I'm curious about what happened. IIRC, as you mentioned too in your reply, that Java was supposed to run in embedded devices. It was supposed to be lean and fast. But I can't imagine the modern Java doing that...
Version 1.0 ran quite smoothly once we upgraded the machines from 4MB to 8MB of RAM (seriously!). But, of course, at those times 8MB was much more memory than the early smartishphones carried, so their Java version was heavily stripped down and almost good-for-nothing.
On the PC Front, James Gosling and cia made were an amazing team, but they pushed for very academic and cumbersome patterns that converged in the EJB architecture. Nobody in its sane mind would fall in love with that.
Two or three years later, the internet bubble fallout affected every technology.
EJB was not invented by Gosling but adopted from IBM. It combined over-engineered concepts from the mainframe world combined with objects and too much XML configuration.
>The background is that Terry Winograd, a professor of Human-Computer Interaction at Stanford University in Silicon Valley, had invited me to lecture on some of my work in 1998. After my talk, Terry invited me to tour his lab and meet some of his graduate students. One of the Ph.D. students was a bright young fellow named Larry Page, who showed me his project to enhance the relevance of web search results.
Many of those lectures are online. I was not able to find the 1998 one he mentioned, but here is one that Jakob Neilsen gave on May 20, 1994 called "Heuristic Evaluation of User Interfaces, Jakob Nielsen, Sunsoft".
He gave another one on October 4 1996 entitled "Ensuring the Usability of the Next Computing Paradigm", but I can't find it in the online collection, although it exists in the inventory of video recordings, however I can't find any 1998 talks by Jakob Nielsen in this list:
Here are some of the older ones that I think are historically important and especially interesting (but there are so many I haven't watched them all, so there are certainly more that are worth watching):
[...]
Bringing Behavior to the Internet, James Gosling, SUN Microsystems [December 1, 1995]:
I also uploaded this historically interesting video to youtube to generate closed captions and make it more accessible and findable, and I was planning on proofreading them like I did for this Will Wright talk, but haven't gotten around to it yet (any volunteers? ;):
This is an early talk by James Gosling on Java, which I attended and appeared on camera asking a couple questions about security (44:53, 1:00:35), and I also spotted Ken Kahn asking a question (50:20). Can anyone identify other people in the audience?
My questions about the “optical illusion attack” and security at 44:53 got kind of awkward, and his defensive "shrug" answer hasn't aged too well! ;)
No hard feelings of course, since we’d known each other for years before (working on Emacs and NeWS) and we’re still friends, but I’d recently been working on Kaleida ScriptX, which lost out to Java in part because Java was touted as being so “secure”, and I didn’t appreciate how Sun was promoting Java by throwing the word “secure” around without defining what it really meant or what its limitations were (expecting people to read more into it than it really meant, on purpose, to hype up Java).
It's not Java, it's the programmer. There are lots of non-hacker types churning out inefficient code using inefficient abstractions.
There are also people using Java for high-frequency trading application with realtime performance needs.
It's true! If memory serves from the Jane Street podcast, the literal NYSE ran for years on a single-threaded Java application. I still struggle to wrap my head around the wizardry that kind of thing must have required.
Actually a few years ago I knew that a significant percentage of option trading routed through a VBA+Access+Excel product developed by a Montreal company (forgot the name) and the earliest code started from 1998.
