On March 24 2023 00:55 {CC}StealthBlue wrote: Maybe that is just the early game inventory.
Also Blizz getting ahead of the storm for tomorrow.
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
I'm sure they're trying to optimize the code to use as little server capacity as possible. I downloaded the entire 85 Gb beta last night without any problems and I can't imagine this thing using that much bandwidth while playing.
They're also making sure they don't have a bottleneck anywhere. Their authentication servers are notorious and I saw an interview where that was the issue last weekend.
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
If they let everyone in all at once, the servers would be trying to process millions of logins, characters, profiles, bnet accounts, etc at the same time. They would crash without the queues to slow down the amount of work they need to do.
A new game means new systems and the introduction of new potential problems as well.
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
Not speaking from gamedev perspective but dev in general where I had to deal with systems that have to handle 50+ million people a day, it's doable, even with tech that's not really designed for that (most of the big systems now started as small systems using tech that has a hard time scaling and they can't really move away from it or switch to something better because it's just not feasible financially).
What surprises me is that they even need queues at all. They could've used all the experience they have with multiplayer, their failed launch of D3 (my friend pre-ordered the game to be able to play early and couldn't really play it until like 1 month into the actual release), all the WoW stuff...
IMO they have enough network engineers and experience to not ever fail at it but they still do. How they do it is beyond me.
Every system I worked with that struggled with net traffic it was mostly due to underlying outdated tech (and they still make it work). If you asked the devs there to get you the same or similar system built from scratch with the knowledge they have now they'd be overjoyed and it would be a system that could easily handle 3-10x the traffic built with new tech.
But underlying tech and architecture are just one thing, there's also the matter of overall design. Why not go the way of GW1? Players meet in cities/hubs that are instanced and you can switch between them freely (almost, unless some instance is full) but when you go out of the hub it's just you and your team, content instanced to a single person/party. Brilliant design from 2005 that works to this day and is even better than most approaches. Especially for an ARPG: want to play solo? No problem, just go out of town and you've got the whole zone/world to yourself. Want to play with friends? No problem either. Want to meet people? Just go to any city/town. Downside? You don't meet randos out in the wild, but I don't think many people care for that (outside of griefers and such).
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
Not speaking from gamedev perspective but dev in general where I had to deal with systems that have to handle 50+ million people a day, it's doable, even with tech that's not really designed for that (most of the big systems now started as small systems using tech that has a hard time scaling and they can't really move away from it or switch to something better because it's just not feasible financially).
What surprises me is that they even need queues at all. They could've used all the experience they have with multiplayer, their failed launch of D3 (my friend pre-ordered the game to be able to play early and couldn't really play it until like 1 month into the actual release), all the WoW stuff...
IMO they have enough network engineers and experience to not ever fail at it but they still do. How they do it is beyond me.
Every system I worked with that struggled with net traffic it was mostly due to underlying outdated tech (and they still make it work). If you asked the devs there to get you the same or similar system built from scratch with the knowledge they have now they'd be overjoyed and it would be a system that could easily handle 3-10x the traffic built with new tech.
But underlying tech and architecture are just one thing, there's also the matter of overall design. Why not go the way of GW1? Players meet in cities/hubs that are instanced and you can switch between them freely (almost, unless some instance is full) but when you go out of the hub it's just you and your team, content instanced to a single person/party. Brilliant design from 2005 that works to this day and is even better than most approaches. Especially for an ARPG: want to play solo? No problem, just go out of town and you've got the whole zone/world to yourself. Want to play with friends? No problem either. Want to meet people? Just go to any city/town. Downside? You don't meet randos out in the wild, but I don't think many people care for that (outside of griefers and such).
They have these random events and world bosses. Those would have to be hub-style as well then. Which would probably make this whole hub/ instance thing quite awkward.
I did meet surprisingly few players out in the wild and at those random events in the closed beta. Curious to see how flooded the open beta will be
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
Not speaking from gamedev perspective but dev in general where I had to deal with systems that have to handle 50+ million people a day, it's doable, even with tech that's not really designed for that (most of the big systems now started as small systems using tech that has a hard time scaling and they can't really move away from it or switch to something better because it's just not feasible financially).
