me being way too anal-retentive again

[shadowdragon]

Ok, so one of the things that irks me (yet again) in the computer design sequences is there is no guideline as to to amount of basic data storage in the computers, as in text, pictures etc...

So- i was thinking about this and came up with the idea that the cpu rating could do double duty here and give us a general guideline over and above the "unless the adventure requires it, don't worry about it" policy I generally use.

What do you think of 1 page of text being roughly equivalent to .001 cpu, a 20cm^2 picture being about one page of text, and a holographic image 10cm^3 being 32x that, for .032cpu.

Too little, too much? why worry about it?
The title of this thread does say it all


thanks
Shad

 

[HG_B]

What do you think of 1 page of text being roughly equivalent to .001 cpu, a 20cm^2 picture being about one page of text, and a holographic image 10cm^3 being 32x that, for .032cpu.

Too little, too much? why worry about it?
The title of this thread does say it all


thanks
Shad

Workable for a ~late 1950's computer.

 

[shadowdragon]

okaaaaayyyyyy......... that's a start, but did you have a reason for your opinion or were you just throwing that out there and expecting me to have an apostrophe?:devil:

IIRC it has been bandied about that Traveller™ computer architecture is not what we use today, so I went with a decidedly low-ball option.

I dont really have anything with which to compare a 1 cpu computer but maybe a wristwatch or something along those lines- what did you have in mind?


Just to clarify, I am working from the T20 design rules.

 

[HG_B]

okaaaaayyyyyy......... that's a start, but did you have a reason for your opinion or were you just throwing that out there and expecting me to have an apostrophe?:devil:

One page of text is ~2k data. So, You said a model 1 should hold 1000X 1 page of text. (1/1000th the model #) For a model 1 = 2MB

 

[atpollard]

I suggest that a non-linear progression may be appropriate ... a Model 2 computer would not store twice as many pictures as a Model 1 computer.

Rather, the basic complexity of the data grows at a rate that renders previous generations of data miniscule in comparison.

For example, in the era of green CRTs and daisey wheel printers, a 'page of text' would have been an ASCII file with only the crudest of graphics at best.

In the era of color CRT monitors and dot matrix printers, a 'page of text' would include fonts and color images of limited resolution. Each new page would require more storage space than most of the old page data storage disks could provide. And all of the old ASCII files in existance could be carried in one hand using the new data storage technology.

Fast Forward to the era of laser printers, liquid screens and laser data disks. A Page could easily contain millions of colors, unheard of resolution (by earlier standards) and embeded video clips or animated illustrations. How many ASCII pages could fit on a Terabyte Hard Drive? How many punch cards would be needed to store even one issue of Stellar Reaches?

So what happens in the near future when a typical 'page' includes embedded encyclopedic hyperlinks, hours of ultra-high resolution video (so you can pause and zoom in for a closer look or watch it from another angle).

Or later on when the 'page' is an interactive 3D holographic presentation, scalable from the micro level to the macro level with no loss in resolution.

Each generation includes such a quantum leap in processing speed and storage density that all earlier systems become trivial files to store. And the 'page' of data expands to fill the new capabilities.

Just my thoughts.

 

[shadowdragon]

One page of text is ~2k data. So, You said a model 1 should hold 1000X 1 page of text. (1/1000th the model #) For a model 1 = 2MB

Now we are getting somewhere!

Actually what I meant was 1 page of text takes up 1/1000th of a cpu, so 1 cpu is 1000 pages. 1000 cpu is 1,000,000 pages, at ~2k each is ~2G of data storage at a model 1 ship computer. Im still thinking this may be a tad low but it is a starting point. One wonders at the number of pages in the current Encyclopedia Britannica?

By leaving pictures at 20cm^2 = ~1 page of text I was deliberately ignoring resolution under the assumption it would be basically determined as needed and by TL. Holo-anything I am starting at 32x text for a 10cm^3 hologram and left resolution out there as well because I was sleepy.

 

[HG_B]

Actually what I meant was 1 page of text takes up 1/1000th of a cpu, so 1 cpu is 1000 pages. 1000 cpu is 1,000,000 pages, at ~2k each is ~2G of data storage at a model 1 ship computer. Im still thinking this may be a tad low but it is a starting point. One wonders at the number of pages in the current Encyclopedia Britannica?

By leaving pictures at 20cm^2 = ~1 page of text I was deliberately ignoring resolution under the assumption it would be basically determined as needed and by TL. Holo-anything I am starting at 32x text for a 10cm^3 hologram and left resolution out there as well because I was sleepy.

Starship computers (even Model 1) are supposed to be equivalent to present day super computers. My TL 7 ball point pen has 4G of data storage.

Up your figures by millions for a computer built at a TL where interstellar star ships exist.

 

[shadowdragon]

Starship computers (even Model 1) are supposed to be equivalent to present day super computers. My TL 7 ball point pen has 4G of data storage.

Up your figures by millions for a computer built at a TL where interstellar star ships exist.

yes, I can see your point, but remember that included in that 1 cpu is the processor, volatile memory for programs, the basic programming for data manipulation (and for starship computers the ability to interface and control a freakin starship!) which is apparently hardwired, and the mysterious ability to hold a multitude of programs of varying PP requirements. Even the vaguely defined "Library Data" program is only a measly 1 PP program, which I really disagree with if it covers so many subjects that it gives a general +2 to gather information checks. Yet another thing to define better, blech!

I am trying to drop in yet another storage medium without adding vl. In addition I am trying to keep it simple for when I actually need to figure it out, and to limit it so my players don't decide their handcomp can in fact store all the relevant data on all the systems in the 3I.

