Hacker News Reader

Comment by jefftk | Hacker News Reader

It looks to me like the design for this was https://github.com/ampproject/amphtml/issues/3133

(Disclosure: I work on ads at Google, and I'm friends with some of the people at the link. Speaking only for myself.)

Thanks for the link. Though as a Google employee you should probably refrain from any further public comments.

There's nothing wrong with a Google employee participating in a Google-related discussion (edit: I mean in an HN thread of course), same as with any company. It's nice if they say so, which jefftk did.

Edit: I misread. Corrected below. Sorry!

I'm pretty sure they mean it's a bad idea from Google's perspective, not Hacker News'

Oh, I think you're right. I'm too used to certain types of internet reflexes. Sorry GP!

I would edit your initial comment because to many it will read as if you understood it as Google’s perspective, and still Okayed

Ok, I did that. Thanks!

Nothing wrong with that, sure, but making comments related to ongoing antitrust litigation against company employing you, while not being any official spokesperson and not having your comments vetted by legal and PR, can be really, really detrimental to one's career if Google decides that it doesn't like your comments. Google is very explicit in its internal trainings to never make any public comments like that unless you're officially empowered to do so. You can say or write anything you want on HN... on the last day at your job.

I worked at Microsoft in Europe and we never had such training. Is this only in the US?

Yes, in the EU employees are more protected and it is much more limited what an employer are legally allowed to do. In fact they have to prove that your intention was malicious.

There's probably better protection against random terminations on average in Europe than in America.

Would whistleblower protections apply in this case?

No, why would they? He's not whistleblowing. The commenter even admits that he has no context on the link in question and just suspects it's what's being implicated.

I've been given formal warning for that, others have been fired. It made it front page of HN. I don't think that's wise advice in all cases dang.

Dan, that's like asking a Google employee to basically get themselves fired. It's not something we should encourage, for their own benefit.

Nothing wrong for HN, but definitely risky as an employee.

It seems possible that in the history of internet forums, some of them might occasionally receive subpoenas because forum members posted something essential to one side's case in some litigation. It also seems possible that some of those forums receiving subpoenas would rather not deal with the burden of responding.

Exactly, wouldn't want to lose those golden handcuffs :)

Do I understand correctly that the 1-second delay existed in both AMP and non-AMP case? And later, someone found a way to remove the 1-second delay for AMP since it could be done without a negative UI impact (unlike for the non-AMP case)?

That seems more likely than complete malice.

So the suit in one paragraph claims that AMP isn't faster, then in the very next paragraph says that it is faster, but that's anti-competitive?

It’s not faster when the browser actually loads them. It is faster when google induces a delay to making it available to start loading.

The first article I found about this was https://timkadlec.com/remembers/2018-03-19-how-fast-is-amp-r..., which claims that AMP is actually faster in practice, probably mostly because it encourages good habits. Where are you seeing that it's not faster? This article is from '18 so perhaps there has been a change in how well AMP performs versus the baseline.

Also worth noting that we should probably judge AMP as a whole product. What if it didn't make things faster at all, but did make them predictable and amenable to prerendering? The product is AMP+Cache+Prerendering. Prerendering unquestionably gives huge speed advantages in many scenarios. If you don't have AMP, you don't have prerendering, which means AMP-as-a-product gives significant speed advantages, since it's on the critical path to prerendering. (All of this is of course predicated on the idea that you can't practically prerender normal HTML pages, which perhaps you can, I don't know. I'm just saying if that justification for AMP is correct, it shouldn't necessarily matter that HTML+AMP markup is not faster than HTML markup.)

> mostly because it encourages good habits. Where are you seeing that it's not faster?

You just answered it yourself. If someone follows good habits already, then AMP isn’t faster for them (aside from Google not allowing non-AMP loads to start sooner).

Did you use Google to search for the benefits of using AMP?

Are you serious? What else would I use? But I just checked, and DuckDuckGo and Bing return the same page as the first result.

This is answered above. It is faster for arbitrary means, nothing substantive.

Where "arbitrary" can be swapped out for "malicious" and "anti-competitive."

So tl;dr: Traditional HTML+JS ads have always needed an extra second delay to load because of cross-domain UX degredation, but because AMP is pre-caching pages, they're able to pre-cache the ads too?

Is that what I'm reading?

I believe that the idea is that if the ad is amp, you can be sure it will load fast and not impede the ux of the page, whereas if it's html/js, you can't. It isn't about whether it is cached or not.

But this sounds like a technical limitation that makes sense? This entire comment chain is mostly people interpreting the lawsuit as claiming this was a deliberate artificial delay for business reasons and getting outraged about that, not a technical issue.

It's a little murky. If the status quo comes with a cost, because of practical uncertainty, and you provide an alternative that eliminates the uncertainty (and the cost), and then you use the elimination of the cost to prefer your own method... at minimum, that feels "not fair."

Why would it be unfair? A lot of UI delays come from uncertainty, it’s often a significant technical issue.

Because you eliminate uncertainty by imposing your own standard, the adoption of which financially benefits you.

An unbiased viewpoint acknowledges that bad people doing bad things for selfish reasons, can have both positive and negative results.

There's further details about it here: https://github.com/ampproject/amphtml/issues/3133

And yeah, it's a technical issue.

So taking websites hostage and demanding a ransom in the form of an amp link. Thanks.

If the top-line goal is to make the page as responsive as possible, I'm a little stumped as to what the alternative would be beyond delaying the loading of arbitrary html and JS. Can you think of a way to load arbitrary HTML immediately and not cause any page responsiveness issues on, say, a three year old Android device?

For the parents interest can we please delete his post...don’t want to see him get into any issues

He can probably manage to do that himself if he wants.

Is Google legal going to extend their protection to you during this?

Reply on news.ycombinator.com