Who has the fastest website in F1? | Hacker News Reader

Who has the fastest website in F1?

jakearchibald.com

230 pointsanacleto99 comments

Wow, that Netflix documentary is reallllly helping grow interest in F1! Enjoyed the long scroll down to see Haas as the winner. I assume the article controls for the physical location of the request (i.e. is Haas fastest because the author is in the US and Haas is the only US-based team and/or hosted site?), but I didn't see it in my perusal.

They should have a test every week from a different country.

They could have team points as well as manufacturer points depending on the web servers.

> Wow, that Netflix documentary is reallllly helping grow interest in F1!

Based on language and phrasing used on the blog, the author is British, and F1 has been a staple on mainstream terrestrial channels for decades, although it has seen a decline as F1 has increasingly became boring and predictable, and almost entirely dictated by team tactics and car reliability.

I grew up watching Schumi vs Hakkinen, those were the times... It's all over the place now.

Massa, Raikkonen, Montoya, Alonso ... those were the days.

What does that mean it's all over the place now? I also grew up with Schumacher vs Hakkinen, but last season and seemingly this one are amazing.

I'm not sure I'd call the last season "amazing". It had some promise at the start, but it was pretty clear before too long that Hamilton was going to run away with it again.

We'll see what this season holds (since at this time last season it still looked like it would be a good fight to the end), but I'm optimistic. If Bottas can retain his form from Australia (I think the wind really threw him off in Bahrain), and if Vettel can get his head on straight, we could see a legit four-way fight for the title.

Last year the midfield was where it was at.

Even more excited for the midfield battle this year. The rookies (esp Norris) look great, McLaren "can fight", and adding Kimi and Ricciardo to the midfield is sure to shake things up.

Fwiw this year's Formula 2 grid has a Schumacher and an Alesi.

There are many father-son dynasties across Formula 1/2/3/E. Pique, Senna, Prost, Verstappen, Rosberg, just to name a few active right now.

Bruno Senna is Ayrton's nephew, but yeah, lots of repeated family names.

'..those were the times...'

If you weren't around in the days of Senna then you really have missed the best of times.

Second that. Senna, Prost and Mansell had some major battles.

Always felt kind sorry for Mansell as you felt that he is on the verge of getting World Championship but never really got there on a consistent basis.

Yes, he wasn't the most stylish driver (or likeable) but his effort was never in doubt.

Coulthard kinda trashed Mansell's effort compared to Senna in this clip:

https://youtu.be/9U_K76vPGYo?t=329

I started getting in to F1 at the very tail end of the 80s and in to the early 90s. It was something else.

Although I rarely watch motorsports these days, I'm much more inclined to watch something like MotoGP than F1.

I watched a lot of F1 in the 80's and 90's. Then stopped for various reasons (around the time that Senna died on the track). But lately have started watching again now that my sons are taking an interest.

I fully appreciate that the heady, noisy, dangerous days of F1 are over, as driver & fan safety, as well as hybrid engines take over the sport. But still, exciting wheel to wheel racing can happen, as evidenced by the race in Bahrain over the weekend.

This year at least, the teams seem on a more even spread, and it is hard to pick a runaway winner for the drivers or constructors championship.

> ...and F1 has been a staple on mainstream terrestrial channels for decades, although it has seen a decline as F1 has increasingly became boring and predictable, and almost entirely dictated by team tactics and car reliability.

Also it's now only live on Sky, which requires a minimum ~£30/m subscription; probably doesn't help viewing figures.

welcome to what football fans have been dealing with for 30 years now

In the UK, that's definitely an overstatement. Many league games, and major cup games remained (and possibly still are) on free to view TV.

In contrast, for the last few years, only half the races have been free to view. As of this year, only one race a year is available to watch for free.

Shame, as the last race (Bahrain 2019) was possibly the best race I've seen in years.

Jake is UK based. https://twitter.com/jaffathecake

Very cool. Had a busy day at work so couldn't check all the details!

