On my test computer on one of Nagorak's projects, it would take 660 ms to update the UI after each change. On Tribble, it now takes 280 ms to update the UI. Once all the optimizations get in, it will take 180 ms to update the UI after each change. (These numbers are all on my computer. A slower computer will run slower.)
Awesome, I'm glad to hear about the improvement. At about 70% increase in speed, I expect the problems should more or less be solved.
Now that a lot of fixes are in the pipeline (one is on Tribble, other fixes should be coming soon), I thought you might like to know what happened.
The new slowdown is primarily the result of changes to the Costume Editor. This means the more costumes you have, the slower your editing would be. Nagorak, your missions performed the worst out of everyone in this thread because you create so many custom costume. (BTW, Nagorak's missions are awesome because of all the custom costumes he makes. You should play them.)
On my test computer on one of Nagorak's projects, it would take 660 ms to update the UI after each change. On Tribble, it now takes 280 ms to update the UI. Once all the optimizations get in, it will take 180 ms to update the UI after each change. (These numbers are all on my computer. A slower computer will run slower.)
Well TRIBBLE. My mission is using all 50 available costumes (unique villagers, soldiers, etc.)
What is this 660ms lag? I'm averaging an almost 10 second wait between every single click now, its simply unusable.
Did I read the notes correctly in that the "speed fix" would be implemented this coming Thursday?
The 500 mb update did nothing to alleviate the consistently 10 second lag associated with selecting an object, moving an object, rotating an object, renaming an object, changing an objects costume, entering text into the Name field of a dialog, entering text into the Text field of a dialog, duplicating an object (which is only successful if the duplicated object is the specific one that was first dropped from the NPC Groups panel), drag-selecting anything, drag and dropping a Detail or Building.
I had big, almost ridiculously eager plans of creating a mission that encompassed an array of my favorite themes encountered as I was growing up, but now I spend about 20 frustrating minutes per day trying to manually rename and "re-costume" an enemy NPC group (who are mysteriously "un-duplicatable") and I eventually just give up, I could spend a hundred hours with the Foundry and only have ten hours worth of progress.
If it functioned as intended it would be one of the most robust user-content generation systems ever devised, but right now it is broken and an exercise in futility to operate.
Please help me tell my story.
0
jfinderdevMember, Cryptic DevelopersPosts: 73Arc User
edited August 2012
Ryandebraal, I tested ST-HHUCJI72A. Your project benefits most from two more optimizations coming down the pipeline that on my computer cut the lag in half. These optimizations were put in the pipeline a bit too late to get out yesterday, but they are both done.
For me I think the lag has gotten better but it's still noticeable. For instance when I try to rename all the individual actors in an NPC group I click on one, type....wait a little while then do the next one.
For me I think the lag has gotten better but it's still noticeable. For instance when I try to rename all the individual actors in an NPC group I click on one, type....wait a little while then do the next one.
Agree.. is better, but still is really bad. I hope this can be fix soon because is unusable.
Ryandebraal, I tested ST-HHUCJI72A. Your project benefits most from two more optimizations coming down the pipeline that on my computer cut the lag in half. These optimizations were put in the pipeline a bit too late to get out yesterday, but they are both done.
Have you guys thought about compartmentalizing the design of the Foundry more? What I mean is the Costume Editor could be completely separate from the mission editor. The map/storyline/dialogue editor probably need to be integrated, but the costume editor really doesn't.
You could design costumes in the costume module that would be its own self contained part of the program, and then save a certain number (I guess mission slots x 50). Then you could load those into each mission, but not have them be editable there. This would accomplish a few things: 1) you could make the costume editor more full featured without having to worry about its affect on the rest of the Foundry (perhaps you could simply use the game's default character editor); 2) It would make it so that we could use the same costumes across multiple missions without having to recreate them each time, which is currently a major pain; 3) Possibly it could allow us to use more than 50 costumes since they could be stored separately and not part of the mission.
I know you've put in some optimizations, but the truth is, the fact that making costumes slows down the whole editor strikes me as a flaw in general.
Also, just in general, it seems to me that when you're not actively looking at a map, the assets there should be unloaded or something. It feels like the more stuff you put in the mission, the more bogged down everything gets. There has to be some way to contain the performance degradation. Even if you want to have everything available at first so it's easy to switch between maps, as more assets are added maybe some additional performance optimizations need to come into effect. Even if it were to take longer to switch between maps, that could still be an improvement if it makes the current map respond faster.
