|
Most likely not safe. The LevelDB isn't really 100% ACID, especially the chainstate directory is prone to corruptions if the node crashes.
Edit: on the other hand if you have a tight backup schedule or the underlying file system supports snapshots then it is OK to run them over NFS because neither the blocks nor chainstate LevelDB databases are really critical and can be always rebuilt after the corruption. It all depends on what exactly is your tolerance profile for outages.
|
|
|
|
As for C++ examples - they will be easy enough to create once we actually have a C++ compiler.
So at the start of the project you are in an endless recursion: 1) you need C++ compiler with cryptocoin extensions 2) you want to produce the design documentation for those extensions using the results of (1) It is a bad PR to do this in public. You seem to be severely underestimating the effort and resources required to retarget GCC/LLVM/Clang. You also seem to be overestimating the possible benefits of having such a toolchain available. The only way I can imagine such retarget to be successful if it is going to happen within the educational institution, meaning the development isn't going to be done for money but for a degree credit. |
|
|
|
I studied JavaFix, but the problem is just, that it's so different from most of the current exchanges.
When you say "current exchanges" do you mean "cryptocoin exchanges" or "established financial exchanges"? In the FIAT financial world every exchange/broker/etc. supports FIX or equivalent session-oriented protocol. When they do not support it its a good sign that it is a bucket shop or other scam. |
|
|
|
Just write some simple, "Hello, world!"-class, example. Without such specifics the whole idea is just too vague and open-ended.
We have already published much more than "hello world" examples of AT ( http://ciyam.org/at) and the first ACCT (Atomic Cross-Chain Transfer) between two separate blockchains (Qora and Burst) has already happened. I could not find any C++ example in your documentation directory. All examples were tabulated assembly listings. I wanted to see your vision on how C++ will improve the situation for a programming user or a financial service developer. I honestly lack this insight. Do you seriously think we don't have a product that works?
What I'm seriously thinking is that you are similar to some newfangled hardware microprocessor developers from Europe that I've met in the past. They had their architecture, they had their assembler and simulations. They even had real silicon implementations in structured ASIC. But when no further investments were forthcoming they decided that they need open source GNU toolchain and an IDE based on Eclipse. And that was where they went bankrupt, got acquired and forced to realign their design with real market demands. Really, instead of posting one-liners, please sit down and write a nice position paper that has your atomic cross-chain transfer recoded in your intended dialect of C++. Something like the listing below will showcase your ideas much better that what you are currently doing. ; Listing generated by Microsoft (R) Optimizing Compiler Version 17.00.50727.1
.model flat
_DATA SEGMENT
$SG1297 DB 'Hello, world!', 00H
_DATA ENDS
PUBLIC _main
EXTRN _puts:PROC
; Function compile flags: /Odtp
; File c:\users\2112\hello.c
_TEXT SEGMENT
_argc$ = 8 ; size = 4
_argv$ = 12 ; size = 4
_main PROC
; 1 : int main(int argc,char **argv) {
00000 55 push ebp
00001 8b ec mov ebp, esp
; 2 : return puts("Hello, world!");
00003 68 00 00 00 00 push OFFSET $SG1297
00008 e8 00 00 00 00 call _puts
0000d 83 c4 04 add esp, 4
; 3 : }
00010 5d pop ebp
00011 c3 ret 0
_main ENDP
_TEXT ENDS
END
Edit: I just realized that a code listing example using the venerable Intel x86 architecture is too readable, people are too familiar with it and it was designed to be hand-programmed in assembly and therefore has nice mnemonic register names. To make a fairer comparison with AT I produced the same listing with ARM/THUMB compiler. This should better reproduce the level of confusion and discomfort that one would encounter with a completely new machine architecture like AT. ; Listing generated by Microsoft (R) Optimizing Compiler Version 17.00.50727.1
00000 AREA |.data|, DATA
|$SG619| DCB "Hello, world!", 0x0
EXPORT |main|
IMPORT |puts|
00000 AREA |.pdata|, PDATA
|$pdata1$main| DCD imagerel |$LN5@main|
DCD imagerel |$unwind2$main|
00000 AREA |.xdata|, DATA
|$unwind2$main| DCD 0x33200011
DCD 0xa8fb02
DCD 0x8802ff02
DCD 0xff03ef00
; Function compile flags: /Odsp
; File c:\users\2112\hello.c
00000 AREA |.text|, CODE, THUMB
00000 |main| PROC
; 1 : int main(int argc,char **argv) {
00000 |$LN5@main|
00000 b403 push {r0,r1}
00002 e92d 4800 push {r11,lr}
00006 46eb mov r11,sp
00008 b082 sub sp,sp,#8
0000a |$M3|
; 2 : return puts("Hello, world!");
0000a 4806 ldr r0,|$LN8@main| ; =|$SG619|
0000c f000 f800 bl puts
00010 9000 str r0,[sp]
00012 9b00 ldr r3,[sp]
00014 9301 str r3,[sp,#4]
; 3 : }
00016 9801 ldr r0,[sp,#4]
00018 |$M6|
00018 b002 add sp,sp,#8
0001a f85d bb04 pop {r11}
0001e f85d fb0c ldr pc,[sp],#0xC
00022 |$M9|
00022 defe __debugbreak
00024 |$LN7@main|
00024 |$LN8@main|
00024 0000 0000 DCD |$SG619|
00028 |$M4|
ENDP ; |main|
END
|
|
|
|
|
Please post a single, concrete example of how you propose to extend C++ to utilize the blockchain-specific functionality of AT. Anyone can make a good guess on how to use the general-purpose computational functionality of AT. But how do you envision the whole GCC/LLVM toolchain integration with the cryptocoin engine? I can't even guess if you think of adding new intrinsics or if you think of a template library.
