I doubt we'll see another leap as huge as from CLI to GUI
Unless you have a particular reason to believe that GUI is the end of evolution, then there's room for massive change.
Circumstances suggest that wimp is not the last word on user interfaces. Some factors to think about:
* The GUI has grown during an era of groups fighting to establish platforms, when there were constant newcomers to the scene which caused a large emphasis on creating systems that are as newcomer-friendly as possible. Those circumstances will change.
* As the scale of data increases, user interface needs change. Consider the progression from flat file structure, to directories on a single machine to searching through large datasets, and the different tooling we use. I deal with large numbers of small files, and have built custom tools to allow myself not to drown in scale. GUI tools and CLI directory systems are inadequate.
* A lot of GUI has been driven by the need to make glossy things that sell, rather than things that are useful. Look at changes to Windows since XP. As free software eats away at territory that is currently commercially viable, this will change. We won't have companies pushing gaudy bells and whistles that have shifted platforms over the last fifteen years. Stability increases in value.
* Plenty of power users have never been happy with the WIMP approach, and consider it a leap backwards from the CLI. 1980s word processing users, unix users, people who knew menu-driven mainframe systems before being "upgraded" to less-useful GUI replacements. There's strong interface loyalty to the Bloomberg interface from people who have also dealt with GUI and CLI tools, despite significant problems in that system.
It's unusual to know about groundbreaking things before they happen. I expect our grandchildren will look back at screenshots of today's candy interfaces and ask how we took it seriously.
Circumstances suggest that wimp is not the last word on user interfaces. Some factors to think about:
* The GUI has grown during an era of groups fighting to establish platforms, when there were constant newcomers to the scene which caused a large emphasis on creating systems that are as newcomer-friendly as possible. Those circumstances will change.
* As the scale of data increases, user interface needs change. Consider the progression from flat file structure, to directories on a single machine to searching through large datasets, and the different tooling we use. I deal with large numbers of small files, and have built custom tools to allow myself not to drown in scale. GUI tools and CLI directory systems are inadequate.
* A lot of GUI has been driven by the need to make glossy things that sell, rather than things that are useful. Look at changes to Windows since XP. As free software eats away at territory that is currently commercially viable, this will change. We won't have companies pushing gaudy bells and whistles that have shifted platforms over the last fifteen years. Stability increases in value.
* Plenty of power users have never been happy with the WIMP approach, and consider it a leap backwards from the CLI. 1980s word processing users, unix users, people who knew menu-driven mainframe systems before being "upgraded" to less-useful GUI replacements. There's strong interface loyalty to the Bloomberg interface from people who have also dealt with GUI and CLI tools, despite significant problems in that system.
It's unusual to know about groundbreaking things before they happen. I expect our grandchildren will look back at screenshots of today's candy interfaces and ask how we took it seriously.