For those (very few) of you who didn't make it to the Steve Jobs presentation today at the Savoy (hi, Scott!), here are my notes. In the main they are taken directly from the slides used (which were produced using Concurrence with the Tekton font, for those who haven't seen Steve recently).

I was sitting with Nigel Metheringham, Malcolm Crawford and Paul Beaumont, so they can fill in any bits that I miss out (please).

The introduction was by Bernhard Woebker, VP of Europe, followed by Steve Jobs with a demonstration of NeXTstep, and closing with a "User Testimonial" from Jean Lequarre of UBS Switzerland.

• Founded 1985
• Global Computing
• primary focus has now switched to large corporates

• CEO: Steve Jobs
• President/COO: Peter van Cuylenberg
• CFO: Marcel Gani
• VP Europe: Bernhard Woebker
• plus 5 other VPs

• Strong background: Canon, Perot, Jobs
• Cash strong (enough for 1 to 1.5 years)
• Continued investment in R&D and sales channels
• Competitive Weapon: NeXTstep

• Chrysler Financial > 2,500 seats
• AIG (American Insurance Group)
• Abbot Labs > 1,000 workstations
• McCaw Cellular 3,000 workstations
• Mobil > $3 million

Dolphin Server Technology

(A spin-off from Norsk Data, probably will also get a NetInfo port)

The new equipment has been shipping for under two years (since January 1991); NeXT embarked on this new business strategy in Summer 1991.

• 550 employees
• in 1992 there were >50,000 installed systems
• 350+ productivity applications
• 100+ ObjectWare objects
• 150+ salespeople

Paul Strassman:

(CIO of the Pentagon, and author of "the Business Value of Computers")

Surveyed the percentage of company revenue spent on IT: for all types of organisation, the amount was roughtly constant at 2%, regardless of the degree of "success" of the organisation. When asking how the budget was spent, divided between management applications and operational applications, less successful organisations spent almost twice as much on management apps; whereas the most successful/profitable organisations spent slightly less on management applications than on operational applications.

Applications for operational productivity correspond to what NeXT has been calling "custom apps".

• '70s: mainframe and terminals
• '80s: superminis and terminals (PCs)
• '90s: server and workstations, and the cost is 10% of a mainframe solution
• BUT: applications still take 2+ years to deliver.

• Create Custom Apps 5 - 10 times faster
• applications have moved from the back office to the front office: the software application is now a vital part of the financial "product".
• "shrink-wrap" quality and reliability
• usable by "mere mortals"
• integrates seamlessly with 100s of great productivity applications; so can have one machine per desk; you can't cut and paste between two different computers

• took 100+ senior programmers and trained them on NeXTstep, then asked them to write the same app on both NeXT and their previous system.
• First application written was written 2 - 5 times faster.
• Savings were 90%
• 83% less lines of code in the NEXTstep version
• 82% said NeXTstep was better in ALL categories
• It isn't faster to code on NeXTstep; you just have to write less of it. The revolution is "getting rid of software".

(long list of "standards" supported by NeXT)

Financial Services, Government, Health Care, Legal, Higher Education (same lists as were shown at NEXTWorld Expo).

• DBKit
• Object Linking
• Novell and AppleShare connectivity
• PostScript level 2
• Renderman
• Global Computing

• Interface Objects
• Access Layer
• Adaptors (one month to write, can be switched without recoding an application)

• Application Software
• ObjectWare
• System Software
• New category of software
• more reliable
• improves "make vs buy" decisions
• > 100 in catalog

"Same as" NeXTstep in all ways

WordPerfect: two hours to port with a simple recompilation, no source code changes (they normally allocate 4 months to a porting project).

a major change in the operating rules.

For the PC market:

1. the PC, 1976 (Apple II)
2. Ease of Use, 1984 (Macintosh)
3. Client/server hardware, 1988 (Sun)
4. Custom Apps, 1992 (NeXT)

end of presentation, on to a live demo. This was basically the same as on the NeXTworld Expo video, a tour of NeXTstep based around Mail, looking at Mail, connectivity (NFS, Novell and Apple), object links, fax and OCR.

Steve did a lengthy demo of DBKit in Interface Builder, effectively prototyping an application on screen. This was also mostly the same as in the video.

Agrees with Jobs 2 years development cycle time. They tried substitiuting spreadsheets as the financial product builder, but they are easy to make miscalculations with. Also need real-time data feeds.