Just write some simple, "Hello, world!"-class, example. Without such specifics the whole idea is just too vague and open-ended.
|
|
|
|
When I did my HFT attempts, the biggest problem were the rest APIs of the exchanges, that had major access restrictions (often like 1 currency pair every15s, or so).
REST is for losers. The whole point of FIX is that there's only one connection long lasting being made for the whole trading session. FIX conceptually hails from the days of analog telephone modems, where establishing a connection took better part of a minute. The problem with FIX is that it requires implementation of the producer-subscriber design pattern and that is way too much for most of the programmers, who never ventured beyond request-response pattern. The main problems with REST (and other communication mechanisms based on HTTP) is that the sudden spike in popularity of the service is indistinguishable from a distributed denial of service attack. |
|
|
|
The 100ms timing seems very consistent.
I was wondering if I was missing anything obvious, any "magic parameter" somewhere.
TCP_NODELAY? Interestingly John Nagle has an account on this forum: https://bitcointalk.org/index.php?action=profile;u=28488Edit: I'm actually posting this from Windows, not my normal machine. It took me a while to locate the place where the OS include files are located. In a search for the correct spelling I found a funny comment in one of the Windows' SDKs: Obviously making a verb out of somebody's last name is a sign of respect. I'm just wondering whether starting it with lower case 'n' means more respect or less respect? |
|
|
|
|
Two questions:
1) according to you which cryptocoins exchanges have an actual working FIX protocol interface with real live data, not some "demo mode"?
2) are you (or your intended programmer/user) familiar with the totally event-driven, non-blocking, programming paradigm?
I'm asking those questions after some experience with non-cryptocoin trading where people and companies claim to support FIX, but don't really do support it and moreover are completely mentally wedded to the request-response paradigm (RPC, REST, etc.)
Edit: And maybe a 3rd question:
3) are you essentially asking for a cryptocoin equivalent of e.g. Interactive Brokers gateway that talks FIX with the exchange but does multithreaded RPC/LPC with the user's/programmer's algorithms?
|
|
|
|
Nah, sha-256 is definitely NOT broken ... not without a fully working quantum computer (then using Shor's Algorithm, NSA or some other powerful state entity could crack it in seconds)
Care to explain how could integer factorization algorithm apply to the cryptanalysis of Merkle-Damagard hash functions? |
|
|
|
I received a call the next day from the CEO of a well-known Bitcoin ASIC company telling me that my idea combined with their own optimizations would make their chips a lot faster.
My experience now shows that a necessary condition for being a "CEO of a well-known Bitcoin ASIC company" is to understand very little about digital logic design and compensate this with chutzpah. All publicly available information points to them having only part-time design talent or to making drastic shortcuts to the design workflow. In particular the statement about "a lot faster" in a "next day" call, without running the simulations, points me to some bullshit artist. I can only commiserate with you about not having any trusted friend to discuss your ideas without being worried about them being stolen. Thankfully I'm not in this position and have some contacts who wouldn't jeopardize friendship for a quick sale. |
|
|
|
The only attack I was thinking of when I wrote the Bitcoin header post, was all mining companies adopting tricks that give them some little advantage, but at the same time they degrade the performance of the network as a by-product. One of such attacks is cited when I posted about using approximate adders, and the danger that a monoculture of approximate ASICs can get stuck in a header that always generates a faulty addition.
