Project, by xpil.eu, przeklad yours truly, czyli dzis troszku inaczej

Oryginal/Original version 

Przeklad rzec by mozna autoryzowany i za zgoda autora.

 

Project, originally planned for three, TOPS four months, has not been released until 2 years later and even then it was a complete failure.  It has consumed 6 times the original budget, and the final Product had to be flushed down the sewer.

Luckily, The Boss has endless patience for us, not to mention funds to go with it.

First problems cropped up at the very start – on the second day, when Gabs decided to add feathers. Feathers, he was arguing, will provide excellent thermo insulation, rain protection, and maybe, just maybe, we can try and implement flying. We can fly so maybe they could too?

Prototyping those feathers took us at least 2 days, only to find out that they were not compatible with the latest-greatest version of the Ribosome, and they had to be substituted by hair.

At that point The Boss came back from his annual leave on Proxima, and barely entering the office shouted that he wanted to see some progress, so we quickly applied hair everywhere. Surplus hair all went on the back, and since Michael was in the middle of configuring the nuts and bolts of the Kernel, even the groin got some of it.

Eyes got some too – which resulted in permanent leaking because they could not shut, so we had to remove all the redundancy, leaving only one pair.

Of course the remaining pair of eyes was at the stupidest location possible – next to the CPU. Balthazar managed to rewire direct connection, but as luck would have it, he fucked up with the cables and ended up cross-wiring them. He’d fried at least four units before succeeding in crossing it all on microcontroller.

To begin with all the riggings were furrowing as planned, but then Raphael, “The Genius”, decided to write his own sorting algorithm, clogging the left rigging with bubbles, which subsequently destabilised the right rigging. Then they both spruced up and had to be redirected to /dev/null, because of course main partition was full. /dev/null immediately got clogged too and we had to call for a ramrod.

Then there was a drama with Octa-Core CPU. I mean Worm-ex Ltd. promotional leaflet of the unit says 8-core, sure thing. What it does not say though is how much power this beast needs. If we were to run all 8 cores on full power it would have to feed on antimatter, and the cost of the power unit would sky-rocket! So we decided to replace the glorified Octa-Core with energy prudent system based on protein axons.

Michael insisted that it will be perfectly sufficient. To be on the safe side though he fitted few test units with a second CPU, roughly in the middle of the body. He insisted it would be a backup CPU which would allow avoiding transactional overload.

Bloody Fool.  

One of the test units ran into a tree, cracked its head and was left with only the backup CPU, the one just above the legs.

It then all got replicated and a batch of those went into production.
Consultants from Worm-ex Ltd. washed their hands stating that we should have used their recommended Octa-Core CPUs.

Adding insult to injury, instead of the classic cube body, Barnaby decided we will go for the Angel Build. The Boss objected, explaining that on majority of worlds where the likeness of the Creator was implemented there were more problems and complications than benefits, blah, blah, blah, you know the usual Boss thing. When you need His input He favours of golden silence but when you need to concentrate in silence and work, He goes all chatty!

So we ignored Him and imported the data from few humanoid libraries to the Ribosome.

Of course something got corrupted at the repository and accidentally production got the current code with the entire submit history starting all the way from that stupid bacteria project. And now the Product thinks it evolved and dedicated entire branch of science to its evolution!

Talk about thinking... we had serious reservations if thinking should be implemented at all. We were debating between AI and leaving it on differentials, but of course Balthazar, Never-ending Improver, started arguing that own intelligence would make the Product autonomous, and more robust against environmental flaws, and nonsense like that. Good half an hour he went on, and on about benefits of Product having awareness.

It was all very inconvenient, because Ribosome didn’t have any drivers for awareness and it would all have to be typed from scratch, but Balthazar took The Boss for a doughnut once or twice and convinced Him.

Next 2 months down the sewer, and still we haven’t managed to complete all DLLs, so if you don’t reset product once a day, productivity drops drastically, then it overheats and resets itself anyways. Every reset takes between 4 and 8 hours so waste of time is humungous, but The Boss said “no” to further months of testing and debugging.

Bloody doughnuts...

Next thing was a power unit.

Original plan assumed cold fusion, which would allow Product to run for several thousand years uninterrupted on one charge. It was all going great.

At least to begin with...

Until that one accident in the Lab when one of the younger technicians confused Palladium with Ruthenium and five hundred of control sample got splashed all over the walls.
It took them full week to clean it off.
At the end it was discovered that some joker swapped labels on the container.