Yep. I'm watching an episode every day and it even made me want to watch a race the past weekend. I kinda miss following it, and I lost touch since 2007-2009.

The Bahrain GP is definitely worth watching!

If you like having your heart broken, yes, it certainly is.

I loved that documentary. But doggone it, the racing scenes were all quick cuts. I could barely focus on it before it cut to another view.

In the movie "Grand Prix", the greatest shot in it was one continuous shot of a lap in the Monaco race, from the driver's POV. No added music, just the engine sounds. I wish that documentary would add such a scene for each track.

The Formula 1 YouTube channel has onboard videos. Here's Charles Leclerc's pole lap from last weekend:

https://www.youtube.com/watch?v=xTezG7LJL5o

Race highlights with some onboard vision:

https://www.youtube.com/watch?v=0VjWY--QdMA

Thank you!

That is a particularly good video of Leclerc's lap. I love that the telematics on his steering wheel are visible as well, and you can get a real sense of the speed and the gear changes needed.

Yes, I like that too. The steering wheel displays are easy to see in night races. It's interesting to see what the different teams are doing with their steering wheels:

https://www.youtube.com/watch?v=nLExSMdjlUg

The onboard lap comparison videos are also good, like this one from last year's US grand prix:

https://www.youtube.com/watch?v=GH-Uv6OKeDc

You can see where time is being gained and lost throughout the lap.

One of my favorite onboards is Ricciardo's pole lap from Monaco last year:

https://www.youtube.com/watch?v=xbrKA6IwgcM

And this one:

https://www.youtube.com/watch?v=dU_xkM7TZR0

I hated the documentary. It was entertaining I guess, but most things presented were blown way out of proportion for increasing the drama. And the wrong sounds were dubbed in for the cars, and the videos were edited in such a way to suggest things that outright did not happen (ex: The Mclaren mechanics laughing at Grosjean crashing).

This is quite sad, because there was a lot of legitimate on field drama last year, that they could have chosen to show instead. I'd even go so far as to say it was a mockumentary rather than a documentary.

Literally, the only good thing was that it seems to have been entertaining enough for a lot of people, that F1s viewership has increased quite a lot.

The other egregious edit was making it look like Grosjean shoved the cameraman in the pit lane. In reality he was reaching out to stop him from running into something (the cameraman was walking backwards)

The requests were all made from Dulles, VA.

It's possible some of the sites would be faster from the UK or elsewhere, but ttfb was the least of these site's problems

Just the exchange between Cyril Abiteboul and Christian Horner before the 2018 Belgian GP regarding Ricciardo going to Renault justified watching the whole series. The gist:

CA: so now you need an engine and a driver. CH: you are spending the money you don’t have for the drivers. CA: I have a lot of money.

The build up to this was Red Bull announcing they are not going to use Renault engines in 2019 but before they knew they’ll lose the driver to Renault.

your description is severely lacking. what was said is largely irrelevant. the body language was the point. horner as cocksure alpha and abiteboul as weak. the 2 are portrayed that way throughout the series. i especially like the highlight / intimation that the highest priority for renault is the quality of their lunch.

I'd love to hear from those of you who do web work for these types of big companies. Do they just not care about performance or are there other issues at play? I've dabbled with some small sites and would get stressed if the page load was more than a couple of seconds. How can these companies with large budgets find it acceptable that page loads can take tens of seconds while jumping content around the page or locking up threads with massive JS?

Most execs have the assets already cached so they dont realize. And if you tell them to clear their cache and revisit the site they will scoff and tell you to "scurry off with your technical nerdistry!" I don't bother trying to rock the boat anymore. It is never appreciated, only scolded as if you are a pedantic whiner.

That's because you're bad at marketing. Imagine how differently you'd be received if instead of saying "here's how slow it is if I clear my cache" you said "here's what a new visitor who has never been to our site before experiences".

