Jacobs and Camelot Unchained showed 1500 bots fighting it out 4 years ago. Here is a piece from their Newsletter. Not sure if they released the video (I think they did) but if have alpha Access you can find it there.
Unofficially, they increased the number of ARCs by quite a bit just or fun.
Pfft, 500, 2800, how about Dual Universe, 30K in one planet...
"It was a major test of our technology, and a world record: we got 30,000 simulated players all together on Alioth, the main planet of Dual Universe, together with members of the community invited to join to be the witnesses. 30,000+ people all together inside the same world, at the same time"
Guess how well that translated over to the actual performance in the current beta?
Just trying to live long enough to play a new, released MMORPG, playing New Worlds atm
Fools find no pleasure in understanding but delight in airing their own opinions. Pvbs 18:2, NIV
Don't just play games, inhabit virtual worlds™
"This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
Yet when FSR launched on Steam last year it had (still has) a player limit of what 10, or other ridiculously small number?
As far as I know, this was an arbitrary restriction because of waiting times (at that time there was no option to create groups; the game starts automatically when there are 4 people in the queue or after 3 minutes), because FSR is based on the CU engine and is capable of much more.
Just trying to live long enough to play a new, released MMORPG, playing New Worlds atm
Fools find no pleasure in understanding but delight in airing their own opinions. Pvbs 18:2, NIV
Don't just play games, inhabit virtual worlds™
"This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
As for the most actual players in a live PVP battle EVE held the title a year or two back, setting two world records at the same time.
Correct. But EVE Online uses time dilation as an anti-lag measurement.
Commands on the server are currently added to a queue and processed in
order. If the load is more than the server can process, this queue grows
at an alarming rate and the server is unable to catch up. Under time
dilation, actions in the game such as firing weapons or moving would be
slowed down to ensure the queue remains short and so the server stays
under its maximum load. Instead of fights becoming laggy and unplayable,
the entire battle would go into slow motion and remain responsive.
Large scale activities in MMORPGs have always had difficulties. Even PvE raids with 70 people were almost impossible in the days when the chat channels only held 50 people. Coordination was a nightmare, and tactics became meaningless as the only thing that really worked consistently for everyone was "stay with the group". I.e., zerg tactics.
A game may have the ability to run 500x500x500 static bots on an internal test. The real test is to add in multiple active chat channels, unpredictable non-scripted behavior, and server (and network) contention. Then factor in the odd graphics/CPU generated lag on various clients on substandard PCs. Adjusting everything to make a smooth, playable experience is a considerable task that few have been able to achieve.
Good luck to those trying.
Steven seems to keep promising big, will be interesting if they can come close to anything above industry standard. As always, the real test is Real World application.
My understanding (as an outsider) is that ARCs are pointed out because, since they are remote connections, they stress test the network communication sufficiently, even though they still are bots.
This. @Torval The fact that they are ARCs is emphasised again and again so that there is no confusion with "normal" bots, like those that can be spawned in masses via console in various first-person shooters.
My understanding (as an outsider) is that ARCs are pointed out because, since they are remote connections, they stress test the network communication sufficiently, even though they still are bots.
This. @Torval The fact that they are ARCs is emphasised again and again so that there is no confusion with "normal" bots, like those that can be spawned in masses via console in various first-person shooters.
I don't know the specifics of *how* they are employing these ARCs. They could be in-building. That is, they may test their proprietary network code through their own internal networks, but network traffic isn't going outside their building (e.g., through a normal ISP connection). A good portion of the time, the ISP providing the external network service is a large contributor to performance.
So, if they are doing bots on machines in their building on their own (presumably controlled) networks, performance tests can have significantly different results once the signals reach the real world.
Is what they are doing a valid test? It *could* be, or it *could not* be. Or both.
Logic, my dear, merely enables one to be wrong with great authority.
Jacobs and Camelot Unchained showed 1500 bots fighting it out 4 years ago. Here is a piece from their Newsletter. Not sure if they released the video (I think they did) but if have alpha Access you can find it there.
Unofficially, they increased the number of ARCs by quite a bit just or fun.
Yes, the testing showed so much potential, but the game lagged like crazy when I played it after the "beta" launch. My takeaway from that was that these tests don't hold much water in a real life environment.
That said, the actual gameplay videos I have seen of AoC are way better than anything I have seen from CU, so maybe they can do better with the lag as well.
I don't know the specifics of *how* they are employing these ARCs. They could be in-building. That is, they may test their proprietary network code through their own internal networks, but network traffic isn't going outside their building (e.g., through a normal ISP connection). A good portion of the time, the ISP providing the external network service is a large contributor to performance.
So, if they are doing bots on machines in their building on their own (presumably controlled) networks, performance tests can have significantly different results once the signals reach the real world.
Is what they are doing a valid test? It *could* be, or it *could not* be. Or both.
Bots are “headless clients” running on separate AWS instances. This
means that on different instances (servers), we are running 1500
separate clients. They are headless, because in our case, they aren’t
attached to real players at keyboards, monitors, etc.
and
Bots are not “NPCs” because those things are usually controlled from the
same server(s) instance(s) and are generally not networked like
players.
and
So, in terms of whether are Bots are the same as players in terms of
Internet/network traffic and also client stability (they can crash too),
etc., the answer is 100% yes!
That said, the actual gameplay videos I have seen of AoC are way better than anything I have seen from CU, so maybe they can do better with the lag as well.
As far as I remember, they had a big problem with the code for tree generation and the block physics (was very unoptimised for a while). Now the worst lag factors should be removed.
