alpenwasser

Since this seems to be nothing more than drama, locked.

@AresKrieger Oh, good point about the common GND.

Yes, I did read it. Bad English does not equate bad circuit design. And it is just an example. Although I would simulate and test any circuit you pick off the internet before hooking it up to something important. If not for safety, then at least to properly understand it.

This will probably not be the same for every surge protector. For example, if a certain surge protector works based on one or several capacitors, your overall capacitance in the chain might indeed sink below what a single one can provide (depending on how specifically the circuits are designed), according to the law for connecting capacitors in series: 1/C_total = 1/C_1 + 1/C_2 + 1/C_3 + ... Good surge protectors should be more complex than a simple capacitor, of course, which makes the overall outcome less predictable. So yes, I'm inclined to agree with whatever people advised you against this. If you want to know with a higher degree of confidence, take a specific surge protector apart and analyse the circuitry (or have somebody do it for you). EDIT: I'm bringing up capacitance because a capacitor's capacitance influences how much energy it can store (0.5 * C * V^2). EDIT 2: Also, here's a DIY homemade surge protector, as an example of one implementation: http://www.brighthubengineering.com/diy-electronics-devices/90219-build-a-simple-whole-house-surge-protector/#imgn_0 One could try to simulate this circuit, then connect two in series and see what happens. EDIT 2: Here's another one: http://www.eleccircuit.com/diy-surge-protector-circuits/

Edited a few posts for aggressive language towards other users. Keep it fucking civil, folks. Thanks.

One of the primary advantages for me is it's usually scriptable, so you can easily automate comparatively complex and/or tedious tasks. If you know the right commands, you can reduce your workload by many magnitudes. Downside: You do need to be aware that those commands exist and roughly what they can do. The details you can always read up in the manual, but if you don't know they exist in the first place, you won't even get that far. Primary downside for me: Much smaller discoverability. When I have a GUI program in front of my eyes, I can start randomly clicking myself through menus, see what's there, without needing to read the manual. That won't really work for most CLI tools, unless they have sophisticated autocompletion facilities.

Dual CPUs, you say? Coppery goodness, you say?

@Stefan1024 Damn! Despite not-so-great internet, I am actually slightly jealous of your new location. I've always had a weakness for spots which lie a bit off the beaten path, so to speak. Plus, I love lakes, rivers, the sea and all that good stuff.

@ShenaniganCoder I've merged your threads together and left the second OP visible since it has one more picture.

Yeah, I can access my CPU and GPU temps from the console in Linux (Intel CPU, Nvidia GPU): I'd probably just need somebody to code the Windows side for me since I haven't worked with Windows in many years. But yeah, should be feasible I think. However, as you can see, the chipset temperature is not displayed, for example (despite my motherboard having one). So you're limited to which sensors your OS supports. I'm also not sure how the AMD side of things is looking on Linux. But I reckon I can always make it a feature which works if the temp sensors are supported, and if not, you can fall back to temp probes.

@LoGiCalDrm Yeah, getting direct temperature readings from the CPU/GPU/Motherboard sensors would be a nice way to do things instead of/in addition to using temperature probes. I suppose if I could read out the temperature sensors from the operating system and then send the results to my device, that might work. The OS would need to support the temp readouts, and somehow offer me a way to tap into that, but if that's feasible, then sending those results to my device shouldn't be too much of an issue. Thanks!

Yeah, I figured something like that. Sounds very hackerish, I like it. You wouldn't really need to know anything about fans specifically. You'd just need to be able to connect to my device somehow and then send the correct commands to it to do what you want (which would be documented in the API docs). So the language you use wouldn't matter much as long as it offers the features which are needed to connect to the device (which aren't defined yet). Lumping these two together because they go kinda in the same direction. I have considered doing basically something similar to IPMI (obviously with fewer features, but at least remote power-up, power-down, and hard reset), and using a RasPi or an Arduino for that. However, I would most likely need to design an add-on board for the Arduino or the Pi anyway, and at that point I'm not yet sure if it's not easier to just make my own solution from scratch. But yeah, doing remote power control would be pretty cool, because I've longed for that as well on systems were wake-on-LAN just doesn't work for one reason or another. Power-up/down and hard reset would probably not be that difficult. It would probably require using ethernet instead of internal USB, which is a bit more of a hassle from what I've been told by people who have experience with both. Doesn't mean it's impossible of course. Alternatively, I could put a GSM module on it and you could send a text message to it to have it power up/down/reset, maybe. Downside of that is that you'd need to get a (prepaid) SIM card specifically for that. WiFi would be another possibility, although I've heard she's a significantly more fickle mistress, so I'm not sure I want to go there. But hey, at this point it's all about finding ideas, thanks!

If I can find somebody to do that for me, or if you can find somebody, it can definitely be done though. The whole point (well, a huge part of the point) of this thing is that it's open and well-documented so that anyone can add more stuff to it. That's the main thing which makes the Aquaero not an option for me. I will add the feature to the list so that it's not forgotten.

