RPKP Update No. 7

September 2nd, 1996

This was originally sent out to members of the RPKP mailing list. If you would like to receive future updates, please see the instructions later in this document explaining how to subscribe to the mailing list.

Time of Transition

Matthew Watkins, founder of the RetroPsychoKinesis Project (RPKP), writes as follows:

For a complex of personal reasons which I needn't elaborate on here, I shall be unable to continue my work on the RetroPsychoKinesis Project. Fortunately, John Walker (creator of the HotBits WWW random number server which went on-line this summer) in Switzerland has agreed to adopt the Website and all that goes with it. This can be thought of as a sort of RPKP/HotBits merger. John is a highly competent programmer (his abilities include Java, conveniently) and we have a very similar vision of where the project could potentially lead, if executed with sufficient care. I'd like to express my thanks to everyone who has contributed to the Project, and urge you all to continue following its progress, as I suspect that at this stage, John will be able to get things done more effectively than I would.
Best wishes,
Matthew Watkins

I (John Walker) would like to be the first to express to Matthew my admiration for what he has done so far with RPKP and my dismay at his not being able to carry on the work himself. When he asked me if I would be willing to become the new host of the RPKP site, I agreed immediately because I thought it would be most unfortunate if the valuable archive assembled so far were to disappear from the Web. I cannot promise rapid progress on developing the project toward its intended goals (at the moment I am involved in several already-underway Web projects in other areas, so the time I'll be able to devote to RPKP will be severely limited for the next few months), but at least the work done to date won't be lost. Like Matthew, I eagerly solicit those with programming skills to volunteer their talent to further the implementation of on-line RPKP experimentation. Later in this Update I will briefly describe a Java toolkit I've made available to facilitate such development, and progress to date on a prototype on-line RPKP experiment.

The purpose of this Update, which I hope will be the first and last devoted primarily to administrative matters rather than substance, is to explain the relocation of the site and mailing list, and what to expect regarding RPKP in the future.

The New Site

The complete RPKP archive which used to be located at http://alethea.ukc.ac.uk/Dept/CPRS/RPKP/ has moved to:

http://www.fourmilab.ch/rpkp/

The administrator of the original server will allow the directory there to remain active for several months; the documents there have been replaced with pointers to the new site. You will find the content of the archive at www.fourmilab.ch to be absolutely identical to that at the original site (other than mechanical changes due to the relocation of the pages and some optimisations so images download more rapidly). If you have made links to RPKP documents in your own pages, please update them so users are sent directly to the new site.

The www.fourmilab.ch site is heavily loaded (about 24,000 accesses per day) with respect to its 128 Kb connection to the Internet, so response may occasionally be slow. The load on the site peaks around midnight Universal time and is lightest in the hours around noon UTC. Since the documents in the RPKP archive are predominantly text and reasonably short, I doubt you'll experience serious delays accessing them here.

Automated Mailing Lists

The RPKP mailing list which Matthew Watkins previously managed manually has been installed as an automated mailing list at this site. All subscribers on Matthew's list are now subscribed to the automated list, which is being used to distribute this message. One can subscribe to the list by sending E-mail to rpkp-request@fourmilab.ch with the word subscribe in the message body (not as the Subject). To unsubscribe from the mailing list, send a message to the same address with unsubscribe in the message body. To distribute a message to all subscribers, send it to rpkp@fourmilab.ch.

The rpkp list is open to all and unmoderated: anybody can subscribe, and anybody can send a message to be distributed to all subscribers. This is the policy for most of the mailing lists at this site; so far it hasn't been abused, and I hope it won't be in the case of this list. If subscribers start getting bombarded with unwanted and/or off-topic messages, I will make the list moderated, but I'd rather not since it's a lot more work for me (having to personally approve every message before it is distributed), and it delays delivery of mail to the list, especially if I'm out of town and can't check E-mail for several days.

RPKP Digest List

In addition to the rpkp mailing list, which forwards messages sent to it on a message-by-message basis, a new RPKP Digest list has been created. This list collects messages sent to the rpkp list and, when a sufficient volume have accumulated or enough time has elapsed, sends a single message incorporating all the individual postings to rpkp since the last Digest. If you do not mind the delay, subscribing to rpkp-digest is an excellent way to reduce the volume of E-mail you receive from the list. If you'd like to transfer your subscription from the rpkp list to rpkp-digest, simply unsubscribe from rpkp as described above and subscribe to rpkp-digest by sending a message containing the word subscribe in the message body to rpkp-digest-request@fourmilab.ch. Every message distributed by both lists contains instructions on how to unsubscribe.

Random Number Generators for Java

As an initial step toward implementing the proposed experiment, I have developed a set of Java classes which implement a superset of the built-in Java pseudorandom sequence generator, while allowing one to easily plug in various underlying generators of both random and pseudorandom byte streams. A variety of pseudorandom generators are included, as well as a package which obtains random data across the Internet from the radioactive decay based HotBits generator. This will be a central component of the forthcoming Java feedback programs, including those which permit independent validation of results and repetition of the experiments by any Web site (complete source code for all programs is available and in the public domain).

HotBits Availability Upgrade

In anticipation of increased load on the HotBits server once it can be accessed directly by Java programs anywhere on the Internet, and especially once the on-line RPKP experiments go live, I have implemented a "HotBits Proxy Server" which mediates between the user and the machine to which the Geiger tube is physically connected. This allows HotBits requests arriving from the Internet to be filled by a high-availability Sun server. I believe that with this upgrade, availability of the HotBits server is limited far more by the instantaneous connectivity of the Internet than the ability of the server at this site to respond to requests.

With the proxy server, HotBits requests can continue to be filled even when the generating machine is off-line for service. From the standpoint of RPKP experiments, the introduction of the proxy server makes no difference whatsoever; it merely lengthens the virtual paper tape between the generator and the experiment which ultimately consumes the bits. Extensive precautions are taken on both the generating machine and the proxy server to ensure that all bytes the client receives have never been examined by, or delivered to, any other person or program.

Experiment Development Progress Report

The central goal of RPKP has always been to make available on the Web a resource allowing anyone with Web access to perform RPKP experimental runs similar to those published by Schmidt. I had hoped to announce, in this Update, the availability of a prototype of such an experiment. Unfortunately, I have encountered serious portability problems among current implementations of Java which will take longer to work around than I want to delay sending this Update. The ability to develop interactive Web content which works identically on a wide variety of computers and operating systems is a central goal of Java, and I expect we will see rapid progress in that direction over the next year. At present, a number of rough edges remain and I do not want to announce even a prototype which may misbehave on one or more widely-used computer/browser platforms.

When posted, the RPKP experiments will be structured to permit critical evaluation by anybody with Internet access and replication by any site willing to install the software, which will be placed in the public domain without any restrictions. Should an apparently statistically significant effect emerge, it is essential that anybody interested in doing so be able to repeat the experiments. The free availability of the RPKP software and its ability to be easily reconfigured to use different random number generators should contribute to independent confirmation or refutation of whatever results the RPKP experiments at this site should obtain.

As ever, questions, comments and suggestions are very welcome.

The RetroPsychoKinesis Project (http://www.fourmilab.ch/rpkp/).