-
Posts
3,736 -
Joined
-
Last visited
Content Type
Profiles
Forums
Developer Articles
KSP2 Release Notes
Everything posted by Nuke
-
i think the forum rules prevent me from saying those words here.
-
whatever the last number was + 1. rewrite the whole game in x86-64 asm and have it only support baremetal os. whatever the last number was + 2. your not allowed to use any nops.
-
good to see more hard scifi films coming out.
-
ive been banned from several game servers for being too good as well. its sad really. couple of noobs say your botting or cheating or whatnot and its enoguh to get you banned these days. i especially hate games that use punk buster. your latency fluctuates and you get bannned. you fire your weapon and you get banned, you look at another player crosseyed and you get banned. id rather play a game full of hackers and cheaters than have to play a game on a server running punkbuster. you might die alot but at least it will be fun.
-
http://en.wikipedia.org/wiki/Binary_multiplier multiplications are done with adds and shifts so i would assume that would be disqualified. but the right binary bit shift, while is equivalent to a multiplication by two is neither multiplication nor addition.
-
if you can export to dae is use that over the max (or .gmax) format. seems the unity editor likes the dae files better, i tried switching over to just loading my max files, and things started getting weird. scales were unpredictable, xform resets dont seem to solve it, and it seemed i never got the same results twice. of course i use max 9. gmax is a little bit too stripped down iirc, but i havent used it in years. i do this: model everything in the (g)max format (you kinda have to), and save it to my wip folder. when the model is done export as dae to my export folder. import into unity do unity stuff exports directy to the my kspDev folder (a second installation of ksp which is kept stock with the exception of my own mods) using the write button as per unity configuration. i set up all my hierarchy and animation in unity. i would prefer doing this in max, so i have less to do in unity (this would simplify model revisions), but ive never gotten that to work. this is probibly the main reason to import directly from the max files.
-
needs point away from face decal and it would be perfect
-
1<<1=2 lol. i like programming in cpu arcitectures that do not have hardware multiply. and this right shift by 1 is equivalent to multiplication by 2. essential operations when dealing with bit fields, also useful for implementing binary fixed point numbers. you shift the integer part and then or in the fractional. that said you could also do this: 1*2=2
-
well the idea was to create a massive explosion and propel the kerbal with the kick. but your idea might work better. i try to limit my ksp playing these days so i wont be board when the game hits final. i play the new version for a couple weeks then go do something else until the next version comes out.
-
i have an idea, use a very large array of very large fuel tanks. point a mainsail directly at each one (think of it as a detonator). and have a platform with a kerbal suspended over it by a platform with a decoupler. set up the action groups to light the mainsails and release the kerbal bearing platform at the same time.
-
on one hand its kind of cheap. but on the other, you built your rocket(s), launched it, docked it with 10 other things, flew it to duna and accidentally hit the planet, im not starting all that over again (especially when it took and entire weekend to do the mission). i say "ok, you got me", then pound that f9 key. my time is valuable to me.
-
since ksp has come out ive hopped from c2q, to an i5, to an i7 (all with 8gb ram). for the later two i had a gtx 560, and the former had a gtx260 (i think). i really havent noticed much difference in performance across those machines. the gpu is way underutilized by the game. and cpu performance is what matters, specifically single core performance, which frankly hasn't been advancing as fast as it used to.
-
i believe the proper astronaut lingo is "roll program".
-
undocumented (or poorly documented) features for the win. not that i would read the documentation. instructions? i dont need no stinking instructions!
-
in retrospect never liked my original x-52 or the x36 i had before it. when i owned them i thought it was the best thing since sliced bread. then i bought ch gear and realized how much saitek sucks.
-
you want boring: any mission involving ion engines. really i think we should be allowed to physics warp when using ion engines.
-
i usually start a slow rotation from 90 to about 35 degrees starting at about 12000m or whenever my first stage runs out (not counting any boosters attached) and i time it so that by the time im out of the atmosphere im at 35 degrees. this manuver is made easy because joystick. then as soon as i cross 100km il flatten out to 0 for orbit. the only exception is if i have a slow upper stage such as ion powered, or a heavy nerva ship. in which case i usually stay at 35 (or sometimes higher) for most of the orbital insertion.
-
dont get me wrong, math is fun, and its taken be 12 years out of highschool to realize it. i do have a crappy 2 year it degree from a diploma mill, but its not worth the paper its written on, i didnt have any math classes to get it either. had i appreciated math more coming out of high school i would have gone for a 4-year engineering degree at an actual university.
-
the reason most people hate math is because schools really suck at teaching it. in high school math was this 1-dimensional abomination where you did a hundred identical problems every day for weeks at a time, and they only got as far as basic geometry and algebra. in the decade after high school, i ended up teaching myself advanced geometry, trig, linear algebra, calculus, etc, on my own, with very little effort. why? well because i kinda had some computer science early on, and it kinda got me into game design as a hobby (starting with simple 2d classic game clones to eventually 3d game engines). i wasnt trying to solve a page full of arbitrary numbers but instead actual problems, like how to draw a line in a framebuffer and how to collision detect a polygon with a ray. math education is too procedure oriented. they show you the procedure and then give you problems ad nauseum until you get really fast at it. this might have been a good way of doing things prior to the 80s and 90s where calculators became dirt cheap. everyone needed to be able to work with money, so you needed your basic operations down, and the construction, automotive and food service industries needed you to be able to operate on fractions, but in this day and age where everybody practically keeps a computer on their person at all times it doesnt make any sense at all. a boss isnt going to hound you for using a calculator to solve a problem like a math teacher wood because they know it will save time and increase profits. math education should focus more on solving actual, practical, problems instead of the monotonous grind it is today (i dont really think its changed that much since i graduated highschool in 2000).
-
id say comparing a finished title to one thats still in development isnt a fair comparison.
-
ive seen at least one crash to a heap fragmentation error. when you allocate memory you get a block from the os. even if its a small amount of memory, its faster to allocate a large block than it is to allocate just what you need. and then subsequent allocations come from that block. so if you allocate space for some data structure, like a 3d model, and then space for a texture, if you free the model you free up some memory, but since textures need to be contiguous, and they happen to be larger than a 3d model, you cant store the texture in the slot the model was in. so instead you allocate a new block. if you keep allocating and reallocating space of various sizes eventually you will reduce your memory space to swiss cheese, a lot of holes, and together it could be a gig or two, but nothing big enough to fit your new data into, and the game crashes.
-
i like em. they are more stable than the carts and modular wheel mods. they stick to the ground better and rovers dont fail catastrophically at high speed. also its a moon buggy, not a ferrari.
-
considering the game is in development and will be for years to come, it makes sense to put off memory optimizations and stream loading until other features are fleshed out. you really cant predict what kind of system the game will need till it has some substance. you can try to make the perfect memory management scheme prior to the rest of the game, only to find out the two just dont get along. its better to implement a basic system to allow further development in other areas, and when the final requirements are better understood then go back and swap out the basic system for a more optimal one.
-
i docked one of those orange tanks, full, to another thing. with a tiny probe docked at one end only equipped with a sas and an ion thruster.
-
its not really an issue with propulsion, the power systems will fail eventually and render communication impossible. solar panels dont cut it that far out, and rtgs die out. really dont think we need to build brpobes built specifically for leaving the system. pretty much every time you do a flyby of an object in the distant part of the solar system, you send the probe out of the system, and so those probes get secondary and tertiary objectives.