Thinking out-of-the-box has both good and bad aspects: +) on the positive side it allows novel and unusual solutions to enter the field, like your idea of intentionally breaking the topmost level in the carry-look-ahead login of a 32-bit parallel adder, which you called "approximate addition" -) on the negative side it disconnects one from the already known solutions in the field. Some EDA tools already can split a 32-bit adder in a critical path into a pipelined pair of 16-bit parallel adders. The general methodology is called "register balancing" or "delay balancing". You've made a far-reaching statements about a possibility or necessity of changing Bitcoin hashing algorithm in the face of your discovery. Have you consulted your discovery in private with somebody knowledgeable with digital logic design? What did they say? |
|
|
|
quite nice mix of math and conspirology, that's very fine, let's continue  I'm not satisfied with calling the psychological aspect "conspirology". My current thinking is that many people here can't really understand either mathematical or software engineering aspects of Bitcoin. For them it becomes a sort of gnostic theological experience, even if they claim to be atheistic. In this thread they will search for a (bad) influence of demiurge who corrupts their ideal. In the nearby threads by no-ice-please we observe a dissatisfaction with insufficiently immaculate conception of SHA256 algorithms. I hate to call them all "trolls". I can settle for "crackpots", because this seems to be the prevailing practice in the scientific fields. But theologians would simply call them "heretics", purveyors of the "bad theology". In very broad term, what we are searching in those threads is a way to communicate with people who'll never have time, knowledge, skills and motivation to completely understand the inner details of Bitcoin. Is there way to maintain the dialogue between the two groups? |
|
|
|
For the love of god, make this thread stop. At the very least, stop calling them nounces. I penned numerous responses to this thread and deleted them every time when I reminded myself of how this thread progresses...
Why stopping? It is fun and educational read about how to become a crackpot. It isn't very interesting technically or mathematically, but it is quite educational from the psychological point of view. The cryptocoin community will soon probably need an equivalent of the well-know "crackpot index" from the physics community. http://math.ucr.edu/home/baez/crackpot.html |
|
|
|
Cryddit gave an estimation on the number of standard gate building blocks required for a Bitcoin ASIC (adders, logic gates) However, adders require more space than OR gates, so generally the number of gates will be dominated by the number of adders. Also adders can be implemented in several ways, with different delay/space trade-offs, so even if there could be a theoretical minimum number of gates, practically all implementations would use much more to reduce the delay. More interesting, you can: - Compute SHA^2 approximately, and get a better practically good SHA^2 ASIC for mining. See https://bitslog.wordpress.com/2015/02/17/faster-sha-256-asics-using-carry-reduced-adders. - Compute SHA^2 asynchronously (e.g. using asynchronous adders) Last, it has not been proven that performing a complete SHA^2 evaluation is required on average to check that a changing header has a SHA^2 hash that is below the target value. In fact, several widely known optimizations have disprove it. It is hard to guess what was the original intention of the question: mathematical/algebraic minimum or some sort of minimum for a defined implementation technology. The typical goal for a implementation using silicon CMOS process would be timing closure, i.e. minimizing the time to compute the result. This is the reason why everyone is using carry-look-ahead adders that add 32 bits in parallel in a single clock cycle. The other approach would be to use serial adders that add 32 bits in 32 clock cycles. Such an adder is just a pair of XOR gates. The problem with this approach is that the resultant clock rate is too high for the practical silicon-based semiconductor manufacturing processes. But if somebody considers e.g. GaAs manufacturing then the very deeply pipelined serial implementation is a viable alternative. |
|
|
|
Vodafone does funny things as well: traceroute to bitcointalk.org (186.2.165.183), 30 hops max, 60 byte packets
-LAN-
2 10.218.128.5 (10.218.128.5) 26.468 ms 29.630 ms 29.621 ms
3 10.218.129.60 (10.218.129.60) 34.503 ms 34.500 ms 34.466 ms
4 10.218.129.34 (10.218.129.34) 40.485 ms 40.447 ms 40.410 ms
5 10.218.130.20 (10.218.130.20) 37.577 ms 40.392 ms 40.359 ms
6 10.218.130.26 (10.218.130.26) 43.195 ms 32.402 ms 49.601 ms
It is perfectly fine to use private/local IP addresses on the internal network of the ISP. So those hops 2-6 are perfectly fine as long as they don't leave the ISP's internal network. Funnies & shenanigans start when one sees the private packets routed (or apparently routed) over the public Internet. Try doing a reverse traceroute, meaning from other public address of another ISP to your IP address on Vodafone. If you then see those 6 hops over 10/8 subnet that would qualify as "funny". |
|
|
|
I am fairly positive that all four are wrong. I always try to be careful when I install something.[1]
I am not afraid, I don't know what to do. This problem is happening to chronicsky too and he is an Indian. It's not only for me. I thought I would get workaround or reason if I post here.
