+ Reply to Thread
Page 3 of 5 FirstFirst 1 2 3 4 5 LastLast
Results 31 to 45 of 72
Like Tree64Likes

  Click here to go to the first Rift Team post in this thread.   Thread: Leveling up the Tech in RIFT

  1. #31
    Rift Chaser Electrah's Avatar
    Join Date
    Aug 2010
    Posts
    329

    Default

    Quote Originally Posted by Metsa View Post
    and maybe being able to turn my graphics up on my quad core machine
    Thing is, when you have a high-end graphics card it seems you get higher performance when you crank up the settings rather than turning them down, as if the graphics card feels more at ease with a higher workload.
    Pulling all the sliders to the right (so Ultra +) and turning on anisotropic filtering and supersampling AA actually increases my framerate. That being said, in raids I still have about 18 fps (in combat) so I really hope these changes will noticeably increase that


    Thanks Snedhepl for your reply. To clarify the above mentioned framerate a bit more: I run a stock i5 3570 (3.4Ghz) quadcore, with 16GB RAM and a GTX 980. The PC has a 64 bit windows 10, and 2 SSDs.
    Last edited by Electrah; 02-05-2016 at 10:04 AM. Reason: added system specs
    Fantasy - 51 Justicar - Eyes of Eternity - Vigil (retired)
    Electrah - 61 Riftstalker - Harshmallows - Gelidra (retired)
    No longer playing Rift

  2. #32
    Plane Walker AtoningUnifex's Avatar
    Join Date
    Jan 2011
    Posts
    491

    Default

    Quote Originally Posted by Snedhepl View Post
    1. Project documents: I can't speak for the team as a whole, as I handle the management for my team of people. I do my best to keep it sane.

    2. We have a variety of machines set up for testing. It's by no means exhaustive (which is essentially impossible) but it does handle things well. This is the sort of testing that very quickly hits the point of diminishing returns.

    3. We have a lot of people on older machines, it's the nature of any mature MMO. I can't share exact figures but there are quite a bit of folks who couldn't support DX11 yet alone DX12 (which also requires Windows 10)
    Thank you Snedhelp for replying!

    1. I realise you can't speak about specifics but I have a professional interest in the area of project management and had visions of 50000 line gantt charts, and some Devs sitting with work packages literally up to their ears. :P

    2. I get that there would be a point where it would quickly go from productive to counter productive but perhaps you could run a post where you ask for players with specific setups to get in touch with you. I run an 8 core FX8320 with 32Gb ram and GeForce 750 so am really looking forward to seeing an improvement in the overall performance and as I run 64-bit windows moving to a 64-bit client will I hope greatly enhance my gaming experience.

    3. I think some other people have responded to this as well, I am almost certain Slipmat suggested having an option on front end, much as you already do with Archeage? Realise that probably adds more development time, testing time, QA and support time into the mix - but would that be a feasible option with the modified engine or would it only support one version of DX at a time?

    Anyway cheers for answering my questions and look forward to more details tonight.
    Last edited by AtoningUnifex; 02-05-2016 at 09:50 AM. Reason: minor spelling and reprhasing of 2.

  3.   Click here to go to the next Rift Team post in this thread.   #33
    Rift Team
    Join Date
    Jun 2014
    Posts
    677

    Default

    Supporting DX11 is a lot more than just adding a switch that says 'use DX11'. Adding support for another API (be it DX11, 12 or Vulkan) is going to be work on a similar scale of what we've just done with multicore. If we were to do it, yeah sure we'd add the UI switch.

    Making a switch isn't the problem. Adding a new API helps -some- who play Rift. Making DX9 work better helps -everyone- who plays Rift.

    Quote Originally Posted by AtoningUnifex View Post
    Thank you Snedhelp for replying!

    1. I realise you can't speak about specifics but I have a professional interest in the area of project management and had visions of 50000 line gantt charts, and some Devs sitting with work packages literally up to their ears. :P

    2. I get that there would be a point where it would quickly go from productive to counter productive but perhaps you could run a post where you ask for players with specific setups to get in touch with you. I run an 8 core FX8320 with 32Gb ram and GeForce 750 so am really looking forward to seeing an improvement in the overall performance and as I run 64-bit windows moving to a 64-bit client will I hope greatly enhance my gaming experience.

    3. I think some other people have responded to this as well, I am almost certain Slipmat suggested having an option on front end, much as you already do with Archeage? Realise that probably adds more development time, testing time, QA and support time into the mix - but would that be a feasible option with the modified engine or would it only support one version of DX at a time?

    Anyway cheers for answering my questions and look forward to more details tonight.

