Report 
by Ehud Shapiro, 
June, 1992. 

I was invited to attend FGCS'92 in order to evaluate its progress and present a 
report in the "Project Evaluation Workshop". Hence I will focus in my report on the 
workshop.  

The workshop seemed like a well-orchestrated psychological warfare against the two 
MITI officials attending, with side-effects for the representatives of the "private sec-
tor", i.e. lab chiefs from the participating companies. The troops used in this warfare  
were the "Logic Programming Mafia", who stood up, one after another, praised the 
Fifth Generation Project, and urged the MITI officials to extend the lifespan of ICOT.  

Even though the different presentations were not coordinated (as far as I know),  
they showed remarkable uniformity of opinion. So perhaps the opinions expressed were  
sensible after all. All speakers praised the project for its achievements, noting espe-
cially (F)GHC and PIMOS as the notable achievements. GHC was praised for being an 
innovative and elegant concurrent logic programming language. PIMOS was praised 
for being a complete operating system built from the ground up using FGHC/KL1. 

Most of the speakers also claimed that the hardware developed by the project was 
of a lesser long-term significance. Some went as far as saying that investing so much in 
hardware development was a mistake. All seem to agree that the fact that the software 
developed by ICOT was available only on proprietary machines diminished its impact. 

As for the future, everyone stressed that without a continued presence of ICOT, 
the research results produced by the project would vanish into thin air. One of the 
main functions ICOT should play is to make the software technology available on stock 
hardware, under stock operating systems (Unix), provide documentation and support 
for the software, and integrate the various changes and improvements the users of the 
software, who will have access to its source code, are bound to make. 

Every workshop participant got a collection of reports by ICOT visitors. Ogawa-
san showed me my 1982 report, which appears there. I must admit I read it with great 
- 93 -

interest, not remembering at all what I wrote there 10 years ago. Reading the report 
caused me to change my presentation. I started by quoting the goals of the project, 
as I saw them then: "[The Fifth Generation Project's] ultimate goal is to develop in-
tegrated systems --- both hardware and software --- suitable for the major computer 
applications of the next decade, identified by the Japanese as 'Knowledge Information 
Processing'. " 

"In addition ... the project is expected to elevate Japan's prestige in the world. It 
will refute accusations that Japan is only exploiting knowledge imported from abroad, 
without contributing any of its own to benefit the rest of the world. Hence the project 
aims at original research, and plans to make its results available to the international 
community. " 

In retrospect, I think that both of these goals --- realizing an integrated, innovate, 
and useful system, and elevating Japan's in the world --- have been fully achieved. 
MITI's recent decision to allow free distribution of software is consistent with the plan 
of making the results available to the international community. However, I agree with 
the other speakers in the workshop, that without continuing support from ICOT, that 
software will simply die. 

I agreed with some of the previous speakers, that the international impact of the 
project was not as large as one hoped for in the beginning. I think all of us who believed 
in the direction taken by the project, i.e. developing integrated parallel computer sys-
tems based on logic programming, hoped that by the end of the 10 years period the 
superiority of the logic programming approach will be demonstrated beyond doubt, 
and that commercial applications of this technology will be well on their way. Unfortu-
nately, this has not been the case. Although ICOT has reached its technological goals, 
the applications it has developed were sufficient to demonstrate the practicality of the 
approach, but not its conclusive superiority. 

This can be partly attributed to the short period available for application develop-
ment, give the software and hardware development schedule. But, more importantly, I 
think that this was the result of the applications being developed in an artificial setup. 
I believe applications should be developed by people who need them, and in the context 
where they are needed. The suitability of the software technology developed by ICOT 
cannot be fully evaluated until such applications are attempted. 

Therefore I made some concrete suggestions for the future direction of ICOT. Some 
of them are in line with what is already planned. I suggest that ICOT now focus on 
making its concurrent logic programming software technology widely available, and 
actively encourage and support research and development groups who will use it in 
real-life applications. Specifically, I suggest that ICOT will: 
- 94 -

- Port KL1/Pimos to stock hardware (both Unix workstations and commercially 
available parallel computers, including the recently emerging small-scale symmetric 
multiprocessors. I believe even a PC version of a mini-system can be quite useful for 
teaching and exploratory purposes.) 

- Initiate some standardization effort for concurrent logic 
  programming languages and systems. 

- Provide specification and documentation for the software. 

- Make the software widely available, with a GNU-like distribution policy. 

- Provide teaching material, tutorials, and consulting services 
  for users who wish to use the software. 

- Provide research grants for research groups who are interested in 
  applying and/or improving ICOT's concurrent logic programming system. 

Since the last point has raised considerable interest, I would like to elaborate on it. 

I think that applications are best developed by people who care about them. There 
are research groups around the world who are interested in developing applications 
using parallel computers. ICOT can more fully achieve its goal of international contri-
bution and impact by supporting such groups and encouraging them to use ICOT's con-
current logic programming system. Fostering an active user's community for ICOT's 
system will have many positive impacts. First, it will require ICOT to clean-up, fully 
specify, and document its system. User's feedback will suggest improvements, and the 
body of knowledge accumulated by users of the system will help the development of 
future applications, and suggest ways to evaluate and improve the system. 

I suggest that research groups involved in such activity will be offered grants for 
periods of 2-3 years each, for the sum of about \10M a year. Grants should be of-
fered to academic and research institutions on the basis of the scientific merit of the 
research proposal. The fact that such grants are available can be easily publicized 
to the international research community through electronic bulletin boards and other 
means, inviting grant applications, which include a research proposal and a budget 
proposal. The applications can be ranked using external referees, and the best ones 
will be selected by a committee. The research agreement could either state the results 
of the research be put in the public domain, or that they belong, independently, to 
the funding agency (ICOT), and the institute carrying out the project. Under such an 
agreement ICOT can, if it so chooses, put the results in the public domain. 

The return on investment for say, 50 such grants, meaning \500M a year, will be 
- 95 -

much higher than any investment in an application development team assembled inside 
ICOT for such a task.  

I ended my presentation with another quotation from my 1982 report: "The even-
tual success of the project will follow not from the amount of money invested in it, 
nor from the number of people working on it, nor even from the individual excellence 
of these people. It will follow from the coherent vision of its leaders, the genuine en- 
thusiasm that they generate, and from the promising path of research that they chose." 

I believe the statement to be as true today as it was 10 years ago. I thank ICOT 
and its leaders for these most exciting and rewarding 10 years. 
- 96 -