I wish that were the reason. In large companies, the communication channels can be as bad as it is to outside customers. Put your energy into things that reward you back for good faith efforts. Large companies often have power structures entrenched by other factors than efficiency and function.

But who is going to say that? The web developer is going to go to the CEO and say "look how slow the site I just made is!"?

The Haas cached second load is slower than the first. You can't blame that one on execs already having the assets...

They are normaly done by boutique development agencies with ZERO thought to the real world.

Well. These are just anecdotal from one job in an agency which also sells websites for some bigger companies... and I hope this stuff is rather not standard... but if I had to guess it's something like:

- bad project management from the client (they often don't care about technical quality or rather can't evaluate it)

- bad project management from the web contractor (they often don't care about technical quality or rather can't evaluate it)

- the web agency has too many projects due and needs to deliver asap

- exhausted & mismanaged dev teams (+ freelancers) which are constantly working on short lived projects, which are a mess because nobody doesn't really care

Also I am amazed over and over again how a lot of managers (or people) will lie straight to your (or the clients) face if they can thereby save their ass.

Code is written by web devs that are more designers than devs with little optimization skills on expensive high end hardware. Work is reviewed by executives on expensive high end hardware. My experience anyway.

That belittles designers -- designers understand what slow means too and can use Google to find a solution.

I used to work at an ad agency several years. We were locked into Wordpress or Drupal with Bootstrap and some base plugins for sliders as well as a ton of stuff for 'SEO' and tracking. The amount of time allotted to deliver the project was actually pretty small. You would get a mild scolding if you even bothered to look at performance. A lot of these things are a part of the theme/plugin out of the box and may not be optimized, but the images and the content was usually uploaded by an intern or whatever and they never bothered with image optimization.

My experience from working at a car company (although not in the F1 area):

This kind of work is typically outsourced, since it's not a core competency for the company. So the company would be contracting some web development agency which builds a homepage and potentially also hosts it. There are contracts and specifications in place around what needs to get built. In the web development domain I don't think the outsourcing aspect is a big plus or minus on execution ability. There will be some good and some bad agencies around - and an internal team might not really be doing a better job.

Regarding performance: They mostly wouldn't even recognize. The work is presented with some fast connection and device, so it's not visible. And even if it would be, the people responsible for it (or even the leads of those companies or teams) are not that tech savy and will just accept it as the current state of the art.

There will be far bigger discussions around other things than performance. E.g. which images and fonts are to be used, how trademarks and the brand are represented, etc.

Well, i’m not a web dev, but as a long time fan I may offer some personal opinion on this. First, who visit these sites? Nowadays fans just follow twitter and Instagram for related info, and would-be sponsors have no reason to take a look at sites looking for interest. This is such a small field with huge public info availability no body needs info from web site to relate. The only time I tried to visit team website is to shop some fan stuff. I guess the teams also know this so they can’t care less

Agreed. I am an avid F1 fan, but I don't think I have ever visited a constructors site in all the decades I have been into the sport. At most I will visit the main formula1.com site, mainly to find out race times or to watch highlights of a race I missed (and a site which I hoped the OP would measure alongside the team sites), or else consume information via Youtube, Instagram and other aggregate sites. No interest in just seeing a fancy moving brochure site that doesn't tell me any more than the others I mentioned above.

Given the amount of analytics scripts in these pages, someone at these places must care about them.

In my experience with non tech clients, the website is built using Wordpress, hosted by the agency, and presented in a board room. The presenter will connect a newish MacBook into the projector after lots of fuss about the right dongle type and proceed to click around. The decision maker at the company will verify items against their memory and usually things that users do not care about, like they will ask who agreed to have a photo of the red car instead of the blue car because if they recall everyone had agreed to the blue car. Oh it was the red car after all. Really, because the blue car looked better. Oh it was a placeholder image of a different racing team. Oh well in that case it’s fine.

+1 for the chuckle. This is exactly how it goes.

