Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

Five workers left Crowfall including Tully?


Recommended Posts

I disagree, there are plenty of ways to make 20vs20 or 30vs30 interesting and thrilling.

My brother and I were talking for hours about what the ideal size of a guild would be and the ideal size of group you could throw together at any given time.  I can't say I liked the chaos of some of the large scale battles I participated in for WoW.  I really hope the sweet spot for groups is somewhere between 20 and 32.

Link to post
Share on other sites
  • Replies 119
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

Some of the people you're taking about left us last year. You're only noticing now because we decided to add an alumni page to the website to honor the work of people who worked on the project and fo

Hey guys,   This is Larissa (can't log into my "official" account so this will have to do) I feel incredibly grateful for all the support and love that this community has given us. Thank you so much

Hey guys, Todd said it best regarding turnover - it happens and its nothing to read too much into. Luckily, I am still very much a part of the Crowfall team:) I helped build the brand and launch the K

Yeah and the scary thing is the Cowfall engine would probably blow up before it got 1/2 way to 100 where as gw2 "worked". Yes it was horrendous with lag and culling and generally considered a failure at launch and yet it is still capable of battles many times larger than what weve seen from Crowfall...

 

Thats scary.

 

Is it? This early in the dev cycle of the game? I just don't think its a fair comparison between a launched game and a pre-alpha one. Yeah we have not seen large battles yet but the current game mode doesn't even allow for it yet. I would expect the next phase "Siege Perilous" would be a better gauge as we might see the concurrent numbers ramping up. So sure the concern is valid but I don't think we have seen enough to be scared about it yet.

 

I also think that we'll have to wait until Siege Perilous to know what are the limits performance-wise. ACE probably already made some tests about this internally (not really hard, just drop hundreds of zombies to stress the engine), but I don't expect them to share this information with us.

Link to post
Share on other sites

If ACE can't get decent performance out of 400v400 sieges, the whole concept of CF isn't ready for consumption.

Now that the money is spent, I guess they could split CWs up into small islands linked by transporters and limit migration between islands (servers) to keep the numbers manageable. But that would be pretty far outside design goals.

 

More likely they would just lower the cap on how many crows can join a CW. Wouldn't cost anything but would be very effective in lowering the size of 'large scale' battles.

Link to post
Share on other sites

My brother and I were talking for hours about what the ideal size of a guild would be and the ideal size of group you could throw together at any given time.  I can't say I liked the chaos of some of the large scale battles I participated in for WoW.  I really hope the sweet spot for groups is somewhere between 20 and 32.

 

Crowfall caters to the PvP crowd though, and PvP in games like Shadowbane, DAoC, Warhammer and Darkfall implied hundreds of players fighting each others on the same battlefield, participating in sieges of epic proportions. Crowfall was aiming for this, we'll see how it plays out with the choices they've made (Unity, Voxels, independent server tech, etc.).

 

"The goal of the series was to create a number of unique images, each of which capture a single moment of gameplay. These images were not intended to be generic; we built these scenes using our characters, in our environments, acting within the scope of our vision. A single Banner Concept is the end result of numerous ideas, discussions, illustrations and sketches compiled over the last year (or so) of pre-production."

 

Crowfall_CitySiegeConcept.jpg

 

Ban_CallBanners.png

Edited by courant101
Link to post
Share on other sites

If ACE can't get decent performance out of 400v400 sieges, the whole concept of CF isn't ready for consumption.

Now that the money is spent, I guess they could split CWs up into small islands linked by transporters and limit migration between islands (servers) to keep the numbers manageable. But that would be pretty far outside design goals.

 

All this speculation could be premature, though.

We really haven't given ACE enough time to see what can be done with the Unity 5 engine.

 

400vs400 mamma mia... If they can get 50vs50 with wall destruction, physics, monsters, pets, siege equipment and mounts, and frame rate stable at 30 FPS on high-end PC, I think it's already going to be an incredible achievement. 

Edited by courant101
Link to post
Share on other sites

And yet, this is the expected size of big sieges post-launch, with the design goals we've seen.

I don't know the limitations of Unity 5, but ACE should have a decent idea by now.

If CF cannot possibly deliver with these stated goals, they need to decide on some limits to some design goals, float the proposed changes out here on the Forum, and let us discuss them like the fools we are.

That way ACE learns what trade-offs are more acceptable to the community.

 

No one ever thought CF would be perfect, let's help them make it as good as they can!

