cantab

That would be of no help in the OP's situation, of expecting F9 to reload the last autosave, nor the common related situation of quickloading without realising how long ago you quicksaved.

Chemical engines are less efficient than in .90 so you'll need a larger ship if you go chemical, and you might want an extra stage. The nuke is as efficient as ever but remember it now uses liquid fuel only. Aircraft liquid fuel tanks are best for it. Rocket fuel tanks with the oxidiser removed are OK but heavier. Rocket fuel tanks with the oxidiser accidentally left in are a disaster. Ions will probably have to be RTG powered, solar panels now behave realistically and give very little power at Jool and Eeloo. Hold Manoeuvre and similar make long burns practical if you're OK with letting the game do its thing. Gravity assists as always can save you a bunch of delta-V at the cost of a long mission time. Jool > Eeloo is effective in dramatically reducing your Eeloo capture burn.

You're not the first, and until Squad change things you won't be the last. And KSP isn't the only game to give you rope to hang yourself like this. To stop it happening again the protection is the same as against any other data loss: backups. It doesn't matter how you do it, it's a good idea to have all your game saves routinely backed up. (I use Spideroak, it's secure, powerful, and cross-platform but it can suffer from performance issues on first installations.)

"Keroperox" is an interesting combination, not one I've known used before. Unlike kerolox it's all storable, with advantages for reigniting upper stages, missiles, and perhaps dry mass savings since peroxide is both denser than LOX and won't require insulation. Unlike hypergolics the fuels aren't as toxic, though HTP is still not something you'd want on your hand. And perhaps only having one fuel (the kerosene) to deal with makes things simpler and cheaper. The big minus is keroperox isn't to my knowledge hypergolic; I wonder if the Chinese have a nice way to make a reignitable engine?

I voted "yes", but promptly had doubts. The advantages are well-stated. The game didn't hand me delta-V on a plate, it wasn't a number spat out of some black box, so I learnt the rocket equation and applied it myself, well before I got KER. On the other hand the game did hand me on a plate the delta-V it takes to get from one orbit to another, it's as easy as setting two manoeuvre nodes, and I've never learnt to do that myself. I think I may be worse off because of it. Then again, the game doesn't tell me how to find an interplanetary transfer window, and I've also never learnt to work those out myself, I rely on mods or external tools. On balance I think I'll still say yes, but it's not an "all good" thing. Career mode unlocks I don't see as that useful, because I don't regard career as appropriate for novice players.

It's not about the car you have, it's about the roads you've driven. I've visited all the stock bodies in some sense and landed on most, and now am playing with multiple planet mods.

I pulled off that rendezvous and delivered the new solar array to my Harvester Station. Did a routine Mun landing from the station to get some more science and data. Then set up the manoeuvres for my initial gravicaptures into the Jool and Titanus systems. I can refine them more with a second course correction. It's gonna be a long time before these arrive, I reckon I'll have an Asclepius mission run by then, and maybe another visit to Serran.

Don't try more power.

Show me where I said it wouldn't be.

To be honest it will be a bit of a shame if mechanical drives go away. Because a five-inch disc of metal spinning 120 times every second is frankly BadS.

I have a new one: Non-existent Squadcasts. I turn up on Twitch waiting for it to start and nothing. Or, sometimes, I find out it's been run about twelve hours earlier than scheduled. Of course there's never any advance information from Squad.

Sent the new solar array for my station towards the Mun. Got my sums a bit wrong and ran out of fuel. I suppose that's what I get for going RCS-only on the little tug that was delivering it. Now how am I going to rendezvous with that? This will be interesting.

Predicting the motion of the target is elementary.Now, the target might opt to be unpredictable. If they expect you to shoot lasers at them they might start jinking around erratically to be hard to hit. But that costs them reaction mass which is finite. True, the fuel to run a laser's power source is also finite, but if your ship uses a nuclear reactor then chances are it might as well be bottomless. So even your laser misses gain you an advantage.

Transfer Window Planner is good, but there's a couple of things I feel are missing from it. Firstly, details of the transfer orbit. Alexmoon's planner gives that, TWP does not as far as I can tell. For me this is the important thing for a transfer. Exactly how you get into the transfer orbit doesn't matter, if you leave at the right time and aim for the given apo and peri round the Sun you'll get your transfer. For example if you want to leave Laythe, snag a slingshot off Tylo, and get to Kerbin Transfer Window Planner is never going to tell you the exact ejection burn from Laythe orbit but if you know the solar apo and peri you want you can work towards that. Secondly, and adding more complexity, the ability to specify the inclination and longitude of the ascending node as well as the altitude of the parking orbit. Alexmoon's planner and TWP both assume an equatorial parking orbit and that's fine in the stock system, but in a system with inclination such as RSS or New Horizons it's plainly an inefficient way to go about things. Thirdly, the ability to actually work out the ideal transfer orbit inclination. I'll grant that will add some considerable complexity. Perhaps one way to do it would be to minimise the normal component of the ejection burn.

The whole idea is it works for things like Earth-Mars, and by implication Kerbin-Duna. In theory it needs no course corrections but in practice you'll never get your path perfect, I reckon you'll need at least one correction per cycle.

Wow. And I thought AMD were the heat generators. But to be fair, I'd say a card catching fire isn't NVidia's fault, it's down to EVGA or Gigabyte or MSI or whoever put the card together. Unless it was actually a reference card, but who buys reference cards.

If you're using ions you probably have lots of delta-V and can afford to use higher parking orbits, that's how I handle things.

With the gizmos you can sort it out anyway.

Check what Norton actually reports. Many antivirus programs will make detections like "Exploit/Generic", basically meaning it's not a known virus but the antivirus software regards it as suspicious. Increasingly antivirus programs are using a "reputation" system which means that programs from big corporations get waved through but lesser known programs get falsely accused of being malware.

You aren't going to be building a solid passive structure much larger than Earth, if that big. There comes a point when no theoretical material can hold up its own weight and the structure will collapse. To go larger it can be a network of some sort, not a single structure. The largest things we have already built are networks - the global telecoms network, the internet, the road network, and so on. The original concept of a Dyson sphere was a network of solar power satellites that would relay power presumably by microwaves, the "solid ball" idea is an invention of science fiction writers. The other way to go larger is with an active structure. The "space fountain" is one such concept, with a powerful particle beam going up and down it, the reaction force to turning the beam at the top holds the tower up. A ringworld might counter its own weight by spinning, though it would still need station-keeping to avoid hitting its star or planet.

The "gotcha" is that in order to catch the cycler you need just as much delta-V as it takes to make the transfer directly. With no need for life support and the relative ease of aerocapturing at either end of the trip there's not much point using a cycler ship.

To keep your capture burn low it's the same as anything else - you want your orbit round Sonnah to just kiss Aptur's orbit, not to cross it at a substantial angle. So far what I've been doing is noting that Aptur appears to be in a retrograde inclined orbit round Kerbin, and trying to launch into that orbital inclination before looking for a transfer. I've been able to get cheap transfers but whether the retrograde inclined thing is necessary I don't really know. Regardless it will be a while before I visit Aptur again, I've explored it fairly well.

That's a Linux thing, where 100% means 100% of one core or the equivalent of that.

In the game settings there's a framerate limit. If that isn't set then KSP will chuck out frames as fast as it can, using a load of CPU even with nothing much happening. With a sensible limit set it won't do that, though it can still use quite a bit of CPU because it's not very well optimised.

SMART data though can be useful with hard drives, provided you monitor it and interpret the data correctly. A few years back Google released the results of a study using their data centres. My understanding is the SMART data is pretty much useless with SSDs.