If you have 100+ developers pushing to a repository, unfortunately many of them don't care about introducing a 100ms regression here and there, and that's how you end up with a slow page.

in our case (a major online retailer), some of the problems are coming from the marketing department. we watch the data thoroughly, as we need each year to prepare for black friday, but after that some guy from marketing inserts through tag manager another tracker or any other crappy script that also heavily affects frontend performance.

we actually care about that, but our options are limited. of course, if the performance is heavily affected we can tell them to buzz off, but usually it's not a single tool that generates the problem, but the sum of all of them.

on the other hand, i'm aware that there's some value added by those tools.

I find it interesting that Google tries to stress website performance and even goes to such extremes as creating a new, stripped down HTML spec (AMP) while simultaneously shipping bloat-ware enabling products like Google Tag Manager.

The performance of a website very often depends on how important the site is to their sales. For many companies who aren't selling direct to consumers, a fast website isn't their biggest priority.

One thing that wouldn't surprise me would be to see that the websites for some of the companies mentioned in the article are running on SharePoint. --If tuned properly, it can be quite fast, but like other solutions, it can also feel bloated and slow if it's not prioritized.

How is it acceptable for any webpage to take 10s+ to render on first load? How did we get here?

Well... Ferrari inlined a 1.7MB BASE64 encoded image of their prancing horse logo, only to render it at less than 1/10th of its size.

That shows it's not always the massive JS libs that are the problem. It's also a lack of knowledge and interest from those who build and maintain the site.

And quite possibly a lack of interest from the customers of their site.

Agreed. It's not like Ferarri is going to lose a (potential) customer because their site loads a bit slow.

It definitely doesn't matter with Ferrari because I doubt people buy cars directly through their site, but I've gotten into spats with art/fashion people who make terrible web stores that shit the bed the moment you're on a spotty connection. Their justification is unironically "we don't make products for poor people, so who cares what our site looks like on a bad signal," which I find really odd—do people with means never travel or something?

Right, nor will they need PR from G like the rest of us do. I mean are people typing in "Fast cars I can't afford" to find Ferarri online?

That sounds less like a lack of knowledge and interest, and more like a lack of time and care. When your boss is yelling at you to upgrade the Ferrari logo because you need to get to work on Dave's Dogfood, and you've got two hours to do it plus nobody remembers the AWS credentials for the account with permission to upload images to S3, you're going to hack some shit together and deploy it then sneak out for a coffee. Your personal levels of interest and knowledge don't come into play when faced with cost-cutting and ignorance at the purse-string level.

Also, the programmer is not the one ultimately responsible. It's the QA team.

Yup, chuck that responsibility over the wall. Not.

True, the QAs job is to test it delivers as requested, UX if it is usable by the end user, ops that is is not heavy on resources etc. But it is a joint responsibility between all and especially the developer and designer that they deliver quality and not cut corners.

If there is an ultimate responsible it would perhaps be the project/product owner who would decide what is important and how much time and effort they should spend on which parts. But still it is not just pass-the-buck, they are all responsible.

The programmer's job is to come up with the best solution in the minimum amount of time. That's an ill-posed/unsolvable problem. Hence they will sometimes come up with a solution that's not perfect, and they will leave it to QA (or the product designer) to see if the solution is acceptable.

It reminds me of my previous company, also a car company. During the Dakar Rally they had an 8MB jpeg as the header of their sale website. It took one week and a message sent through a "Contact us" page (public) to get them to reduce it to 700 KB. I suppose it happens because the web devs never tested the website with a realistic internet connection. Everything loads fast when the server room is less than a km away with gigabit bandwidth. It's either bad project management if they forgot to test correctly/change placeholders or as you said lack of knowledge.

>I'm going to put each site through WebPageTest to gather the data on Chrome Canary on a Moto G4 with a 3g connection.

I guess the results would be very different when using 4g? and that the real problem is that the author is using a unrealisticly slow connection.

