kilemall
SOC-14 5K
My company years ago ceased messing around with individual machine problems that were software related and will just reimage the PC rather then waste time troubleshooting.
embracing retro 'puters
- Thread starter BwapTED
- Start date
My company years ago ceased messing around with individual machine problems that were software related and will just reimage the PC rather then waste time troubleshooting.
kilemall
SOC-14 5K
Management justification is job 1 for many of them.
That's why you get data kingdoms, where they try burrowing into the system by holding some piece of the company to critical processes only their team knows.
kilemall
SOC-14 5K
Well I disagree with most of your commentary on the topic, but this was always implied in the software dev rules, that you could save Big Bucks writing your own and possibly start Microjumpsoft.
The hidden bug problem however should NOT be handwaved away, there should always be a risk for getting advantage. Again, this is gameplay way more then sim.
And real business software that people do spend real millions on has bugs.
But our software isn't for business, they are life and death.
Getting to the point where it's nuclear power plant/avionics safe actually does take years of time and millions of dollars. For the level of performance we need, the cost structure is actually more of a sim and the dev rules are way too lax.
The other interesting angle are the more advanced programs that require increasingly skilled experts to help write them. A Skill-3 should be expensive to obtain, the Skill-4 through Skill-6 guys should be beyond expensive and require goodwill, blackmail, favors owed, etc. to get their services for the few weeks it will take to ensure success.
So what's the chance that a dirty stinkin' anarchist (sic) has decided to hack that element of the starship economy and put out there a shareware version for anyone to use? Or trade war between megacorps sees one firm release freeware that cuts into the income of their opponent?
Referee's discretion. There's no other definitive answer.
Try equity in the company, a fridge full of soft drinks and a ping pong table. You also have to have a project interesting to them. You're not going to get developers of that calibre working on your poxy little navigation program.
Hacks are two a penny but talented developers are a lot rarer than folks give them credit for. Really good devs are hot property in Silicon valley because there are very few of them. I doubt the whole U.S. university system would produce more than a couple of thousand top notch developers in a year.
Some back-of-envelope stats
I'm not sure about the U.S. but I saw some stats that suggested that the UK produces about 5,000 computer science grads per year, or about one per 12,000 population. This is post dot-com boom statistics, which has computer science enrolment settling about 2-3x pre boom levels.
Before the dot-com boom, about 10% of computer science undergrads (from my observations doing computer science) could have been characterised as 'really talented'. That would imply about 1 'really talented' developer per annum per 300,000 (or so) population. During the boom, the absolute number of talented students didn't shift much, but stage 2 enrolments roughly quadrupled (50-190) from 1997-2002.
If you assume a 15 year career on average before they go on to softer work (some love it and carry on but many burn out or go onto softer work) then you're looking at one 'really talented' (computer-4+) developer per 20,000 population or thereabouts, about 10x that number of decent devs (computer-2 or computer-3) and about another 10x that number of self-taught hacks who can spell HTML on a good day (computer-0 to computer-1). That's from an industrialised economy with a large middle class and a mature education system.
But it gets better - that 1 in 20,000 is just folks who are really good programmers. Maybe 1 in 5 or 10 of those folks would have a strong background in some other discipline like maths that you might need to understand the computations behind calculating a jump course. Maybe 1 in 100 would also be a top notch mathematician, physicist or engineer. Combining that skill with (say) Gunnery would be rare outside of R&D departments specialising in naval weapon technology.
You're getting up to a population 8 or 9 world before you've even got a fighting chance of finding someone with two level 4 skills to write something like Generate. Chances are that this person is already a distinguished engineer on a nice salary. In fact, folks at this level only ever come up on the job market once - when they graduate. After that, most job changes are from word of mouth. This is why companies go to a lot of trouble to recruit bright grads. Once they disappear into the system they never come up on the job market again.
You're a bunch of shifty looking adventurers. What have you got to offer someone like that?
Last edited:
There are quite a few plausible ways to solve that - most are adventures of their own.
- Their life. As in, you won't snuff them
- their family's and/or pets' lives.
- their retirement savings (bank hacking)
- nothing at all except an end to their scopolamine† dosing
- some hobby item they've been searching for
- Not releasing the scopolamine-induced blackmail behavior you holo'd
- Putting a good word in with the high nobles. Or the mob boss. Or some other VIP.
- Blackmailing their employer, instead of them.
- Addicting them to a drug, which you can then provide a lifetime source of. but only enough to keep working, until it's done, then a container full of it.
†Scopolamine is noted for inducing both a highly suggestible state and preventing long term memory formation, at the same time. It's evil stuff. Can make people do crimes and have no memory at all, not even of one blowing it into their face.
kilemall
SOC-14 5K
Er.
