Big computers are no problem

[kilemall]

1970s machines, whether the very first personal computers or mainframes, took a few minutes to load programs especially if they were on some sort of tape media. Booting up also took a few minutes, and might be as slow as running through a card deck.

The swapping between active and 'storage' programs should be considered instant as that's really more like mulitple programs, all loaded per se but many 'swapped out' in virtual memory, ready to run the instant it's given enough resources to actually execute.

Of course this is about what the CT system was modeled on. I'd assume this is avionics or better grade machines (better be for millions of credits and risking the whole ship), so they likely have an instant-on capability, extensive self-diagnostics running on startup and doing internal resource swaps on error/damage detection.


Which gets to another point as to why they could be bulky and/or spread out, their damage resiliency.

By CT ship damage standards, the ship computer can largely shrug off damage that would cripple a Type A engineering section. By 7 hits it is a very flaky machine, but still serviceable whereas a smaller ACS would be dead in space and the larger ones painfully damaged.

HG doesn't really capture that, but the computers still take beatings that leaves weapons damaged and drives crippled.

So, a LOT of redundancy, distributed failover nodes, backup consoles etc. that take damage- and take space.


Also was reading an editorial from Grognard where Wiseman was answering the critics of the Big Machine by saying the large computer spec was also intended to cover crewing space for controlling the machine. So, that too, and the larger models may have 2 crew stations.

The 1990s mainframes I worked often did, and I saw earlier generation machines that had the computer equivalent of a flight engineer station, with direct external controls on the functioning of the machine. Even today mainframes will typically have hardware and a console that controls how the system boots with what resources separate from the control/execution main console.

 

[whartung]

The boot process can be more so tied to the diagnostics and waiting for peripherals as it is waiting on a CPU. Disk drive spun up? Check. Network card initialized? Check. Video stable? Check.

Heck, my car takes about a minute to "boot" to the main screen, and even longer to connect to the phone.

 

[aramis]

The boot process can be more so tied to the diagnostics and waiting for peripherals as it is waiting on a CPU. Disk drive spun up? Check. Network card initialized? Check. Video stable? Check.

Heck, my car takes about a minute to "boot" to the main screen, and even longer to connect to the phone.

A lot of "fast boots" — a topic currently relevant to Linux and MacOS X 10.5+ — are done by doing post, then immediately launching the GUI, and running the rest of the initiation process in the background.

It takes 30 seconds to get to a login on my macbook air. It takes 5 minutes more to be useable without slowdowns if I let it be; if I try to do anything processor intensive, it's 15 to 20 minutes before things speed up. If I'm kerning fonts in Fontographer, it's more like 30 minutes...

 

[wellis]

I apologize for bumping this but I've been wondering something: are most of the complaints on Traveller computers being too big possibly referring to the terminals instead of the mainframes?

Like how most people think of computers as being the monitor.

Or do a lot of complaints in general notice something about Traveller's computers being too slow (because I think a ship's mainframe isn't going to be small. It's not a personal computer)?

 

[tjoneslo]

Or do a lot of complaints in general notice something about Traveller's computers being too slow (because I think a ship's mainframe isn't going to be small. It's not a personal computer)?

This is closer to the truth. The question usually was something like "I have an iphone 11 which fits into my pocket and capable of doing everything that is listed in the software list, so why do the Traveller ship computers take up Tons of space and still can't do everything my phone can? " without completely understanding that about half of what the phone does is on an array of computers in a massive data center half a continent away.

The better way to look at this is two fold. First the Model/1 computer is TL-5, larger and less capable than an Apple II. If you want a good comparison look at the Apollo mission computer.

Second, in 1980 few people understood the impact of Moore's Law. The original sizes of the Traveller Computers thought up in 1977 have remained, even if the capabilities have increased over the 40+ years since then. I think these days gamer's in general have a better idea of the level of automation required to run a starship with only a few crew. So of course they're going to be big.

 

[Condottiere]

Which probably explains why in the MongoVerse, computers become virtual, requiring no volume allocation.