  4.   Click here to go to the next Rift Team post in this thread.   #34
    Rift Team
    Join Date
    Jun 2014
    Posts
    677

    Default

    These are snippets of screenshots that are up on the web version of this post. They're taken literally seconds apart and represent a best case scenario. But what you won't see in the screenshot is that the pictures were taken during "The Ravenous Devourer" event. I forget which shard precisely but it was on the NA cluster.

    Without Multicore.


    With Multicore


    Turning it on requires super-secret special sauce and thus won't affect anyone else around you or be turned on accidentally.


    Quote Originally Posted by Necalli View Post
    ............

    Ocho.... if you are wrong....
    Last edited by Snedhepl; 02-05-2016 at 10:45 AM.

  5. #35
    Plane Walker AtoningUnifex's Avatar
    Join Date
    Jan 2011
    Posts
    491

    Default

    Quote Originally Posted by Snedhepl View Post
    Supporting DX11 is a lot more than just adding a switch that says 'use DX11'. Adding support for another API (be it DX11, 12 or Vulkan) is going to be work on a similar scale of what we've just done with multicore. If we were to do it, yeah sure we'd add the UI switch.

    Making a switch isn't the problem. Adding a new API helps -some- who play Rift. Making DX9 work better helps -everyone- who plays Rift.
    Fair enough, sounds entirely reasonable to me.

    The last 18 months must have been very...busy ...much kudos to you all for getting this to a point where you can even announce it.

    I am I right in interpreting it this way:

    If adding API's effectively doubles the burden of work and therefore increases the burden in terms of staffing resources, funding requirements, and lengthens delivery timescales then at some point it would become counter-productive in terms of the benefit to players versus cost to complete, and that a new engine would be the only remaining option, which from what you say above sounds like your least desirable option in terms of amount of recoding that would be required.

    Is this a fair assessment of what has been happening or am I barking up the wrong tree?

  6.   Click here to go to the next Rift Team post in this thread.   #36
    Rift Team
    Join Date
    Jun 2014
    Posts
    677

    Default

    I'd say that's pretty close to correct. One thing that's missing is that with multicore we can more readily do further optimizations to improve performance that would be difficult or impossible when trying to support multiple APIs.

    Quote Originally Posted by AtoningUnifex View Post
    Fair enough, sounds entirely reasonable to me.

    The last 18 months must have been very...busy ...much kudos to you all for getting this to a point where you can even announce it.

    I am I right in interpreting it this way:

    If adding API's effectively doubles the burden of work and therefore increases the burden in terms of staffing resources, funding requirements, and lengthens delivery timescales then at some point it would become counter-productive in terms of the benefit to players versus cost to complete, and that a new engine would be the only remaining option, which from what you say above sounds like your least desirable option in terms of amount of recoding that would be required.

    Is this a fair assessment of what has been happening or am I barking up the wrong tree?

  7. #37
    Ascendant Samuraiko's Avatar
    Join Date
    Feb 2011
    Location
    Filming on the Internet
    Posts
    2,355

    Default

    Okay, I have a legit reason then to hold off on making more videos in RIFT if I can finally get smoother graphics. I will spend the interim scripting stuff and plotting so when this drops, I'll be ready.

    TALES OF THE RIFTWALKER is on YouTube!!
    Part IV - Marked by Fate is now live!

    Subscribe to my YouTube channel and stay tuned for chances to win stuff via the Trion Worlds Creators Program!
    ** The Forum Drinking Game (to be played during server downtimes) **

  8. #38
    Prophet of Telara MisterWibble's Avatar
    Join Date
    Mar 2011
    Posts
    1,108

    Default

    1% overhead for the rift debug tooling. That's not bad ;)

    Will I be able to restrict Rift to a certain number of cores (or even more fun, assign specific cores to Rift)? While it's never that awesome to tell an OS what to do with the workload, I kinda want to make sure there'll be sufficient room for OBS to do its job when I'm streaming.

  9.   Click here to go to the next Rift Team post in this thread.   #39
    Rift Team
    Join Date
    Jun 2014
    Posts
    677

    Default

    This is a good idea, I've passed it on, hopefully we can squeeze it in.

    Quote Originally Posted by MisterWibble View Post
    1% overhead for the rift debug tooling. That's not bad ;)

    Will I be able to restrict Rift to a certain number of cores (or even more fun, assign specific cores to Rift)? While it's never that awesome to tell an OS what to do with the workload, I kinda want to make sure there'll be sufficient room for OBS to do its job when I'm streaming.

  10. #40
    Plane Touched
    Join Date
    Feb 2011
    Posts
    261

    Default

    Completely serious question, and I think it might have been asked before: Would F2P players be granted the ability to use multicore support / 64-bit clients?

  11.   Click here to go to the next Rift Team post in this thread.   #41
    Rift Team
    Join Date
    Jun 2014
    Posts
    677

    Default

    It's going to be for everyone.

    Quote Originally Posted by Juxtaposed View Post
    Completely serious question, and I think it might have been asked before: Would F2P players be granted the ability to use multicore support / 64-bit clients?

