taob@pnet91.cts.com (Brian Tao) (12/01/90)
[moved from comp.binaries.apple2] From ART100@psuvm.psu.edu (Andy Tefft): > This is not a binary but I thought it might be useful for people > to see what this program does before downloading it. It is > *very slow* - took about 3 hours on my //c - and it doesn't > print anything except for the occasional "SMASH!" on the screen > (There were 8 of these in the 2 hours I paid attention). The program > ends by writing an output file like the following: > > seeds: 0 0 -8286 -27406 > SYSTEM CHARACTERISTICS > Mass of central star (in solar masses): 1.21 > Luminosity of central star (relative to the sun): 2.51 > Total main sequence lifetime (in million yrs): 4824.979 > Current age of stellar system (in million yrs): 2530.437 > Radius of habitable ecosphere (AU): 1.585 I also downloaded this simulation, and it is certainly interesting to read through the reams of data it pours out. Have you been able to find seeds which will create a system with an Earth-type planet (a "class M" planet)? I've done 5 runs so far and most of my planets are barren ice worlds or cloud-covered balls of hot rock. If you find any interesting seeds, please post them here! This type of simulation is GREAT for space RPG's like Traveller where you need a realistic star system for your campaign. I just need some guidelines for choosing the seeds to produce habitable planets. I'm surprised it takes so long on a //c I'm running it on my GS, and it takes about 30 minutes on average per simulation. This is at the standard 2.8 MHz clock speed. From that figure, a 1-MHz machine should finish it in about 90 minutes or so. \/\/\/\/\/\/\/\/\/ | Brian T. Tao | UUCP: torag!pnet91!taob | / \ | University of Toronto | INET: taob@pnet91.cts.com | \ The Apple II / | Scarberia, ON | taob@pro-micol.cts.com | / Lives On!! \ |:::::::::::::::::::::::::::::::::::::::::::::::::::::::| \ / | "Computer guru? Someone who got their computer a | /\/\/\/\/\/\/\/\/\ | couple of weeks before you did." (Alvin Toffler) |
neufeld@physics.utoronto.ca (Christopher Neufeld) (12/01/90)
In article <238@generic.UUCP> taob@pnet91.cts.com (Brian Tao) writes: > > I also downloaded this simulation, and it is certainly interesting to read >through the reams of data it pours out. Have you been able to find seeds >which will create a system with an Earth-type planet (a "class M" planet)? > >If you find any interesting seeds, please >post them here! > I found an interesting one. The seeds: 3040, 12042, -9948, 31225 forms a system with a few "Mercury"s, a "Venus", an "Earth" (pressure's a bit low, but easily in the liveable range), and oodles of gas giants, some quite huge. Question: I noticed that the first eight simulations I ran came out pretty well the same. In particular, the primary's mass never changed. I found out that the program always came up with the first seed either -17095 or -17094. Subsequent seeds may have been similar from run to run also. I've taken to running a one line random number generator to produce the seeds, then running ACCRETE. I've noticed this with other programs also. Is my GS unusual in failing to reseed correctly? Has it something to do with the keyboard buffer being active, and so the GETKEY sequence doesn't get a chance to increment the random number seed in zero page, since the keyboard is never "idle"? > I'm surprised it takes so long on a //c I'm running it on my GS, and it >takes about 30 minutes on average per simulation. This is at the standard 2.8 >MHz clock speed. From that figure, a 1-MHz machine should finish it in about >90 minutes or so. > That's the same time as it took on my machine (an unaccelerated GS). -- Christopher Neufeld....Just a graduate student | neufeld@helios.physics.utoronto.ca Ad astra! | S = k log W cneufeld@{pnet91,pro-micol}.cts.com | Boltzmann's epitaph "Don't edit reality for the sake of simplicity" |
bill@pro-gateway.cts.com (Bill Long, SysOp) (12/02/90)