Gratifying that other recognize how rare the Skill-4+ people are.
But I'm not referring to the Computer-3/4 people, but the Navigation/Pilot 4/5/6 types.
I don't know that your numbers are right empirically, but they do illustrate the difficulty of getting both a Nav and a Pilot-4 together along with the programmer to dev this thing.
And those rolls won't happen in a week on average.
The sort of thing that's going to require gameplay to happen.
It smells more like a quest, with each program slamming into place making for a more and more powerful ship that punches waaaay above it's weight.
Oh wait, reread your post, and I think you are under a misunderstanding of the rule. The Engineer/Electronics/Gunnery/Pilot/Navigator skills are not required to be the programmer too, you can have a separate expert help write the program. They will be doing the specialized definition of the problem set and solution mechanisms while the programmer codes it.
Alternatively in a cyberpunk/transhuman direction you could decide the programmer is capturing an engram of the expert and converting it to code.
Whatever, it still is difficult to put together the right people to stay in that coding basement until it's done, and possibly losing a ship or two before the bugs are worked out.
Gratifying that other recognize how rare the Skill-4+ people are.
But I'm not referring to the Computer-3/4 people, but the Navigation/Pilot 4/5/6 types.
I don't know that your numbers are right empirically, but they do illustrate the difficulty of getting both a Nav and a Pilot-4 together along with the programmer to dev this thing.
And those rolls won't happen in a week on average.
The sort of thing that's going to require gameplay to happen.
It smells more like a quest, with each program slamming into place making for a more and more powerful ship that punches waaaay above it's weight.
Oh wait, reread your post, and I think you are under a misunderstanding of the rule. The Engineer/Electronics/Gunnery/Pilot/Navigator skills are not required to be the programmer too, you can have a separate expert help write the program. They will be doing the specialized definition of the problem set and solution mechanisms while the programmer codes it.
Alternatively in a cyberpunk/transhuman direction you could decide the programmer is capturing an engram of the expert and converting it to code.
Whatever, it still is difficult to put together the right people to stay in that coding basement until it's done, and possibly losing a ship or two before the bugs are worked out.
kilemall
SOC-14 5K
Randyb has the right answer, figure out how much you want the computer software subgame in your system or not. Just be prepared to work some other money sink in if they start regularly piling in those MCr10 speculation hauls.
Or they start buying ships with cash, once that loan monkey is off their back the trade game gets a lot more profitable even hauling the milk runs.
In my case I make it a bureaucracy/safety thing- the MCr programs and computers are the safest and so are the only ships that can get the Cr10000 or Cr1000 per ton standard rates and pass the inspection related to the yearly maintenance.
They cut corners, they end up uncertified and working the illegal/shady junk ship circuit, where you can move illegal goods and desperate people on the cheap with third-hand equipment you never would have jumped into in your service days.
The cheaper computers allow the mercs to have military-grade capabilities- but it can break more easily.
And that cheap software you got off the Spectrum isn't so cheap when you slap in that used fusion gun and find out you don't have a Turret program compatible with it.
The 'fully licensed' version gets support and a new driver/firmware interface installed as part of the shipyard work, but now you have to pay for or write a new Turret program to work with all the risk that implies.
If you don't want to do the detailed CT game, just figure the software to take advantage of the computer is equal to the cost, then divide by 10 for each cheaper level and roll more bugs per cheapness.
That's my solution to give my players options. Yours' may be different.
Last edited:
I can go along with the idea of using inspections and regulation to minimise the extent to which alternative software is used by vessels. Big government looking after big corporations, it's not like that hasn't happened before.
But after a thousand years of just this iteration of the Imperium, other alternatives would have to be out there. Other system or multi-system governments could offer kosher versions of software, and hence certification, in order to attract shipping to their backwoods or off-the-main-shipping-lanes locations. If their equivalents weren't acceptable within the 3I, then trade from outside the Imperium would be stifled unless it was carried out solely by 3I-licenced carriers...
kilemall
SOC-14 5K
See, there you go. Much better if it isn't spelled out in saluting the canon cannons and you can work out your own consequences and gameplay generation.
mike wightman
SOC-14 10K
Solution to the skill required problem:
build a robot with skill 4 in computer and the other skill you need to write the program - machines telling machines what to do, scary
get yourself a wafer with skill 4 in computer and the other skill you need to write the program - discorporate skills and memories taking over your body and if anything goes wrong with the program after you have taken out the wafer... equally scary
build a robot with skill 4 in computer and the other skill you need to write the program - machines telling machines what to do, scary
get yourself a wafer with skill 4 in computer and the other skill you need to write the program - discorporate skills and memories taking over your body and if anything goes wrong with the program after you have taken out the wafer... equally scary
Here are my experiences with software.
