First superficial numbers given on OCS performance gains, from this German interview.
They still can't give a percentage in performance gain, but they gave numbers that demonstrate how impactful this technology is:
If you are at Port Olissar, the game calculates 70.000 entities.
With OBS it gets reduced to calculate 9.000 entities.
The performance improvement is also used for more NPC's, hence not 100% better performance, but a mix of both: Better performance, more NPC's and in general more background activities.
Alpha 3.3 progress around 11% towards its completion since last week, from calculating the number of tasks stands 56% completed at this point.
Network Bind Culling is alive and progressed this past week, with the OCS sub-tasks nearing 100%.
I don't think that completion % is based just on number of tasks, it seems to use some other more intelligent formula because it is able to take items which have no tasks into account.
The formula could be something like: sum (Item_Completion_Percent) / Number_Of_Items
I don't think that completion % is based just on number of tasks, it seems to use some other more intelligent formula because it is able to take items which have no tasks into account.
The formula could be something like: sum (Item_Completion_Percent) / Number_Of_Items
Each item has its own % based on its number of tasks, so the total should be the medium based on the completion % of every item on the roadmap, without individual weight.
It'll likely be mission givers tied to factions, as there has been mentions of factions, the form of progression is reputation, so makes sense, unless it's a specific story type thing that the game should have, not just the generated stuff.
Here's a visual representation of the big optimization undertaking Object Container Streaming:
This level of culling is what will allow the great complexity of a capital ship to exist without overwhelming the client, the whole point of OCS is those containers dynamically loading and unloading in the bakground as you move around instead of what happens currently where you load everything in one go.
In the screens you notice the highlighted boxes, as shown the game dynamically determines which will you render as you move.
A wave of Evocati invites for Alpha 3.3 had just happened, the first testing phase of 3.3 is expected to start within the next 2 weeks, as per the norm.
Ships are only about one third complete, so it wouldn't be a wonder if one or more of them got delayed. But on the other hand they might also act like last patch and decided it's good enough to push out.
I don't think that completion % is based just on number of tasks, it seems to use some other more intelligent formula because it is able to take items which have no tasks into account.
The formula could be something like: sum (Item_Completion_Percent) / Number_Of_Items
Each item has its own % based on its number of tasks, so the total should be the medium based on the completion % of every item on the roadmap, without individual weight.
Its very simple. 54 tasks, 6 complete. 6/54 is 11%. Its a 0-100 percent approach. And that's OK since only completed tasks will make it into the next build.
Edit: other types of calc can be done from %complete to first release to calcs that take into account (money, time, value added - lots of calc can be done. And in different ways.
The 11% is focused on 3.3 and is "transparent" in how it has been arrived at. They are using the same approach for the line item tasks as well. There is a fraction at the end of each e.g. 17/32 which becomes 53%. The fraction is the number of "sub-tasks" that have to be completed for each task.
Note though that some of these "sub-tasks" are common to a variety of tasks. So 5 tasks with 10 sub-tasks each may not translate to 50 sub-tasks. If e.g. 4 of them are common across all 10 and the rest unique there will be 34 unique sub-tasks.
And the calc is so simple that sometimes you also see "negative progress" - the result of a sub-task being added!
Ships are only about one third complete, so it wouldn't be a wonder if one or more of them got delayed. But on the other hand they might also act like last patch and decided it's good enough to push out.
Ships are "repeaters" not "new" tasks. Low risk chance of them not being finished. Same deal with weapons, armour, planets it seems after this patch. Now stuff like Gameplay, AI and Core Tech is different. 3.1 and 3.2 suggests that they have gotten better at estimating however. (Maybe last year replaced unbridled enthusiasm and a wish to please by hardnosed realism.)
"We all do the best we can based on life experience, point of view, and our ability to believe in ourselves." - Naropa "We don't see things as they are, we see them as we are." SR Covey
Ships are only about one third complete, so it wouldn't be a wonder if one or more of them got delayed. But on the other hand they might also act like last patch and decided it's good enough to push out.
