Miscellaneous Bugs And Questions That Don't Deserve Their Own Thread Thread

NickTheNick wrote:

Also, we can't use magic to just make the game load an earlier point that wasn't permanently doomed just because we want it to. First of all, that would mean saving the game literally every second, and secondly the computer somehow would need to be able to determine when, and in which saves, the player is permanently doomed to lose, which is just not possible.

Altough i'm a new thrives programmer, that issue does not sound hard to fix! Simply creating save-points at significant events or every hour or so, and when you go extinct, we just iteratively check gamesaves backwards until we find one where the population is above a certain threshold and can breed.
The species could still be doomed in that state tho for whatever reason. Alternatively just give the player access to a number of previous save-points upon extinction, and let him choose which he can survive from.

Personally i don't mind having a hard-core game-over-when-extinct aspect, however, but i'm also not casual gamer representative.

I'm not sure if this has been brought up before but I thought of something after the underwater civilization thread closed. I was wondering how realistic thermodynamics would be? I mean if an under water civilization manages to get a piece of metal to it's melting point over a mold then would it smelt or does the specific mode of heating have to be coded in?

The issue wasn't so much thermodynamics, it was getting smelting to work underwater without killing the creatures. People weren't coming up with any new ideas so that was why the thread was closed. Underwater civilizations cannot advance past the stone age without assistance from surface species or a space-faring race uplifting them.

So thermodynamics does play into that a little, since heating metal underwater to its melting point would also heat the surrounding water and would kill the creature trying to smelt it.

EDIT: Now my two cents on game overs.

In my opinion, the player should be given some way of figuring out where they went and [expletive deleted] up. How that should be done I don't have a specific clue about. Maybe a view of the timeline since it is a game over and so they can see what choices made their species decline, like a bar graph?

crovea wrote:

Altough i'm a new thrives programmer, that issue does not sound hard to fix! Simply creating save-points at significant events or every hour or so, and when you go extinct, we just iteratively check gamesaves backwards until we find one where the population is above a certain threshold and can breed.
The species could still be doomed in that state tho for whatever reason. Alternatively just give the player access to a number of previous save-points upon extinction, and let him choose which he can survive from.

Personally i don't mind having a hard-core game-over-when-extinct aspect, however, but i'm also not casual gamer representative.

I know that you could save on a less frequent basis, but just from estimation I would presume each save would take a whole lot of space, even if the frequency was less than once per second. However, with no finished game it's purely speculative, and you're the programmer here so I would trust your estimates more.

Also, I wasn't referring to the population being doomed, which is calculable due to population dynamics, I mean the factors other than that, like drastic climate change, natural disasters, outcompetition, etc.

Immortal_Dragon wrote:

The issue wasn't so much thermodynamics, it was getting smelting to work underwater without killing the creatures. People weren't coming up with any new ideas so that was why the thread was closed. Underwater civilizations cannot advance past the stone age without assistance from surface species or a space-faring race uplifting them.

So thermodynamics does play into that a little, since heating metal underwater to its melting point would also heat the surrounding water and would kill the creature trying to smelt it.

I am aware of this what i meant was that supposing that a player managed to heat it up without killing his or her creature or they use strategy mode to use another creature then would it work or does how it is heated have to be coded in?

Yes, the approach would have had to be coded in, but since there was no feasible approach decided on, it is pretty much moot.

NickTheNick wrote:

I would presume each save would take a whole lot of space

Depends on how accurate you want to be with the gamestate, but i don't it should be much of an issue :)[/quote]

NickTheNick wrote:

I wasn't referring to the population being doomed

Yeah i see what you mean, but if you give the player a choice of saves, he could make a guess, and it would be a lot less frustrating.

