research Letters 38(5) 2010
-----------------------------------------------------
p.397, second column, near the middle, the paragraph that starts with
"We show by induction on time t...": "rejected by OPT" should be
"scheduled by OPT".

Also, the next few paragraphs may be confusing since it was implicitly
assumed that an F-charge is associated with the packet in OPT
that creates the charge, so for example |F_t| is the number of packets
generating those F-charges in F_t. Similarly for G_t, F^1_t etc.

--------------------------------------------------------------------
Online Preemptive Scheduling with Immediate Decision or Notification
and Penalties, COCOON 2010
--------------------------------------------------------------------

There is something not quite right in the proof of Theorem 1, although
I can't remember what it is now... it is fixed in the journal version.

----------------------------------------------------------------
Online interval scheduling: randomized and multiprocessor cases,
Journal of Combinatorial Optimization 16(3) 2008
----------------------------------------------------------------

p.254, line 5, the formula should be

f_{k+1}(y) - f_k(y) = \frac{1/\alpha^{k-1} - y/\beta^{k-1}}{...}

i.e. the + in the numerator should be -. The conclusion in the
following two lines remain correct as this is even more obvious that
it is a decreasing function.

------------------------------------------------------------
Online scheduling of unit jobs with bounded importance ratio,
Int. J. Foundations of Computer Science 16(3), 2005
-------------------------------------------------------------

p.597, Lemma 12, instead of 2-1/(B+1) it should be 2-2/(B+1).
The graph in Figure 6 already shows the correct values.