-
Posts
28,868 -
Joined
-
Last visited
About manikyath
- Birthday Dec 26, 1994
Profile Information
-
Gender
Male
-
Location
commonwealth in the former soviet banana republic of belgium
-
Interests
Breaking stuff, to make it better.
-
Biography
Microsoft Certified Ass..ociate.
-
Occupation
Not causing lithium fires.
-
Member title
keeper of the magical rowenta space heater
System
-
CPU
Ryzen 7 5800x
-
Motherboard
Asus ROG STRIX X570-E GAMING
-
RAM
32GB DDR4 3200 (corsair vengeance LPX)
-
GPU
asus GTX970 strix
-
Case
Fractal Design Define R5 (black, no window)
-
Storage
2TB Samsung 970 EVO PLUS M.2
-
PSU
corsair RM750
-
Display(s)
2x BenQ BL2420PT, Dell UP2414Q
-
Cooling
Noctua NH-D15 chromax.black
-
Keyboard
logitech G512
-
Mouse
logitech MX master 2S
-
Sound
hyperx cloud | pioneer vsx-c300 | PE HSV 40T
-
Operating System
MSX brickware edition.
-
Laptop
Asus ZenBook flip 15 | Asus transformer book T101HA
Recent Profile Visitors
155,310 profile views
manikyath's Achievements
-
how are you testing this speed? it's not because you *have* a direct 2.5G link, that it'll absolutely use the direct 2.5G link.
-
How much ram is enough for a minecaft server
manikyath replied to ThePCNerd129's topic in CPUs, Motherboards, and Memory
first having to add my usual ramble about "dont home-host for anyone outside your direct group of friends, essentially only those you'd invite for a barbecue".. i cant be arsed to keep writing a paragraph for that. that said, unless you plan on doing some forced chunkloading stuff or very high render distance stuff, allocating 2 gigs is plenty for a handful of players. -
Upgrade from Windows 8.1 to Windows 10/11
manikyath replied to IndirectX's topic in Operating Systems
there's a number of things that could be wrong there, but i'm not here to disprove you with wild theories i cannot verify on your hardware anyways. i *have* gotten quite handy with keeping both win10 and win11 in check. usually when they run like ass it's to do with some windows update process (usually drivers), or just some insane misfortune with platform specific drivers. i have one AMD based netbook that runs both win10 and win11 better than the win7 it shipped with, other than a bug with the latest release of the mouse driver, which crashes and takes the keyboard driver with it because someone at HP was truly a genius that day... so it's basicly unusable. -
let's break down what your cold unfeeling robot has to say... your AI misquoted their own source.. they're on about how creating an engine with fewer parts simplifies the manufacturing process. again.. a misquoted source, it's a stackexchange discution (of all places to pull data from...) about the pros and cons of both sides, and the reasons to choose one or another. it's a shame your AI misquoted it because it's a pretty good read. my job is safe, AI will not replace me.. because nothing in the quoted source evn talks about TWR, instead it talks about the reliability benefit of having more engines, and how modern computer technology allows us to control more complex rockets more accurately. also - kinda funny.. the source YOU are quoting here.. is elon musk. same source as 2, same misquoting.. the only reference to combustion stability is a reference to car engines. this is exactly what we have been telling you.. and you now blatantly use it to somehow try and prove your own point? have you any idea how ridiculous this makes you look? have you not proof-read this at all? so.. NASA then? because the artemis contracts are historically tight timing-wise, presumably because of the 6-year delay from NASA's own internal dealings. i think you mistake "being cautious with optimism" for being critical. now, i've only watched the final thoughts of the video you linked.. but i dont think they sounded critical at all. apollo essentially got a bunch of "test subjects" to the moon and back in a tin can, as a physics person you ofcourse understand that if you want to get more than "just another tin can" to the moon, you'll need more energy. the goal of artemis isnt "put human on moon", the goal is to put ACTUAL science down on the moon, and be a pathfinder for further human spacetravel. or to put it in a vaguely quoted clarkson quote from earler in the thread: enough time and money.. should i bring up the cost difference between starship and SLS again? and why do you think NASA isnt flying the whole party with SLS? BECAUSE THEY CANT AFFORD IT. wether you, or anyone else at NASA likes it or not, fact of the matter is that any chance the US has got at winning the modern era space race has to include their commercial launch partners, because they simply cannot do it on their own dime. NASA is too slow, too expensive, and too complicated of an entity to do revolutionary things. the reason why SpaceX is actively blowing up starship prototypes is because it's cheaper and faster to develop that way. two things NASA sorely needs, but cant do themselves for political reasons.
