Zwift Chat

[IrishAl]

Great work alan really. The base plate in particular looks absolutely flawless. Did you / why did you cut the holes into it?

Cheers fellas. I figured cutting the holes out would save a bit of weight - not that I plan on moving it around too often. But it also did come in useful to help with assembly although that would still have been possible without.

I cant see, turbo feet blocks. Beware of full power usage, you could fall off the whole board- dont ask

None fitted Andy...I was thinking the weight of me plus the turbo would make it unlikely to shift in use but I’m guessing you’ve experienced some fun and games there 😄

Fore aft movement im guessing is around 50 to 60mm travel though I’ve yet to put any pressure on it to date. I’ve used a 5inch inflatable ball either side for the rocker cushioning.

As for going into business making them I’ve got a spare chrome plated rod as the steel stockist would only sell me a 9ft length but apart from that I’d say it would be a long wait for anyone wanting one as I’m not the most proficient with the tools and the paint was cheap as so it took about 4 or 5 coats to get the coverage right 😴 I’d say I’ll be able to sell on the rod to someone local.

 

[CXRAndy]

I'm on the 5pm TDZ event 1 today, if anyone wants to wave me off. Wish me luck


Note to self, login early, do all updates

 

[bobinski]

I'm on the 5pm TDZ event 1 today, if anyone wants to wave me off. Wish me luck


Note to self, login early, do all updates

Good luck cuddles 👍😀

 

[Whorty]

I'm on the 5pm TDZ event 1 today, if anyone wants to wave me off. Wish me luck


Note to self, login early, do all updates

Which cat you signed up in?

 

[JuhaL]

I'm on the 5pm TDZ event 1 today, if anyone wants to wave me off. Wish me luck


Note to self, login early, do all updates

I will do my stage 1 too. I ride Greater London Flat.

 

[CXRAndy]

Which cat you signed up in?

I will do my stage 1 too. I ride Greater London Flat.

Im Going A multi category. Im expecting 1hr 30 to complete. Im probably lower category D really

 

[mjd1988]

Queue on zwiftpower growing every day...

With TDZ adding quite a lot more results, unless they get a miraculous fix I can't see how wtrl are going to filter out out of cat racers for the ZRL?

 

[JuhaL]

Im Going A multi category. Im expecting 1hr 30 to complete. Im probably lower category D really

So do i. Now there is over 5600 participants, there suppose to be enough draft. Before that i will do Ascenders group ride. Anyway i try to ride easier that TdZ.

 

[bobinski]

Queue on zwiftpower growing every day...

With TDZ adding quite a lot more results, unless they get a miraculous fix I can't see how wtrl are going to filter out out of cat racers for the ZRL?

It really is absurd;

 

[Aleman]

It really is absurd;
View attachment 567220

To be fair, without seeing the code, and perhaps most importantly the data structures, what Flint has posted may very well actually be correct. We have several ETL jobs (Extraction, Translation, Load) that run overnight, and I am repeatedly asked how they can be made quicker ... My LAST response is always "Throw more hardware at it!!", more normally it comes down to redesign the database(s), and refactor the query code. If the core code has been written to process data / files in a sequential manner, then it's probably going to be quicker to junk the lot of it, and refactor the design to handle multiple files in parallel, which would also fit in really nicely with an AWS/kubernetes approach.

Having said that, "GET IT DONE, IT'S REALLY STARTING TO P!SS ME OFF!!!"

Planning on doing the TdZ C Cat tonight at 19:00, B Cat on Wednesday, and A Cat on Thursday

 

[BurningLegs]

If the core code has been written to process data / files in a sequential manner, then it's probably going to be quicker to junk the lot of it, and refactor the design to handle multiple files in parallel, which would also fit in really nicely with an AWS/kubernetes approach.

Tweak it so that every metric processed is a Lambda function at AWS, then never worry about scaling again. If .fit files are involved then it would be as simple as dumping the .fit files into an S3 bucket, moving the processing code to Lambda and then sit back while Lambda does the rest! If we see this miraculously fixed overnight then I would place a sportsman's bet that this is how it was done.

The BBC just moved their news site to Lambda, so every single page load is a lambda function. Seems a bit odd to me when most pages probably only need updating once every few minutes at most, but I'm sure they have a particular reason to essentially re-write the entire page for every single request.

 

[Whorty]

Queue on zwiftpower growing every day...

With TDZ adding quite a lot more results, unless they get a miraculous fix I can't see how wtrl are going to filter out out of cat racers for the ZRL?

I still have a race from last Wednesday that's not been processed yet. Very poor. Queue now up to 79,000 and, as you say, growing.

 

[Whorty]

To be fair, without seeing the code, and perhaps most importantly the data structures, what Flint has posted may very well actually be correct. We have several ETL jobs (Extraction, Translation, Load) that run overnight, and I am repeatedly asked how they can be made quicker ... My LAST response is always "Throw more hardware at it!!", more normally it comes down to redesign the database(s), and refactor the query code. If the core code has been written to process data / files in a sequential manner, then it's probably going to be quicker to junk the lot of it, and refactor the design to handle multiple files in parallel, which would also fit in really nicely with an AWS/kubernetes approach.

Having said that, "GET IT DONE, IT'S REALLY STARTING TO P!SS ME OFF!!!"

Planning on doing the TdZ C Cat tonight at 19:00, B Cat on Wednesday, and A Cat on Thursday

I'm like you, done a lot of ETL work, and the queue of 79,000 there at the moment doesn't 'feel' too big. I'd have processes running millions of lines of data daily. The Zwift logic to run these processes must be pretty flawed somehow if they can't process a few thousand a day - so yes, I'm probably with you on the need to re-write what they have ... but then us consultants always come up with that solution when we join a project

It is very frustrating but hopefully the Zwift team are now realising how key ZP is to their racing platform.

 

[bobinski]

To be fair, without seeing the code, and perhaps most importantly the data structures, what Flint has posted may very well actually be correct. We have several ETL jobs (Extraction, Translation, Load) that run overnight, and I am repeatedly asked how they can be made quicker ... My LAST response is always "Throw more hardware at it!!", more normally it comes down to redesign the database(s), and refactor the query code. If the core code has been written to process data / files in a sequential manner, then it's probably going to be quicker to junk the lot of it, and refactor the design to handle multiple files in parallel, which would also fit in really nicely with an AWS/kubernetes approach.

Having said that, "GET IT DONE, IT'S REALLY STARTING TO P!SS ME OFF!!!"

Planning on doing the TdZ C Cat tonight at 19:00, B Cat on Wednesday, and A Cat on Thursday

😁
2 men and their cat were doing it before so if nothing else they could have carried on doing in the same way pending the lambada or whatever it is 😁

 

[<Tommy>]

😁
2 men and their cat were doing it before so if nothing else they could have carried on doing in the same way pending the lambada or whatever it is 😁

More of a Hokey Cokey man myself...