Greetings, Thrive community!
Maybe you remember (probably not, but that's ok) that ~5 months ago (wow, 5 months already? Time goes by fast) I asked if you could use a Java programmer. In the meantime, I kept watching this project, though I have yet to read through the forums (regarding the LOT that has been discussed here, I hope I'm at least a bit excused. At least I read though this misc topic and also through its closed predecessor).
Currently I'm doing a C++ tutorial, but in my experience, you don't really learn programming languages by watching, but rather by doing. Which leads me to the question I want to ask: As C++ is fairly pointer-prone, it seems especially crucial to me to get experience about that part of the programming language, as there are no pointers in Java. However, when used inappropriatly (and I am the kind of noob that does experiments to get a deeper understanding of things), they might, instand of performing as intended, just "do-the-shredder" through your OS (afaik).
As I have only one PC with personal data, I am not really willing to regularly format my hard disk and reinstall Windows because of C++ code that went havok.
How do you deal with that? Do you just meticulously avoid doing pointer mistakes (because I know that I won't always be of full possession of my thoughts, such things will happen to me sooner or later) or did you set up some extra "defences" against OS corruption by enraged pointers?

HariboTer wrote:

Greetings, Thrive community!
Maybe you remember (probably not, but that's ok) that ~5 months ago (wow, 5 month already? Time goes by fast) I asked if you could use a Java programmer. In the meantime, I kept watching this project, though I have yet to read through the forums (regarding the LOT that has been discussed here, I hope I'm at least a bit excused. At least I read though this misc topic and also through its closed predecessor).
Currently I'm doing a C++ tutorial, but in my experience, you don't really learn programming languages by watching, but rather by doing. Which leads me to the question I want to ask: As C++ is fairly pointer-prone, it seems especially crucial to me to get experience about that part of the programming language, as there are no pointers in Java. However, when used inappropriatly (and I am the kind of noob that does experiments to get a deeper understanding of things), they might, instand of performing as intended, just "do-the-shredder" through your OS (afaik).
As I have only one PC with personal data, I am not really willing to regularly format my hard disk and reinstall Windows because of C++ code that went havok.
How do you deal with that? Do you just meticulously avoid doing pointer mistakes (because I know that I won't always be of full possession of my thoughts, such things will happen to me sooner or later) or did you set up some extra "defences" against OS corruption by enraged pointers?

Wow. Well first of all, we use lua about equally much as C++ now, which may be more familiar to you.
Regarding pointers and the dangers of C++, the operative system will protect you using what's called virtual memory, meaning your application thinks it has all the memory of the machine and can do what it wants but in reality it only has a small part. So you can't mess up unless you start trying to do really complicated tricks and the worst that will happen is your program crashes. Furthermore even if you could mess up, you won't be using pointers like that at all. The way you use pointers is you ask c++ to get some space in memory for an object

Code:

MyObject* myThing = new MyObject("hello");

the new keyword returns a pointer to where in memory the object is stored. So you never really specify where the pointer points to yourself, C++ and the operative system will handle that!
You can change what a pointer points to manually by normal arithmetic but you'll never want to do that.
So mess around with pointers all you want, i promise you won't mess up your machine! Here's a post about it
What's more is that the guys behind C++ (the C++ standard committee) are trying to get us away from using pointers all together as you have to always delete what you obtain with new.

Code:

delete myInteger;

If you forget this, the memory will stay occupied by your application causing a "memory leak" (unreleased memory). (theoretically memory will stay occupied even after your application stops, however this will never happen on modern operative systems which will take the memory back when your application closes)  . Java also deals with "pointers" whenever you use 'new' but handles the deletion automatically (which is often costly on performance and creates lag spikes in demanding applications).

The alternative to using pointers as seen in the above myThing example is to allocate the integer on the part of memory called the "stack" instead of the "heap":

Code:

MyObject myThing("foo");

. This will automatically be deleted using some magic (called RAII) nomatter what happens, when you go out of scope (exits the brackets { } that you are between). When you need memory to persist outside scope you can use things called smart pointers that will also automatically be deleted.

If you have more questions i'll be happy to answer them! Might be a better idea to ask in this subforum i think tho.

Thanks for your answer, I feel a bit safer now^^
However, what if you do things like

Code:

MyThing *pointer = new MyThing();
delete pointer;
pointer->myMethod();

or

Code:

MyThing[] array = { MyThing(), MyThing(), MyThing() };
int arrayLength = 3;
for(int i=0; i<arrayLength; i--) //This might seem banal, but in more complex algorithms such stuff could probably well happen
  delete &(array[i]);

? In the first example, the disposed object was occupying memory in your own program, which was unoccupied or allocated for other data of the program itself, so a program crash is raised, that should do no harm and not access memory that might have been allocated by the OS in the meantime, right?
But in the other example, the pointer (represented by the "array") seems to me to be running unstoppably through the RAM, invalidating whatever it gets into its greedy cold pointer fingers, or is that also being prevented somehow?

On another note, which IDE is being used / recommended for the development of Thrive? I have been hacking C++ code in Microsoft Visual C++ 2010 Express, but I'm not really happy with it, as many features of Eclipse I'm used to seem to be missing there (e.g. auto-completion, javadoc-on-mouse-hover, line-enumeration, message-overview on the right, line-jump-per-click-on-stack-trace).

Btw, if I am going too offtopic, then I'm sorry, but I couldn't find a thread of which I'd be certain for my questions to fit in there :S

In your first case you give the memory back to your memory manager, allowing calls to new to use that space. However in your case, if thats the only thing going on, your method will actually work (had to go try and test this myself). C++ doesn't know the object is deleted and will interpret whatever is at that place in memory as a MyThing, and since noone else has claimed and overwritten the memory the data/information is still there for it to function (delete doesn't reset the memory or anything, that would cost too much), if someone had overwritten it, you could get a really wierd acting MyThing or more likely a program crash.
Regarding your concern in this case, the pointer still points to the same place (within your applications virtual memory), so no it will never point anywhere harmful.

Your second case is a little more curious. First of, your code doesn't work the way you think, i'll explain why.
What would happen is what is known in the c++ world as "undefined behavior" meaning wierd stuff could happen, but your virtual memory still protects you from harming anything but your own running application.
Regarding raw arrays like that, you will want to call

Code:

delete[] array;

as this deletes/frees the memory of all the elements in the array and not just the first one, i think your example (with ++)  would work for deletion, there is no reason to ever do it like that.
Let me run through what your second example actually does

Code:

MyThing[] array = { MyThing(), MyThing(), MyThing() };

This wouldn't compile as you need the brackets on the right side of the identifier: MyThing array[] = ... But you could write the same thing as: MyThing* array = ...
as an array is simply a pointer, nothing more. (delete[] will make the compiler remember how much space was allocated and free the correct amount of space.
So your first line will allocate three MyThing's in the heap and give you a pointer to the first one.
Allocates memory for an element of 3 objects on the heap.

Code:

for(int i=0; i<arrayLength; i--)

would run until integer underflows or until your program crashes yes (it will crash).

Code:

delete &(array[i]);

What you really mean here is just: delete array[i]; which would start freeing the memory of your application from the location of the array in the heap, all the way backwards through your programs stack, through the enviroment memory all until it reaches 0. deleting 0 will run and array[-x] for some x will point to memory address -1 (0 is the start of your program, and -1 is outside your program) your operative system will crash your program, throwing an exception. You can try it yourself with this:

Code:

int* array = 0;
 array--;
 std::cout << *array;

To clear up: array[5] means take the pointer array (pointing to the first element) and add 5 times the size of the type to it, giving us a new pointer. Could also written as: array+5.
So doing:

Code:

delete &(array[i]);

doesn't do what you think, as the address of operator (&) in this case gets the address of the temporary pointer (array+i) which doesn't actually exists in memory.

HariboTer wrote:

Btw, if I am going too offtopic, then I'm sorry, but I couldn't find a thread of which I'd be certain for my questions to fit in there :S

I'm fairly new here myself, so i'm not sure but a continued series of questions probably deserves it's own thread, so i recommend you create a new thread over in the subforum i linked above or find an existing thread and ask there, i'll find your questions.

Sorry to interrupt, but I must ask a quick question: how will sensory organs work? Will they work kind of like Spore where there is a library of pre-modeled parts for the player to select from? Or will the player be made to create their own sensory organs? Regardless, I'm assuming they will affect gameplay. For example, an organism without eyes will be blind, an organism without ears will be deaf, etc. But what about scent? Touch? Taste?

I know scent works by having some sort of colored cloud, I think.

I think things like eyes are pre-made parts, and start as primitive eye cups and all. You can actually choose which spectrum the creature will see in, and it will change how you view the world in the first-person.

Touch and taste, I think that will have some kind of text, hearing is just the speakers of the computer I think. Parts for these I am not sure about.

Taste is all about the taste buds and the tongue, so I don't know how that will work in Thrive, since that seems to have had little discussion.

So eyes will be pre-made parts then; good. That's something I could work on. They should have plenty of modifiers/morphers, such as eyelid heaviness and pupil size.

I was hoping there would be a scent mechanic! I envisioned it as working like the mole creature from Avatar the Last Airbender, which as it turns out we can agree upon. As you gain better scent organs, the colored particle trails become clearer, allowing you to find prey or others of your own species.

As for taste, some reptiles "taste" the air in a way that works a lot like an acute sense of smell. Perhaps the scent mechanics could be applied to taste buds as well?

I'm glad my answer was satisfactory, I'm just parroting the conclusions I've read. I know there is a senses integration thread floating around here somewhere.

Taste Buds, it really depends on what the creature is using them for. I am not sure how much discussion senses other than sight and smell have gotten on here.

I found the wiki page on senses but I'm not allowed to post it for another few days apparently.

It says taste will enable the organism to detect what effects foods have on them. Considering there's talk of a toxicity mechanic, it seems like it could be very useful.

As for the rest of those senses I don't even know what the hell. Maybe magnetoceptory organs will be used to form migratory paths, like birds?

The more primitive the creature's vision is, the blurrier/darker the screen should appear.

I just had a thought, what about a creature having some sort of sound-based attack. This is outside of the normal sonar-based attack I talked about in the Organ Design thread with spermaceti.

Here's my thought, a large creature has big lungs, a reinforced throat, as well as some kind of resonance chamber to create a sound that is extremely low frequency at a high enough volume that a creature caught in it without protection will be disoriented or even killed at close range.

Immortal_Dragon wrote:

I just had a thought, what about a creature having some sort of sound-based attack. This is outside of the normal sonar-based attack I talked about in the Organ Design thread with spermaceti.

Here's my thought, a large creature has big lungs, a reinforced throat, as well as some kind of resonance chamber to create a sound that is extremely low frequency at a high enough volume that a creature caught in it without protection will be disoriented or even killed at close range.

I'm already terrified, but I guess it makes sense.

Immortal_Dragon wrote:

Yes, the approach would have had to be coded in, but since there was no feasible approach decided on, it is pretty much moot.

Interesting, what if a amphibious civilization tried to do this and measured the temperature to be past the melting point could they show the save to a higher up and have it updated to work?

Well, an amphibious civilization can simply work the metal above the water, so I don't understand what you're getting at.

Immortal_Dragon wrote:

Well, an amphibious civilization can simply work the metal above the water, so I don't understand what you're getting at.

What I am geting at is a civilization of amphibians are controlled by a player who wants an underwater civilization so after reaching a high tech level they will build an underwater base and experiment with raising the temperature of metal to it's melting point safely by measuring the temperature using their tech but raising it with metheds available to a tribal race of mermen. Then when one is found the player will show the save to the higher-ups and have the game updated.

The problem is that the approach you are talking about would have to be coded into the game before it can be used in the game. So that's why the approach needs to be brought up on this forum first. A completely underwater civilization can trade with a surface civilization and get tech that way.

Immortal_Dragon wrote:

The problem is that the approach you are talking about would have to be coded into the game before it can be used in the game. So that's why the approach needs to be brought up on this forum first. A completely underwater civilization can trade with a surface civilization and get tech that way.

I think you misunderstand. I don't want to melt the metal I just want to heat the water around it enough that were it coded in it would melt so that I can prove it's possible.

Yes, I did misunderstand, but in order to shape metal it has to be heated to temperatures where it would not only melt, but it also means the surrounding water will be heated to temperatures that would kill life around it. Tribal creatures simply won't have the technology to protect themselves from that if they could at all.

The only life that would still be safe at those temperatures would be extremophile bacteria.

Immortal_Dragon wrote:

Yes, I did misunderstand, but in order to shape metal it has to be heated to temperatures where it would not only melt, but it also means the surrounding water will be heated to temperatures that would kill life around it. Tribal creatures simply won't have the technology to protect themselves from that if they could at all.

The only life that would still be safe at those temperatures would be extremophile bacteria.

While that is true there is no harm would come to them if after they discovered this they placed it over a funnel into a mold or some other way of shaping it remotely.