Jump to content

Od1n

Clubhouse Member
  • Content Count

    653
  • Joined

  • Last visited

  • Days Won

    4

Reputation Activity

  1. Like
    Od1n reacted to Netduma Admin in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    We are making a lot of improvements to QoS - we will be revealing more on this soon. This feature won't be out in the 3.0 release, but we expect it to be the next major release. More on this soon.
  2. Like
    Od1n got a reaction from Shmicle0909 in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  3. Like
    Od1n got a reaction from Squizz in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  4. Like
    Od1n got a reaction from Phantomblu in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  5. Like
    Od1n got a reaction from Tripper in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  6. Like
    Od1n got a reaction from Locosano in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  7. Like
    Od1n got a reaction from Kaks07 in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  8. Like
    Od1n got a reaction from UK_Wildcats_Fans in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  9. Like
    Od1n got a reaction from fakkel in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  10. Like
    Od1n reacted to Netduma Fraser in NEW FEATURE ANNOUNCEMENT: GEO-FENCING   
    With more time freed up after release we should be able to sort out more cloud updates when they're needed.
  11. Like
    Od1n got a reaction from WalkedDave in NEW FEATURE ANNOUNCEMENT: CONNECTION BENCHMARK   
    When will we be able to change the congestion algorithm that the R1 uses? In the last few seconds of this Netduma announcement video you can see the download graph and it looks like Cubic Congestion Control is used. This is sub-optimal for gaming since it doesnt consider roundtrip time or other metrics that are relevant for gaming devices in your network. And using Cubic is probably also the reason for why we are forced to set our connections to 70% to prevent congestion.
    Ages ago Iain claimed to have created a congestion control algorithm that in his lab allowed him to max out his complete connection while his ping only spiked by 5 milliseconds. Googles BBR Congestion Control is very close to what he promissed, is out for several years now and could probably be implemented very quickly by you guys.
     
    For those who dont fully understand what i mean here is a picture from google:
    In the last few seconds of the Netduma Video above you can see the bandwidth graph which basically looks like left graph below in the picture. This is bad and causes lag. What you want is the right graph it gives more bandwidth AND less lag. It has some downsides but those could be adressed by the R1 only assigning BBR to gaming devices for example.

  12. Like
    Od1n got a reaction from kymadn in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    Even if things are still confidential you can give out monthly updates that contain informations like
    - very persistant bug has been found that gave us troubles for weeks now
    - new feature we cannot unveil yet is currently being tested, things are looking very good
    - for next month we want to refine our congestion interface and discuss new features
    All it requires is a 5-10 minutes meeting once a month with the netduma staff to define the boundaries of what will be shared and what not. It really isnt a huge effort but will greatly please your community. You will get much less ETA questions that way. Overall it might require 1 hour per month to pull this off, and its not really hard to keeping certain things confidential while still sharing the general development progress.
     
