Diagnostics is a unified framework for code annotation, logging, program monitoring, and unit-testing.

The annotations of diagnostics provide simple trace messages, tracing method entry and exit, checks (which potentially fail), asserts (which should never fail), invariance checking on method entry and exit, and a number of annotations for testing.

A program built with diagnostics can be compiled either at audit, debug, or production level. At audit level, each component is also checking its own correctness, e.g., each class checks its own invariance. At debug level, each component checks that it is used correctly, e.g., a method is checking its preconditions but is not asserting its loop invariants. At production level, only those checks remain which can fail on corrupt external input.

The annotations provide a stream of logging events which are delivered to number of loggers. A logger can select a number of logging events and write them to file, or it can be used by a software package to monitor failures and to react on them. In particular, a unit-testing logger checks for erroneous behaviour during unit-testing.


For Debian and Debian-based distributions see also http://packages.debian.org/diagnostics

Latest News

Helmut Veith Stipend

The first recipient of the Helmut Veith Stipend for excellent female master’s students in computer science will be presented on March 14 at the following event: "More female students in computer science. Who cares?" Panel discussion with renowned scientists about diversity in STEM Studies March 14, 5:30pm, TU Wien The Helmut Veith Stipend is dedicated […]

Continue reading

WAIT 2016 in Vienna

The third WAIT workshop on induction is held between 17-18 November at the TU Wien. Details are available on the workshop page.

Continue reading

Two papers at POPL’17

Two papers co-authored by researchers from our group have been accepted for POPL’17: “Coming to Terms with Quantified Reasoning” by Simon Robillard, Andrei Voronkov, and Laura Kovacs; and “A Short Counterexample Property for Safety and Liveness Verification of Fault-tolerant Distributed Algorithms” by Igor Konnov, Marijana Lazic, Helmut Veith, and Josef Widder

Continue reading

Helmut Veith Stipend

Outstanding female students in the field of computer science who pursue (or plan to pursue) one of the master‘s programs in Computer Science at TU Wien taught in English are invited to apply for the Helmut Veith Stipend

Continue reading

Full news archive