.

The history manager

The processing history

Every time you execute an algorithm, information about the process is stored in the history manager. Along with the parameters used, the date and time of the execution are also saved.

This way, it is easy to track and control all the work that has been developed using the processing framework, and easily reproduce it.

The history manager is a set of registry entries grouped according to their date of execution, making it easier to find information about an algorithm executed at any particular moment.

Figure Processing 30:

../../../_images/history2.png

History win

Process information is kept as a command-line expression, even if the algorithm was launched from the toolbox. This makes it also useful for those learning how to use the command-line interface, since they can call an algorithm using the toolbox and then check the history manager to see how that same algorithm could be called from the command line.

Apart from browsing the entries in the registry, you can also re-execute processes by simply double-clicking on the corresponding entry.

Along with recording algorithm executions, the processing framework communicates with the user by means of the other groups of the registry, namely Errors, Warnings and Information. In case something is not working properly, having a look at the Errors might help you to see what is happening. If you get in contact with a developer to report a bug or error, the information in that group will be very useful for her or him to find out what is going wrong.

Third-party algorithms are usually executed by calling their command-line interfaces, which communicate with the user via the console. Although that console is not shown, a full dump of it is stored in the Information group each time you run one of those algorithms. If, for instance, you are having problems executing a SAGA algorithm, look for an entry named ‘SAGA execution console output’ to check all the messages generated by SAGA and try to find out where the problem is.

Some algorithms, even if they can produce a result with the given input data, might add comments or additional information to the Warning block if they detect potential problems with the data, in order to warn you. Make sure you check those messages if you are having unexpected results.