Hi Zaphod
Oops, You are right I won’t do it again!
Mike
No harm meant, so no problem. Don’t forget to remove the link at the bottom of your signature ! 
P.
suncover, did my post answer your question?
A looong time ago, I had a really bad experience with lucid dreaming.
I was to attend an exam (in 1986, I believe) on a subject I was really bad at, so I guess that subconsciously, I really didn’t want to attend it.
The exam was at 8.00 in the morning and the University was more than 1 hour away from where I lived in Paris, so I must have set the alarm clock at 6.15 or so. And I still clearly remember how I heard the buzz of the alarm clock, how I stopped it, how I woke up and took the metro, how I waited on the metro to get to university and this was as real as if I had done it. I had the feeling that my eyes were open and I was really feeling all this…
But at some point, it turned out that… I was still in my bed! And when I came out of this strange lucid dream, it was something like 9.00 am and it was too late for me to get there and be accepted to sit the exam, that I therefore failed by simply not attending it (this is the only time in my life this happened to me).
Since then, sometimes, when I wake up, I can open my eyes in total darkness and still see my dream going on. If I don’t stop this consciously (still remembering what happened to me with my exam nearly 30 years ago), I could go on dreaming with my eyes open, while I know I am awake. In fact, I have a very fertile imagination, which is really helping for my work (I am a researcher, so dreaming about how to solve problems or come up with new solutions is something important).
Now, for a much more positive example, last week, I was talking with musicologists who are starting a research project on Ligeti’s 10 pieces for wind quintet http://www.youtube.com/watch?v=ysuddlMbW9U (they are trying to understand how sounds mix together to create complex atmospheres and trigger feeling in listeners).
As for me, one of my project is on massively parallel evolutionary computing on the Cloud, where we are using possibly hundreds of computers to try to optimise a single very large problem.
One of the problems with Cloud computing is to monitor and control hundreds of computers that are working together. A part of the project is to create new scientific visualisation tools to tell if the run is developing well. If you have followed the thread on sleep, you may have seen that I typically sleep for 4 hours a night when I am loaded with many projects (which is the case now) meaning that sometimes, my attention can drop a bit if I am not actively participating to a meeting or working on a challenging task where I need to muster all my awake neurons.
So, a couple of hours after this discussion with the musicologists, I had a drop in attention at around 2.00 am, where my brain wandered into a strange mode in which I suddenly clearly saw each computer in the Cloud emit a note of its own, with a result sounding like a choir that would allow to monitor a run by simply listening to the cloud of computers.
It is this talk on Ligeti that triggered this thought. In my half dream, I clearly saw Ligeti’s Kyrie http://www.youtube.com/watch?v=JWqxPp6SvMw but not with singers, but with computers that were all singing a different note, but following a harmony similar to Ligeti’s.
Ligeti’s Requiem was used in 2001 a Space Odissey http://www.youtube.com/watch?v=GgqI32JX_jY&feature=fvwrel , which featured HAL, an intelligent super computer that was autonomously running the Discovery One spaceship. Even though I saw this movie quite a long time ago, its connection between HAL and Ligeti’s music was very probably a factor in triggering this semi-dream of having the cloud of computer emitting a complex sound that would allow to monitor the run.
Parallelizing evolutionary computation over hundreds of machines can be done through an island model where each computer looks for a solution to a problem independently, in what is called an ecological niche, and periodically, computers exchange individuals to avoid getting stuck in local optima.
If migration rate between islands is too large, all the computers will unite their populations of solutions into what is called a panmictic poopulation, and all computers will be in tune. This is detrimental to finding good solutions as it is important to maintain diversity during the run for exploration to allow to always find new local optima.
Conversely, if there is not enough migration between islands, all islands will converge individually on a different area of the serach space and the Cloud of computers will not converge towards a global best solution, so a too smal migration rate is also detrimental to the search.
Migration probability must therefore be tuned depending on the problem, the number of computers and the size of the population on each computer (which itself depends on the complexity of the problem to solve).
Migration probability is therefore an essential parameter that needs to be very finely tuned, but unfortunately, it depends on too many parameters to be able to tune it well for each different problem and computer configuration and topology.
Having each computer emitting a continuous sound (maybe with a voice spectrum for a more eery sound as in Ligeti’s requiem) will definitely allow to finely tune this essential migration rate: too many migrants and the sound produced by the computers will be too homogeneous, and too few migrants will mean that all the computers will emit very heterogeneous notes.
