If you use a non-standard UI, your app does not look good. Nobody wants to learn a new interaction paradigm for every application they touch; that simply doesn't scale.
Yeah, and its a pain in the arse. While native widgets aren't perfect, I'm much happier to deal with them instead of a reimplementation of one. Even when the web guy tries, they never work exactly the same, and they hardly ever try. The browser is a box where users have to throw away all the assumptions they have about the way controls behave because everything is done wildly different everywhere.
Users are forced to return to figuring everything out as they go, and then start over again at square for the next website. Visual cues are frequently sacrificed for the visual mess someone called a "design". For people highly familiar with native widgets, they just can't navigate around as easily in a webapp as they can in a first-class gui citizen. Not only because everything is different, but because the extended functionality that comes baked into real toolkits just isn't present in the web reimplementation. Things like the history api are even further eroding what assumptions can be made about gui behavior on the web.
Comments
If you use a non-standard UI, your app does not look good. Nobody wants to learn a new interaction paradigm for every application they touch; that simply doesn't scale.
You are wrong. Have a look at Tweetdeck, it has a non-standard UI, looks good, and is successful.
Moreover, the fact that you use non-standard UI does not necessarily imply that the user has to learn a new interaction paradigm.
Used the web?
Yeah, and its a pain in the arse. While native widgets aren't perfect, I'm much happier to deal with them instead of a reimplementation of one. Even when the web guy tries, they never work exactly the same, and they hardly ever try. The browser is a box where users have to throw away all the assumptions they have about the way controls behave because everything is done wildly different everywhere.
Users are forced to return to figuring everything out as they go, and then start over again at square for the next website. Visual cues are frequently sacrificed for the visual mess someone called a "design". For people highly familiar with native widgets, they just can't navigate around as easily in a webapp as they can in a first-class gui citizen. Not only because everything is different, but because the extended functionality that comes baked into real toolkits just isn't present in the web reimplementation. Things like the history api are even further eroding what assumptions can be made about gui behavior on the web.
Yes, the web is like that. How does that affect my argument?