Leaderboard

Those aren't typo's and errors. They are tests to see if we are paying attention.

I've just spent an hour arguing with an LLM (Deepcoder). TL;DR A.I. is useless at programming. I had a bug. I'd spent about an hour trying to figure it out and not succeeding but it shouldn't be that hard-I'm just missing something obvious. So. Ideal scenario for a clever AI to show dominance and help out a poor old flesh-bag programmer, right? Just point out the mistake or mistakes and laugh at my stupidity like a real coder. The bug was that the address from recvfrom would be 0.0.0.0 instead of 127.0.0.1. The problem was either bind wasn't binding to a specific address (which it was supposed to) or the address translation was not working quite right. I gave it 3 functions where I thought the bug was and explained that IPv6 seemed to be reporting correctly but IPv4 was in error: one function (Listen) had the bind function one function had IPv6 and IPv4 address translation from a SOCKADDR_STORAGE structure to strings and the other function was the read function with the recvfrom. This is an abbreviation of the conversion: - Round one. The following code has a bug that returns 0.0.0.0 as the IPv4 address from "recvfrom" when it should be returning "127.0.0.1". Find the reason that the code is returning 0.0.0.0 instead of 127.0.0.1 and rewrite the code to fix the issue. It told me how to bind to a specific address and how to use the recvfrom function. It then basically wrote the example on the MS page for recvfrom but within the Listen function. - Round 2. I'm not using the "sockaddr_in" structures; I'm using SOCKADDR_STORAGE structures. Please modify your example to use SOCKADDR_STORAGE structures. It apologised made an excuse about misunderstanding and then proceeded to create an example, still using sockaddr_in structures, but this time hardcoding "127.0.0.1" in the bind address. - Round 3. No, no, no. It needs to support user entered IPv6 and IPv4 but there is a bug with the IPv4 address. It also needs to use SOCKADDR_STORAGE as in the code I supplied to you. Reevaluate the code and write a version of the code that fixes the bug. It apologised, said that it understands why it got it wrong, and then proceeded to create an example, still using sockaddr_in structures, but now hardcoding "::1" in the bind address. -Round 4. Look. It needs to support both IPv6 and IPv4 and that's the reason I'm using SOCKADDR_STORAGE. If you don't have enough information then ask for clarification but you have the code that the bug is in so find the damned bug FFS! Another apology, said it can understand my frustration and then proceeded to spit out the example from MS again. This went on for an hour. No code I could actually use in my functions, never pointed out the bug in my code and the prompts just got longer and longer as I tried to head-off it's stupidity. This was one of the functions. int Addr2Address(SOCKADDR_STORAGE addr, PCHAR Address, int *Port, int *IPvType) { int err = 0; *IPvType = 0; switch (addr.ss_family) { case AF_INET6: { if (Address == NULL) {return 46;} *IPvType = 2; char strAddress[46]; inet_ntop(addr.ss_family, (void*)&((sockaddr_in6 *)&addr)->sin6_addr, Address, sizeof(strAddress)); break; } case AF_INET: { if (Address == NULL) {return 16;} *IPvType = 1; char strAddress[16]; inet_ntop(addr.ss_family, (void*)&((sockaddr_in6 *)&addr)->sin6_addr, Address, sizeof(strAddress)); break; } default: {err = WSAEPROTONOSUPPORT; break;} } *Port = ntohs(((sockaddr_in6 *)&addr)->sin6_port); return err; } The bug is in the AF_INET case. inet_ntop(addr.ss_family, (void*)&((sockaddr_in6 *)&addr)->sin6_addr, Address, sizeof(strAddress)); It should not be an IPv6 address conversion, it should be an IPv4 conversion. That code results in a null for the address to inetop which is converted to 0.0.0.0. I found it after a good nights sleep and a fresh start.

I once went for an interview where they gave me a coding test and asked me to modify it. It was a very long time ago so I don't remember the exact modification they wanted (nothing to do with memory leaks) but I do remember the obtain queue and read queue inside a while loop with the release queue outside. I asked if they wanted me to also fix the memory leak as well as the modifications and they were a little puzzled until I explained what you have just said. I must have seen (and fixed) this while-loop bug-pattern a thousand times since then in various code bases. I also created this VI which I generally use instead of the primitives as it intialises on first call, can be called from anywhere, and prevents most foot-shooting by rolling them all into a single VI and ensuring all references but 1 are closed after use. Queue.vi

Some people might be tempted to use Obtain Queue and Obtain Notifier with a name and assume that since the queue is named each Obtain function returns the same refnum. That is however not true. Each Obtain returns a unique refnum that references a memory structure of a few 10s of bytes that references the actual Queue or Notifier. So the underlaying Queue or Notifier is indeed only existing once per name, BUT each refnum still consumes some memory. And to make matters more tricky, there is only a limited amount of refnum IDs of any sort that can be created. This number lies somewhere between 2^20 and 2^24. Basically for EVERY Obtain you also have to call a Release. Otherwise you leak memory and unique refnum IDs.

I cannot look at your file, but I suggest save the data to TDMS or any binary format of your choice. Once the file is saved, then you can convert it to text.

There is no typos and errors in your posts. Only pearls of wisdom and oracles of truth that we mortals can't understand yet...

Drat, and now my typos and errors are put in stone for eternity (well at least until LavaG is eventually shutdown when the last person on earth turns off the light) 😁

@Rolf Kalbermatter the admins removed that setting for you as everything you say should be written down and never deleted 🙂

If the child classes are statically linked in the code (via class constants, or whatever other mechanism you use), then this approach should always work, because the child classes will always be in memory.

There is a "best practices" document (this too) but I suspect you are looking for a less abstract set of guidelines.

@Ravi Beniwal, I have the same problem, VIPM reports that missing dependency for me too. If it isn't required could it be removed ? If it is required could it be included ? I suspect it is included in version 1.7.028 which does not report the error. I spent(wasted !) time looking for the dependency and downloading it, only to realise it might not be needed as it launches OK from the LV IDE menu Tools\LabVIEW Task Maanger.... Peter