You could laminate this post and wave it around the whole history and future of computing.
Success is the magic combination of “it’s time” and great marketing. Too big a jump with too many compromises like performance, quality, novelty or cost, and it’s dead on arrival. If what you’re doing is genuinely novel then prepare to be a footnote in someone else’s success story.
Nokia had a cell phone that could browse the web and edit spreadsheets in 1996. Microsoft had Windows CE on smartphone-sized 'personal digital assistants' in ~1998, and Windows XP tablets in ~2002. Meanwhile, the iPhone didn't come out until 2007, and the ipad until 2010.
It's only with hindsight we realise the importance of things like capacitive touchscreens and stylus-free operation; and the fact spreadsheet editing wasn't actually a must-have feature for a smartphone.
But you don't really need it to use it. One son bought the latest one, was handing the old one to his brother to use when the stylus popped out. He said "Huh. I didn't know that was there". Had used it for 2 years without noticing the lack.
Also, I found that stylus makes a world of a difference in case of 2-in-1 devices. A touch-enabled laptop is a gimmick; add a stylus, and you may just start using the touch screen quite a lot.
Dr. An Wang designated that his son to run the company when he (Dr. Wang) died. His son was a terrible businessman, and quickly drove the company into the ground. This was right around the time that most other computer companies in the 495 and 128 belts around Boston died off, as speculators moved on to the California/SV gold rush.
Passing on the business to sons and daughters is the habit of the East Asia, as well as the reason for businesses to fail. I am from Hong Kong and we keep laughing at a handful of local banks of staying in the 1960s for this reason. Similarly in Japan, but they find a brilliant way to work around this tradition: Business owners adopt adults as sons so that they can pass on the company to the right hands.
I had voice dictation software for my Windows 3.1 PC in the early 90s. Most of this tech has been proven for a long time, but it's taken decades to figure out how to make them usable and useful.
For instance, voice dictation software still has an error rate, so it has to be very easy to edit on the fly. You can't go through three different directories to start voice transcription, and then stop it and go back and so forth.
It's interesting how these trends have come and gone for the entire history of computing.
A constant shift between "computer-origin" and "real-world origin". At first, actual paper - you could say this is "real-world" but the interface is 99% because of how the computer works. TTY is again 90%+. But then we start introducing real-world components like folders, pages, and paper-based analogues like cut+paste.
For a long time the design focus seemed "let's make this as much like the real-world workflow, within the limits of technology". Freestyle is an excellent example of the nadir of this: a very "real-world" interface, as much as they could possibly do constrained by technology.
But there was a point where the limits reversed: "a computer can do this, but we have no 'real-world' way to display it". Design hammered against this problem culminating with the Skueomorphism of the mid 00's.
Casting off those shackles led to Flat design, but maybe that threw the baby out with the bathwater? Without any real-world grounding a lot of the basic affordances went with them, relying entirely on a user's previous experience with computers. A real ouroborous :)
Maybe with "soft skueomorphism" there's a reasonable middle ground for now?
But soft skueo doesn't approach some of the great things this demo represents.
Effortless human-first HCI, enhanced by computers, while ignoring "real-world" limitations. That's the dream.
You just need good abstractions that the user can understand. Skeuomorphism borrows from the real world but then your abstractions are essentially random, Unix has directories and files, smalltalk has objects etc.
Modern UIs have (?) I don’t know what the basic abstractions iOS is supposed to have in its UI that the user is supposed to manipulate, I guess web documents?
We had a Wang 2200S in junior high school marine science class - in the 70s. What a wonder it was! Easy to program- at least most BASIC keywords had shortcuts on the keyboard keys, and this is in a junior high school so we're talking self taught kids - and letter quality output to a Selectric daisywheel. We used it for marine science projects a little bit but I think the real reason we had it was to whet our appetite for the PDP-8A waiting for us in the next trimester's math class.
In the late 1970's, high school, I used a Wang 2200, PCS I believe it was. The BASIC was wonderful compared to anything else around (TRS-80 Level II, PET, Apple).
Disks were an out of reach luxury at this time, but Wang's 9-track cassette tape drive was great. Nothing like using an audio cassette recorder on microcomputers of the era. More like using a mainframe tape drive. You could seek for files by name. Back up or skip forward by files at a time. The software controlled the tape drive forward and backward. The 9-track head was across the entire tape, so you could not "flip" the cassette tape over. This was a custom drive, NOT an audio cassette tape drive. And it ran the tape considerably faster than audio drives.