  13. Like
    Od1n got a reaction from Sunaikinti in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    But does that vision go hand in hand with what the customer wants?
    It took 3 years to come up with DumaOS which is bascially just a nicer looking modular OS but with pretty much 0 new features. The post release roadmap was to deliver modular feature add-ons in short succession. But that has long been droped and now (for whatever reason) everything has to be bundled into one new DumaOS 3.0 which totally works against the initial modular approach. So why again have we waited 3 years for DumaOS then?! I start to get the feeling that the main driving factor for DumaOS was Netgear and that there MUST be a reason for why those "over-promissed" features havent been released individually yet. The original R1 OS may have had some expansion bottlenecks, but at least it was WAY more responsive and easier on the CPU. I mean the new DumaOS isnt even mobile compatible, and its 2020!
    And to come back to my first sentence about what the customer wants, i could care less about another overhaul of the netduma interface or its traffic monitor. The R1 is a gaming router and not a device for network engineers who need to track down traffic sources. So why would a gamer need a traffic history in the first place? People who need that buy different routers with far more monitor capabilities. The idea of the R1 is to activate congestion control and forget about lag and all the complicated stuff.
    And talking about congestion control, last year some really impressive progress has been made in terms of congestion control. And it will not take that much longer until you can natively use those with Linux. So the R1 will soon get some severe and free competition that doesnt require you to reduce your bandwith to 70%. Im curious to see who is faster with the release, but deep down i already know the answer. Plus im not even sure if the R1 congestion control can beat Googles congestion control from 2016, Iain didnt include it in the benchmarks from 4 years ago.
    Netgear isnt stupid, when taking the congestion control out of the equation there isnt all that much that separates the R1 from the competition. Geo filtering becomes less and less important as more and more games use dedi servers plus dont enforce world wide lobbies. And the VPN function kills the R1 CPU plus most competitive shooters straight up ban you for VPN-ing anyway.
    After 4 years you guys better have some REALLY good features in the oven. Luckily it is no longer our job as the community to make pressure but for the first time the competition will now do that. And maybe you will listen more to what they have to say than what we had to say over the last 4 years...
  14. Like
    Od1n got a reaction from AnnonymouS in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    But does that vision go hand in hand with what the customer wants?
    It took 3 years to come up with DumaOS which is bascially just a nicer looking modular OS but with pretty much 0 new features. The post release roadmap was to deliver modular feature add-ons in short succession. But that has long been droped and now (for whatever reason) everything has to be bundled into one new DumaOS 3.0 which totally works against the initial modular approach. So why again have we waited 3 years for DumaOS then?! I start to get the feeling that the main driving factor for DumaOS was Netgear and that there MUST be a reason for why those "over-promissed" features havent been released individually yet. The original R1 OS may have had some expansion bottlenecks, but at least it was WAY more responsive and easier on the CPU. I mean the new DumaOS isnt even mobile compatible, and its 2020!
    And to come back to my first sentence about what the customer wants, i could care less about another overhaul of the netduma interface or its traffic monitor. The R1 is a gaming router and not a device for network engineers who need to track down traffic sources. So why would a gamer need a traffic history in the first place? People who need that buy different routers with far more monitor capabilities. The idea of the R1 is to activate congestion control and forget about lag and all the complicated stuff.
    And talking about congestion control, last year some really impressive progress has been made in terms of congestion control. And it will not take that much longer until you can natively use those with Linux. So the R1 will soon get some severe and free competition that doesnt require you to reduce your bandwith to 70%. Im curious to see who is faster with the release, but deep down i already know the answer. Plus im not even sure if the R1 congestion control can beat Googles congestion control from 2016, Iain didnt include it in the benchmarks from 4 years ago.
    Netgear isnt stupid, when taking the congestion control out of the equation there isnt all that much that separates the R1 from the competition. Geo filtering becomes less and less important as more and more games use dedi servers plus dont enforce world wide lobbies. And the VPN function kills the R1 CPU plus most competitive shooters straight up ban you for VPN-ing anyway.
    After 4 years you guys better have some REALLY good features in the oven. Luckily it is no longer our job as the community to make pressure but for the first time the competition will now do that. And maybe you will listen more to what they have to say than what we had to say over the last 4 years...
  15. Like
    Od1n got a reaction from slickrick6922 in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    But does that vision go hand in hand with what the customer wants?
    It took 3 years to come up with DumaOS which is bascially just a nicer looking modular OS but with pretty much 0 new features. The post release roadmap was to deliver modular feature add-ons in short succession. But that has long been droped and now (for whatever reason) everything has to be bundled into one new DumaOS 3.0 which totally works against the initial modular approach. So why again have we waited 3 years for DumaOS then?! I start to get the feeling that the main driving factor for DumaOS was Netgear and that there MUST be a reason for why those "over-promissed" features havent been released individually yet. The original R1 OS may have had some expansion bottlenecks, but at least it was WAY more responsive and easier on the CPU. I mean the new DumaOS isnt even mobile compatible, and its 2020!
    And to come back to my first sentence about what the customer wants, i could care less about another overhaul of the netduma interface or its traffic monitor. The R1 is a gaming router and not a device for network engineers who need to track down traffic sources. So why would a gamer need a traffic history in the first place? People who need that buy different routers with far more monitor capabilities. The idea of the R1 is to activate congestion control and forget about lag and all the complicated stuff.
    And talking about congestion control, last year some really impressive progress has been made in terms of congestion control. And it will not take that much longer until you can natively use those with Linux. So the R1 will soon get some severe and free competition that doesnt require you to reduce your bandwith to 70%. Im curious to see who is faster with the release, but deep down i already know the answer. Plus im not even sure if the R1 congestion control can beat Googles congestion control from 2016, Iain didnt include it in the benchmarks from 4 years ago.
    Netgear isnt stupid, when taking the congestion control out of the equation there isnt all that much that separates the R1 from the competition. Geo filtering becomes less and less important as more and more games use dedi servers plus dont enforce world wide lobbies. And the VPN function kills the R1 CPU plus most competitive shooters straight up ban you for VPN-ing anyway.
    After 4 years you guys better have some REALLY good features in the oven. Luckily it is no longer our job as the community to make pressure but for the first time the competition will now do that. And maybe you will listen more to what they have to say than what we had to say over the last 4 years...
  16. Like
    Od1n got a reaction from santa78 in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    Thats nice but you should still consider to upgrade your communication channels. The biggest backlash that netduma received wasnt about the over-promissing or missed release dates but about giving only 2 statements within 3 years. Everyone was completely left in the dark!
    I used to regularly read and post in this forum when Iain and Luke used to post here too and shared their thoughts and development progress (regardless of it they made progress or not). Nowadays i read the forum maybe once every 5-6 month as i can be certain not a single announcement on the progess has been shared since my last visit.
    This could easily be changed by giving out monthly update announcements...
     
