Hacker News new | past | comments | ask | show | jobs | submit | eloycoto's comments login

I need to say that I'm very impressed with the PDL project, a lot of things can be done in there.

https://github.com/IBM/prompt-declaration-language


I have something like this, and I'm super happy with anythingLLM, which allows me to add a custom board with my workspaces, RAG, etc.. I love it!


100%, Embassy is great and I'm in love with it. If you add the PIO interface for RP2040 the setup makes code super simple and beatiful and difficult to achieve with another processor.


Nowadays, some riders use glucose monitoring devices during training with relation to how long it takes from ingesting and getting into blood. They are now measuring more and more things, and food is a crucial role.

Glucose monitoring is banned during races, but I guess that is why you don't see some cyclist collapsing like as 15 years ago is because they understand their glucose levels.

For sure, doping it's still a fear, but these guys cannot climb Galibier at that speed only with storoids/drugs, it's insane the effort that they did over the year.


I haven't been a cyclist since about the time that the market of energy drinks first exploded. And at the time there were a number of articles that said that really all you need is Gatorade, that most of these new drinks were advertising more calories per liter, but Gatorade was already tested out on absorption rates and dialed in to the maximum calories per liter of water that the average human stomach can absorb.

The tricky word there is average. If Tadej can absorb an extra 5 grams per liter, then you should give Tadej an extra five grams per liter. If Jonas can absorb 2 grams less per liter than the average then you should give him that serving.

These race results come down to mere seconds per hundred miles, for cyclists that are averaging 20 miles an hour. Any 1% difference is going into the training regimen.


I still use Gatorade because I'm a cheap bastard.

The limits aren't wrt water absorption, but gut tolerance of sugar.

I get about 100g of sugar into a bottle by doing roughly 50:50 Gatorade and maltodextrin, and then throwing in some extra Na an K salt if it's going to be hot. Although with the super hi carb stuff you should make sure that you have some plain water as well - it sucks to be super thirsty but only have carb drink on the bike.


Sibling comment explains about the maltodextrin. I'll have to read up more.


that is outdated advice. for one, some people sweat more than others, the salt levels in the sweat is totally different too. this means that if you're really serious about sport or are doing endurance races (ie 2 hours+) you should really not just use gatorade, but something where the mix of salt/water is closer to whatever you're sweating.

if i personally use gatorade for endurance exercise, i'll just cramp up after 90 minutes and not be able to ride normally. if i use a high salt mix instead, this isn't an issue whatsoever. I'm sure the exact same is true about food itself. remember that for the tour de france, last years time difference between the winner (Vingegaard) and the second (Pogadcar) was 8 minutes out of 82h 05' 42", aka only 0.16% faster overall. every single sub-percentage matters here.

there are tons of products that cater to this. the one i've been using is https://www.precisionhydration.com/ which is cheaper and more tailored than gatorade (i have no affiliation to them).


I basically just add salt to Gatorade and cut it with extra sugar and malto. It's a couple extra steps but I train regularly enough that I make a concentrate (100ml=200kcal) that I put in bottle and dilute as needed.


One of the runners I mined for training advice was using gatorade to take salt tabs. That was not the advice I took from him, though. I will probably avoid at least one injury from other things he said.


> that is outdated advice. for one, some people sweat more than others, the salt levels in the sweat is totally different too.

People forget that. They just follow "salt is killing you" advice, meanwhile I sweat salt like it's water. Last race I did had the sense to add salty potato chips to their rest stops... I didn't cramp up.

Meanwhile Gatorade is absolute trash for electrolyte maintenance for someone like myself.


The people drinking a gallon of water (not fluids, just water) a day are often flushing out electrolytes. There are a lot of things you might expect a doctor to yell at you about in an ER but from what I understand coming in with heart palpitations triggered by low potassium due to over-hydrating is a good one.