Ships are "repeaters" not "new" tasks. Low risk chance of them not being finished. Same deal with weapons, armour, planets it seems after this patch. Now stuff like Gameplay, AI and Core Tech is different. 3.1 and 3.2 suggests that they have gotten better at estimating however. (Maybe last year replaced unbridled enthusiasm and a wish to please by hardnosed realism.)
They pushed some ships back just last patch. Even if they're low risk that doesn't mean there's no risk.
Ships are only about one third complete, so it wouldn't be a wonder if one or more of them got delayed. But on the other hand they might also act like last patch and decided it's good enough to push out.
Ships are "repeaters" not "new" tasks. Low risk chance of them not being finished. Same deal with weapons, armour, planets it seems after this patch. Now stuff like Gameplay, AI and Core Tech is different. 3.1 and 3.2 suggests that they have gotten better at estimating however. (Maybe last year replaced unbridled enthusiasm and a wish to please by hardnosed realism.)
They pushed some ships back just last patch. Even if they're low risk that doesn't mean there's no risk.
Yes they did. And yes there are always risks. What they did last last patch though was - imo - "rescheduling" rather than "slippage". Probably to manage resources.
When you have done something many times before you get a good idea for what is needed and you can plan accordingly. Hence they believed before 3.3 got underway - and contiunue to believe since they haven't at this point in time moved anything out - that they will complete them for 3.3.
Ships are only about one third complete, so it wouldn't be a wonder if one or more of them got delayed. But on the other hand they might also act like last patch and decided it's good enough to push out.
Ships are "repeaters" not "new" tasks. Low risk chance of them not being finished. Same deal with weapons, armour, planets it seems after this patch. Now stuff like Gameplay, AI and Core Tech is different. 3.1 and 3.2 suggests that they have gotten better at estimating however. (Maybe last year replaced unbridled enthusiasm and a wish to please by hardnosed realism.)
They pushed some ships back just last patch. Even if they're low risk that doesn't mean there's no risk.
Yes they did. And yes there are always risks. What they did last last patch though was - imo - "rescheduling" rather than "slippage". Probably to manage resources.
Imho there is no rescheduling that is not slippage, not unless there's a large natural catastrophe or something else that counts as force majeure on the scale of their company.
Imho there is no rescheduling that is not slippage, not unless there's a large natural catastrophe or something else that counts as force majeure on the scale of their company.
I'm not sure if it was a ship or not but seen push backs on a dev getting sick, the programmer working on the feature for some weeks, on this scale it would be quite common as one does not simply put someone into someone else's code and expect it to work, so that hit will always be taken.
Imho there is no rescheduling that is not slippage, not unless there's a large natural catastrophe or something else that counts as force majeure on the scale of their company.
I'm not sure if it was a ship or not but seen push backs on a dev getting sick, the programmer working on the feature for some weeks, on this scale it would be quite common as one does not simply put someone into someone else's code and expect it to work, so that hit will always be taken.
Yes, and schedules continuously slip because of stuff like that.
Preproduction started on the Arccorp moons, currently on the roadmap schedule for 3.4 (Q4 2018).
Remember back in 2016 when we were supposed to get full Stanton system + Delamar planet at the end of 2016? Now it's 2018 and RSI has started looking at Microtech beyond the roadmap.
Production roadmap was updated again, but very little changes. Improvements to Retrieval and Delivery Missions, Improvements to Mining on Planetary Bodies, and FPS Combat: Search Behavior are now in polishing stage.
Remember back in 2016 when we were supposed to get full Stanton system + Delamar planet at the end of 2016? Now it's 2018 and RSI has started looking at Microtech beyond the roadmap.
From what was said later, and I was on the chats when the dev was talking about this, on the original announced 3.0 the flagship landing locations of the 5 planets were still localized areas (aka arccorp integrated) and not full seamless planets with cities on them as what we now expect Hurston in 3.3.
Soping up the main locations would explain the enormous pushbacks by pulling out all flagship landing zones from the update.