Anyhow, the power unit was nearly complete when the new Health & Safety regulations were implemented. Turned out we were not certified to use Borane, and some of our protons were missing ISO documentation. Then on top of that Boss forgotten about deadlines for standardisation request and the almost ready cold fusion went to PlaityRollers in Ursa Minor, while we had to settle for regular oxygen burning.

Obviously that came with side effects – toxic leftovers that have to be removed using separate ventilation system.

Oxygen delivery to the CPU required pioneering approach and we would have failed profoundly if not for the crapper on 8^th floor that decided to break down and local plumber gave us idea on how to solve the distribution problem!

It was all a bit trigger-happy and cowboy-style, but initial tests came out pretty well, so there was a chance of fitting at least in the budget. But of course there was Barnaby made and his SNAFU in his energy calculations and we found out that product would have to be pancake flat to ensure CPU gets consistent power supply. So we had to add an additional pump, which of course increased the mass of the system and energy consumption.

Chicken and Egg, it’s all fecking Chicken and Egg. And we were already well delayed, and everybody was clocking 17-hours per day.

Finally, somehow, the energy was balanced, which required every Product to be equipped with long vertical pipe. Additional algorithm conditioned the system to output all high-energy waste via the top exhaust, rest was taken care by few simple switches so remaining waste was going down, and everything seemed hanky-dory.

Well, it would have been if not for our team chemist who was totally wasted after a particularly fun weekend and missed out occasional gas waste coming through the bottom exhaust. Hence the bottom end of the pipe was occasionally emitting weird whirring and tooting sound, because of methane and sulphuric mixtures.

We have tried to patch it but yet another issue reared its ugly head.

Beta testing customers decided they wanted the product to be able to replicate. So that they don’t need to buy new one when the current one starts wearing off but instead have it “somehow”, perhaps “self-replicate”.

We checked the BRD, but of course there was nothing on self-replication there.  The Boss said “customer’s wish is our desire”, and we had to improvise again.

There was number of ideas. The most promising were nanobots, but of course – as usual – costs... Boss strongly suggested coming up with something simpler yet effective.
Preferably ASAP because we are already way behind, so much behind in fact that we may lose an important customer, and then the future of The Firm will be more than questionable... Et cetera, you know The Boss... Blah, blah, blah...

Besides nanobots and few rather bizarre ideas with a plant pot and a brush, we had pretty much nada. Despite of three days of head banging, heavy brainstorming, cisterns of coffee, sea of RedBull and guarana tea, we were stuck. Requirement was impossible to meet unless we completely rebuild the entire product. And there was absolutely NO time for that.

On the day four Gabs got major case of runs, and managed to completely block the crapper, so we had to ask the local plumber again, who upon hearing of our self-replication dilemma, gave us an idea based on hydro-dynamics.

Sure, some calculations were required, mainly on friction and sliding parameters, but in the end it worked out rather nicely.

Only requirement was that we would need 2 significantly different types of the Product, which when required would be in position to initiate process of self-reproduction.

Hardware wise it all worked well, but as usual, with software there was a major SNAFU. Running 2 projects in parallel, one for each Product’s type, ended in a total fiasco.

In the end we came up with 2 pretty much separate Models, and all that because an idiot-customer forgotten to mention the self-replication in the BRD.

In the mean time the Headquarters upgraded the Ribosome and of course bang in the middle of our tests, so not only had we to upgrade from the four core histone to the double helix, we have also lost very convenient rack for genetic material exchanger. Finally, The Boss got so pissed off He threatened that if we don’t sort it out here and now, we can forget about our annual bonus right away.

At the end of the day we managed to squeeze exchangers between the bottom exhausts (YES – there were TWO bottom exhausts by then, because some ‘genius’ miscalculated something, and we had to frantically install a splitter for the liquid waste), which annoyed most of the team (Genetic material mixing with the waste!! Outrageous!!), but The Boss decided – Not Broken – DON’T Fix It.

And we launched the production.

Just before releasing the first batch of the Product, Raphael messed up config files, and one of the Models got E.G.O. parameter maxed out and the I.Q. capped, while the other Model got expanded chatterbox module, and destabilised emotional module.

Outcome was such, that customers after 3 months were going ballistic. Customer Service System was stretched beyond reason, from all the complaints. For example: after self-replication, new Products were much, much smaller than the final Product and all functions wouldn’t activate for ages – nobody knew for how long exactly. Some users were reporting leaks. Others were asking for additional CPUs because the default ones were too slow.

At this point, all Products have been recalled and refunded; customers have been given profound apologies and (as compensation) a cat each.

The production site was ditched it in some mouldy corner of the Galaxy, where it still is, most likely covered in cobwebs and gathering dust.

We’ll probably recycle it during the next Purge...