 

[tjoneslo]

That makes the most sense. Computers do take space but it could easily be distributed, and there would be good reasons for doing it that way. You no longer have one sabotage incident render the whole ship dead.

GT: Starships design did the same thing with the power plants. Ships are designed with modules, each one contains the power plant requirements to run it. There was some confusion in the system of how big the plant in a ship really was.

 

[wellis]

Which probably explains why in the MongoVerse, computers become virtual, requiring no volume allocation.

That makes the most sense. Computers do take space but it could easily be distributed, and there would be good reasons for doing it that way. You no longer have one sabotage incident render the whole ship dead.

GT: Starships design did the same thing with the power plants. Ships are designed with modules, each one contains the power plant requirements to run it. There was some confusion in the system of how big the plant in a ship really was.

So does this mean computers wouldn't takenup volume/dtonnage because the mainframe is distributed throughout the ship?

Or is the "computers become virtual" a reference to holographic terminals?

Actually I kind of feel Model 1 terminals or computers or whatever shouldn't be so heavy or take up so much space due to the fact components have gotten lighter and thinner.

And even with rad hardening and such, that adds dtonnnage to a vessel overall IIRC and isn't something specific to computers or terminals. Assuming one says computers being rad-hardened makes them so heavy.

Because a Model 1 computer shouldn't be something out of the 1970s. In a modern, interstellar spacefsring society, new materials and such have probably slimmed down and lightened stuff.

 

[Straybow]

I apologize for bumping this but I've been wondering something: are most of the complaints on Traveller computers being too big possibly referring to the terminals instead of the mainframes?

Like how most people think of computers as being the monitor.

Or do a lot of complaints in general notice something about Traveller's computers being too slow (because I think a ship's mainframe isn't going to be small. It's not a personal computer)?

The CT rules state that the computer is separate from the bridge. Nothing says the computer space has to be accessible, so all the terminals would be in the bridge. It's just the computer, based on the size of '70s era mainframes.

I'm also convinced that setting the basic unit of ship size measurement as large as an 8' cube is unworkable, and when they were looking at sizes as integer numbers they weren't thinking of how big the units really were.

 

[Straybow]

So does this mean computers wouldn't takenup volume/dtonnage because the mainframe is distributed throughout the ship?

Or is the "computers become virtual" a reference to holographic terminals?

Actually I kind of feel Model 1 terminals or computers or whatever shouldn't be so heavy or take up so much space due to the fact components have gotten lighter and thinner.

And even with rad hardening and such, that adds dtonnnage to a vessel overall IIRC and isn't something specific to computers or terminals. Assuming one says computers being rad-hardened makes them so heavy.

Because a Model 1 computer shouldn't be something out of the 1970s. In a modern, interstellar spacefsring society, new materials and such have probably slimmed down and lightened stuff.

In the Official Traveller Universe that's what you get. In Your Traveller Universe it can be different. One can simply ignore the space requirements, with the assumption that the computer is distributed around the ship. Fire control nodes are near weapons, drive control nodes near drives, other functions built into bridge stations.

As mentioned elsewhere, tube tech can be substantially miniaturized and be EMP resistant. Optical data, and perhaps optically based circuits for some applications, are comparatively large and very EMP resistant. Maybe that would add some space requirements.

 

[Cryton]

In this day and age, I'm less likely to argue that the "Big Computers" in Traveller are a single huge computer, but more akin to Blade Servers, large cabinets full of linked processors, capable of the (I presume) huge number of simultaneous calculations required for running a starship, all its components, and handling the extra-dimensional, quanta and gravity based mathematical calculations and functions required for Jump that would fry the most powerful desktops or portable computers (or at least slow than down so much as to be useless).

What I base this on, is that oftentimes, a seemingly outdated computer system, that takes up a room sizes area can still outperform a modern computer for certain applications, and for those reasons those old monster sized machines are STILL used today.

Ie, while your cell-phone can out perform in computational power, it can in no way handle the 1000+ simultaneous communications connections of an old room sized beast. I figure Traveller ship computers huge for similar reasons.

Just my thoughts.

 

[flykiller]

I just say the "computers" are in fact sensor suites - the output is never better than the input.