marketing discussion -- understanding: we face two independant decisions: what we do on hardware we might possibly have an option. what we do on software No option. We must get to the marketplace immediately, if not sooner. competive situation: Hardware prices are crashing. We just ordered 8 MACII's, 3 with 19 inch high res color, 3 Laserwriter Pluses, with ethernet cards, all for 51.3K (developer prices) TI etc. We have been aware of this for four and a half weeks. TI is about 3 months ahead of where we earlier assumed they were. There is no doubt that TI is making a major market moving announcement which will affect all segments of the Lisp Machine market. Can we survive in this market environment? two possible strategies: (1) try to find a "secure" niche away from TI (2) try to carve out a niche at the top of TI's market. --------- RG: path 1 is hopeless in the long run as well as being uninteresting. There is no real security of medimum or long term differentation in a attempted Niche such as VME bus systems. (1) People wont buy VME unless it is reasonably competitive with other options. (2) TI and others will put their frob on the VME if it represents a significant market. (3) VME is "old" technology at this point, and is not one of the two primary viable boxes. [SUN, arguably, represents a third viable box, but their new, cheap stuff is not suitable for the K]. path 2 is also difficult. We face a number of obstacles to acceptance: (1) TI is a big company with service, etc etc (2) TI's is contained within the MACII box, etc estimating roughly, TI has a 2K versus 5.5K cost advantage. (3) TI has been doing things right in that they have been building compatability with their installed base, evolving their software, We should have been doing this too, but that's water under the bridge now. TI is actively involved with COMMONLISP standards and has a effective human wave which is toiling away polishing, reorganized, documenting, etc. (4) Market Timing. TI has demonstrated their machine to a few customers. On the other hand, TI moves slowly and we may be able to match their eventual announcement with a demonstratable prototype of our own in the same time frame if we do things right and work extremely hard. (5) TI has a network of third party agreements and industry contacts. We have some assets which give us a chance: (1) Processor speed. This is extremely critical to our chances. we need a sizable factor. (2) Some remaining advantage in software structure, in some respects. We are down to our last chance. We have to stop doing stupid things immediately. We are up against a large, well financed company which is basically doing intelligent things. If we cannot achieve "war-time" type unity and work intensity levels, we have no chance and should quit now. technical discussion -- what has to get done Our only chance to produce a salable software environment in a resonable time is to port the Lambda stuff. (side note: it is highly compatible with TI, which is a plus). Making various goals compatible. -old discussion- -NEW software versus Lambda Compatible, etc- Technical means at hand to do this, namely the package root feature. RG will explain this here and now to the satisfaction of all concerned. Then, we must go off and implement it without further hassle. Actually, its already implemented, the only thing that needs real work in the regard is the K loader.