This topic contains 4 replies, has 1 voice, and was last updated by 
josh November 20, 2020 at 8:49 am.
-
Posts
-
joshThe organizational concept about has “experts” assigned to area by project management. Their views about quality are not external to the project. A particular expert might be tasked with reflecting the quality perceptions of a teen user, for example. The development process should involve argument & disagreement about conclusions where that is rational. The concept is for extra & quickly engaged eyeballs at the vague fringes of design & current efforts to meet up there.
-
joshQ: Individual developers, in practice, are continually making implicit micro-level decisions about the design that are subject to later inspection & revision. In contrast, early opinions about current design details & quality of implementation are likely to generate early focus on conflicts of opinion which can be distracting. At the same time, developers & managers do not wish to map frequent calls for expert-hat opinion to an implicit authority for an outsized weight in overall micro-level design decisions. How can conflicts be handled in a way that minimizes political distraction & preserves overall design ethos at the org level?
A: One idea is an active, standing design committee that automatically reviews any conflicts & responds by updating design docs & describing the mechanisms for arbitrating quality of implementation issues (a broader poll, test, whatever).
-
joshRhetorical Q – what part of software design evaluation is theoretical & what part is empirical? Most will agree there are lots of empirical parts. Is it better to get early feedback on the empirical parts? Yes, as long as it’s not too hard to do that or too much of a distraction. So how do we design the methodology for getting early feedback on the empirical parts in a way that’s not too costly or too much of a distraction? Cost compared to what – redoing a lot later vs. not too fussy with immature mock up? The proposed system leaves it to managers to judge when it is worthwhile to get empirical feedback on which parts of the system. Try to make that as smooth as possible & see if can be a general methodology win.
-
joshWe hope that some of the work being done on expressive languages for visual display will also be beneficial for more easily bring visuals into early design phase & transitioning to prototype & production code – words in design => visual mockup => comparison with current system, some of which is not mock up.
You must be logged in to reply to this topic.