@Vrika hey you missed on Network Culling, it progressed 22% this week stands at 81% what is great news that is now considered feature-complete with the upcoming OCS, this should mean a big improvement on the client-side performance with both game and player entities being loaded only in range.
This likely will be is a lot of bugs and stuff to sort out, being OCS something that required everything in the game to be changed to work with it, we're surely up to some turbulence before stability complements performance.
The weekly graph, as said already the biggest boost this week was the network bind culling that is finally considered feature-complete, it's all up now to the ongoing implementation of OCS in terms of tech.
The update is scheduled for Oct 10 release or at least open-PTU by then.
Comments
They still can't give a percentage in performance gain, but they gave numbers that demonstrate how impactful this technology is:
If you are at Port Olissar, the game calculates 70.000 entities.
With OBS it gets reduced to calculate 9.000 entities.
The performance improvement is also used for more NPC's, hence not 100% better performance, but a mix of both: Better performance, more NPC's and in general more background activities.
Alpha 3.3 progress around 11% towards its completion since last week, from calculating the number of tasks stands 56% completed at this point.
Network Bind Culling is alive and progressed this past week, with the OCS sub-tasks nearing 100%.
The formula could be something like: sum (Item_Completion_Percent) / Number_Of_Items
When you do enough "good/bounty hunting" missions you will be directed to Miles Eckart that gives more missions.
If you do "bad/pirating" missions you will get Ruto calling you for more mischief missions.
I can see CIG putting other hero npc's for the specific roles like mining/trading/exploration etc
This level of culling is what will allow the great complexity of a capital ship to exist without overwhelming the client, the whole point of OCS is those containers dynamically loading and unloading in the bakground as you move around instead of what happens currently where you load everything in one go.
In the screens you notice the highlighted boxes, as shown the game dynamically determines which will you render as you move.
Ships are only about one third complete, so it wouldn't be a wonder if one or more of them got delayed. But on the other hand they might also act like last patch and decided it's good enough to push out.
Edit: other types of calc can be done from %complete to first release to calcs that take into account (money, time, value added - lots of calc can be done. And in different ways.
The 11% is focused on 3.3 and is "transparent" in how it has been arrived at. They are using the same approach for the line item tasks as well. There is a fraction at the end of each e.g. 17/32 which becomes 53%. The fraction is the number of "sub-tasks" that have to be completed for each task.
Note though that some of these "sub-tasks" are common to a variety of tasks. So 5 tasks with 10 sub-tasks each may not translate to 50 sub-tasks. If e.g. 4 of them are common across all 10 and the rest unique there will be 34 unique sub-tasks.
And the calc is so simple that sometimes you also see "negative progress" - the result of a sub-task being added!
"We all do the best we can based on life experience, point of view, and our ability to believe in ourselves." - Naropa "We don't see things as they are, we see them as we are." SR Covey
With close to 6 weeks to go, they need close to 10% progress per week, together with the whole bug-fixing phase.
When you have done something many times before you get a good idea for what is needed and you can plan accordingly. Hence they believed before 3.3 got underway - and contiunue to believe since they haven't at this point in time moved anything out - that they will complete them for 3.3.
Could they be wrong? Of course.
Preproduction started on the Arccorp moons, currently on the roadmap schedule for 3.4 (Q4 2018).
https://youtu.be/Z-3YBuFI3iI?t=1555
Production roadmap was updated again, but very little changes. Improvements to Retrieval and Delivery Missions, Improvements to Mining on Planetary Bodies, and FPS Combat: Search Behavior are now in polishing stage.
Soping up the main locations would explain the enormous pushbacks by pulling out all flagship landing zones from the update.
This likely will be is a lot of bugs and stuff to sort out, being OCS something that required everything in the game to be changed to work with it, we're surely up to some turbulence before stability complements performance.
The weekly graph, as said already the biggest boost this week was the network bind culling that is finally considered feature-complete, it's all up now to the ongoing implementation of OCS in terms of tech.
The update is scheduled for Oct 10 release or at least open-PTU by then.