People have been looking for this one for a long time now. Given that the constraints aren't obviously technological in nature, we might be waiting a lot longer. Incidentally, this is something of a minor trope in sf, the idea that the use of machines would have a profound influence on the spoken and written forms of communication. Niven had a bit in one of his Leshy Circuit stories where thousands of years of computer use eventually resulted in one civilization adopting a base eight number system, for example.
I don't really buy this one . . . but I'm old. And inflexible. Not plastic. My 17-year-old, otoh, has no problem with weird abbreviations, using swype on her phone, etc.
]]>Bellinghman @ 106 32k RAM is perfectly adequate for a Nuike power-station. PROVIDED you think clearly about what you want to do and control Sorry, as may be obvious from my posting abopve, I still tend to think that way, as computer controlling specific tasks - and very well, if the code is written carefully and economically.
SM @ 112 Are there not " Alt+numeric" codes for German / Greek characters available? I use them for the odd bit of German, especially the Umlaut-carrying ones....
Generally, in spite of their coming economic dominance (again) Chinese glyph-symbols won't take over the world. Too inflexible. Alaphabets are flexible. Anyway - who wants to go back to cuneiform (which is where chinese originated) - though they won't admit it - not being invented in the "Central Kingdom"
]]>(I tend to do so via the Demon webmail interface about once a month.)
That permits me to get the same emails on this work machine, on my home machine, on my laptop, on my wife's laptop, ...
]]>Greg, you understand cacheing?
IMAP: your email lives on the server. But when your IMAP client connects to the server it pulls in a local copy of the email. Then whenever you do something with it (mark it as read, file it in a mailbox, reply, mark it as spam ...) the IMAP client tells the server to update the master copy. There is some locking, so that while your IMAP client is tweaking the state of a file other IMAP clients can't do so simultaneously.
This means that multiple IMAP clients can maintain a relatively coherent view of the contents of the same mailbox.
The old client-server email protocol, POP3 ... well, the mailbox lived on a server; when the client connected to the server it slurped down the entire mailbox, and, if and only if you'd set it to do so, deleted the original. POP3 lacked the synchronization capabilities of IMAP4, so is a lot less useful to those of us who read email on multiple devices.
]]>Lack of tactile feedback is a big problem (although I can make do at a pinch with the audio "click" that an iOS device produces when you hit a key). A chord keyboard would be fine ... but it'd have to work with everything I currently use; providing bluetooth HID compatibility and a full set of key mappings would go a long way towards this.
]]>A big problem with Microwriting is that it presupposed the user was right-handed; AIUI no left-handed version was ever produced (the market was probably too small at the time).
]]>(And a bit before I encountered a mouse.)
]]>Pros over the ZaggMate keyboard/cover I've been using hitherto: it's a folio-style case so protects the back of the iPad, the keyboard is somewhat improved (better battery life), feels more secure when used on your knees.
Cons: it's about 200 grams heavier than the earlier version. (See "added protection".)
All told, it's about as good as it gets if what you want to do is to turn your iPad into a netbook. My complaints about cursor movement remain, but it's solid enough to tempt me into trying to do some serious writing on it ...
]]>if you have each app running atop a single-process OS instance, handled by a VM layer, how do you manage a GUI with multiple windows? And the simple answer is, you don't.
Sure you do. X11, which is now more than 25 years old, has a good part of the necessary architecture: it's built on a network architecture, with the server (read display) connected via socket to it's clients. And the window manager process is a client, so it doesn't have to be on the server machine (though not having it there raises performance issues). The point is that each VM has a network address and so you can put clients and server in separate VMs and they can communicate easily. I'm not advocating using X; there were design decisions in the X architecture that I was skeptical of at the time, and am still dubious about. Starting fresh with multi-VM architectures would be a good idea. But there's nothing about a multi-VM architecture that rules out a GUI with window management.
]]>No, certainly not. I have a lot of respect for the people who built X; I worked with one of them for several years, and used him as a source of advice on the system I was working on1. But there's a lot of art in the field that's come along since, and there's no reason not to take advantage of it.
One of the things that I think OS X did right, that Linux did not, was to start with the Mach microkernel OS. In the long term the security and dynamic subsystem plug-in capabilities the microkernel gives far overshadow the performance hit, and much of that can be gotten back in multi-core systems. A microkernel GUI and window system would be a interesting way to go.
1. I still say that if it hadn't been for absolutely brain-damaged management and some personal problems among one of the engineering teams, we would have pushed Sun Microsystems right out of the marketplace by 1986 (we had a Sun-3 competitor ready to ship months before the Sun-3 shipped), and Tektronix would have been the major force in the graphic workstation market. Certainly our network and graphic solutions were both superior to Sun's initial offerings, and I think our virtual memory architecture was better than anything available in a small computer at that time (including VAX-11/780, which we used as a our development machines).
]]>I might be about to eat my words over the title of this blog entry, because it's official: Scrivener for iOS is under development.
]]>Oh, and I mustn't forget the work on hypertext software, or the distributed test and measurement equipment (probes here, signal processing electronics there, and display and user interface a few miles down the road. That was my last project at the Computer Research Labs in Tektronix before they shut the lab down.) Now you know why I gnash my teeth when I think about working there.
]]>