Given the limited capabilities given to computers, comparing a model 1 to a modern super computer is IMO inaccurate. There is no way to build a "modern" computer in the design sequence- its a handwave to begin with and its a handwave based on late 60's / early 70's architectures, which are no where near what is in your TL7 4G ballpoint pen. I am working within the established rules as written (mostly) to define it a little better.

In that vein, I can agree to up the .001 cpu to holding 10 or even 100 pages of text, but I thinks that 32x text for holo storage may still be too small.

 

[HG_B]

Given the limited capabilities given to computers, comparing a model 1 to a modern super computer is IMO inaccurate.

But, that is what a Model/1 was meant to be.

 

[JustinInOz]

But, that is what a Model/1 was meant to be.

Just curious, where in the rules does it say that?

Considering most spaceflight activities have been done with way less than a supercomputer.

 

[Gray Lensman]

But, that is what a Model/1 was meant to be.

You do remember that it was 1977? I have went with MT Comp rules ever since it came out, NOT perfect for the current R/L even then but WAAAAAAAAAAY better than the Original or 2nd Ed CT rules.

Just a thought, YMMV

 

[HG_B]

Just curious, where in the rules does it say that?

Considering most spaceflight activities have been done with way less than a supercomputer.

It was in the original rules I believe. But, no we haven't even left Earth's gravity well with "way less" much less attempted interstellar flight... The Apollo had guidance from ground computers for navigation. It wasn't a self contained space craft in that sense.

 

[Gray Lensman]

dang shuttle used computer tech that was about 1977 at best, look it up

 

[HG_B]

dang shuttle used computer tech that was about 1977 at best, look it up

Right and it can't even leave orbit. It also uses ground based computing to plot courses and changes. Apples & sand.

 

[JustinInOz]

Right and it can't even leave orbit. It also uses ground based computing to plot courses and changes. Apples & sand.

You might want to look at the following: http://history.nasa.gov/computers/contents.html

I am not too sure what you are talking about in your reference to leaving orbit. I think you are implying that it is easier, hence requires less computing, than operating away from a large body. Operating within an orbit is "harder" then moving freely. The acceleration of the large body imposes additional computational difficulties.

The early spaceflights used alot of guys on the ground looking at screens to track instrument readouts. Much of the computing of courses et al was done by other rooms filled with nerds with sliderules rather than computers. Even the large computers that took up whole rooms, did not do that much in the day. Ground based or not, you don't need anything near a super computer to run the calculations needed for a starship.

The thing which is not accountable in this are the calculations for a jump. Given that this, at its core, is condensed handwavium, coming up with anything about the computing required for this is arrant speculation. You do need a person, or there could be fully automatic ships.

 

[HG_B]

You might want to look at the following: http://history.nasa.gov/computers/contents.html

No thanks. I have much better sources of info. I might go to dig out the Apollo program Tech manuals (for the entire system) out of storage. A close relative wrote ALL the tech manuals for the entire system (rocket to capsule). I am EXTREMELY familiar with the program.

Thanks though.

 

[aramis]

Now we are getting somewhere!

Actually what I meant was 1 page of text takes up 1/1000th of a cpu, so 1 cpu is 1000 pages. 1000 cpu is 1,000,000 pages, at ~2k each is ~2G of data storage at a model 1 ship computer. Im still thinking this may be a tad low but it is a starting point. One wonders at the number of pages in the current Encyclopedia Britannica?

One DVD-rom. <2GiB compressed; 3.55GiB full install.

And we know that memory can be at up to 20GiB per cc at present. (DVD is actually LESS info-dense than current memory sticks!)

 

[shadowdragon]

s'ok, so we all know the various design sequenses for T20 are beyond broken. Let's not focus on ship computers if you please. I am talking in general as a sort of guideline. ~20g per cc works for me but is that per cc of dedicated memory or cc of computer that you calculated that from? I am just thinking base memory integral to an item - additional memory can be added and since it is just memory without all the bells an whistles, it would have a significantly higher density. What I need is a good way to determine the base number. Again I toss out there to base it on CPU - anyone have an idea for a good and relatively easy formula to use to come up with a reasonable base number? It can even be modified by TL, probably should be actually.

thanks

 

[HG_B]

What I need is a good way to determine the base number. Again I toss out there to base it on CPU - anyone have an idea for a good and relatively easy formula to use to come up with a reasonable base number? It can even be modified by TL, probably should be actually.

thanks

Pure Silicon has introduced a 2.5" 1 Terabyte solid state drive at 2009 CES (consumer Electronics Show). The drive is higher density than magnetic hard drives of that size. The 1TB Nitro SSD is the most compact SSD per gigabyte: 15.40GB per CC (cubic centimeter). Tech Level 7

Multiply ~1000X at each successive TL

 

[JustinInOz]

base it on CPU - anyone have an idea for a good and relatively easy formula to use to come up with a reasonable base number? It can even be modified by TL, probably should be actually.

Basing it on the computer model number is as good as anything to base it upon.
Another might be the total number of "slots" available to each computer.
Another might be the volume of the computer. Your call really.

If the computer model number availbility is limited by tech level, and I think that it is, don't use a tech level modifier. It is not necessary. When you buy a low model computer from a high tech level source I don't think that the computer statistics alter. Buying a model 1 computer from a TL-14 source doesn't get you a better computer. This is a game rule artifact. It is completely divorced from reality that we all occupy.

Just remember that if you start adding too much reality, you might not be playing Traveller anymore

As an aside, there is a limit to how close atoms can be and not interfere with each others. This is an ultimate limit of density of information. This is the wall that Moore's Law will soon be rubbing up against. I think that Aramis has mentioned this one before.