    Also curious to see if we will ever get those over-promissed features like anti-spike or anti-jitter, last time we heard about those was like 2+ years ago...
    Iain showed lab results of +5ms spike with 100% maxed out up- and download...
  17. Like
    Od1n got a reaction from raf1212 in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    But does that vision go hand in hand with what the customer wants?
    It took 3 years to come up with DumaOS which is bascially just a nicer looking modular OS but with pretty much 0 new features. The post release roadmap was to deliver modular feature add-ons in short succession. But that has long been droped and now (for whatever reason) everything has to be bundled into one new DumaOS 3.0 which totally works against the initial modular approach. So why again have we waited 3 years for DumaOS then?! I start to get the feeling that the main driving factor for DumaOS was Netgear and that there MUST be a reason for why those "over-promissed" features havent been released individually yet. The original R1 OS may have had some expansion bottlenecks, but at least it was WAY more responsive and easier on the CPU. I mean the new DumaOS isnt even mobile compatible, and its 2020!
    And to come back to my first sentence about what the customer wants, i could care less about another overhaul of the netduma interface or its traffic monitor. The R1 is a gaming router and not a device for network engineers who need to track down traffic sources. So why would a gamer need a traffic history in the first place? People who need that buy different routers with far more monitor capabilities. The idea of the R1 is to activate congestion control and forget about lag and all the complicated stuff.
    And talking about congestion control, last year some really impressive progress has been made in terms of congestion control. And it will not take that much longer until you can natively use those with Linux. So the R1 will soon get some severe and free competition that doesnt require you to reduce your bandwith to 70%. Im curious to see who is faster with the release, but deep down i already know the answer. Plus im not even sure if the R1 congestion control can beat Googles congestion control from 2016, Iain didnt include it in the benchmarks from 4 years ago.
    Netgear isnt stupid, when taking the congestion control out of the equation there isnt all that much that separates the R1 from the competition. Geo filtering becomes less and less important as more and more games use dedi servers plus dont enforce world wide lobbies. And the VPN function kills the R1 CPU plus most competitive shooters straight up ban you for VPN-ing anyway.
    After 4 years you guys better have some REALLY good features in the oven. Luckily it is no longer our job as the community to make pressure but for the first time the competition will now do that. And maybe you will listen more to what they have to say than what we had to say over the last 4 years...
  18. Thanks
    Od1n reacted to AnnonymouS in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    If so, why have we not seen anything come out for a long time. I have been here on this forum from the beginning of the R1 and you know that. I wished you guys would have stayed on your own instead of teaming up with Netgear . That way you would be in control to do want you want to do. All you guys needed was to have your own line of routers and more staff to carry it out.     
  19. Haha
    Od1n got a reaction from raf1212 in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    So you play with a 200€ controller that has features such as super precise analog sticks, trigger stop and paddles giving you a HUGE advantage over regular controller players yet you complain about XIM users?
    Guess you re one of the people who always find something to complain about unless they have ALL the advantages for themselfs.
     
    And sure go buy the XIM, youd not be the first to realize that a mouse doesnt turn you into a Shroud. Just like with a controller you have to grind for years to become good.
  20. Like
    Od1n got a reaction from dpk_ydv in ANNOUNCEMENT: The Future of DumaOS: Version 3.0   
    So you play with a 200€ controller that has features such as super precise analog sticks, trigger stop and paddles giving you a HUGE advantage over regular controller players yet you complain about XIM users?
    Guess you re one of the people who always find something to complain about unless they have ALL the advantages for themselfs.
     
    And sure go buy the XIM, youd not be the first to realize that a mouse doesnt turn you into a Shroud. Just like with a controller you have to grind for years to become good.
  21. Like
    Od1n reacted to Netduma Jack in R1 DHCP instead of PPPoE   
    It doesn't necessarily matter if the R1 gets a new address occasionally, though it could mean you need to keep accessing the UI through different IP's. I'd just give it a static one
  22. Like
    Od1n got a reaction from Netduma Jack in R1 DHCP instead of PPPoE   
    Notification for those also reading this topic:
    DCHP reduced my R1 CPU load from PPPoE @ 70-90% on bandwidth load to DHCP @ 30-50% on bandwidth load
    Connection is currently 50 down 10 up so not even that fast. Overall a huge difference, when i switch to 100 down 40 up in the next month or two im quite confident that the R1 will deliver those speeds without any problems :).
  23. Like
    Od1n got a reaction from Netduma Fraser in NEW FIRMWARE OUT (Netduma R1): DumaOS Development Milestone 1.3   
    Thanks Fraser, then i still seem to be on the latest version. As i mentioned earlier this has been the most stable firmware i have had so far (early bird r1 user), you can tell Iain that he did a great job!
  24. Like
    Od1n reacted to Netduma Fraser in NEW FIRMWARE OUT (Netduma R1): DumaOS Development Milestone 1.3   
    Unfortunately not, there have only been two versions of DumaOS on the R1 .10 & .29 with the latter being the 1.3 milestone. We will try to be more detailed when we release the next version on what has changed.
  25. Like
    Od1n got a reaction from Netduma Alex in NEW FIRMWARE OUT (Netduma R1): DumaOS Development Milestone 1.3   
    Are there any patchnotes that you can read what the latest DumaOS fixed over the last one? Not refering to the 1.3 version but the very latest one that i think is discussed here in the more recent comments.
    Still using the second firmware release version i believe and im really happy with it, no crashes or anything. So far the most stable netduma firmware that i had. Though its of course interesting to see what has been added since then ;).
×
×
  • Create New...