-
let's go trough the rest then: - Atlas V: it has flown 99 times, has 17 more scheduled flights, and ULA is not selling any more flights.. it is therefore just barely "still in operation". it is also HUGELY aided by SRB's, supporting my previous notion that the number of engines might be more related to wether they want to rely on SRB's or not. - Delta IV heavy: has flown 15 times, with one more flight scheduled before it's announced retirement. so.. again, just barely "still in operation". - H-II is JAXA's (japan) SRB-only first stage rocket, i'm willing to debate if this should even be considered a "few engines" rocket, because it's basicly an advanced firework. cool tech, but irrelevant in the debate of number of engines, because clearly "no plumbing at all" hugely reduces the chance of plumbing issues. this is also both less in number, and SIGNIFICANTLY less in number of launches than the "many engines" rockets.. do you want to correct your hilareously poor autopilopt response, or should i just assume that many engine cluters are 'at least just as viable' as few engines aided by SRB's? as far as i'm aware, past the first stage just about everything is one or two engines.. aside from starship, which is unique in this regard because they are hoping to develop soft landing capabilities. so, not a fan of SLS then? because you must have missed the previous douzen times i've mentioned that Artemis 1 launched 6 years late. i know you're just ignoring this fact for the benefit of your own argument, but i'm not. so whatever the case is, if the NASA contract says "moon by 2026", they have until 2032 before they're later than Artemis 1. i'm not saying this is a "free pass" on a 6 year delay, but if you are criticizing the chance for delays on starship, you cant then use SLS as an example of being on time. that makes two of us.. i just happen to enjoy rockets that happen, i'm in it to watch cool technology happen, preferably more than once every few years. i like SpaceX because they launch a lot, and their development is very "visible" which really vibes with my engineer mindset. watching artemis 1 fly was cool.. but it was one flight. likewise i really enjoy rocketlab for their "style" of rocketry. their broadcasts are almost like they're playing KSP in real life, down to the gaming chairs and sponsored by logitech headsets.
-
Upgrade from Windows 8.1 to Windows 10/11
manikyath replied to IndirectX's topic in Operating Systems
you know... this performance thing is utter marketing nonsense. win11 is just as heay as win10, and i've ran it just fine on a wide range of shitboxes, including an i3-540, and a j5005. i do support the notion of going for an ssd, win10 and win11 both rely on disk IO far more that win8 and before. -
name them, because between SpaceX, rocketlab, and soyuz, i'm pretty sure i caught the great majority of launches in the past 5 years. souyz is also a 60 year old design, the treshold of "cost of complexity" has changed A LOT in that timeframe. besides the obvious technological advances made over the past 60 years, soyuz is also essentially an ICBM adapted for human and freight transport. it was *a* solution that worked, and has been extremely reliable in any weather, as you'd expect from an ICBM.. but that doesnt mean plumbing all those engines into a single thank then cant also be a viable option.. which.. i'm gonna reiterate again: and likewise.. i'm gonna suggest you look into N1, because it really appears that the number of engines had nothing to do with the problems it had. you can keep shouting nonsense, but if i can come with examples and you cant, i'll just have to assume that what you're shouting is in fact nonsense. and just to make sure you dont accuse me of talking nonsense: in 2023: - 96 falcon 9 family launches - 9 electron launhes - 2 ariane 5 launches - i'll count the one SLS launch in november of 2022, because otherwise NASA isnt on this list at all. - 17 souyz launches - 3 "long march" launches (that's china) - 9 ISRO launches, on a variety of solid rocket booster configurations. - 2 JAXA launches fact of the matter is that every big ticket item on this list is either a 9 engine cluster design, pure SRB's, or the undecided quantity that is souyz. while a cluster of 9 engines to a cluster of 33 engines is a big jump, SpaceX has clearly proven that the plumbing into a single tank isnt an issue, because for all 3 test flights none of the problems were caused by said plumbing. - in fact.. all i can find about N1's plumbing is the engine manufacturer blaming the plumbing for their engines blowng up. as for the complexity of many rocket engines.. let's talk about the simplicity of many rocket engines.. or as someone responded to this photo below: forget your special transports for spacecraft parts, having the engines this size and none bigger means you can actually deliver them "by the truckload".