In-Reply-To: message from neufeld@physics.utoronto.ca > Question: I noticed that the first eight simulations I ran came out >pretty well the same. In particular, the primary's mass never changed. I >found out that the program always came up with the first seed either >-17095 or -17094. Subsequent seeds may have been similar from run to run >also. I've taken to running a one line random number generator to Same here. Except mine seems to always come up with three 0's and then a non-zero number. >> I'm surprised it takes so long on a //c I'm running it on my GS, and it >>takes about 30 minutes on average per simulation. This is at the standard >>2.8 MHz clock speed. From that figure, a 1-MHz machine should finish it in >>about 90 minutes or so. >> > That's the same time as it took on my machine (an unaccelerated GS). Yep, 30 mins sounds about right for my GS. I also ran it on a IIe (unenhanced, unaccelerated) and it took about 1:30. Sounds good to me... | ProLine: bill@pro-gateway |Internet: bill@pro-gateway.cts.com | UUCP: crash!pro-gateway!bill | ARPA: crash!pro-gateway!bill@nosc.mil | BITNET: bill%pro-gateway.cts.com@nosc.mil +----------"Maturity is overrated" - Garfield-------->Pro-Gateway 214/644-5113
dpalermo@aludra.usc.edu (Dan Palermo) (12/03/90)
In article bill@pro-gateway.cts.com (Bill Long, SysOp) writes: >>> I'm surprised it takes so long on a //c I'm running it on my GS, and it >>>takes about 30 minutes on average per simulation. This is at the standard >>>2.8 MHz clock speed. From that figure, a 1-MHz machine should finish it in >>>about 90 minutes or so. > >Yep, 30 mins sounds about right for my GS. I also ran it on a IIe >(unenhanced, unaccelerated) and it took about 1:30. Sounds good to me... > Just for kicks I ran Accrete on an 8MHz "Zipped" //c and it finished in just under 30 minutes. Maybe this Accrete could be a new Apple // benchmark for accelerator comparisons - intesive memory/loop/comparison/ integer and floating-point operations with very little I/O. Anyone run this on a //gs with one of those Zip GSXs? Dan Palermo dpalermo@usc.edu
cbdougla@uokmax.ecn.uoknor.edu (Collin Broad Douglas) (12/03/90)
In article <13434@chaph.usc.edu> dpalermo@aludra.usc.edu (Dan Palermo) writes: >In article bill@pro-gateway.cts.com (Bill Long, SysOp) writes: >>>> I'm surprised it takes so long on a //c I'm running it on my GS, and it >>>>takes about 30 minutes on average per simulation. This is at the standard >>>>2.8 MHz clock speed. From that figure, a 1-MHz machine should finish it in >>>>about 90 minutes or so. >> >>Yep, 30 mins sounds about right for my GS. I also ran it on a IIe >>(unenhanced, unaccelerated) and it took about 1:30. Sounds good to me... >> > > Just for kicks I ran Accrete on an 8MHz "Zipped" //c and it finished >in just under 30 minutes. Maybe this Accrete could be a new Apple // >benchmark for accelerator comparisons - intesive memory/loop/comparison/ >integer and floating-point operations with very little I/O. Anyone run >this on a //gs with one of those Zip GSXs? > > Dan Palermo > dpalermo@usc.edu Well, seeing this message, I ran it on my GS with a Transwarp GS. also, to make it go as fast as possible, I put the program in and wrote the output to the a RAM drive. I ran the program twice. Both times it was under 10 minutes (I forgot to time the first one very exactly) but the second one was around 8:35. I'm sure the RAM drive helped a little here. After all, the output came to be 4 single spaced pages when I printed it out. -- Collin Douglas | "We demand rigidly defined areas of doubt | and uncertainty." cbdougla@uokmax.ecn.uoknor.edu | America Online: CollinD | -Douglas Adams from Hitchhiker's Guide
bchurch@oucsace.cs.OHIOU.EDU (Bob Church) (12/03/90)
In article <13434@chaph.usc.edu> dpalermo@aludra.usc.edu (Dan Palermo) writes: > > Just for kicks I ran Accrete on an 8MHz "Zipped" //c and it finished >in just under 30 minutes. Maybe this Accrete could be a new Apple // >benchmark for accelerator comparisons - intesive memory/loop/comparison/ Where and how can I get Accrete. I have a 10 mghz //c and would like to give it a shot. Bob Church bchurch.oucsace.cs.ohiou.edu