  12. #42
    Fia
    Fia is offline
    Champion of Telara Fia's Avatar
    Join Date
    Apr 2012
    Posts
    1,348

    Default

    If multicore support would get me around 10fps more while infight in 20man raids, that would make a world of a difference to me! * - * I really hope we can test this soon! :D

    Can't wait to see the increase when 64-bit hits and further improvements are made over time. Being finally able to raid with smooth fps, yaaay! * - *
    Fiia@Typhiria | Asthea@Hailol
    Cleric-Heal/Tank/DD | Lvl70

    Now with new Prime Content! -> http://youtube.com/astheatv


  13.   Click here to go to the next Rift Team post in this thread.   #43
    Rift Team Gingers's Avatar
    Join Date
    May 2013
    Posts
    668

    Default

    Quote Originally Posted by AtoningUnifex View Post
    1. I realise you can't speak about specifics but I have a professional interest in the area of project management and had visions of 50000 line gantt charts, and some Devs sitting with work packages literally up to their ears. :P
    I can give you some insight into this. (Yay someone interested in the organizing side of things!!!) When I first became the producer on RIFT, I did make a Gantt chart in Project with every task and it's status and I quickly realized "This will not work for us." Using a scheduling chart that is too detailed, that actually tracks every individual task would be too time consuming to maintain for a team so large when we need to remain flexible. I quickly scrapped that.

    Instead each team lead makes their team schedules showing estimated blocks of time for features rather than every single task. Then I take the team charts and put them all together into one. We do have individual tasks, for example a large feature might be 10-20 smaller tasks. But these larger tech features such as multi-threading and 64 bit, we prefer to give our engineers the goal of "accomplish X" and we trust our engineers to figure out the best way to get to X and how long they think it will take. The larger the project (such as these), the more flexibility there is in the schedule which is why we're so reluctant to give an exact date.

    In terms of huge amounts of individual tasks, I'd say the Wardrobe revamp, making a new Calling and especially going free-to-play had the largest amount of tasks. But Snedhepl is really good at not overloading the engineers so no need to worry there!

  14. #44
    Rift Disciple
    Join Date
    Apr 2014
    Posts
    156

    Default

    Quote Originally Posted by Snedhepl View Post
    It's going to be for everyone.
    But would there be an option to pay for "early access" or patrons receiving access first? That would have been a better option than screwing over non-patrons to make up for costs imo.
    Oh and fyi I've been patron for quite some time now so I'm not complaining because they affect me directly but will affect me indirectly through people quitting.
    It just seems more logical to make up for costs in the same area.

  15. #45
    Plane Walker AtoningUnifex's Avatar
    Join Date
    Jan 2011
    Posts
    491

    Default

    Quote Originally Posted by Gingers View Post
    I can give you some insight into this. (Yay someone interested in the organizing side of things!!!) When I first became the producer on RIFT, I did make a Gantt chart in Project with every task and it's status and I quickly realized "This will not work for us." Using a scheduling chart that is too detailed, that actually tracks every individual task would be too time consuming to maintain for a team so large when we need to remain flexible. I quickly scrapped that.

    Instead each team lead makes their team schedules showing estimated blocks of time for features rather than every single task. Then I take the team charts and put them all together into one. We do have individual tasks, for example a large feature might be 10-20 smaller tasks. But these larger tech features such as multi-threading and 64 bit, we prefer to give our engineers the goal of "accomplish X" and we trust our engineers to figure out the best way to get to X and how long they think it will take. The larger the project (such as these), the more flexibility there is in the schedule which is why we're so reluctant to give an exact date.

    In terms of huge amounts of individual tasks, I'd say the Wardrobe revamp, making a new Calling and especially going free-to-play had the largest amount of tasks. But Snedhepl is really good at not overloading the engineers so no need to worry there!
    Thanks you for the reply, it is what I was expecting, at a certain point, maintaining the documentation becomes a project in its own right...or at least so time consuming that there is no time to do anything other than maintain documentation.

    I work in eHealth in the UK and we use project extensively but use a very similar programme management approach so that we have that macro-management view and you can click through down to individual teams and their projects and then from there to stages or tasks if necessary in case we feel the need to do a bit of micro-management.

    It sounds like you have an excellent team and being able to have that trust in them so that you can "point and shoot" them at a task is actually really good news for everyone that plays this game.

    One last question if that's okay? Do you maintain a project office to maintain documentation, prepare test plans, version control, monitor risks and issues and to help to revise plans and schedules where necessary? Or is this again done within the teams.

    Sorry to be so nosey but I love trying to see parallels between very different industries and their use of project management.

    Thanks again

+ Reply to Thread
Page 3 of 5 FirstFirst 1 2 3 4 5 LastLast

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts