In the previous section the business processes of a small department, developing a user manual for their software were mapped.
Hence firm decisions (with fatal impacts on the system, including its requirements) are to be made, archetypal users mentioned in business process model are discussed.
This section is a brief user persona definition. It is intentionally drawn back from the lengthly formal persona definition from the usability testing [@usability:persona], since a detailed persona definition proves less useful for a one man team.
The users are described with regard to their role and their knowledge and skills regarding the system. They are given names for further convenience of the thesis, but the part of persona definition regarding their personal life and the related features of the archetype personification is neglected.
The head of department is a lady called Hump and she just organizes the team. Her job related to the system is very limited, but includes potential access control setting. She is also a developer. In a scenario where Hump manages several teams access rights, it might be useful to keep the settings in a configuration file with the scripting potential.
The developers are programmer gentlemen called Dump and Lump. They are working in an IDE or coding editor and they are very efficient using it. It is best to let them work in their natural environment. Dump and Lump know the selected LML by heart and they are skilled at reading it from the source code, as well as writing it without additional visualization or preview tools. They are familiar with core VCS principals and use a VCS on their daily bases, when coding in the team.
Reviewer Rump is not a developer and just checks for readability, typos, etc. He makes subtle changes in the manual. Rump knows the LML syntax, but prefers an easy to use tool with a rendered preview to read the texts in a formatted document. He knows of VCS and its basic principles, but does not use it often.
The publisher named Pump makes no changes to the pages and just downloads the repository to his workstation and produces the desired outputs using a set of maintained scripts.