On Thu, Dec 15, 2016 at 6:51 PM, Pine W wiki.pine@gmail.com wrote:
My proposal would be that proposed UI changes which affect large proportions of the user base should be announced 3 months in advance. This would provide plenty of opportunity for discussion, synchronization, and testing of proposed changes.
A much finer definition is needed here. "Proposed UI changes which affect large proportions of the user base", to my mind, includes basically any UI change that affects any public user page, e.g., public Special Pages, article pages, etc. It does not include any specification about the significance of the change. (I understand this is intentional based on your replies in the previous thread.)
I am still of the opinion that small UI changes, regardless of how many users will be affected by the change, should not require a three-month holding pattern while they are discussed. The amount of developer time and productivity lost far outweighs the other consequences. Something like minor shade changes to existing colors, slight movements or alignments of specific containers, minor padding or margin adjustments, etc., are not significant, even if the change is on an article page, which would affect every millions of users. Something like a major color shade change, or a major refactoring of the design of a popular page, is of course another story.
My motivation for this is simple: I don't think software UI changes should ever be based on community consensus, nor do I think Wikipedia users (or rather, the subset of users that take interest in a Tech News announcement or the like) are a suitable group of individuals for making decisions regarding software development. Hell, not even the entire software development team of MediaWiki (volunteer or WMF) should be making such a decision. There's a reason why we have a UI standards group, composed of experts who know at least a thing or two about UI design. Does that mean these changes should occur opaquely without any communication? No, but the opposite extreme of forcing a three-month comment period is possibly even worse.
I do understand that there is a significant business requirement when it comes to announcing significant changes, i.e., those beyond the shred of a definition of what I consider a minor change. Be it community backlash and a resulting decline in community engagement by users who consider the volunteer software development team to be their mortal enemy, or instructional videos that will become outdated and need funding to recreate (which I don't view as a blocker for UI changes, but merely a consideration that should be acknowledged before making cost decisions of UI changes), there are some valid reasons why community notice should be given for a change. But it certainly should not be for *every* change. There should be a case-by-case decision of whether the given change would have a valid requirement for prior notice.
*-- * Regards,
*Tyler Romeo* 0x405d34a7c86b42df https://parent5446.nyc