Creating a choir of computers will therefore allow a scientist to finely tune the run and get optimal results, in pretty much the same way a mechanics can tell how to tune an engine so as to maximise its efficiency.
Then, if a single computer goes crazy (because of a bug, for instance), it will also be possible to immediately detect this in the global sound of all computers because it will emit a sound of a totally different pitch than the rest of the choir. Here again, the analogy with a mechanics listening to a car to diagnose a problem is very good.
I contacted the musicologists again and they are very interested to help with this idea, so we will ask the post-doc whose task was initially to design the cloud scientific visualisation software to spend time implementing a musical monitoring of the cloud, because beyond the advantages described below comes yet another one that is absolutely major: with a visualisation monitoring tool, you need to watch a computer screen to see if everything goes well. This can be very tiring if you must monitor a 30 hours run like I sometimes do, to find critical solutions. I usually restrict the number of computers I watch to 20, and have 20 terminals on my screen that are outputting numbers, and the last time I did this from home, we had friends around and I periodically went to watch my 27" monitor with thousands of numbers flowing in 20 different terminals. Our friends were totally puzzled by this as it really looked like I was the guy monitoring the Matrix in the movie (but somehow, this is not very dissimilar to what must have been happening in Matrix). Where the guy in Matrix could see a women with a red dress in the numbers of his monitor, I could sometimes see that something was going wrong on a particular computer by seeing how individuals were or were not migrated between computers, as this translated into strange numbers flowing on terminals… I would then modify some parameters on the ill behaving computer (or restart it) but monitoring 20 computers out of 200 was not very efficient…
Now, rather than tediously observing a computer and trying to catch abnormalities through glimpses, with a choir of computers, I could now monitor a run while having dinner at a restaurant, by simply calling the cloud on the phone and listening to the run!!!
Or we could work on something else and answer our mails (or write on a forum) while monitoring a run by simply listening to it as if it were music, and if at some point, the music changed, we would know that something was happening in the run.
I very much believe that getting a Cloud of hundreds of computers to output sounds as a choir is going to give experimenters very new perspectives on fine tuning and monitoring a run.
And all this thanks to a conversation on Ligeti with musicologists, followed by a drop in attention because I was not doing something that was challenging enough. Since last week, I now found out that we should connect our computers to PureData http://puredata.info/ to create the sound environment, and the post-doc will probably start on this very soon.
So lucid dreaming can produce both good and bad outcomes, such as missing an exam by vividly dreaming you woke up and caught the Metro, while you are still in your bed…
Pierre
Thanks both Mike and Amodia - you have answered my queries 
That’s a very interesting idea. It’s also the kind of thing that would bring media attention, because it’s a good story…
I wonder if you could even add a visual aspect, maybe not exactly like this, but some sort of visualization:
https://code.google.com/p/logstalgia/
(I wish I could run logstalgia on my server, but there isn’t enough traffic yet. )
Something like that used to happen to me regularly when I was in elementary school. I would wake up, eat breakfast, get on the schoolbus – then get woken up by my dad, realize that it had just been a dream, and then have to do it again. I had no control over it though. 
Old thread, but I just had a lucid dream yesterday. Funny thing is, I converted all the numbers I saw with my PAO list, intent on memorizing them to remember the dream better. I kept seeing clocks, numbers on mailboxes and things and just thinking what would that number be. I was aware I was dreaming, but not thinking sober, was still mulled over.
Just a tip for lucid dreams; when youre in one, rub your hands together, or spin in a circle, or both; this greatly increases the chances of your lucid dream continuing. Focus on the sensations. It makes lucid dreams last longer and have greater clarity.
But I just found it extremely funny how I have it so ingrained to turn numbers into pao. Mailbox number 1528… time was 11:17 to around 11:46… For me, unlike other people apparently, numbers seem to stay stable. But that could be because of memory training and converting numbers into people. Usually for “normal” untrained in mnemonics people, numbers change when they look away and look back in dreams.
A way to get started for anyone wanting to have lucid dreams; draw a square of a color, or anything really, on your hand. Then, every time you see it, do a reality check. Pay attention to everything around you. Or look at your hands, does everything look normal? Or you could do this every time you enter a room, or sit down, or use the bathroom. When this becomes automatic, it will seep into your dreams and you’ll do a reality check there too.