Yeah I wasn't talking about electrolytes, just calories per minute. The top level was about glucose monitoring not electrolyte monitoring.

It's funny that Precision uses almost exactly the same bottle as Nuun.


On the Netflix TdF documentary, they mentioned that the difference between the winner and last place is 2%. There was one stage where the winner won by 4 inches.

It also said that doping made a 20% difference, so either none of the current competitors are doping or all of them are.


> so either none of the current competitors are doping or all of them are

The drafting effect in cycling means that a clean cyclist can finish very closely behind a doped cyclist


That only works on the flat stages. A small difference in power to weight ratio makes a huge difference in finishing times on the mountain stages where drafting is less of a factor.


Don't forget the GC rules for assigning times.

from https://en.wikipedia.org/wiki/General_classification:

    Riders who finish in the same group are awarded the same time, with possible subtractions due to time bonuses. Two riders are said to have finished in the same group if the gap between them is less than three seconds. A crash or mechanical incident in the final 3 kilometres of a stage that finishes without a categorised climb usually means that riders thus affected are considered to have finished as part of the group they were with at the 3 km mark, so long as they finish the stage.
For this year's 5th stage of the Tour de France, the first 155 riders were assigned the same time.

see https://www.procyclingstats.com/race/tour-de-france/2024/sta...


And then there’s Pidcock who has the reflexes of a fighter pilot and can scream down the hills at speeds that would make you blanch.


I'm not sure their reflexes are better. It's more like they're the master of the feel and minutia of the airplane. They're also masters of managing the energy of their airplane, constantly trading off between altitude and speed. John Boyd was a fantastic example of that. There's a fascinating biography of him, "Boyd".

https://www.amazon.com/Boyd-Fighter-Pilot-Who-Changed/dp/031...


Do you suppose there are stats on number and success of breakaways out there? Could be my imagination but this year seems less dramatic than the ten years or so around Armstrong’s rein.

Did anyone else feel like there was more drama around the end of the Indurain era?

I feel like doping could show up as more intense struggles for first without necessarily increasing the spread of placing. Also people who drop out don’t count right?


> for cyclists that are averaging 20 miles an hour

I don’t think modern professional cyclists even can ride at 20 miles an hour ;-)

20 miles an hour didn’t win you the Tour de France in 1950. Nowadays, with better tech, better training, and shorter stages, you have to do over 25 miles an hour to win it, and the last rider in the general classification easily is within 10% of that.


The last rider that finished and it’s not like if you or I go out and try to ride a hundred miles at 25 mph solo. Drafting and team dynamics play a significant role is energy expenditure.


I'm not a pro cyclist but I've seen one stop before. By the intermediate value theorem, they can ride at 20mph.


They count it as gram per hour. If I remember well, pro cyclists can absorb 2-2.5 times as much carbohydrate in an hour than beginners.


I wonder how much of that is adaptive and how much selective. Does your body learn to absorb more, or do you have to have a high absorptivity to make it as a pro.


Both. So they adapted themselves to it, they didn’t born like that. It’s an intentional procedure. However, not everybody can handle that.


Maximum calories per liter sounds like something you really want to avoid unless you are engaging in extremely intense exercise.


Your body can burn calories faster than you can absorb them. And if you’re concerned about what sports medicine says the limits of human absorption are, you’re either a nerd or engaging in extremely intense exercise.

The Tour de France did 124miles on its first day this year. Through mountains.

Many amateur cyclists have a 100 mile single day even or 150-300 mile multi day event as their yearly goal. And they can’t push the sort of power for the duration that a pro can. Something less than 75% as much power for less than half as long per day. And for days instead of weeks.


Sustained cycling is intense exercise. Your body burns through its glucose reserves rapidly, and if you run out of energy the effect ("bonking") is awful. On the other hand, trying to eat a bunch of energy and having too much in your stomach will make you sick and is also awful. So you do in fact want to ingest the most you can process (but no more) per hour, particularly for extended events.