-
did you read my post at all? at higher bitrates ah.. so this is what it's about.. look.. i'm sorry to bust your FOSS bubble here... there's benefits to FOSS, there's benefits to OGG in particular, but fact of the matter is that music files are very small in today's standards, so outside of streaming platforms where that bandwidth really matters, it's either 320k MP3, or FLAC for the people who spend 4 digits on their headphones. since MP3 has kind of become a de-facto standard it makes no sense to migrate away if there is no clear benefit. low-bitrate quality is not a benefit in the majority of usecases. oh - and for streaming services.. they use what they deem is the most interesting format to use. they have control over the backend, they have control over the frontend, they can stream in .txt for all i care, as long as the quality is good and the bitrate is acceptable. believe it or not.. my entire music library is stored as both MP3 and lower bitrate OGG, because i used to have a phone with very little storage and it supported OGG so i used the 'more quality at lower bitrate' format there.. literally havent touched any of that since i've upgraded my phone.
-
most rockets in operation? name them. i'll name some "many engine" rockets: - rocketlab electron uses 9 *tiny* engines on the first stage. like actually "you could put one on your desk" tiny. - SpaceX falcon 9 uses, as the name implies, 9 merlin engines on the first stage. - SpaceX falcon heavy uses 9x3 merlin engines, for a total of 27 engines. - debatable if soyuz counts, they have quite an exotic design, see sidenote below. as for upcoming stuff: - rocketlab's upcoming "neutron" design currently also has 9 engines on the first stage - SpaceX's upcoming "Starship" has 33 raptor engines on the super heavy booster. now.. all of these have a pretty major design choice they all share: no solid rocket boosters. it's almost as if some things are related here.. maybe SLS can get away with 4 liquid fuel thrusters because the SRB's carry over 75% of it's thrust on the pad (that's on NASA's fact sheet). that means that each of the 4 engines on SLS provide less than 6% of the total thrust on the pad. sidenote.. what the heck is soyuz even.. on paper they're 5 engines with 4 combustion chambers each.. and a collection of 12 smaller thrusters for steering. but really.. depending on how you count, we're looking at somewhere between 17 and 32 thrusters here.. and soyuz is pretty much *the* most proven launch vehicle on the market. and all of this is completely irrelevant because.. news flash.. many thrusters isnt a problem, superheavy booster has made it up to hot staging twice with zero issues, the uphill portion hasnt been an issue since that one time they rapidly disassembled the launch pad. and the downhill portion is of no effect to artemis. big rocket go up is the goal, big rocket come down is a cost cutting bonus. is this the place where i remind you that SLS is 6 years late, and work on starship didnt begin until after that 6 year delay in SLS's first launch? as NSF has talked about before (and no, i'm not gonna delve into their hours upon hours of artemis ramblings) atemis's reliance on starship is NASA's "fault", SLS is way late, SLS cant provide all of the launch capabilities they want, SLS cant reach the budget goals they want. if you really delve into the time scale of "NASA returning humans to the moon" the most obvious answer is that using a launch partner is almost an "afterthought" because it eventually occurred to them that given the budget and timeframe they had no other option. just to remind you; SLS is quoted to cost about 4 billion per launch, which is more than the 2.9 billion they awarded to SpaceX for the entire HLS programme. oh, and while researching i found this statement, that nicely explains what these IFT flights are: (hint: "T" stands for "Test")
-
this is based off a quick google; yeah.. who knew.. making an application specific card will outperform everything else in that specific application. if only the crypto world could have told them this... oh wait. but yeah.. outperforming nvidia's A100 in LLM's, but gets beat by the A100 in other AI tasks.. the revolutionary thing here is that they made it, not what it does. let's just hope this market for AI ASIC's (that's a horrible name..) grows extremely quickly and nvidia has to go back to making GPU's for other things, it's almost like the crypto boom all over again..
