Engineer on the Tiles team here.
One of our core tenets is "user agency", or putting the user in control. The user should always have the choice. That option will not go away.
417 karma · joined March 10, 2008
One of our core tenets is "user agency", or putting the user in control. The user should always have the choice. That option will not go away.
Revenue is not our primary objective. We really want to change how the industry works. At the moment, revenue is our way to say that we can make something that is viable and will work for everyone else.
It is something that is hard to condense into blog posts, because the topic is so touchy. That said, our primary goal is really to make a big change and to prove that it can be viable for others.
We're rethinking how digital advertising works. It is what makes the internet free for everyone. We're hoping to make a system that is effective, yet less annoying for users, and does not violate their right to privacy.
More on my reply on another thread: https://news.ycombinator.com/item?id=9588451
That is not exactly true. We do not obtain the list if top sites.
What we receive is the list of non-history tiles that has been shown to the user. i.e. if the Tile is a "known" tile, we'll capture the impression, or click. If it's not, we will receive an almost empty placeholder for that site, barring for the frecency score.
This is along our philosophy of least amount of actionable data collected.
For instance, if a user had one Tile we know about at the 5th position, we'd get something like:
{"tiles":[{"score":42},{"score":32},{"score":22},{"score":8},{"score":3},{"id": 500, "score":2},{}],"locale":"ja","view":6}
The id allows us to count impressions, clicks and whatnot for that Tile, the "score" allows us to gauge frecency, i.e. how important is the tile for the user, together with the position in the array.
It's not nothing, but it's not your top sites.
Edit: formatting
Source code available at:
https://github.com/mozilla/onyx
We're trying to create a new way for ads to be targeted. In the classical model, the server tracks wherever you've been on the internet.
Basically, to show you relevant ads, at least one entity needs to know where you've been.
What we're trying to achieve is similar, except there is no tracking. Most of the decisioning (e.g. which sites similar to the target group have you been on before?) is made in Firefox.
The ad server will send many ads based on a user's geo (as determined by IP address) and locale (browser language, e.g. en-US). This package will include more Tiles (some are sponsored, some are not) than Firefox will decide to show.
While we do get data based on the impressions and other interactions with the Tile, we only get the strict minimum needed to compute our counts.
And on the topic of IP addresses, we consider that sensitive information. We only keep the raw data for a very short while (7 days).
The only thing that is kept for longer is the aggregate data, e.g. how many impressions tile X did on day Y.
It may sounds strange, but this is really our honest goal. We want to change the ad industry.
The ad industry in its current state is built on foundations we think don't make sense. For instance, the whole idea of abusing cookies, a useful technology, to track where users go around the internet so that the data can be traded, so that others can make guesses about what ads to show... sounds a bit in need of a change.
We know for a fact that many of the players in the business, the ones that matter, don't really care about intruding on people's privacy. For them, it is what they need to do to achieve their goals.
We have to face it, the internet wouldn't thrive without ad-tech. Not many people are willing or able to pay for content. The digital ad industry is important and is here to stay.
That said, we think we can make a change... for the better. We can think about how to do this from first principles, to be the first customer of our tech.
Frankly, no one will be willing to play the new game with us if we can't prove that it works at least as good as the current way they are doing things: the old tech may be clunky, not that effective and there may be a lot of middlemen, but there are 2 decades of investment in the way its built.
Users are affected. Users care. We know we can make for a web with less annoying ads. They don't need to be nagging, vying for your attention the same way they are now. They don't need to be creepy. And you know what? They may not even be ads as you know them today.
We thought about this a lot, and it's a very touchy topic, one which would cause controversy any way we'd broach it . Why do you think we're not being honest?
There are more articles on the topic on the net. This identifier has been dubbed "AdID". Read more here: http://m.spectrum.ieee.org/tech-talk/telecom/internet/google...
Is that a problem with the cable, terminals on the devices, the mini-displayport dongle or either?
The problem also only seems to occur on Retina Macbook Pro's
If you've got the bandwidth, could you please post your monitor model and/or snap a video?
The reason it's got the "We are watching you!" is because i noticed the problem while viewing https://optin.stopwatching.us/
Edit: hosted here: https://people.mozilla.com/~oyiptong/retinadisplayproblem/
While it is true that there are less companies you would expect would pay Bay Area size salaries, things have changed in Canada.
potatolicious told me to write this for the public good.
I work for Mozilla.
New grads, admittedly with co-op experience, get paid at least 90k+ with the same 30% bonus as all Mozilla employees, putting them in the 6-digit range mentioned in the thread.
Amazon recently opened an engineering office in Toronto, and Zynga has been here for a while, Google is in Kitchener, 1 hour away from Toronto.
A friend running a startup was complaining about the war for talent going on and was willing to pay salaries well in the NYC/Bay area range.
Food for thought?
While its true that employees tend to be less promiscuous as they can be in the peninsula, it is not true that developers do not have interesting work to do in Canada or that they won't be well paid.
While Mozilla has offices in Vancouver, I do not know enough about the city to talk about the situation there.
edit: missed part of a sentence. oops
The editing portion is easily handled: since your code lives locally, you could be using any editor or IDE of your choosing.
Were you thinking about automated app restarts? What else?
It is free, pretty well done and they haven't stolen my AWS credentials yet.
Blueprint is a nice step in that direction: make a change and dump out the environment.
Perhaps what we need is a system that compiles to that manifest, or to chef/puppet.
Setting up systems is a complex task that neither puppet nor chef solve, they make it less daunting, but it still is a real pain to do so!
We don't have windows right now, but I'm sure that will be a major driver for user acquisition.
I am just spitting out ideas without deep thought but a discussion might prove fruitful.
It could come down to certain types of data, some to be ignored, some to be careful of (like data or code), others to blindly overwrite (critical security updates).
The point would be treat the data as "dumb", but to keep in mind that some data are "dumber" than others.
Perhaps if we had our own disk image format, we could mark certain types of data to be ignored for instance.
Our VM would know what to mark as ignored for say POSIX systems and that aspect could be configurable.
We've thought about going pure diff at the beginning and decided that to implement that method will give us less flexibility in the short term.
I speculate that do so, we would need to implement our own virtual machine (we are using virtualbox) and/or disk image format.