Perhaps giving us the option of "loading" or "unloading" maps depending on whether we plan to work on them now or not. Unloaded maps would have to be loaded before we could switch to them, but the assets associated with them would not slow down the Foundry. Even if it took 30 seconds to load a map, it would still be worth it if everything else performed better.
Ryandebraal, I tested ST-HHUCJI72A. Your project benefits most from two more optimizations coming down the pipeline that on my computer cut the lag in half. These optimizations were put in the pipeline a bit too late to get out yesterday, but they are both done.
That's good to hear.
As of now I've stopped using the Foundry because it has become an exercise in futility, the frustrating Map interface makes me less likely to write thoughtful dialog because I just see it all as placeholders that I'll comeback to once I can actually work with the interface and not fight an uphill battle against it.
I admit the scope of my project is beyond what most missions probably entail. I've used all 15 possible map transitions, half of the maps are 1 building or 1 NPC group away from "exceeding" the maximum memory alotment and I am using 50 costumes to give variety to non quest NPCs that would just wander around.
So I'm not expecting a miracle but after populating one more additional map with trees/buildings/NPCs the lag jumped from 3 seconds to easily 10 seconds, and it's rage inducing if you misclick and suddenly not only failed to perform the action you wanted but you must now dedicate atleast the next minute to putting right what has gone wrong before you can even make the initially intended change.
Honestly, if you're hitting the limits on all your maps and also using all 15 maps, you might want to reconsider. It sounds to me like your mission might either be too long, or have a lot of unnecessary map transitions (ex: warp into a system to scan one thing and then warp out).
My missions are all about 1.5-3 hours long and each one only uses about 5 maps. Granted, it depends how much each map holds too, but using all 15 maps sounds a bit worrying to me.
That being said, you can always copy and split your mission into 2 or 3 parts later, so it's not the end of the world. But when you finally finish and go to play it, I would be prepared to do that if it turns out to be too long.
Either way, the both of us are locked out of the Foundry for now due to the poor performance. I'd like to get my missions back up, but I'm not willing to fight with the lag.
Hopefully the next update will improve things a bit more, so I can get back to work.
In my humble opinion I think it has nothing to do with the costumes, and if I did not explain the reason for the game if casting a mission seeking the appropriate menu of missions I slowed the entire game.
Ryandebraal, I tested ST-HHUCJI72A. Your project benefits most from two more optimizations coming down the pipeline that on my computer cut the lag in half. These optimizations were put in the pipeline a bit too late to get out yesterday, but they are both done.
Is today's patch going to include the two optimizations that reduce approx. 50% of the lag?
I logged on today, and there was indeed a very noticeable increase in reaction time!
Thank you so much, I know it must be very frustrating to work so hard and then have issues crop up which cause your user base to become upset with you.
Then when the issues are corrected no one even acknowledges your hard work.
Well, I for one am very thankful for your dilligence in this manner. It's much appreciated!
I just wanted to say that I finally got back into the Foundry last night. It seems like with the changes the Foundry is at least as fast as it was before, possibly faster even, but definitely no worse.
So, nice job. Now get back to work fixing some of the other bugs!
Comments
Awesome, I'm glad to hear about the improvement. At about 70% increase in speed, I expect the problems should more or less be solved.
Click here for my Foundry tutorial on Creating A Custom Interior Map.
Division Hispana
www.divisionhispana.com
Well TRIBBLE. My mission is using all 50 available costumes (unique villagers, soldiers, etc.)
What is this 660ms lag? I'm averaging an almost 10 second wait between every single click now, its simply unusable.
Did I read the notes correctly in that the "speed fix" would be implemented this coming Thursday?
I had big, almost ridiculously eager plans of creating a mission that encompassed an array of my favorite themes encountered as I was growing up, but now I spend about 20 frustrating minutes per day trying to manually rename and "re-costume" an enemy NPC group (who are mysteriously "un-duplicatable") and I eventually just give up, I could spend a hundred hours with the Foundry and only have ten hours worth of progress.