I discovered that it was possible to use inexpensive Radio Shack C-15 tapes (15 minute) because the tape was thick enough. This was vastly cheaper than buying the cassette tapes sold by Wang.
For the couple years I used this in high school, I had lots of fond memories. But soon grew out of it when I used bigger better machines and languages in college.
Worked on the Wang 2200 seriers (MVP, other?) in the early 80's. Stationed in Panama and they were used at the local Army Communications Command for telephone operations and perhaps other uses. It supported 8 users, had 64K (?) and started with 10 MB hard drive - 5 MB fixed and 5 MB removeable. Eventually added a 72 MB drive.
I used it to reconfigure some Army communications software. Cut my teeth on BASIC with me, and it prepared me for full-time, non-military programming.
Off-topic: I really enjoy seeing online articles before 2010 or so that have modern stylesheets. It shows that writing is timeless, and the power of separating presentation and content.
The best thing -- well, one of -- about HN is that people who worked on the product or used it are around to chime in.
This article reads like an advertisement, so I would be curious to know what it was like to actually use this thing, and what "small" features that seem like minor details prevented this from being as successful as it sounds like it ought to have been.
28 comments
[ 3.7 ms ] story [ 74.5 ms ] threadI remember the pink eraser on the back actually erased.
https://www.youtube.com/watch?v=FRKzmFH7-cM
So why did the product (and the company) fail?
Success is the magic combination of “it’s time” and great marketing. Too big a jump with too many compromises like performance, quality, novelty or cost, and it’s dead on arrival. If what you’re doing is genuinely novel then prepare to be a footnote in someone else’s success story.
It's only with hindsight we realise the importance of things like capacitive touchscreens and stylus-free operation; and the fact spreadsheet editing wasn't actually a must-have feature for a smartphone.
Also, I found that stylus makes a world of a difference in case of 2-in-1 devices. A touch-enabled laptop is a gimmick; add a stylus, and you may just start using the touch screen quite a lot.
https://en.wikipedia.org/wiki/Wang_Laboratories#Decline_and_...
For instance, voice dictation software still has an error rate, so it has to be very easy to edit on the fly. You can't go through three different directories to start voice transcription, and then stop it and go back and so forth.
But finding the right UI is labor intensive.
A constant shift between "computer-origin" and "real-world origin". At first, actual paper - you could say this is "real-world" but the interface is 99% because of how the computer works. TTY is again 90%+. But then we start introducing real-world components like folders, pages, and paper-based analogues like cut+paste.
For a long time the design focus seemed "let's make this as much like the real-world workflow, within the limits of technology". Freestyle is an excellent example of the nadir of this: a very "real-world" interface, as much as they could possibly do constrained by technology.
But there was a point where the limits reversed: "a computer can do this, but we have no 'real-world' way to display it". Design hammered against this problem culminating with the Skueomorphism of the mid 00's.
Casting off those shackles led to Flat design, but maybe that threw the baby out with the bathwater? Without any real-world grounding a lot of the basic affordances went with them, relying entirely on a user's previous experience with computers. A real ouroborous :)
Maybe with "soft skueomorphism" there's a reasonable middle ground for now?
But soft skueo doesn't approach some of the great things this demo represents.
Effortless human-first HCI, enhanced by computers, while ignoring "real-world" limitations. That's the dream.
Modern UIs have (?) I don’t know what the basic abstractions iOS is supposed to have in its UI that the user is supposed to manipulate, I guess web documents?
Thanks you, Steve.
Disks were an out of reach luxury at this time, but Wang's 9-track cassette tape drive was great. Nothing like using an audio cassette recorder on microcomputers of the era. More like using a mainframe tape drive. You could seek for files by name. Back up or skip forward by files at a time. The software controlled the tape drive forward and backward. The 9-track head was across the entire tape, so you could not "flip" the cassette tape over. This was a custom drive, NOT an audio cassette tape drive. And it ran the tape considerably faster than audio drives.
I discovered that it was possible to use inexpensive Radio Shack C-15 tapes (15 minute) because the tape was thick enough. This was vastly cheaper than buying the cassette tapes sold by Wang.
For the couple years I used this in high school, I had lots of fond memories. But soon grew out of it when I used bigger better machines and languages in college.
I used it to reconfigure some Army communications software. Cut my teeth on BASIC with me, and it prepared me for full-time, non-military programming.
Good memories. Thanks.
This article reads like an advertisement, so I would be curious to know what it was like to actually use this thing, and what "small" features that seem like minor details prevented this from being as successful as it sounds like it ought to have been.