What surprises me is that they even need queues at all. They could've used all the experience they have with multiplayer, their failed launch of D3 (my friend pre-ordered the game to be able to play early and couldn't really play it until like 1 month into the actual release), all the WoW stuff...
IMO they have enough network engineers and experience to not ever fail at it but they still do. How they do it is beyond me.
Every system I worked with that struggled with net traffic it was mostly due to underlying outdated tech (and they still make it work). If you asked the devs there to get you the same or similar system built from scratch with the knowledge they have now they'd be overjoyed and it would be a system that could easily handle 3-10x the traffic built with new tech.
But underlying tech and architecture are just one thing, there's also the matter of overall design. Why not go the way of GW1? Players meet in cities/hubs that are instanced and you can switch between them freely (almost, unless some instance is full) but when you go out of the hub it's just you and your team, content instanced to a single person/party. Brilliant design from 2005 that works to this day and is even better than most approaches. Especially for an ARPG: want to play solo? No problem, just go out of town and you've got the whole zone/world to yourself. Want to play with friends? No problem either. Want to meet people? Just go to any city/town. Downside? You don't meet randos out in the wild, but I don't think many people care for that (outside of griefers and such).
They have these random events and world bosses. Those would have to be hub-style as well then. Which would probably make this whole hub/ instance thing quite awkward.
This is one of the main reasons the devs went away from GW to make GW2 because the system limited them in what they wanted to do with a living word, event chains, word bosses etc. I have spent quite a lot of hours tinkering about this because I much prefer GW's instanced system over an open world because of aforementioned reasons but also due to the fact that it gives devs the ability to solve a lot of MMO related game design problem (a lot of which GW had solved in 2005). I think events, moving fronts, world bosses etc. can still work in an instanced game, especially when they are on a timer like in GW2 anyways. It would be a bit different, sure, but if you make use of NPCs and the map to indicate spawns. To encounter a world boss, players could meet at certain times in adjacent outpost, build a group and move out for example. Not that different than porting to a waypoint at 2 pm to slay the dragon with 100 other players - the only difference being that you don't 100 players but a certain group size, which takes away from the large scale encounter but allows for much better balancing and boss design.
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
Not speaking from gamedev perspective but dev in general where I had to deal with systems that have to handle 50+ million people a day, it's doable, even with tech that's not really designed for that (most of the big systems now started as small systems using tech that has a hard time scaling and they can't really move away from it or switch to something better because it's just not feasible financially).
What surprises me is that they even need queues at all. They could've used all the experience they have with multiplayer, their failed launch of D3 (my friend pre-ordered the game to be able to play early and couldn't really play it until like 1 month into the actual release), all the WoW stuff...
IMO they have enough network engineers and experience to not ever fail at it but they still do. How they do it is beyond me.
Every system I worked with that struggled with net traffic it was mostly due to underlying outdated tech (and they still make it work). If you asked the devs there to get you the same or similar system built from scratch with the knowledge they have now they'd be overjoyed and it would be a system that could easily handle 3-10x the traffic built with new tech.
But underlying tech and architecture are just one thing, there's also the matter of overall design. Why not go the way of GW1? Players meet in cities/hubs that are instanced and you can switch between them freely (almost, unless some instance is full) but when you go out of the hub it's just you and your team, content instanced to a single person/party. Brilliant design from 2005 that works to this day and is even better than most approaches. Especially for an ARPG: want to play solo? No problem, just go out of town and you've got the whole zone/world to yourself. Want to play with friends? No problem either. Want to meet people? Just go to any city/town. Downside? You don't meet randos out in the wild, but I don't think many people care for that (outside of griefers and such).
They have these random events and world bosses. Those would have to be hub-style as well then. Which would probably make this whole hub/ instance thing quite awkward.