I bought an Amiga computer about 1990. The store in a nearby town is where my computer caught its first virus. So I stopped going there, they did apologize.
I discovered the Fred Fish, yes thats its name, public domain and shareware disks.
Some of those programs were substantionally better than the ported commercial software being sold for the Amiga computer.
So that is one thing I am looking at for starships in my game universe. Some 'not off the shelf' software is going to be better than commercial software.
I realize its not a direct equivalent, but it could make things interesting.
I bought an Amiga computer about 1990. The store in a nearby town is where my computer caught its first virus. So I stopped going there, they did apologize.
I discovered the Fred Fish, yes thats its name, public domain and shareware disks.
Some of those programs were substantionally better than the ported commercial software being sold for the Amiga computer.
So that is one thing I am looking at for starships in my game universe. Some 'not off the shelf' software is going to be better than commercial software.
I realize its not a direct equivalent, but it could make things interesting.
I remember early computers had key clicks so new users wouldn't be frieghtened of them.
So I can see older machines, landing on a primitive planet that doesn't have starships yet, would have older clunky computers and interfaces.
If someone uses old interfaces, looking up how they worked would be a good idea.
I don't know what the url is, but bitsavers has manual and photos on the old punch card systems.
Ah, here it is: http://www.bitsavers.org/
That web site is rather 'old interface' as well.
You can find web sites, and youtube videos, that show how the older machines looked and sounded. Like the Amiga, C128, Ataris, Apple ][+, etc.
So I can see older machines, landing on a primitive planet that doesn't have starships yet, would have older clunky computers and interfaces.
If someone uses old interfaces, looking up how they worked would be a good idea.
I don't know what the url is, but bitsavers has manual and photos on the old punch card systems.
Ah, here it is: http://www.bitsavers.org/
That web site is rather 'old interface' as well.
You can find web sites, and youtube videos, that show how the older machines looked and sounded. Like the Amiga, C128, Ataris, Apple ][+, etc.
An interesting proposal: getting someone under the influence of mind-altering drugs to write a piece of safety-critical software for you single-handedly (Balmer peak notwithstanding: https://xkcd.com/323/).
So, you've managed to blackmail a 1 in 10 million genius into writing a piece of software for you. He's writing code that you have no hope of understanding the nuances of. Your veiled death threats mean that your party has just become an existential threat to someone with an IQ of 150+ who is writing you a piece of safety-critical software that you're depending on not to navigate your ship into a convenient black hole. Maybe you can intimidate them, but maybe he (probably correctly) surmises that you intend to dispose of him once he's completed the task. Maybe they can do something like Ken Thompson did.
Reflections on trusting trust: https://www.archive.ece.cmu.edu/~ganger/712.fall02/papers/p761-thompson.pdf
In practice, any piece of non-trivial software would require multiple parties to write it. The myth of the lone genius hacker is exactly that; you don't build safety-critical software in your mom's basement. There's a reason that the avionics software for the space shuttle had a staff of 100+ developers to write a roughly 350 kloc system. You're now in the business of assembling a team of crack developers - and over at the Daily WTF we can see many stories of the outcomes of non-technical management who thought they had the nouse to do that.1
Then there's the problem of getting it flight certified - which is likely to be needed if you want to flog it. Our audit trail of 'we doped this guy up with scoplomine and he wrote it' isn't going to be so easy to get past the certification authority - with a bit of luck the party also has someone with Bribery-4+ and Admin-4+ on the books.
I guess I've been working in I.T. long enough for this to totally fail my verisimilitude test. I had a couple of profs in uni who were into algorithm design and formal proofs - one still does proofs on security and safety-critical systems and is probably the smartest person I ever met.
1 See also In Search of Stupidity (available from Amazon or other outlets). This tells the story of 9 leading PC software companies from the 1980s - and Microsoft. The author's thesis is that the other 9 other companies were run by non-technical management who made noddy mistakes that would have been obvious to anyone with a background in software. Microsoft survived because their management didn't make any catastrophic screwups.
The author used to be head of marketing for MicroPro - the outfit that made Wordstar and one of the other 9 also-rans.
This is actually more plausible - the Type-S has been around long enough that the software for it could easily have landed in the public domain. In fact, quite a lot of state-produced IP winds up in the public domain today.