4g wouldn't change the main thread issues.

Also, it's a mistake for developers to assume that everyone is on 4g, and that those on 4g are experiencing good speeds.

Even in London, if you're at a train station that's hit by delays, connectivity grinds to a halt as everyone starts looking for alternative transport.

In the big cities in India, 4g is pretty much always like that.

Where on earth do you live where 3G is "unrealistic?"

There are a fairly significant number of users without LTE. Bear in mind too that some of the names at the bottom of the list, like Brazil, are countries hosting a Grand Prix this year.

https://www.opensignal.com/reports/2018/02/state-of-lte

He's using a cheap phone and turned off 4G.

It makes me kinda sad how the article talks about a five second loading time as something that's "relatively good". I guess between cheaply made, slow sites and huge sites like reddit not giving two fucks about performance the goalposts have shifted dramatically toward the bad.

Personally for simple content-based sites I find anything above ~half a second is already quite bad, assuming a reasonable underlying connection to the server (ping < 30 ms).

reddit is slow? I've never experienced that?

The new SPA site is slower than https://old.reddit.com by a fair margin.

On mobile it's an artificial delay to upsell the app.

Please elaborate with technical details.

Their mobile site is even slower than the desktop site if you try to load both on desktop.

That would explain it. I avoid the new SPA redesign. Thanks!

As a long time F1 fan, I love the premise of this article. Perhaps we should also investigate the correlation between team marketing budgets vs web site speed, as that seems to be a major talking point about their R&D budgets vs track speed these days...

When I started out I used to contact companies and try to sell them a site upgrade, generally focusing on something that was better looking. I wonder if now the better opportunity is to demonstrate how much more efficiently you can code their site to increase performance. Reinforcing that with data on how this affects sales and surely you could snatch quite a few contracts.

Hey look, Williams is fast again...

Haha, good one

Bloat. Just bloat. Bring back the motherf... website that is just plain old html.

yeah, and if you really need those scripts, stop loading them from 10 different domains...

Testing F1 race team websites for speed is a neat idea.

I'm a cofounder of PacketStream.io

We'd love to see you re-run these tests from our network of global residential proxies to see results from different locations around the globe. Our product lets you select a geolocation to exit from as part of the request.

If you're interested in free credits to run some more speed tests send me an email: ronald at packetstream io

This is one of the metrics I use when voting

Cool article, read the mercedes example. That stuff is super basic, why would they ship such a terrible product?

They've been too busy winning the F1 championship every year for the last five years.

But apparently not too busy to make a website about it.

Mercedes got 500 folks employed just to make the F1 engine[1], I'm sure they can afford to hire someone to get a couple of websites made. That it's not better is probably because, does it need to be?

[1]: https://www.mercedes-amg-hpp.com/about-us/

I think your comment is a lot more insightful than it might appear on the surface.

Companies do a ton of market research, if there were some edge to optimizing your company’s site, I suspect there would have been 10 books on the subject by now.

Truth is they don’t do it because they don’t need it. Plain and simple. What exactly would they get other than HNs approval?

Your comment struck a cord with me because it’s a conversation I have had in IT millions of times.

The real wisdom in any type of endeavor is knowing when it’s ok to cut a corner and when there is absolutely no room for compromise.

These kind of websites are just a brand box-ticking exercise. They're farmed off to a low-bidding contractor or agency who have no incentive to care about performance. So long as the site runs fast on the client's iMacs using office wifi that's good enough.

Would be cool to see what the performance would be with Cloudflare in front of those sites.

Aside from Williams who suffer from HTTP/1, the performance bottlenecks are little to do with server architecture or TTFB.

They do all kinds of caching, image compression, etc. I think most of those are in paid plans though.

Just mentioning it as it's quite a simple, fast, free change that doesn't cost any developer hours.

> I mean, it makes sense right?

Yes.

Reply on news.ycombinator.com