conversations
one
conversations
one
Here's my latest take: three routes --
1) corporate web slave: talent salaries start at about 50k plus bonuses and go up to around 100 or 125k.
2) start up: talent salaries start at about 50k plus stock and go up to about 100 or 125k plus significant stock. vesting times range from 2 to 5 years.
3) consulting: talent earnings run from a low of 65 or 70/hr to 135/hr. People are getting as much work as they can handle. No stock, bonuses or benefits.
A memento from the first World 3 party, Los Angeles, July 1995.
15 June 1995
>Top of the totem poll right now are generalists, who don't have any specific area of "super" skill, but no weaknesses. They must know Perl, UNIX, CGI, HTTP etc. And the going rate is $45-60K per year.
But finding anyone who’s any good is like finding needles in a haystack. There are 50 companies out there after every one person with talent. Everytime I take my team anywhere I have to fight people off with clubs.
4 May 1995
>>Those quotes are LOW. Prices have exploded.
4 May 1995
>What's your take, off the record?
5 May 1995
Minimal sites (not a lot of pages or graphics and very little scripting) are about 50k to start (and that's if they're running the site themselves and have a competent staff)
Middling sites (more cgi-stuff, some cutting edge programming, more pages) begin at about 75k
The big sites are beginning at 100k
Revenue generating sites (secure transactions etc) are beginning at about 250k
5 May 1995, Stanford University
This afternoon, Marc Andreessen, inventor of Mosaic and now co-founder of Netscape, gave a lecture to a grad class at Stanford. Attending the lecture was a literal who's who of human-computer interface design.
What follows is an embellished version of my shorthand scrawl:
Netscape estimates 6 million people use their browser; no market percentage was claimed
According to protocol analysis, the majority of IP packets being sent over the internet contain http, having surpassed email a few weeks ago
Lots of old metaphors are now being used to display information, those metaphors will break down very soon
Major Netscape customers are looking to the internet for salvation because they really don't know what or where their businesses are, anymore
Providers like Compuserve, AOL, Prodigy, etc. are in big trouble if they don't adopt an infrastructure that uses the internet model
There is still a place for services like usenet, where the same information only needs to be transmitted, once, instead of over and over again as with http
Some of Netscape's first, large customers were Penthouse, Playboy and Hustler
Proxy servers are a key software technology
Actual internet/web business application software is the growth market, not just browsers
A change of the page metaphor is imminent
VRML and Hot Java will support this change
Have computers become "geek-free" or have we all become geeks?
Privacy is still an issue, though not as big as before
Netscape is now accepting advertising on it's site, but is not leasing space on its server farm for other external content
He forsees custom protocols being developed for interactive sessions
HTML and PDF are complimentary technologies
Emphasized use of "push-pull" facilities as the basis for crude interactivity and background "multimedia" experience
Lost in hyperspace is still a big problem and lots more research needs to be done to solve this issue
--
by: William Barr
- statistics were provided by a study done at MIT (no other reference)
- the majority of users now access the internet via the web
- other methods (ftp,gopher,telnet) are sharply down in use
- most sites are electronic versions of shopping malls or newsstands
- these are limited in a hyperlinked environment
- on the web, you can jump from site to site, seamlessly; you can jump from a retailer's page to a manufacturer's page ... breaks down barriers
- they know what their businesses are not (ask a phone company if they can tell you exactly how much a local, party to party call costs them; they can't tell you the answer)
- communications and telcos
- publishing
- financial (banking, investment, insurance)
- computer/software (sales/service)
- Global Fortune 2000 companies
- currently, their backbones can't handle the increased level of traffic
- many users complain of slow connections and poor browsers
- even slow email response
- glass house, closed shop mentality
- most users want internet access, NOW!
- what took the providers so long, anyways?
- bandwidth requirements will not increase exponentially because browsers/clients will start doing much more work and custom, compact protocols will be designed
- Netscape commerce servers
- a server providing a key, single point of egress to the internet
- without them, large companies won't hook up
- security
- content control
- traffic control
- complete business applications that not only support electronic transactions, but re-define corporate-customer communications
- more than just a database, a complete electronic retailing/warehouse/inventory management system with a WWW front end
- HTML 3 will be the launchpad
- interactivity will be responsible for new metaphors
- springboard for new metaphor design
- interactivity will be the ultimate user control for page layout
- interactive browsers will let users redefine the layout of a site on the fly, at will
- indexing, navigational aides and content organization will quickly supercede current layout and design issues
- VRML = Virtual Reality Markup Language
- Hot Java allows video/audio and more interactivity
- "Doom!" like interfaces will be the next model for browsers
- current VRML does not support views of other people using browsers on the same page, Java will change that
- 3D scenes will be "commonplace" by the end of the year
- Hot Java is actually about 6 years old
- ultimately, user will have complete control over how content is viewed
- he suspects the latter, especially in light of the average user trying to network Windows 3.1
- Windows 95 will change some of that, but there are better examples
- current, publicly available encryption technology will require about 64 mips years of CPU time to crack a message
- it's easier to make existing encryption technologies harder to crack than it is to design faster CPU's; current encryption is good enough
- this form of sponsorship will be commonplace, soon
- user connects to site, browser downloads protocol for interactive session, after session is complete, browser forgets protocol, user jumps to another site
- PDF = Portable Document Format; a way of saving documents that are readable without needing to own the software that created it
- PDF will be the mechanism where providers can retain their control of content and layout
- soon there will be more browser improvements that will have little to do with HTML or page manipulation, but will facilitate data retrieval
- browsers will be able to retrieve info from databases (SQL) and other standard formats
- pass off more work to helper applications, designed to do specific jobs more efficiently (i.e. Would you use a browser to read your email?)
- a short term solution
William Barr, Stanford Computer Forum phone: 415-723-6632 ERL 448/450, Stanford, CA 94305-4055 fax: 415-725-7398
wbarr@leland.stanford.edu http://www-forum.stanford.edu/Multimedia.html
"My opinions are mine and only mine."
--