What is more plausible is that for most computer specifications (say 1 to 4 or 5) there is actually a design that is effectively in the public domain for the hardware or software - after all, any design of TL11 or lower could date back to the first Imperium and any design of TL12 or lower could date back to the second. These are a matter of archaeology as the parties holding the rights to the IP ceased to exist millenia ago. The designs are thousands of years old so it's entirely plausible that they could be well documented by dozens or hundreds of scholarly research projects and available essentially royalty-free to third parties - both hardware and software. The Imperial authorities might even sanction this as a means to encourage interstellar commerce and shipbuilding in outlying regions.
Alternatively, some movement analogous to today's Free Software Foundation could have developed designs for equivalent systems and released them under an open licence. Certification is left as an exercise for the reader - certain long-term support versions could have standing certifications but your specially tweaked version might require a flight certification program costing some tens of millions.
You're not going to get the schematics for a Model/9fib through this route but anything up to (say) TL11-12 tech might be available in this manner.
Straybow
SOC-14 1K
Sorry, let me clarify. I'm saying that MM didn't use or study project management to establish man-hour requirements, it was just made up whole cloth. I believe he did know enough about computers to know that being able to write a program in one week with a chance that it debugs successfully means it was a very small project.
No. For example, when laying out requirements for stewards an 8-hour day is specified, without detailing any allowance for overtime, or the fact that there is nowhere else to go and nothing else to do in jump.
I'm pointing out the glaring error in assuming there is a huge difference between "professionally produced software" for which the poor schlubs are paying kidnap ransom price, and "self-written software" when the requirement is one guy with Comp-3 (which is equivalent to a doctorate, or maybe even postdoctorate level of work), and one week of work time.
As several posters have pointed out, quite aptly, the benefit of "professional" programming isn't the management or the marketing. It is the budget for testing and the size of the work force. On the latter, the requirement for writing Generate clearly removes the work force size because a single worker can produce a finished product in as little as a week, or in four months with 95% confidence. The variance there is the debugging process.
Jump calcs are a math product. Plain and simple. You feed it into the computer once and then you jump. Jump calcs aren't jump engine control stuff, which is a separate program and probably would require vastly more programming to work with any jump engine produced to the given standardization.
When there are bugs in enterprise software it is usually because they're trying to make it do too much. As somebody pointed out, MS Word change tracking can destabilize large documents. But when all you're doing with Word is formatting your document to look nice, it's bulletproof.
Churning out jump routing doesn't even have to look pretty on any printer type and paper dimensions, it's not a graph that has to be arranged to present the data to the eye, it's just a string of numbers for the jump program. Where are we now, where are we going, where are the gravitational bodies are near the path.
Maybe it requires some really high-falutin, multidimensional math... that would be available in a professional programming library. Granted, it might require Nav-4 to understand the math enough to properly use it. But it is math, and it has a right answer and a wrong answer.
What are you proposing would be the trigger for precipitating this Critical Flaw? Being in a hurry doesn't matter to the program. That might lead to feeding the Generate program some incorrect data, but user error doesn't disappear depending on who wrote the program.
Oh, maybe it has sophisticated input error detection algorithms that the user-written version might not be able to implement. Again, why would input error detection (out of range values being the #1 problem) not be in professional programming libraries? Or in the Comp-3 programmer's personal library of thoroughly tested, absolutely reliable code?
Straybow
SOC-14 1K
Something tipped a memory here. I recall there was a guy on this BB going by the handle Thrash. He had all kinds of inside stuff. Somewhere around 2008ish he took offense at something and went through the entire board and deleted every single post he ever made. That may have been the source of the "jump calcs can be made with high school math on a hand calculator" thing I recall. But I'm just speculating on that source, not on whether I read it. But it's obviously not "canonical," so...
I'm back to relying on the one-man-working-one-week-can-do-it defines the scope of work position.
Note that this isn't a Management Estimate, notorious for overestimating the efficiency and success rate, this is a Canonical stipulation. Therefore, it really does only take one guy one week to write it, and beating out the tricky bugs takes the variable amount of time. If in debugging the programmer rewrites some code, that doesn't change the fact that the amount of code required could be written in 5x8 hours.
I would argue that decompiling the official program to see what tricks they used would eliminate the Critical Flaw, and enable the programmer to complete it in 5x8 with a very high likelyhood. It is, of course, contrary to the EULA to decompile and learn their secrets, but that's not your problem if you're buying the discount program off the grey market.
I reckon you're right in the last line. The concept could be the same as illustrated with an Amiga, but I imagine that even a Model 1 being used to run a large (we know they start around 747 size) space-going vessel run off a nuclear fusion reactor and capable of Jumping into an alternate dimension in order to undertake FTL travel is going to have a few more complicating factors than what we started to learn programming on as kids.