That said, the actual gameplay videos I have seen of AoC are way better than anything I have seen from CU, so maybe they can do better with the lag as well.
As far as I remember, they had a big problem with the code for tree generation and the block physics (was very unoptimised for a while). Now the worst lag factors should be removed.
I see, so you are saying that the bot tests you have been raving about were a meaningless marketing gimmick?
Havent really been following the game so much, but this discussion is interesting nevertheless. It reminds me a bit of Planetside 2. A pretty old game, but I cant recall much lag despite battles often being very huge. Several hundres on each side and 3 factions? No problems. And, without lag. If they could do it back then, why cant they now?
Havent really been following the game so much, but this discussion is interesting nevertheless. It reminds me a bit of Planetside 2. A pretty old game, but I cant recall much lag despite battles often being very huge. Several hundres on each side and 3 factions? No problems. And, without lag. If they could do it back then, why cant they now?
Planetside 2 uses SOE's ForgeLight engine which was later acquired by Daybreak and used in the infamous H1Z1. It is therefore a proprietary product for a specific "use case". Commercially available more general engines such as Unity and Unreal are not designed for mass battles and must therefore be laboriously "reworked". So the answer is: yes, it can be done, but it is too time-consuming and too expensive for a mere afterthought, like PvP in theme-park MMORPGs.
The Forgelight engine was one of the more notable gaming achievements that never got the full credit it deserved in how well it handled large scale fights.
Just trying to live long enough to play a new, released MMORPG, playing New Worlds atm
Fools find no pleasure in understanding but delight in airing their own opinions. Pvbs 18:2, NIV
Don't just play games, inhabit virtual worlds™
"This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
Havent really been following the game so much, but this discussion is interesting nevertheless. It reminds me a bit of Planetside 2. A pretty old game, but I cant recall much lag despite battles often being very huge. Several hundres on each side and 3 factions? No problems. And, without lag. If they could do it back then, why cant they now?
Planetside 2 uses SOE's ForgeLight engine which was later acquired by Daybreak and used in the infamous H1Z1. It is therefore a proprietary product for a specific "use case". Commercially available more general engines such as Unity and Unreal are not designed for mass battles and must therefore be laboriously "reworked". So the answer is: yes, it can be done, but it is too time-consuming and too expensive for a mere afterthought, like PvP in theme-park MMORPGs.
Steven has said they are writing their own back end code. They also have people who worked on Planetside 2 writing it I believe. When he showed off this test he specifically said it was client side only. He never pretended it was simulating connections.
I will say though that as someone who was played the alpha I dont think their back end code is something you need to worry about. We had pretty big battles with no issues and they have made quite a number of improvements since then reducing the number of draw calls and such. I am no game developer but I am confident.
Either way 500 v 500 battles will probably rarely, if ever, happen. I am glad they are pushing for as high of a goal as possible though.
Comments
"It was a major test of our technology, and a world record: we got 30,000 simulated players all together on Alioth, the main planet of Dual Universe, together with members of the community invited to join to be the witnesses. 30,000+ people all together inside the same world, at the same time"
Guess how well that translated over to the actual performance in the current beta?
Not much.
https://medium.com/@jcbaillie/dual-universe-redefines-the-meaning-of-massively-multiplayer-with-over-30-000-a04c0e8b4106
"True friends stab you in the front." | Oscar Wilde
"I need to finish" - Christian Wolff: The Accountant
Just trying to live long enough to play a new, released MMORPG, playing New Worlds atm
Fools find no pleasure in understanding but delight in airing their own opinions. Pvbs 18:2, NIV
Don't just play games, inhabit virtual worlds™
"This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
The two GUINNESS WORLD RECORDS titles now held by EVE Online are:
Performance did suffer, and I imagine the participants felt much like this
https://www.ccpgames.com/news/2020/eve-online-breaks-two-guinness-world-records-tm-titles-in-one-day-with
"True friends stab you in the front." | Oscar Wilde
"I need to finish" - Christian Wolff: The Accountant
Just trying to live long enough to play a new, released MMORPG, playing New Worlds atm
Fools find no pleasure in understanding but delight in airing their own opinions. Pvbs 18:2, NIV
Don't just play games, inhabit virtual worlds™
"This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
Logic, my dear, merely enables one to be wrong with great authority.
That said, the actual gameplay videos I have seen of AoC are way better than anything I have seen from CU, so maybe they can do better with the lag as well.
I see, so you are saying that the bot tests you have been raving about were a meaningless marketing gimmick?
Planetside 2 uses SOE's ForgeLight engine which was later acquired by Daybreak and used in the infamous H1Z1. It is therefore a proprietary product for a specific "use case". Commercially available more general engines such as Unity and Unreal are not designed for mass battles and must therefore be laboriously "reworked". So the answer is: yes, it can be done, but it is too time-consuming and too expensive for a mere afterthought, like PvP in theme-park MMORPGs.
"True friends stab you in the front." | Oscar Wilde
"I need to finish" - Christian Wolff: The Accountant
Just trying to live long enough to play a new, released MMORPG, playing New Worlds atm
Fools find no pleasure in understanding but delight in airing their own opinions. Pvbs 18:2, NIV
Don't just play games, inhabit virtual worlds™
"This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
I will say though that as someone who was played the alpha I dont think their back end code is something you need to worry about. We had pretty big battles with no issues and they have made quite a number of improvements since then reducing the number of draw calls and such. I am no game developer but I am confident.
Either way 500 v 500 battles will probably rarely, if ever, happen. I am glad they are pushing for as high of a goal as possible though.