This is one of the main reasons the devs went away from GW to make GW2 because the system limited them in what they wanted to do with a living word, event chains, word bosses etc. I have spent quite a lot of hours tinkering about this because I much prefer GW's instanced system over an open world because of aforementioned reasons but also due to the fact that it gives devs the ability to solve a lot of MMO related game design problem (a lot of which GW had solved in 2005). I think events, moving fronts, world bosses etc. can still work in an instanced game, especially when they are on a timer like in GW2 anyways. It would be a bit different, sure, but if you make use of NPCs and the map to indicate spawns. To encounter a world boss, players could meet at certain times in adjacent outpost, build a group and move out for example. Not that different than porting to a waypoint at 2 pm to slay the dragon with 100 other players - the only difference being that you don't 100 players but a certain group size, which takes away from the large scale encounter but allows for much better balancing and boss design.
Good thinking! It could easily be implented say to have a NPC or sign and click it to port to the instance where the event happens. No need for premade parties. You could just fill instance up to a certain amount and then open new instances. Those would scale dynamically to the amount of players. Only thing I can imagine is people bitching about break of immersion or sth along those lines
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
Not speaking from gamedev perspective but dev in general where I had to deal with systems that have to handle 50+ million people a day, it's doable, even with tech that's not really designed for that (most of the big systems now started as small systems using tech that has a hard time scaling and they can't really move away from it or switch to something better because it's just not feasible financially).
What surprises me is that they even need queues at all. They could've used all the experience they have with multiplayer, their failed launch of D3 (my friend pre-ordered the game to be able to play early and couldn't really play it until like 1 month into the actual release), all the WoW stuff...
IMO they have enough network engineers and experience to not ever fail at it but they still do. How they do it is beyond me.
Every system I worked with that struggled with net traffic it was mostly due to underlying outdated tech (and they still make it work). If you asked the devs there to get you the same or similar system built from scratch with the knowledge they have now they'd be overjoyed and it would be a system that could easily handle 3-10x the traffic built with new tech.
But underlying tech and architecture are just one thing, there's also the matter of overall design. Why not go the way of GW1? Players meet in cities/hubs that are instanced and you can switch between them freely (almost, unless some instance is full) but when you go out of the hub it's just you and your team, content instanced to a single person/party. Brilliant design from 2005 that works to this day and is even better than most approaches. Especially for an ARPG: want to play solo? No problem, just go out of town and you've got the whole zone/world to yourself. Want to play with friends? No problem either. Want to meet people? Just go to any city/town. Downside? You don't meet randos out in the wild, but I don't think many people care for that (outside of griefers and such).
They have these random events and world bosses. Those would have to be hub-style as well then. Which would probably make this whole hub/ instance thing quite awkward.
This is one of the main reasons the devs went away from GW to make GW2 because the system limited them in what they wanted to do with a living word, event chains, word bosses etc. I have spent quite a lot of hours tinkering about this because I much prefer GW's instanced system over an open world because of aforementioned reasons but also due to the fact that it gives devs the ability to solve a lot of MMO related game design problem (a lot of which GW had solved in 2005). I think events, moving fronts, world bosses etc. can still work in an instanced game, especially when they are on a timer like in GW2 anyways. It would be a bit different, sure, but if you make use of NPCs and the map to indicate spawns. To encounter a world boss, players could meet at certain times in adjacent outpost, build a group and move out for example. Not that different than porting to a waypoint at 2 pm to slay the dragon with 100 other players - the only difference being that you don't 100 players but a certain group size, which takes away from the large scale encounter but allows for much better balancing and boss design.
Good thinking! It could easily be implented say to have a NPC or sign and click it to port to the instance where the event happens. No need for premade parties. You could just fill instance up to a certain amount and then open new instances. Those would scale dynamically to the amount of players. Only thing I can imagine is people bitching about break of immersion or sth along those lines
You'd rather have your immersion broken slightly for a couple of events but have a stable game client/servers or have your immersion broken on a regular basis (randos can ruin your immersion greatly) and an unstable client/servers?
One of my gripes with GW2 as opposed to GW1 is that while GW2 has this super cool and extensive transmog system (being called Fashion Wars for a reason) it is super immersion breaking when you're fighting some dark demons inside an active volcano and Winnie the Pooh shows up...
