“We started experimenting with AMP in our Bing App last May and have noticed that AMP pages load, on average, approximately 80% faster than non-AMP pages” says Marcelo De Barros, Group Engineering Manager in charge of the AMP integration at Bing.
https://blogs.bing.com/search/September-2016/bing-app-joins-...
People talk a lot of trash about AMP, but until there's another solution that allows me to go to a news website and actually read it, I'm an AMP supporter.
There's similar blogs for every small community, run by people who care about that community. Usually unfunded, and only covering stories of interest to that community.
Shills and Kool-aid-drinking zealots mostly - they are the ones usually motivated enough to do 'journalism' for free. Don't expect any professionalism, depth or any form of investigative journalismbeyond tweet-storms.
We outsourced our opinions and fact-finding to journalists for a few hundred years. Now that time has past. Journalists themselves have destroyed their own industry by producing a poor quality product. Maybe in future journalists will re-invent themselves as purveyors of pure factual reporting, and those that care about that kind of thing will pay for it. But it is a niche market. Democracy, decency, enlightenment and society will suffer of course. But those things were never guaranteed with any kind of real safeguards.
If you think this was ever the case, rose colored glasses are in effect.
Unless you’re doing tree shaking, which AMP can’t do duplicate, compression has rendered “minification” useless long time ago anyway.
But still you are loading ton of unnecessary content. Ad blocking on iOS still sucks (no custom filters).
If anything I prefer this way of working - Apple alone are responsible for maintaining a high performance adblocker in Safari, the wider app ecosystem just provides the block lists, rather than rely on app developers to write performant browser plugins.
I'm not aware of anything AMP does differently than non-AMP with regard to tracking, except that it's declared in a way that allows it to be cached+preloaded without falsely triggering analytics.
Repro: 1. Search for some BBC news on google.com, click on an AMP link and load a BBC page 2. Notice that you're still on google.com and not on bbc.com
That said, I use StartPage, who have a contract I believe with Google. It's Google's search results, minus the tracking. It's as much as I am willing to compromise on something as fundamental as search.
At least it's honest about not having results whereas Google presents 5 billion, all of which are missing one of the three keywords (which it notes in a small, light grey text, which you only notice after the first three results were completely unrelated and you were wondering what went wrong).
For example, "new double c++" is the query I did most recently and in the top 3 there are 2 results that answer my question.
Making something up at random like "torrent clients" gives me as top hit the Wikipedia article "comparison of bittorrent clients", which is better than expected.
I can't seem to think of a vague query right now. "audio books" gives me sites with audio books; "psychology books" gives me articles of 'the best 50 psychology books' and such; and looking in my query history, "draw unicode" seems vague but the top hit (shapecatcher.com) is the one I was looking for.
Something localized then: "drankwinkel echt" (where Echt is a place and drankwinkel a liquor store) indeed gives terrible results. The store name, surprisingly, works though: "gal & gal echt" gives similar results to google.nl.
Yes, that's really bad; it's what killed Altavista and could really be Google's undoing.
Still, Google is miles ahead of the competition.
Small experiment: searching for "movie old man balloons" on Google and Bing.
On Bing there is a first line of 4 videos, none of them related to the movie "Up" in any way. The second link is to Up on Imdb (good). The 3rd link is to a crazy religious fanatic site page titled "Disney PIXAR's, 'Up' - The Sugarcoating of Pedophilia!" (WTF??!? - but at least related to the movie). The 4th link is again to a youtube video with no connection to the movie.
On Google, the first 8 links are to the movie. There is a line of images, all from the movie / movie poster. There's a list of 4 questions "People also ask" that shows questions about the movie ("How many balloons would it take to lift a house?"). To be fair, the crazy Baptist site does show up on Google too (God has good SEO!), but way down below the fold.
Anyway, my point is, when answering the question, Google is certain you're looking for information about Up, and tries to give it to you.
Bing seems to have doubts and tries to guess if maybe you're looking for a funny video of a man in the subway wearing a balloon hat (??!? it's not a "movie"!!) or the hit song "99 Luftballons" from 1983 (not a "movie" either!)
