I am not sure with whom to share this idea: my current contact with the computing world is mainly via this forum. Nor am I sure that this has anything original so I assume the risk.
The current way to overcome a difficulty with an application, which usually manifests itself by the emission of a machine-message to the user, consists in following non exhaustive actions:
1) Read the message, correct your environment (inputs, whatever) accordingly and re-try.
2) Read a related help page, if any, to find out what you did wrong, correct, and re-try.
2) Neither the message nor the help can guide you, read the source -assuming it’s handy and within your competencies- to figure out what’s the expected environment the program does not find and hence complain, correct and retry.
4) Use the message as an index in the web, read other manifestations of the problem and fixes found by other users; if any. Try the fixes for problems more closely related to your environment.
5) Register to a forum dedicated to the project / program with which you encountered the difficulty -usually these resources will surface in step 4- and post a question, read the answers and try them.
If you are lucky, you will have a fix for this difficulty, and will use the program until the next difficulty. If you are not, you will do without the means offered by the program or, will try other programs offering equivalent means.
Every project is granted from a centralized authority (as IEEE for MAC addresses) a bunch of “unique machine message identifiers” (ummi) for the project.
Whenever the programmer prints a user message, it associates it with one ummi out of the lot, to be printed together with the message. Automatic means are provided to insure exclusivity of usage of any given ummi in one project
This ummi can be used by the user to point directly into the “machine messages data base” (mmdb) where the author (maintainer) of the program writes clearly the discrepancy between the expected state and the actual state that the programs complains about: this is the “developer view’ filed of this ummi in the db and it belongs to the author/maintainer alone. In addition to the developer view, there may be “user views” field which is filled as wiki. The user view is either completely free or partly guided; e.g. must enter version / build / platform etc, followed by how did the user resolved the difficulty. Should the user, helped by these resources, get to solve his problem, he will no for sure where to add his own input for the benefice of all other users.
It’d be nice if the ummi scheme and the mmdb were common to all operating systems and more generally to any platform such as car’s or washing machine’s computer but maybe Debian could pioneer this concept.
Good weekend everybody.
Discussion about development of the Debian OS itself
1 post • Page 1 of 1