Nutrition doesn't matter for rides up to 15 miles, and you can improvise almost anything for rides up to 40 miles. More (or particularly hilly courses, etc.) and it becomes very important to get the amount you eat and drink right.


I was referring to how Gatorade is a popular beverage for people not doing any kind of intense exercise, not realizing it's just as bad or worse than soda for weight gain in that case.


There are multiple kinds of carbohydrate and the differences can matter.

Gatorade’s ingredients list shows sucrose (glucose+fructose) and dextrose (glucose).

Maltodextrin is isotonic in a 6x stronger solution than fructose and glucose. This means you need to drink meaningfully less water during a race to digest maltodextrin. So do we want all maltodextrin? Nope! Fructose can be absorbed by a separate pathway, which is less efficient, but increases your total intake of carbohydrates into the bloodstream.

So you’re probably not going to see Tour de France teams giving their riders Gatorade. You want more maltodextrin and less glucose.


Powdered sugar is 50/50 fructose and glucose, a close to perfect ratio for absorption. A dash of Gatorade powder for taste and you got yourself a drink that's inexpensive and gets you all the carbs you need.


Tour de France teams don’t care about the cost savings from cheaper sugar, they do care about the time and weight effects of drinking more water to absorb the glucose.


I think that goes without saying. My advice is for the layman.


I forgot about the Fructose pathway. Yeah that's been known for ages.

It turns out though that maltodextrin didn't exist in manufactured quantities until the mid 1970's, which was after Gatorade was a national brand. That they never changed their formula is disappointing but maybe not surprising (especially after Quaker Oats bought them in '83)

But none of the 'sports drinks' that Coca Cola and friends were pumping out in the 90's had maltodextrin in them to my knowledge. It was just more sucrose and glucose, so its star fell a bit more recently.

Do you know when the isotonic research was done? I don't have any recollection of that being even mentioned in the tests I talked about.

Also maltodextrins are made from wheat in Europe and a problem for celiacs.


That is indeed why you don´t see them collapsing and also why the finals are much longer nowadays. Riders train their guts to be able to eat up to 120g of carbs per hour, for the whole duration of the race.


I am far FAR from the tour de france but I do train for and run ultramarathons.

When I started the general guidance seemed to be about 200 calories (50g of carbs) per hour during training and races. I followed that or a little above for a long time, and constantly had problems both physically and cognitively near the end of races.

It took about a year to get used to fueling heavier (for me...about 400 calories per hour or 100g). Lots of figuring out digestion and timing and sources and other factors. But once I got there it was transformative...I could push much harder for much longer without a deep bonk and recovery was faster. More importantly, my ability to think/plan/make decisions at the end of races was orders of magnitude better.

I'd bet if you went to an aid station around 80 miles on a 100mile ultra marathon you could pretty reliably identify who had been on the 200cal end of the spectrum and who had been on the 400 + end.


At 250w average during a 5 hour ride I'm burning like 1000 calories per hour or something mething close. Glycogen reserves are 2000 calories from a quick Google search, so that's 3k calories I need to make up for or 600 per hour. I'm sure there's a bunch of stuff I'm missing but 400 definitely makes more sense than 200


What you're missing is fat metabolism. If you're metabolically healthy (which you must be if you can hold 250W for hours) then you can make up the difference by converting stored fat into glycogen.


what is also missing is the ability to process calories in.

900-1000 calories might mean that in is optimal - but think about eating a full meal from mcdonalds every hour while massively extering yourself.

Training your gut matters and I've never seen anything that suggests much beyond 400 calories per hour is possible during meaningful exertion.


Yup, part of my long distance cycle training was somewhat getting "fat adapted".