I will probably not be writing any major windows software for it myself because I just don't use Windows enough to make something decent on that front. If I do make something which goes beyond the command line, it will likely be a web interface, because that runs on pretty much any platform as long as you have a browser. However, it will have an open API, so anyone could make additional software for it. Why not both?

Hm, that actually might not be the stupidest of ideas. Adding to OP. Edited title because apparently it was too ambiguous. Who'd have thunk? But yeah, think Aquaero: But probably without the display, at least for starters, so it would look more like the headless version:

Title pretty much says it. Basically, I'm looking into making my own device. As capable as the Aquaero is (as one example), the Aquasuite is not of much use to me on Linux or *BSD, nor can I control it from the command line (which would be very handy for a server, for example). So instead of trying to reverse-engineer Aquacomputer's USB protocol (and needing to adapt every time they change something), I'd like to build my own cooling controller (not necessarily just for water cooling, which is why this thread is in the GD section). Besides, it should make a nice project to get some practice in electronics development. Ideas which have been thrown around so far: Accessible from terminal, suitable for headless machines Open comm protocol, so other interfaces (e.g. web interface) can be easily added (it's not going to be Linux-only) 4-pin PWM output temp sensors, obviously voltage, current and power monitoring, possibly programmable fan curves via mathematical expressions for coupling temp sensors to fan/pump speeds (thanks to @Ryan_Vickers for that idea) connection to computer (probably) via USB should provide enough power for multiple pumps and lots of fans (haven't really decided on a number yet) possibly expandable in a later revision (adding more pumps/fans/whatever else) lighting control, emergency shutdown, ...? can spoof temp and tach signals (courtesy of @manikyath) Windows taskbar thingy ( @AluminiumTech) IPMI-like features temperature readouts via OS and internal component sensors instead of temperature probes ( @LoGiCalDrm) Since it's easy to get tunnel vision when looking at a problem for a while, I thought I'd ask for some input on what others think. I'm not really looking into manufacturing and selling it at this point, although I will make the entire thing open, so anyone could have it made or make it themselves if they felt so inclined. Aside from that, it's a rather long-term thing, not going to be done in the next few months. Oh, also: If anyone has any bright ideas regarding names, feel free to throw them out there.

Terribly sorry for the long delay; this semester's been kinda crazy, and I had the great idea to fall ill twice, which didn't help with having spare time to hang around on the forums at all. Anyway: Added/updated. You pretty much got it. Just make a new post with the updated info. I'm not yet 100% sure what your drive config is though. You now have two machines, each with six 2.6" drives? You wouldn't happen to have the vendor for those drives by any chance? I'll update you ASAP once I'm clear on what's what. Thanks! I think we can tolerate some off-topic chit-chat. Besides, it's sorta on topic. Anyway, the way you do that is by getting HTML editing privileges. Sadly, can't really enable that for regular users for security reasons. ^This. Updated/added. Thanks for participating, everyone!

The Sueddeutsche Zeitung (Newpaper from Southern Germany) was approached about it about a year ago and the thing then got going. But while the data was being worked on, new data kept coming in until late 2015 as far as I can tell. And then the results went public yesterday. The data itself goes back to 1977. I presume that's why at least parts of it needed to be OCR'd so that it could be indexed and processed electronically. It was probably just scanned handwritten or typed documents. But maybe we'll find out for sure at some point.

Well, if you siphon it off over a few months and pay attention to not use too much bandwidth (and vary how much bandwidth you use, not just same load 24/7) it would still be doable. I've uploaded quite a few terabytes within somewhat reasonable timeframes over a 50 MBit/s upload line. Plus, the victims might just have shitty security. I'd still lean towards an internal leak, personally. But either way is just guesswork in the end.

Unless it was an internal leak. Haven't seen any info yet on that.

I doubt it will fundamentally change how the world works. BUT: I would bet that heads will roll. Not all of them, but some. If we think long-term, things like this might actually cause a slow shift. If people can't rely on their secret machinations being kept secret anymore, it might affect their behavior for the better eventually. Not all of them, and not from "totally corrupt" to "holy savior", but if it has the effect of some people not going quite as far as they would have otherwise, that's change for the positive. It won't save the world, but it might improve it slightly. And in the end, through the millennia of history, improvements usually only happen in small increments over long periods of time. But still, the long-term trend is upwards. If you compare the life of an average human today with that of an average human 3000 years ago, we're probably better off today, all things considered. Not all of us, but on average, things are better. It was from an anonymous source who handed it to the Sueddeutsche Zeitung. How admissible it is in court might depend on court, country, document... probably not going to be decided for everything at once. But even if it's not, it could still end more than one career, or at least severely damage it. In any case, I think this should make for an interesting combination of disillusion and schadenfreude in the coming weeks.