I think the K-Mart of MMO's already exists!  And it ain't us!   :)

 

Link to post
Share on other sites

 

Crowfall caters to the PvP crowd though, and PvP in games like Shadowbane, DAoC, Warhammer and Darkfall implied hundreds of players fighting each others on the same battlefield, participating in sieges of epic proportions. Crowfall was aiming for this, we'll see how it plays out with the choices they've made (Unity, Voxels, independent server tech, etc.).

 

"The goal of the series was to create a number of unique images, each of which capture a single moment of gameplay. These images were not intended to be generic; we built these scenes using our characters, in our environments, acting within the scope of our vision. A single Banner Concept is the end result of numerous ideas, discussions, illustrations and sketches compiled over the last year (or so) of pre-production."

 

Crowfall_CitySiegeConcept.jpg

 

Ban_CallBanners.png

 

I am not here for PvE trust that.  Let me be more specific.  We were just more concerned with the kind of hardware needed to see it and play it all smoothly.  I always run the best of the best hardware there is and upgrade everything every other year.  Even doing that with large scale battles video lag was an issue at times even with the best of the best Hardware.  I would rather have a no frustration experience at the scale that allows that.  Whether it be 16v16, 32v32, 50v50, 100v100, or even 400v400 is fine.  I don't think my hardware will be an issue.  So as big a scale as they can do while it is frustration free is fine with us, speaking for my brother.

Link to post
Share on other sites

Instead of limiting the numbers in battle, I think they should make is advantageous to split into smaller groups. You can have a big group of say 100 players storming a castle. But you should separate them into smaller strike teams. I would think you would want to storm a castle from all sides to spread out the defending forces, also you spread out your own team and hopefully don't take a lot of AOE damage. The same goes with field battles. If you have 100 people that want to group up tight, get a bunch of ranged classes behind a small group of tanks to AOE them down and out flank them. Make it punishing for the team that doesn't use proper battlefield tactics and instead just try to overwhelm by numbers. Big battles can be great when done right. But they wont be any fun if its just a zerg fest.

Link to post
Share on other sites

400vs400 mamma mia... If they can get 50vs50 with wall destruction, physics, monsters, pets, siege equipment and mounts, and frame rate stable at 30 FPS on high-end PC, I think it's already going to be an incredible achievement.

 

Other games did it, often with some lag. Crowfall hopefully will to. I've always been skeptical of the real physics and voxel destruction. Seems like a lot of potential for lag that only adds marginal benefits to the tactical pvp. Fake physics (e.g. cast power, target moves 10 feet back) is just fine. For sieges, the wall can either be standing or not. Voxels not needed.

 

Whatever though. Decisions were made. We get to see the results in March.

Link to post
Share on other sites

Other games did it, often with some lag. Crowfall hopefully will to. I've always been skeptical of the real physics and voxel destruction. Seems like a lot of potential for lag that only adds marginal benefits to the tactical pvp. Fake physics (e.g. cast power, target moves 10 feet back) is just fine. For sieges, the wall can either be standing or not. Voxels not needed.

 

Whatever though. Decisions were made. We get to see the results in March.

 

True, some games achieved that, we'll see how it plays out in March!

Link to post
Share on other sites

On Topic :

 

Glad I just did leaway, if something goes wrong at least it's no big lose ;P

And seriously, rotation of 1/3 of the squad is no big deal ? :)  I guess some ppl do live in another world.

 

On Numbers Topic :

Realistic and tech things aside for a moment, I hardly ever can feel immersed in siege thats ... 40 vs 40 or something.

That's just not the scope.  Where are siege engines, multiple routs of attack, stealth attackers, tunnel attackers, trenches, barricades and other things.

That is the siege.

40 vs 40 or so... just please no, that would be dissapoitment.(for me at least)

Link to post
Share on other sites

The question has been answered. Please lock this thread.

 

For the record, I still have faith in this game and I wish the best for all who left Artcraft Entertainment.

 

Who gives a poorly made socks? If you don't want to read the thread anymore don't read it. Quit trying to stifle conversation because you are buttmad about the direction the thread has gone. 

You are so incredibly helpful, CYT. I don't know how I ever managed to do anything before we met. I was just bumbling my way through life, all lost-like. Thank you. My blessing cup runneth over.