Right! That was a big missing chunk :)


Since my first ultra, Ive been fascinated by the idea of adapting to increased fueling and have tried to apply periodization principles to how I eat. 23 miles into a 33 race, I completely stopped being able to process food at around 150-250 cals/hr (mostly carbs and some fats)


speaking from experience - I puked during a fair number of runs in the process of getting there. I mix calories in my hydration with gels just to try and get the balance right. Its not stable but its better after about 4 years of work. I have ound that if I don't do a good job of eating, what you describe happens about 6-10 miles AFTER I stopped caloric intake.


Yeah, some of the modern nutrition stuff like gels work really well. I did a 115 mile road bike ride on Sunday and kept eating gels and... I'm not going to say I was super fresh or anything by the end of that, but I was still feeling pretty good.


There is usually a point where you can't get those gel inside. It is nice to mix with other stuff.


After my longest ever bike ride, 300 km, I didn't touch a gel for years, heh. Got so sick of 3+ gels per hour for those ~8 hours.


I dropped gels pretty fast. Hard to get if you're not somewhere with sports stores, and expensive.

These days I cram haribo and peanut butter M&Ms, although I think the latter are more useful for multiday things.


Yeah, I bought a huge bulk pack online, and then it was okay. But when you need a couple a week it quickly adds up in cost. Same with bars, which I sometimes prefer over gels for longer rides to not just have gels in my stomach. I've tried to bake some bars myself, or make some rice cakes from Velochef ( https://www.rouleur.cc/blogs/desire-journal/velochef-henrik-... ) instead.


Thanks for the recipe! Reminds me... A friend of mine made a megagel that was some kind of chocolate concoction that he carried in a 300ml soft flask. It had sugar, salts, caffeine, actually tasted great, and I believe was pretty cheap to make. I have to get the recipe from him.

Edit: I did the math on that risotto cake recipe and it comes out to about 2500kcal for the batch, for me that comes out to calories for ~5-6 hours of riding.


I don't understand how you manage to do 300km only with gels. That would have been atrocious.

I did a similar ride a few years ago, I mixed gels, energy bars with bananas, nuts, sandwichs, nuts and pastries bought along the way.


> For sure, doping it's still a fear, but these guys cannot climb Galibier at that speed only with storoids/drugs, it's insane the effort that they did over the year.

The way I heard it, the don't use doping to improve their performance. They use it so they can train longer and more intensely than they would be able to without it.


That's part of it sure, but anything which improves oxygen transport like EPO or blood transfusions will absolutely improve climbing performance regardless of training quality. For anyone with good fitness the main limiting factor is amount of oxygen delivered to the muscles.


I moved to a Dygma keyboard early in the year, and I'm not looking back. The software to record macros, remap is just great, and I'm no longer suffering pain because of my position.


I hadn't seen this keyboard before, but it looks great. I like the idea of having a lot of keys on the thumb (8 vs the 4 on the Moonlander).


There is also this project, that it's pretty cool to know who is consuming what. https://github.com/sustainable-computing-io/kepler


Thanks for sharing, a nice K8s integration :)


I'm super excited about this!


Woww, this is great, thanks for sharing


And this post didn't mentioned custom functions, where things are getting more interesting :-) https://acalustra.com/jq-functions-modules-and-dotfiles-tips...


At that point why not write a Python script?


jq is a much, much more expressive language than Python at applying similar operations to the list of inputs (with fanout!).


How about combining the best of both worlds: https://github.com/mwilliamson/jq.py


Yeah, jq has some of the benefits that APL users like to talk about.


I'm not disagree with you, but in this case all stored in custom dotfiles, as functions, easy to use IMHO than a custom Python script.

But I use a lot JQ, so it's super easy for me ;-)


Shameless plug, but I created jello[0] to give you the ease of parsing json in bash, but using python syntax[1]

[0]https://github.com/kellyjonbrazil/jello

[1]https://blog.kellybrazil.com/2020/03/25/jello-the-jq-alterna...


This make me super happy, jsonnet is something that should be used by default in Kubernetes.

Hope this project get a lot of traction!


Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: