The Devs Listen to Us

[Katish]

I remember they showed us a lot of random statistics one time, such as how many times someone had said Yoshi-P in the game and how many times the squadron NPCs had squat. They particularly record the battle log across all players so that they can gather mass statistics from it, such as the win and loss rate of each job in PvP.

They tend to have a good idea of how many people main each job even though this game involves switching jobs a lot and how popular content is. They also know how popular FATEs are, because they told us the one in Azim Steppe was popular around the world and they wanted to copy the concept for Shadowbringers.

All I can say about the housing is that they used an external server and I think I remember it being caused by a bottleneck. So the math did not function because it had no data with which to use the math on because the data had not reached the players at the expected time, or something like that.

There should be no math done on a boolean check if there is math, they are doing it wrong lel. They should be checking yes or no, that's it. Assuming houses are objects and players are objects, which they should be...
this will be somewhat pseudo, as this is pseudo will not follow an actual language but should be understandable...:"
bidUI(); <---opens UI..handles which buttons are shown on UI generation...hides bid button if not isBiddableHouse...

void houseConfirm(object house, object player){
//fetch dynamic assigned housing dates and confirm it is a biddable house...would need to declare function that does that
bool houseBool = isBiddableHouse(house);
if(houseBool && buttonPress(specificBidButton)){
onBid(player);
} else {
//fetch arraydata as you should implement a function to manipulate and get the player winner
object winner = houseWinner();
---> Assign houseWinner to a newArray that will check if the player already owns a house...
}

}

object houseWinner(){
//this is where we would fetch our array data so let's assume we made that function...fetch data handles manipulation to read back into a new array;
array myFetchData = manipulateData("...locationOfData");
int rand = newrand(myFetchData.GetSize()-1); <--- this would get an int at random by our array size (i.e the maximum number), as arrays start at zero we would need to -1
object playerWinner = myFetchData[rand];
return playerWinner;
}

void onBid(object player, int date){
//closeDate will need to be the date at which you can bid until, after bidding allow the houses to do nothing
int closeDate;
if(date < closeDate) {array bidArr = arrBid(player);} else {player.ShowMsg("The bidding for this house has passed");}
}

array arrBid(object player) {
array myArray;
//add player object to array if you cannot do that find player name and instead hold strings in an array then find player by that string...
myArray.Add(player);
return myArray;
//convert array to string if you plan to save it to file, you will need to make a separate fetch function
}

this outlines the general idea...the logic might not be 100% sound as I have no way of testing unless I made a program and followed the instructions but again this was done out quickly.
I don't trust the numbers they show if they can't do this much...

[Koros]

There should be no math done on a boolean check if there is math, they are doing it wrong lel. They should be checking yes or no, that's it. Assuming houses are objects and players are objects, which they should be...
this will be somewhat pseudo, as this is pseudo will not follow an actual language but should be understandable...:"
bidUI(); <---opens UI

void houseConfirm(object house, object player){
//fetch dynamic assigned housing dates and confirm it is a biddable house...would need to declare function that does that
bool houseBool = isBiddableHouse(house);
if(houseBool && buttonPress(specificBidButton)){
onBid(player);
} else {
//fetch arraydata as you should implement a function to manipulate and get the player winner
object winner = houseWinner();
---> Assign houseWinner to a newArray that will check if the player already owns a house...
}

}

object houseWinner(){
//this is where we would fetch our array data so let's assume we made that function...fetch data handles manipulation to read back into a new array;
array myFetchData = manipulateData("...locationOfData");
int rand = newrand(myFetchData.GetSize()-1); <--- this would get an int at random by our array size (i.e the maximum number), as arrays start at zero we would need to -1
object playerWinner = myFetchData[rand];
return playerWinner;
}

void onBid(object player, int date){
//closeDate will need to be the date at which you can bid until, after bidding allow the houses to do nothing
int closeDate;
if(date < closeDate) {array bidArr = arrBid(player);} else {player.ShowMsg("The bidding for this house has passed");}
}

array arrBid(object player) {
array myArray;
//We assume server has a method to return to player name as this is quite a staple.
myArray.Add(player.GetName());
return myArray;
//convert array to string if you plan to save it to file, you will need to make a separate fetch function
}

this outlines the general idea...the logic might not be 100% sound as I have no way of testing unless I made a program and followed the instructions but again this was done out quickly.
I don't trust the numbers they show if they can't do this much...

honestly people don't realize that outside of silicon valley and a select few top companies (of which game companies are not) most programmers are really bad.

se can't hope to beat L5 salaries at Meta or Google.

the median coder at a video game company never took algorithms and thought intro to python was a hard class, like no offense, the disparity is huge.

of course, they have other comparative advantages like passion or having a good intuition for what makes good game design, but if management keeps overriding their ideas and if they get no funding, they can't do much

[Katish]

honestly people don't realize that outside of silicon valley and a select few top companies (of which game companies are not) most programmers are really bad.

se can't hope to beat L5 salaries at Meta or Google.

the median coder at a video game company never took algorithms and thought intro to python was a hard class

Unfortunately that is the case.

[Jeeqbit]

There should be no math done on a boolean check if there is math, they are doing it wrong lel. They should be checking yes or no, that's it.

Again, the lottery was conducted on an external server. The external server did not communicate the data at the expected time, so it defaulted to no (you didn't win the house) even though you did.

It's not about booleans but about how normally-working code can be inefficient on a large scale (like an MMO) and create a bottleneck.

The issue was https://eu.finalfantasyxiv.com/lodes...d233a2ab8e17b6

2. The lottery was conducted successfully, but the winning number was not communicated to the housing server

After investigating this issue, we discovered that, while the server responsible for conducting the lottery is functioning as intended, the results are not being properly relayed to the servers responsible for managing housing-related data.

Once this issue has been resolved, and lottery results can be relayed to the housing data servers without issue, it will be possible to confirm the lottery winners.

The lottery data that exists on the server responsible for processing results will be used to overwrite the housing server so that the correct data is reflected.

[Raikai]

You know, I watched a Mr. Happy video today about Criterion dungeons and he put out a point that makes a lot of sense:

Mount Rokkon's poor reward structure is the same as the previous V&C because it already 'sailed' (production wise) when the first one was out. He mentioned that the developers explained about a similar situation in Eureka, that only saw major positive changes in Pyros, because Pagos was already ahead at a development stage and they couldn't improve by then.

That's probably a symptom of how formulaic they work... The features across the expansion are consistent in what they deliver, so to change something on the go might not be as simple, sadly... But hey, I'm not a game developer, so that might not even be the case.

[Koros]

to change something on the go might not be as simple, sadly... But hey, I'm not a game developer, so that might not even be the case.

ok yeah, make criterion dungeons drop loot that unlocks a third materia slot, yeah i'm sure that's so hard to implement, gonna take five years

they're just giving excuses, or maybe they're so bad that is unironically impossible for them

[Sindele]

You know, I watched a Mr. Happy video today about Criterion dungeons and he put out a point that makes a lot of sense:

Mount Rokkon's poor reward structure is the same as the previous V&C because it already 'sailed' (production wise) when the first one was out. He mentioned that the developers explained about a similar situation in Eureka, that only saw major positive changes in Pyros, because Pagos was already ahead at a development stage and they couldn't improve by then.

That's probably a symptom of how formulaic they work... The features across the expansion are consistent in what they deliver, so to change something on the go might not be as simple, sadly... But hey, I'm not a game developer, so that might not even be the case.

I am! And this can be how it works in some less agile companies. Given SE's rigid patch cadence and the size of an MMO project, it really wouldn't surprise me to know that that design was signed, sealed, and delivered months ago, and everyone involved was too deep in their next work to really go back and iterate - but it also suggests the outcry wasn't really big enough to force their hand. We've seen them move quickly when they think it matters.

ok yeah, make criterion dungeons drop loot that unlocks a third materia slot, yeah i'm sure that's so hard to implement, gonna take five years

they're just giving excuses, or maybe they're so bad that is unironically impossible for them

Unless there's precedent I can't think of for something like that already functioning in game, that'd be a significantly taller order than you might think it is. It's one thing to have an item step up from one amount of slots to another with no respect for its previous state, but just adding a third slot to an existing item without blowing up the item and creating a new one in its place would be a pretty significant systems-level change.. and with that comes a whole lot of QA.

[Koros]

Unless there's precedent I can't think of for something like that already functioning in game, that'd be a significantly taller order than you might think it is. It's one thing to have an item step up from one amount of slots to another with no respect for its previous state, but just adding a third slot to an existing item without blowing up the item and creating a new one in its place would be a pretty significant systems-level change.. and with that comes a whole lot of QA.

I mean they can just put it in the shop, duplicate the item and call it "Augmented Ascendant X of X" and make it have three materia slots and use the same model, it's quite literally just adding entries to their database. Takes literal minutes to make such an item.

Then again they managed to screw up even basic database management with the Diadochos bug earlier this patch so lmao, can't expect much from them.

[Jeeqbit]

You know, I watched a Mr. Happy video today about Criterion dungeons and he put out a point that makes a lot of sense:

Mount Rokkon's poor reward structure is the same as the previous V&C because it already 'sailed' (production wise) when the first one was out. He mentioned that the developers explained about a similar situation in Eureka, that only saw major positive changes in Pyros, because Pagos was already ahead at a development stage and they couldn't improve by then.

Now you see why I mentioned earlier that this post contained a similar talking point to Mr Happy's video.

Although this is a good point in some cases, it is not a good reason for the rewards to be unmodified. The arguments being made are to make it useful for gear, which is as simple as making it drop a right side or (maybe) left side augment once a week. These items already exist and require no extra development such as model or icon design or concept art.

If something involves concept art it makes sense why it won't change fast but this isn't the case when it's just adding augments.

[Catwho]

They listen to us, sometimes.

The problem is they listen to other people a lot harder, and the people who they listen to want completely contradictory things from what we want.