Hi Don, You pretty well sum up what is wrong at PG. You begin by describing what a user can do that is the tool offered for editing to the user. Simple and idiot proof. What you have forgotten is that the to with its user friendly interface is actually muscling databases and vcs-system behind the scenes. From the discussions going on here on the list, you can see the worry about the right tools to use and except the average JADU (just another dumb user) to master a vcs. The paradigm at PG is flawed. they expect user to understand and master concepts they simply have no idea about. Allow me to go off topic to explain. A friend of mine manages the rental of spaces for boots in a building for the landlord. He ask If I could help him send out the bills and create a report for the landlord. Well, I sat down at his computer found numbers and pages. so I set up a database in numbers and wrote up a template for pages in around two hours. The crux was he continuously kept asking me what the hell are you doing and why. I tried to explain. he kept saying he could do it quicker and the hoops I was going through were not needed. I would reply just wait and see. When I was finally finished he said what I did was to much work for the task at hand for the time being. But wait I said! For next year all you need to do is change the data here. Go to pages and us the to templates and VOILA! everything Done in ten minutes at the most! Well, he sure was happy, und could master what I did, but he still said why I had to go through the through the troubles I had. THE LESSON to be learned is the making of good tools is al ot of work, but the have to be easy to use and get the job done quickly. PG says we want tools fast and does not care about ease of use. the exact opposite of the way things should be done. regards Keith. Am 01.02.2012 um 23:39 schrieb don kretz:
This exercise with Al's errata sheet turns out to be a useful context for thinking about the difference between thinking about the problem from the reader's point of view and thinking about it from a programmer's point of view.
Here a reader spent a bunch of time documenting changes he felt should be made to the text, and then, in his own words, from his own point of view, he described what the correction should accomplish.
Note that he doesn't mention checking out anything. He doesn't mention versions. He tells us a) where the problem is located in the text by page, b) a line or two of context; b) what it says now; and c) what it should say.
Which when you think about it is probably a pretty good description of how the user interface could work that would be simple, precise, and provide the feeling to the user that, yes, they can handle this.
It should be possible to display the current PG text, allow them to browse to the position of interest, display an edit box with the general area, allow them to make the proposed correction, click "Submit", and display the text as it would then appear to give them closure and complete the transaction.
I'm not sure where in there one would find a requirement for a version control system. Certainly not as the starting point for a design discussion.
Don