If it functioned as intended it would be one of the most robust user-content generation systems ever devised, but right now it is broken and an exercise in futility to operate.
Please help me tell my story.
Foundry Mission Database
Check out my Foundry missions:
Standalone - The Great Escape - The Galaxy's Fair - Purity I: Of Denial - Return to Oblivion
Untitled Series - Duritanium Man - The Improbable Bulk - Commander Rihan
Agree.. is better, but still is really bad. I hope this can be fix soon because is unusable.
Division Hispana
www.divisionhispana.com
Have you guys thought about compartmentalizing the design of the Foundry more? What I mean is the Costume Editor could be completely separate from the mission editor. The map/storyline/dialogue editor probably need to be integrated, but the costume editor really doesn't.
You could design costumes in the costume module that would be its own self contained part of the program, and then save a certain number (I guess mission slots x 50). Then you could load those into each mission, but not have them be editable there. This would accomplish a few things: 1) you could make the costume editor more full featured without having to worry about its affect on the rest of the Foundry (perhaps you could simply use the game's default character editor); 2) It would make it so that we could use the same costumes across multiple missions without having to recreate them each time, which is currently a major pain; 3) Possibly it could allow us to use more than 50 costumes since they could be stored separately and not part of the mission.
I know you've put in some optimizations, but the truth is, the fact that making costumes slows down the whole editor strikes me as a flaw in general.
Click here for my Foundry tutorial on Creating A Custom Interior Map.
Perhaps giving us the option of "loading" or "unloading" maps depending on whether we plan to work on them now or not. Unloaded maps would have to be loaded before we could switch to them, but the assets associated with them would not slow down the Foundry. Even if it took 30 seconds to load a map, it would still be worth it if everything else performed better.
Click here for my Foundry tutorial on Creating A Custom Interior Map.
That's good to hear.
As of now I've stopped using the Foundry because it has become an exercise in futility, the frustrating Map interface makes me less likely to write thoughtful dialog because I just see it all as placeholders that I'll comeback to once I can actually work with the interface and not fight an uphill battle against it.
I admit the scope of my project is beyond what most missions probably entail. I've used all 15 possible map transitions, half of the maps are 1 building or 1 NPC group away from "exceeding" the maximum memory alotment and I am using 50 costumes to give variety to non quest NPCs that would just wander around.
So I'm not expecting a miracle but after populating one more additional map with trees/buildings/NPCs the lag jumped from 3 seconds to easily 10 seconds, and it's rage inducing if you misclick and suddenly not only failed to perform the action you wanted but you must now dedicate atleast the next minute to putting right what has gone wrong before you can even make the initially intended change.
My missions are all about 1.5-3 hours long and each one only uses about 5 maps. Granted, it depends how much each map holds too, but using all 15 maps sounds a bit worrying to me.
That being said, you can always copy and split your mission into 2 or 3 parts later, so it's not the end of the world. But when you finally finish and go to play it, I would be prepared to do that if it turns out to be too long.
Either way, the both of us are locked out of the Foundry for now due to the poor performance. I'd like to get my missions back up, but I'm not willing to fight with the lag.
Hopefully the next update will improve things a bit more, so I can get back to work.
Click here for my Foundry tutorial on Creating A Custom Interior Map.
A greeting.
https://www.youtube.com/user/NenfisGirfud
Is today's patch going to include the two optimizations that reduce approx. 50% of the lag?
Thank you so much, I know it must be very frustrating to work so hard and then have issues crop up which cause your user base to become upset with you.
Then when the issues are corrected no one even acknowledges your hard work.
Well, I for one am very thankful for your dilligence in this manner. It's much appreciated!
Foundry Mission Database
Check out my Foundry missions:
Standalone - The Great Escape - The Galaxy's Fair - Purity I: Of Denial - Return to Oblivion
Untitled Series - Duritanium Man - The Improbable Bulk - Commander Rihan
Slow and steady wins the race in this instance. Great job guys and gals.
15 seconds each time I have to look at the map is loooong time. Slow down the game every time I open a menu of foundry is looooong time.
But if you want me to thank something I do, thanks for having the foundry operation.
https://www.youtube.com/user/NenfisGirfud
So, nice job. Now get back to work fixing some of the other bugs!
Click here for my Foundry tutorial on Creating A Custom Interior Map.