SWrkfdj.jpg

Link to post
Share on other sites
  • 2 weeks later...
  • 3 years later...
On 2/22/2016 at 3:13 AM, Urahara said:

By December 2016 the core module of Crowfall will be out for a soft launch.

What does that mean?

The core module wont be the Dregs plus the EKs, thats what their originally minimum goal for their crowdfunding campaign was.

They surpassed this minimum by far by now and are fully able to create the game they want/wanted to.

So from a developing structure standpoint its more likely, or the better approach, to start into the campaign worlds with creating the faction based CWs instead of starting with the guild-based CWs.

So the core module will be the EKs plus the faction based CWs God's Reach and The Infected, most likely.

Soft launch means that in December 2016 wont be the official release of Crowfall, instead you will be able to play the game in early access with the first impressions of what Crowfall will be like.

I am pretty sure they will grant people with "Beta 2 access" in December 2016 access to the game. As this is the last stage of access before open beta, as well as the first access level granted to their backers/early supporters. So they will make sure to let everyone of their backers in at least at the release date, they originally communicated at start of their crowdfunding campaign. To not discourage their early supporters.

Thats pretty much how its gonna be.

I dont think that is how it ended up

Does anyone know if ace is still using tornado.

Edited by Marth
Link to post
Share on other sites

This is not the right kind of necromancy Marth.

CF has always been a Unity project, I'm not sure where you got this "tornado" idea, or even what it is. I can find no mention of a game or unity plugin called "tornado" anywhere. Are you guerilla marketing your own middleware?

PopeSigGIF.gif

Rub rock on face and say "Yes food is eaten now time for fight"

Link to post
Share on other sites
12 minutes ago, PopeUrban said:

This is not the right kind of necromancy Marth.

CF has always been a Unity project, I'm not sure where you got this "tornado" idea, or even what it is. I can find no mention of a game or unity plugin called "tornado" anywhere. Are you guerilla marketing your own middleware?

"Thomas "Dreadflame" Sitch here.  I worked with Todd back on Shadowbane and then later on Wizard 101, and built the backend server technology ("Tornado2 Engine") during my own independent game developer time.  We're using that technology on Crowfall, which cuts off about 2 years of server development."

 

Link to post
Share on other sites
38 minutes ago, Marth said:

"Thomas "Dreadflame" Sitch here.  I worked with Todd back on Shadowbane and then later on Wizard 101, and built the backend server technology ("Tornado2 Engine") during my own independent game developer time.  We're using that technology on Crowfall, which cuts off about 2 years of server development."

 

**Permit me to Ramble in place of having a definite answer to your question, as I know a bit about the field but not the specific subject.**

Chances are they're still using the same server backend for everything that isn't the simulation (e.g. account systems, banking, etc.) and using networking calls from Unity to interface with it.

I did this on a project where we had to interface with a GIS framework and dynamically fill a city map with data from that data base. The actual database back end that held that data was a separate system (which we had read-only access to), while the navigation, GUI, etc. was all unity. I never touched the actual database code but I had to write a Unity interface so I could talk to it.

What the network back end would be responsible for in terms of lag in an MMO if its built like most MMOS would be stuff like how long it takes to load the proper appearance, name, guild, etc. of a new player, or how long things like looting or banking take. This is because the sim simply says "hey Client #149982 is in sight range of client#568879, so I need to know who client#568879 is so we can send the right instructions to #149982 to have his PC build the correct name string and load the right models and (maybe) cache the right power effects"

So the sim, the thing in charge of making sure your movement and hotkeys and jumping and so on feel good is basically just configuresd to send a simple command while the database backend software does the heavy lifting of keeping track of who these clients are, what characters they're playing, what their inventory is, what accounts they're logged in as, and in some cases (for instance crafting or generating mob loot) doing math that isn't critical to the pace of combat or movement.

This prevents all that heavy lifting from impacting things like  movement lag, ability timing lag, and anything happening in the world as that's going to be done on the simulation server so the sim can focus on making sure everyone isn't suffering 9000ms lag and displays and animated in the right place, with some help in prediction/movement smoothing from the client.

 

So your Crowfall Architecture MIGHT look like:

Unity Client<>Unity Sim Server (Tornado I/O plugin)<>Tornado DB handler (Administrates SQL or its own DB format)

 

I'm banking on Unity as the sim server based on what we've seen of their dev process, as there's otherwise very little need to do stuff like build a visual AI decision tree IN unity (as you could build that in whatever UI you wanted to if your sim was running on proprietary code)

 

Your DB handler and sim are built hand in hand, and Unity is a popular free tool for indie devs so it is a sensible platform if you're just building this kind of tech in your spare time, as it seems dreadflame did.

