Spending 30 minutes in Escalation Protocol just to realize that your victory wasn't registered is just but infuriating. I understand, what you were trying to achieve but, why punish people like that? (bungieplz)
First time helping out with protocol writing. They want to do subgroup analysis with their test to show that it doesn't perform especially poorly with certain sub-groups (gender, race, age, several others).
We all know subgroup analysis is poor practice when trying to see where a test or therapy performs well, so I'm a bit concerned about plans to do subgroup analysis to show that things don't perform poorly. It's entirely possible that the test will perform "significantly worse" (or better) for one of those groups completely due to chance. Should/can I mention that we will do an alpha/p correction where p = # of subgroups to account for multiple testing?
I’ve been following the project and team members since day 1. I’m sure Alex would recognise me here as we had few exchanges last year when I tried to join the team :)
As a community member of a decentralized project where governance is still not fully implemented, I was thinking that a good initiative would be to regroup individual concerns, feedbacks & improvement proposals with arguments of course. Topics could be discussed with the team and possibly result lead in positive unplanned actions.
It is clear that we don’t have all cards in hands as we don’t have access to any insiders info that would push the team to behave in such or such way, but the worst this can do is to let the team know about how the community is understanding/seeing the project which is a good indicator of their communication success.
As an ex startup founder, I believe that it is always extremely positive to get constructive positive & negative remarks which could make the team see things from other angles especially coming from people who care.
What in your opinion could be improved or done in a different way in the protocol life at this stage?
Basically, it's a state-less wifi without all the authentication and access points. You add (up to 20) peers via their MAC address and then you can send single messages with up to 250 byte payload to those, without any need to connect to anything first.
I've tested it on two ESP32 (the Wemos LOLIN32 and D32), but it also works with the older ESP8266.
Here's the official documentation: https://docs.espressif.com/projects/esp-idf/en/latest/api-reference/network/esp_now.html
Examples are already in the Arduino IDE if you've added ESP support, under Examples -> ESP32 -> ESPNow
Pros: Cheap, easy to set up, high range, quick messaging (so for example, you can keep your ESP in deep sleep, and send a message on an event, and go back to sleep. I've read online that people manage all that in <50ms).
Cons:
- Limited data transfer rate. I've managed 70kbyte / sec (up to 1000 packets per second). UDP over Wifi can go more than 1 Mbyte / sec.
- Can't use wifi while using ESP-NOW. So if you want to have a node that talks to the internet and to other ESP-NOW nodes, it'd be tricky.
- No error checking. You'll need to test manually if messages have been received. There's an easy call-back if the message was received on MAC level (by the wifi antenna), but if you want to make sure the other ESP has actually processed it, you'll need to write your own code that sends back an ok.
- My esp32 heats up a lot when I continuously send messages, over 60C. Don't think it's an issue, but I guess it's using a lot of power to become so hot.
Hope that's useful for some of you!
For those of you less familiar- we will be using enterprise grade point to point networking hardware and technology to build out and distribute direct access internet to consumers in our market. We will be building our network on structures like cell phone and water towers. Consumers will be able to purchase our product and use a modem in their home to connect directly to our towers or use an external professionally installed exterior unit.
Edit* removed name of a tower company to prevent it from seeming like we are pointing people at a business. this is not an ad.
immediately? Is it at my discretion and I do it, or don't, to make a point? Let's assume that it is possible for me to receive field promotions as appropriate -- we have subspace comms so maybe I get made LT Cmdr or Commander. Does rank affect the uniform etiquette?
Thanks.
... if Brood Queens can instant kill Swann's Siege Tanks with the Brood Queen "Spawn Broodling" ability, bypassing the whole idea of Swann's Immortal Protocol?
Immortal Protocol does not work against the one ability that causes Siege Tanks to die? That seem a bit unfair and unnecessary considering few Swann's pick Immortal Protocol mastery over Structure Health mastery.
😡
#edit: A presentation of the Brood Queen ability: https://i.imgur.com/sRFBXS9.jpg
Tripio comprises of its dApps, as well as a set of Tripio protocols that forms the basis of a new ecosystem. One of these Tripio protocols, Tripio Points, has been developed. We will launch it with some dApps that leverage new protocol in the near future.
So i have a few bulbs connected to my Conbee, using deCONZ, HA and Node-RED
Do i group my lights in deCONZ and call that group from HA/NR, or do I group in HA only.
Ultimately i'll be using Node-RED, and I could "group" using that too I guess as a third unlikely option.
What is most efficient in terms of how lights are called and the network, is there any difference? My gut instinct is deCONZ is a little closer to the metal, so is probably better the group there, and call elsewhere. But it is a whole lot easier and more flexible to just expose lots of hardware and group in HA, and use Node-RED that way.
Anyone have any experience of large groups and setups? As mine grows i'd rather not make a decision now i regret!
...and my conclusion is that most judges suuuuuck.
Non-shitpost version of this post in a week or two once I'm done with Project Judges Suck version 3.0.
In the meantime does anyone have a way of standardizing judges' scores so they are comparable across competitors, segments, and disciplines? Currently I'm converting each judge's score to a z-score based on the mean and standard deviation of the scores for each competitor. Eg. suppose the scores of 9 judges for skater X's free skate average to 180 with a standard deviation of 5. Judge A gives a score of 185. I record a z score of 1. Not sure if this is the best way to do it, but I really need some way of standardizing scores cuz not having that was what caused problems the last time I did this.
the modal difference between how judges score their own skaters and how they score other skaters is like 1 full standard deviation lmao
