[Wikimedia-l] Let's have the courage to sit down and talk about VisualEditor

[Federico Leva (Nemo)]

Erik Moeller, 31/07/2013 07:28:
> We can't just work through a
> mountain of feedback in a waterfall development model and hope that
> all our assumptions about how to fix this or that complex issue will
> work out in practice.

+1
Also, such an important feature cannot be based on biased feedback from 
a subset of users and projects.

Erik Moeller, 30/07/2013 18:03:
 > The steady stream of feedback has
 > been invaluable, and I think the changelog of the last few weeks
 > demonstrates that beyond all doubt.

I think it would be helpful, if possible, to give some guesstimates of 
this, i.e.: how longer a wait it would cost us to reach some rank of 
quality if the deployment was downscaled; or, what would be the 
"deadline" for feedback on aspects X and Y to be actually able to be 
processed and worked on, before previous development decisions become 
irreversible or the developers move on to something else.
	We have seen some other products stuck for a few weeks or months in 
semi-ready state, which have then been deployed and have experienced 
some problems that were not predicted; or other products which have been 
used only on en.wiki (with dozens of WMF staffers involved in processing 
the feedback from one single wiki) and which are later deployed to other 
wikis when the product is already in maintenance mode, so that those 
wikis will never have a chance to influence the development.
	If de.wiki users or other users discussing/voting on whether to delay 
wider VE deployment could do so knowing that "delaying by x months will 
make us wait y months more to get use case w to work", or "if we delay 
after day z, feedback from our community will not influence the 
deployment of w", the conclusions would be more meaningful. If one 
assumes that the cost of delaying is 0, as it's what we've been using 
for 12 years, of course the benefits will always seem to outweigh the 
downsides.

Nemo