It is hard for me to take your statements on their face value. I'm an experienced engineer, but the teenage computer-gaming-kids I know all are well aware of their ping times. So when somebody posts a multi-second ping times like if this is nothing special all I can think of is http://www.urbandictionary.com/define.php?term=space+cadet . It is very hard for me to really put myself into your frame of mind, when the teenage kids near me (about 6th grade) all know their ping times and discuss them in the breaks of their soccer matches. Post your trace will you? Mine shows two class A IPs as well, just before "reaching" the DDoS protection, my ping times are significantly better though. See redsn0ws post above, they also have similar IPs 10.10.48.5 and 10.10.68.11 traceroute to bitcointalk.org (186.2.165.183), 30 hops max, 60 byte packets
3 83-169-183-73-isp.superkabel.de (83.169.183.73) 16.311 ms 24.876 ms 26.022 ms
4 ip5886c9ad.dynamic.kabel-deutschland.de (88.134.201.173) 26.888 ms 27.741 ms 28.585 ms
These IPs do not show when I use an VPN (university line). What is the point? I don't have routing problems. I always have at least 2 ISP available when at home, and all of them properly filter http://en.wikipedia.org/wiki/Martian_packet . Is "Superkabel" a reseller of "Kabel-Deutschland"? That setup looks like a typical reseller agreement where the reseller has no technical knowledge to maintain their network, only to do the billing. Perhaps you should ask you ISP why they don't filter martians? Your uni seems to have competent network engineers. Edit: The only thing of that may be of interest is that from my Eastern European ISP I see a hop to the Western Europe via bundle-ether9.ffttr4.FrankfurtAmMain.opentransit.net (193.251.255.201) before reaching ddos-guard.net which seems to be located in Russia. Edit2: Since more people posted the traceroutes with 10/8 subnet, I think it may be worthwhile to ask their ISP if they distinguish between bogons and martians ( http://en.wikipedia.org/wiki/Bogon_filtering & http://en.wikipedia.org/wiki/Martian_packet) and why don't they use RPF ( http://en.wikipedia.org/wiki/Reverse_path_forwarding). theymos should probably also ask ddos-guard.net what is going on. I have no way of testing that because all the ISPs I have access to properly filter martians and my access to the internal ISP traffic engineering tools is very limited. |
|
|
|
I think I understand what you are trying to proove. I posted exact lines(except I removed first 3) I got after running the command "traceroute bitcointalk.org". I didn't want to make changes as it may lead to more complications. So...?
I apologize in advance for being blunt. I don't otherwise will achieve any sort of positive communication. M.Z. doesn't know what he's doing. He's afraid to admit it. He had either: 1) managed to get his computer&gateway working by installing some malware or shitware that screws up routing 2) buys his Internet access from some black market supplier, equivalent to earlier Indian dishwallahs who frequently were resellers of cracked-encryption satellite TV receivers 3) steals his Internet access from someplace with significant restrictions 4) is completely honest and forthright, but pays for Internet access to somebody who does 1)-3). The situation you are with is typically called "painted yourself into a corner". You are now afraid to make any changes, like a floor painter who didn't think ahead and didn't plan a way to leave the room he had to paint. |
|
|
|
Okay. Thanks for telling! But is it easy to plant malware with "minimum" information? What is considered as "minimum" information?
I don't know what is the minimum information required to rip-off somebody. Probably the most common rip-offs are done with a help offer via Teamviewer: http://en.wikipedia.org/wiki/TeamViewer#Fraudulent_uses but I'm not really up to speed in that racket. As I have mentioned earlier, my knowledge in this is very limited. I looked more about locally and globally-routable IP. I think I understood basics/theories.
Personally, I don't think you understand the basics. 13 10.10.48.5 (10.10.48.5) 2463.142 ms 2974.825 ms 5016.395 ms
In the previously posted traceroute you have a non-globally-routable IP address (10/8 subnet) at position 13. I don't know if there's an simpler and more obvious way to detect routing shenanigans than that. |
|
|
|
Sorry for lack of details. What all details are necessary? As I have mentioned earlier, I have tried with another ISP.
The details are necessary for anyone to really help you. The only people who will not ask for details are the scammers who will need the minimum information required to plant a trojan/exploit on your computer, especially if you seem to have just one machine loaded with coins. The first detail I always ask: do you know what you doing? Like the basics of networking: local vs. globally-routable IP? My experience shows that it is the best to assume that somebody having weird problems simply has very minimal understanding of networking and pretends to be knowledgeable for reasons related to the social status. On another forum I've watched someone posting about "trying different ISPs," with no success. Finally somebody posted who knows the particular city. It turned out that the "different ISPs" were just a different resellers for the same ISP, just selling in different languages and with different marketing/payment plans. |
|
|
|
|
Show Posts