It wouldn't surprise me if that stuff is still in operation as the basic architecture of most MMO needs tend to stem from very similar basic concepts (account, character, inventory, stat points) which is where you're seeing a lot of time saving. No need to reinvent your database handlers for stuff like "account is logged in, keep track of this hash and use it for further lookups" or "what's an inventory" when you've already got them built.

Otherwise you'd have to build them from scratch and make sure the whole thing is efficiently queing the hundreds or thousands of requests it gets every second.

Interestingly enough, if this is the case it would also be the backend tech barrier that needed updating to support things like the new decentralized banks. If dreadflame's previous indie dev use didn't need non-character inventories, it makes sense he wouldn't have developed them in the back end.

Edited by PopeUrban

PopeSigGIF.gif

Rub rock on face and say "Yes food is eaten now time for fight"

Link to post
Share on other sites
On 8/29/2019 at 12:08 PM, PopeUrban said:

**Permit me to Ramble in place of having a definite answer to your question, as I know a bit about the field but not the specific subject.**

Chances are they're still using the same server backend for everything that isn't the simulation (e.g. account systems, banking, etc.) and using networking calls from Unity to interface with it.

I did this on a project where we had to interface with a GIS framework and dynamically fill a city map with data from that data base. The actual database back end that held that data was a separate system (which we had read-only access to), while the navigation, GUI, etc. was all unity. I never touched the actual database code but I had to write a Unity interface so I could talk to it.

What the network back end would be responsible for in terms of lag in an MMO if its built like most MMOS would be stuff like how long it takes to load the proper appearance, name, guild, etc. of a new player, or how long things like looting or banking take. This is because the sim simply says "hey Client #149982 is in sight range of client#568879, so I need to know who client#568879 is so we can send the right instructions to #149982 to have his PC build the correct name string and load the right models and (maybe) cache the right power effects"

So the sim, the thing in charge of making sure your movement and hotkeys and jumping and so on feel good is basically just configuresd to send a simple command while the database backend software does the heavy lifting of keeping track of who these clients are, what characters they're playing, what their inventory is, what accounts they're logged in as, and in some cases (for instance crafting or generating mob loot) doing math that isn't critical to the pace of combat or movement.

This prevents all that heavy lifting from impacting things like  movement lag, ability timing lag, and anything happening in the world as that's going to be done on the simulation server so the sim can focus on making sure everyone isn't suffering 9000ms lag and displays and animated in the right place, with some help in prediction/movement smoothing from the client.

 

So your Crowfall Architecture MIGHT look like:

Unity Client<>Unity Sim Server (Tornado I/O plugin)<>Tornado DB handler (Administrates SQL or its own DB format)

 

I'm banking on Unity as the sim server based on what we've seen of their dev process, as there's otherwise very little need to do stuff like build a visual AI decision tree IN unity (as you could build that in whatever UI you wanted to if your sim was running on proprietary code)

 

Your DB handler and sim are built hand in hand, and Unity is a popular free tool for indie devs so it is a sensible platform if you're just building this kind of tech in your spare time, as it seems dreadflame did.

It wouldn't surprise me if that stuff is still in operation as the basic architecture of most MMO needs tend to stem from very similar basic concepts (account, character, inventory, stat points) which is where you're seeing a lot of time saving. No need to reinvent your database handlers for stuff like "account is logged in, keep track of this hash and use it for further lookups" or "what's an inventory" when you've already got them built.

Otherwise you'd have to build them from scratch and make sure the whole thing is efficiently queing the hundreds or thousands of requests it gets every second.

Interestingly enough, if this is the case it would also be the backend tech barrier that needed updating to support things like the new decentralized banks. If dreadflame's previous indie dev use didn't need non-character inventories, it makes sense he wouldn't have developed them in the back end.

Tldr; nerd copy pasta. 

Marty must have finished the necro tree.  Most people don’t have the power to necro thread this old. 

 

Edited by mandalore
40 minutes ago, Andius said:

W/HoA were held up as like these mystical forces of highly skilled players with legendary theorycrafters chained to a desk in some deep dungeon holding all the arcane secrets we could use to win if only we knew them.

wiDfyPp.png

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    No registered users viewing this page.


×
×
  • Create New...