I logged in to a 53 min queue but it got halved soon after. Managed to get in after around 25 mins. After tinkering with the menus and customizing character appearance, I tried to start game and it got stuck. No queues or anything. Had to force close and go back to the queue.
Okay summons is a lot different than what I was expecting. Golem isnt available till lvl 25. Now I'm wondering if you can have skels and a golem at the same time... because you can only have 1 type of skeleton summoned during one time so far.
edit: And just got d/c after defeating dungeon boss.
if ppl are not waiting in queue for 2+ hours at a time, i will be impressed. and to that note, why even bother doing an open beta just to piss off your entire player base when u know its just going to be a shit show?
im certainly not a network engineer, but what would be so different from running a server for diablo 3 to diablo 4. like i cant imagine it to be so much different. but again this is blizzard, every game launch is as if they're a brand new indie company who's never done it before.
Because even if you literally copy&paste the code (you won't) some way it hooks into another process somewhere will cause a problem and cause it all the come crashing down. That's why you test, better to have a beta with an issue that people get mad about then it happening at the actual launch.
Not speaking from gamedev perspective but dev in general where I had to deal with systems that have to handle 50+ million people a day, it's doable, even with tech that's not really designed for that (most of the big systems now started as small systems using tech that has a hard time scaling and they can't really move away from it or switch to something better because it's just not feasible financially).
What surprises me is that they even need queues at all. They could've used all the experience they have with multiplayer, their failed launch of D3 (my friend pre-ordered the game to be able to play early and couldn't really play it until like 1 month into the actual release), all the WoW stuff...
IMO they have enough network engineers and experience to not ever fail at it but they still do. How they do it is beyond me.
Every system I worked with that struggled with net traffic it was mostly due to underlying outdated tech (and they still make it work). If you asked the devs there to get you the same or similar system built from scratch with the knowledge they have now they'd be overjoyed and it would be a system that could easily handle 3-10x the traffic built with new tech.
But underlying tech and architecture are just one thing, there's also the matter of overall design. Why not go the way of GW1? Players meet in cities/hubs that are instanced and you can switch between them freely (almost, unless some instance is full) but when you go out of the hub it's just you and your team, content instanced to a single person/party. Brilliant design from 2005 that works to this day and is even better than most approaches. Especially for an ARPG: want to play solo? No problem, just go out of town and you've got the whole zone/world to yourself. Want to play with friends? No problem either. Want to meet people? Just go to any city/town. Downside? You don't meet randos out in the wild, but I don't think many people care for that (outside of griefers and such).
They have these random events and world bosses. Those would have to be hub-style as well then. Which would probably make this whole hub/ instance thing quite awkward.
This is one of the main reasons the devs went away from GW to make GW2 because the system limited them in what they wanted to do with a living word, event chains, word bosses etc. I have spent quite a lot of hours tinkering about this because I much prefer GW's instanced system over an open world because of aforementioned reasons but also due to the fact that it gives devs the ability to solve a lot of MMO related game design problem (a lot of which GW had solved in 2005). I think events, moving fronts, world bosses etc. can still work in an instanced game, especially when they are on a timer like in GW2 anyways. It would be a bit different, sure, but if you make use of NPCs and the map to indicate spawns. To encounter a world boss, players could meet at certain times in adjacent outpost, build a group and move out for example. Not that different than porting to a waypoint at 2 pm to slay the dragon with 100 other players - the only difference being that you don't 100 players but a certain group size, which takes away from the large scale encounter but allows for much better balancing and boss design.
Good thinking! It could easily be implented say to have a NPC or sign and click it to port to the instance where the event happens. No need for premade parties. You could just fill instance up to a certain amount and then open new instances. Those would scale dynamically to the amount of players. Only thing I can imagine is people bitching about break of immersion or sth along those lines
Apparently it's doable to a pretty big degree even without instancing...
EPT
![[image loading]](https://cdn.discordapp.com/attachments/1019713478900076615/1082234895352483860/image.png)