-
they have made it work just as many times as sls, there is no reason to suspect SLS is now a proven known value while in the same breath you would suspect starship to turn out deeply flawed. and again... "the space focused creators you have watched" - the people i watch have been covering launches since the days of a written blog covering shuttle launches. they ofcourse have concerns about what goals starship stillo needs to rach, but they see no reason why spacex would not be able to reach them. I'm sure that if you went to look up any of these outlets back when they were covering the early days of falcon, they would have been highly critical of this unproven commercial space entity and their wild ideas. fact of the matter is (and NSF does refer to this occasionally) that SpaceX has proven they can trailblaze radically new ideas, bring them to market, and turn them into highly reliable services. people being critical at an early stage of development doesnt have any implications for how the design would be somehow deeply flawed, and you've yet to show me any evidence that it would be flawed past your own opinion on the engine count. which.. about that... so.. it was a factor when the rocket was going down vertically trough atmosphere, but wasnt a factor during the flip, y'know.. where fuel slosh would be an issue. now.. i could be wrong, but my limited education in fluid dynamics says that fuel slosh happens when an object is making large movements, not essentially going vertical at terminal velocity. ascribing the failed landing burn to fuel slosh without any sort of reasoning behind it is nothing more than your shallow preferences for one solution over another. wich, just to remind you... landing the booster is a SpaceX thing to cut costs per launch down to that "0.1% of the cost of SLS", it is in no way required for the HLS contract. to just remind you again, in a big font this time, nice and separated out, just so you get why SpaceX is tossing out conventional knowledge to "re-invent" a better wheel; 0.1% the cost per launch of SLS.
-
i find that a rather harsh line.. for platforms like spotify a theoretical cut of 30% bandwidth is a HUGE cut in their costs, because while one stream is a very small amount of data, they have potentially millions of concurrent streams, they have to store all that, they have to have worldwide infrastructure for that. people like you exist, i'm not gonna question the validity of your actions, what i'm saying is that people like you are not the target audience for those platforms that *actually have to think* about the size of their music library, because it is a very large impact to their bottom line. it's all good talking about how they treat consumers, but if spotify suddenly had to be a $60/mo subscription because "only serving high bitrate flac is the right thing to do", that would be prohibitively expensive, and actually be problematic for people's data plan. fact of the matter is 99% of their target audience doesnt hear a difference between 320k mp3 and your flac files, and i'd dare wager a good 90% of said target audience wouldnt even notice of it were suddenly 128k mp3 again, because even my "very gamer" headset is better quality than what the majority of spotify users listen to their music on. or to put it very short and blunt... to you it's art, but for the majority of spotify users it's "not silence". in the same way you might spend hours to appreciate a fine painting, but the majority of paintings are used as "not an empty wall". at work we have a good old FM radio, and when we plug in too many chargers it starts to get static-y.. that's the market spotify is in.
-
oh lord.. this is some pent up beef you got here.. from what i recall from back when i cared.. what it comes down to is when limiting bitrate or filesize is important, ogg is superior. but if peak quality is what you're after, MP3 does better at the higher bitrates.
-
i understood that as 'only the ISPs that do FTTP offer such options", not "FTTP is always symmetrical". the reasoning behind this is that both coax and copper infrastructure is essentially half-duplex, meaning the more upload speed they offer, the less download speed they can offer. this isnt an issue with FTTP, hence FTTP is the place to be to get symmetrical. but yes.. could be more clearer in the video. every time i see an example of this i wonder why text books even bother to include information like this.. what a way to date yourself. there's even a brilliant gem like this burried in a cisco exam..