在校内上再次看到这个视频,很早以前Bumptop的演示片,没想到现在被Google收购了。视频里TechMessager的评论讲得很振奋,在校内上的流行度也跟那个烂了大街的six sense有一拼。其实吧,手势操控除了看着炫,它的真正效用还是要加个问号的。前几个月大佬Don Norman也在Interaction杂志上发表了”gestural interfaces are unnatural“的观点。我最近刚刚重新整理完相关的研究综述,其中几个关键论文很适合用来分析Bumptop、多点触摸、以及手势操作。
先挖挖祖坟。第一个多点触摸系统早在1982年就在多伦多大学的Input Research Group实现了。可是此后的20年间,这个地球上也许没有多少人知道multi-touch是啥玩意儿,直到苹果抛出了iPhone,multi-touch才登上舞台。想起Bill Buxton说的,这二十年what the heck are we doing? 这是闲话。
First computer first telephone etc. were not from design innovation
Design research doesn’t lead to new product categories
Technology first, application next, needs last: there was no fundamental need for indoor toilet, but now it is.
Creative minds in technology, design ,science: Ideas first, justification later
Who is the audience for design research (esp. human-centered design): end user? store? distributer? Engineers, programmers, marketing, sales.
There is a gulf between the product and the research
Research output: publications, conference talks. Manufacturing, marketing, etc. are absent
Company product wants practical results, specifics, its output is shippable products
Steve Jobs fired usability people and designers at his return, just kept some graphics designers, but led to better products
Researchers are good at creativity but not producing reliable shippable products
The gap is not bad nor good, but fundamental
The need of a group between research and product – the translational engineering
Two types of innovation
Radical innovation
Incremental innovation
what human-centered design does
Technology drives innovation – the y dimension, meaning changes driven by design – the x dimension: e.g. Swiss mechanical watch and Japanese electronic watch and swatch; computers shifts from for work to for entertainment
另一种方法是依照任务类别分类(从底层到高层):注意力(attention),感知和识别(perception and recognition),记忆(memory),学习(learning),阅读、说和听(reading,speaking and listening),解决问题、规划、推理和决策(problem solving, planning, reasoning and decision-making)。
人们更擅长识别,而不是回忆。George Miller的7±2理论即指出,人的短期记忆中只能同时保存7±2块信息。这个原则可以应用在很多地方。比如Mac OS X的搜索spotlight,根据用户的一些模糊回忆如关键词等,列出可能符合的各种文件,这样就能利用让用户利用更为有效的识别能力(而不是回忆能力)从中找到正确目标。这被称为recognition-based scanning。