• must be very careful with design
• "trees and grass" - create lots of small objects
• Multiple links (important, but hard to create)
• static instantiation (in IB; prefers dynamic under program control)
• monster links (whatever he meant?)

UBS now do a two-phase development: they have one developer who works very closely with a trader (in a team), and produces prototypes. A further development team takes the initial work, and scans for reusable code for the library. The initial prototype never gets redeveloped, but can be recreated from the library code if much work is required.

NeXTstep development Improves Productivity

(Even with NEXTstep and OO programming) Prototypes are not the same as Production programs.

Q1) Why is JP Morgan still on the customer list?

Because there were two NeXT projects; one went well, the other (reported in the press) was cancelled. Reasons were really political, they wanted it to run on Macs and Suns.

Q2) Is NeXT a hardware or software company?

Both. NeXT would like hardware sales to account for 25% of all NeXTstep 'seats', if possible; but it' up to the customers to decide.

Q3) Will there be NeXTstep on SPARC?

No plans for it, but that could change.

Financial services companies typically change hardware every three years; a lot of Sun purchases are coming up to two years old.

Q4) What processors will NeXT use in future?

NeXT is working on RISC. They need faster and better video support than Intel (PC) boxes. The processor doesn't matter.

Q5) How many platforms will there be for NEXTstep?

Three by the end of the year (NeXT, 486 PC, RISC), maybe more.

Q6) I heard that the cost of a Intel box suitable to run NeXTstep will be around 10,000 pounds?

You were misinformed. It's about $5 - 6,000 in the USA.

Q7) Something about OMG and standards.

OMG stopped trying to set standards some time ago; now they are defining bridges for everyone's objects to talk to each another.

Q8) Is it objects all the way down the OS?

Steve mentioned various levels that were OO; said that Driver Kit will be released with NS486, which was OO.

Q9) Something about competition.

Taligent: thinks that APple will reject it, it is competing with System 8 and against the jobs of the Apple programmers working on it. How will Taligent sell its product? Taligent is "one of the best things to happen to us"; NeXT's approach has been endorsed by Apple and IBM.

Q10) When will NS486 be released?

Beta will be at the end of December, with a long (6 month) beta. Not much pressure from clients to release until Q3, which is when their own apps (developed on NeXT hardware) will be ready. Ship in early Q3 '93. 25 beta sites in January, 100 in April, and 250 in May/June.

Q11) Something about Windows NT.

Going for 100s of seats per order, not one and tens.

"Sun, USL are not factors in this equation"; about reduced devlopment timescales.

"We don't look at these people as real competitors".

"NT is better plumbing for Windows" (reduces the number of crashes; agreement from audience).

Cairo: OO file system technology; may be a competitor in the mid-'90s.

"The language is ten percent; the environment matters" (the appkit).

"NeXT has an order of magnitude more productivity apps than Sun, for example".

NeXT has 45 - 50% growth rate, meeasuring the second half year '91 against '92. "Our losses are behind us". NeXT could be profitable next year, but want to grow the sales force and R&D, which costs money. Will prefer to invest and breakeven rather than show a profit.

Go public? Maybe just this side of 1994, December 1993?

12) What will be in NeXTstep 4.0?

We are getting feedback from 50,00 customers, our best resource. Yes, definitely.

The exhibition included stands from: Trimac (PasteUp, Virtuoso and Calliope), Kapiti, Cambridge Animantion Systems (Animo), Insignia (SoftPC), DEC, White Crane (servers), Auspex (servers), and Data General.

In conclusion, and only interested readers will have read this far, so I don't expect any flames, the show was as great as normal from Steve Jobs, despite no new announcments. No mention of the Nitro, and there were no reactions to any comment about release dates of NS486 or the NRW. The audience reaction was interesting, the first time that I have seen Steve speak before a British audience. He didn't get responses where he expected it, but some points went down very well. The demo of DBKit delighted the (mostly financial) audience, in particular. Even more of that would have been very well received.

However, I was totally astounded that NeXT fielded only German speakers (apart from Jobs). Do they somehow imagine that the UK is part of Europe? Given the ease of finding decent speakers in the UK, there was no need to import a Swiss user. It would seem that, apart from van Cuylenberg, who wasn't there, there are no British NeXT employees who could present an introduction.

Paul