Bing tries hard, but is obviously more than a little clueless.
Your experiment's results are not repeatable. When I do that here:
* Bing gives me 8 pages about Up (including that spoof site), one about Danny Deckchair, and a page about a magician who re-creates old movies with balloons.
* Google gives me 13 pages about Up (also including that spoof site), a book of best movie scenes on its page for The Third Man, and an article from The Rotarian from 1948.
Of course, some knowledge of how these things work teaches that this is a terrible methodology, given that it does not account for the fact that both Bing and Google tailor their search results to the searcher. One should at the very minimum log out of one's Google and Bing accounts, which you made no mention of doing.
FYI, Landover Baptist Church (which that article is from) is a very old, very well known spoof site.
A perfect search engine would not return this on the first page of results about the movie, because it's not about the movie.
That's a spot-on description of Altavista :)
Other search engines feel like a noob salesperson who needs to be told 5 times in 5 different ways what you need -- and it's really simple stuff too.
I like experienced salespeople.
I got relevant results in my first day there. I didn't notice a drop in result quality.
Google is just that good even if I hate to admit it.
It seems to expect more precise queries, while Google is geared toward "close enough" searches. I've found that DDG is often preferable when I have an exact phrase in mind. Google is sometimes too helpful, correcting errors that are not actually errors and failing to take some queries literally enough.
The main issues I have with DDG are that it doesn't seem to prioritize recent results (which can be good, but usually isn't) and it fails at local results, which is basically by design.
While I am still personally a huge fan of Google's work in this are and many others, sometimes I long for the days of straight boolean search queries in search engines. I could often find exactly what I'm looking for, and get the same result each time.
Are there any other major or effective engines that allow for boolean queries beyond AND/OR?
nowadays Yandex uses its own index even for “foreign” (non-Russian) sites.
Also, if it is using Bing on inside, I'm not then sure why the results are still different than that(even at the places where there's no scope for personalization). For example type this line exactly in both bing and ddg:
difference between std list and std set
The first result DDG shows is of a difference between set and a vector - not only that, it also highlights that result in a Google Card inspired fashion (which is worse because Google does that only when they have a certain amount of confidence in the answer), while the first result Bing shows is an actual difference between list and a set. So either Bing is cheating DDG, or DDG is doing something stupid on top of its results. (Also, if you cover the above string in quotes, Bing still shows 'some' search results, while DDG doesn't).
May be after using the '!' ninja techniques it'll be same as Google, but then I have also come to love the fact that I can type in 'my next flight' or 'Show me emails from ....' or 'remind me to ... ' and other personalized things on Google so may be my priorities are different here.
PS: DDG redirected image searches to bing, but not generic searches. Which honestly seems reasonable as doing image search well is incredibly hard, with minimal payoffs.
It sounds like Bing is personalizing your results either by location or by adtech tracking, which is kinda antithetical to the point of DDG.
My contract with a search engine is that they should provide links to other sites in a relatively painless way. In exchange for this the search engine can track me while I am on the site.
AMP breaks this relationship by refusing to let me leave the site and extends tracking to other sites. The speed difference is negligible to me and not worth this violation of privacy.
Just like AMP are fundamentally websites built on a bunch of guidelines with performance and accessibility prioritized which, once you sprinkle them with enough metadata, get treated specialty by third party like google.
Yeah, I don't think I misunderstood. AMP's guidelines are just more strict before the 3rd parties start treating you specially. And in both cases, the original website/app is just fine, until it starts getting treated specially.
Update what? What do you mean by "app store"?
I think it's reprehensible for google to push this so relentlessly and beyond simply stealing links it makes google into the "internet."
This (AMP) could easily be a standard, in fact it's mostly just common sense (good lightweight HTML/CSS/JS). Instead of Google forcing its way on users and creators it could just lower the page rank of the offenders.
One other thing about AMP that pisses me off as a user and an engineer is it's one more place to maintain meaning one more shitty neglected experience. As a user I hate it when AMP pages are broken and I somehow can't get to the non-AMP version. I don't blame the developers because we have enough on our plate. My anger is solely directed to google for making the damn mess in the first place.