There is decentralised finance (defi). One example of application is lending your money (in stablecoin or token format) in a lending protocol of your choice. You can expect a very good yield compared to traditional finance. Aave and Compound are the most famous. This is one type of the many financial applications developed on Ethereum's smart contracts.
Indeed I think they don't ship the whole browser package.
What they might actually do is to locally launch a HTTP server which renders the app if you call it on any browser (localhost:8080 in the examples). The chrome of the app is actually a thin layer that use the default OS web browser. That is why the memory footprint is much less higher that alternatives like electron.
What a surprise: developers getting screwed by big centralized companies. Nowadays wasting time and money building anything on the shoulder of another company is crazy. It's not if but when are you getting screwed: Twitter, Facebook, Apple, this is history repeating itself plenty fold.
I'm just surprised nobody mentioned any decentralised, blockchain-based platform where spending time and money building on it is just worth it for the future.
Have a look at the Steem blockchain: social network like publishing and commenting are free and even rewarded with crypto-currency payouts. And as a developer it is very elegant to develop on it.
> I'm just surprised nobody mentioned any decentralised, blockchain-based platform
What about decentaralized platforms that are not blockchain-based? The Fediverse (a part of which is write.as) achieves this goal of being decentralized without using any kind of blockchain technology.
I think the /idea/ of IPFS is very useful. On the other hand, I think the creators are way too interested in bikeshedding and writing new white papers for less useful things. They just aren't interested in providing real value to end users. They have this early adopter bubble that seems to be quite popular but I don't think it's going anywhere.
AFAIK it is possible now to get the ICP, but to integrate Wechat / Alipay or any other meaningful service you need a Chinese legal entity. A foreign company can register WFOE which is a Wholly Foreign Owned Enterprise, you do not need any Chinese partners or anything, but might be costly to set up and in paid-up capital depending on the industry. E.g. for a fashion brand I know of, they asked to have ¥10 million ($1.5 million) to be deposited within 1 year.
Any content hosted outside China is very slow for chinese users. So if you're serious about serving content in Mainland China, you will first need to obtain an ICP license. Only then you will be able to use the Mainland CDN and cloud instances.
And pretty much everything useful is blocked. We rolled out an update that broke our app for all of China, because cloudfront was blocked. That was a new one for me.
Having spent a painfully slow 10 hours in the airport in Beijing let me appreciate just how much of the Internet breaks when you block the biggest sites. Serve up jQuery through googleapis.com? Not in China, you don't. Hope it wasn't doing any heavy lifting for you.
That is categorically false. Chinese communism is strongly influenced by nationalism and protectionism comes from that but make no mistake that their goal is censorship.
They could easily force foreign companies to operate under a strong censorship regime if that were their main goal. Facebook basically bends to the will of every market it operates in.
However they clearly prefer local startups and makes it generally difficult for foreign companies to purely exploit the Chinese market. Joint ventures are always required at the least.
Sources tell me that the azure infrastructure there are not actually operated by microsoft, but by a subcontractor, and merely licenses the name from microsoft. I would be quite wary of assuming that microsoft policies apply there.
Where: We are getting people to register on our own website through various incentives (mainly discounts).
How: Well, we are spending money to reach our page likes through FB ads and make a compelling case to have them go over to our site and register. It's not easy. My guess is that it might take up to six months to get all the people who are willing to get off FB for a better experience.
It's not about spamming them but it is about being able to reach them and engage with them. So, at some level, we only want the people who are actually willing to click away from FB and go through the trouble of registering on our site. If they do that chances are they will actually engage and they'll also be more receptive to our message. That makes it all far more efficient.
If you have a FB page with 50,000 likes you might not reach more than 5% --if that-- with any update you post to your page. There is no way to blast an update to your entire list. Not even in a staged manner. Meaning: Send this update to people who liked this page at a rate of 1,000 per day for 50 days. If you want to reach those who liked your page you have to pay. Over and over and over again with every single post you want to reach your audience through the "boost" process.
FB is clearly telling you they own your audience. I have to deeply disagree with that. FB does not own the audience that likes, say, the Ford, Audi, Coca Cola or Victoria's Secret pages. Those companies worked very hard and spent lots of money to build and capture an audience. FB is effectively hijacking your audience in an attempt to make money with every message you want to send them. I really hope companies recoil at this in a big way and the whole thing backfires on them. It's wrong.
Beyond that, list management is horrid. For example, we wanted to delete nearly 1,000 likes we got from Bangladesh, India and a few other places that had no real relevance or connection to our products. Not sure how they got there because our advertising for page likes clearly defined a US-only audience. One theory I read about is that page like click farms tell their employees to like a wide variety of pages in order to hide the likes-for-hire activity. Anyhow, the ONLY WAY to maintain your list is to use the banning interface and ban each and every user you don't want ONE AT A TIME. You can't even click a bunch of check boxes and then apply an action. You can't export to CSV, mark them and then upload to clean-up your list.
So, you write a post, it only reaches 5% of your audience and, by random chance, a bunch of them are from corners of the world that have no relevance whatsoever to what you are doing. It's a royal mess and a total waste of money.
I am sure there are many who are doing well with FB ads. I am sure these are very specific use cases that might be so lucrative that paying for noise in your likes and paying to reach the audience over and over again might not be a problem.
Oh yes, and then there's the issue of the dumb-ass rules for ads and the even dumber software that checks against those rules. My favorite being the 20% text rule. It is so unreliable that you can actually run an ad, have it approved and half way into your campaign get a message telling you that the ad was disapproved due to violating the 20% rule.
The software they are using to check against this is using some kind of crazy voodoo that works half the time. The other half of the time it confuses text on, for example, a product with text outside of it. That's why you end-up seeing ad after ad on FB with stupid stuff like a landscape or a bird flying or whatever, because you can actually try fifteen different generation of ads with text and have every single one of them rejected when the text doesn't even approach 20% usage.
Then there's the absolutely terrible feedback when ads are rejected. No reason at all given in most cases. So you have to guess. Is it the 20% rule? Is it the nature of the image? Is it the text in the comment? Is it the link? Is it the call to action? Is it the page it links to? Is it the subject matter? Is it the color of the background? Is it the capitalization? What is it? So you are left guessing with no feedback and NOBODY to contact, even with smoke signals.
We had one horrible case of launching a single day discount offer through an ad on FB. A website was built and all related systems and software were in place weeks in advance. The FB ad was put together and staged a week in advance. Approved. The day of the offer the ad turned on and started to run. No more than 15 minutes into the 24 hour run it got kicked out for "violation of TOS". What? We sat there trying to figure out what the heck was going on. We submitted updates at a furious rate to try to shotgun the problem and see where it was. We changed the description, the link the text, the words, the image, the background. We changed everything. We took a beautiful and turned it into the most disgustingly bland thing you've ever seen. Nothing. Nothing could make the ad gain approval. We burned sixteen hours chasing after this. The promotion would have been a complete bust if we hadn't reached out to AdWords and run a bunch of ads about six hours into this mess.
We somehow filed a support case with FB. FOUR DAYS LATER we get a message basically saying "Sorry, the ad should have been approved. There was nothing wrong with it. Probably a glitch somewhere. Try again next time."
To say I blew a fuse is the understatement of the year. FB Ads might just make this work because of their scale. I don't know many companies in the real world that could behave like this, have such shitty customer service, performance and reliability problems and survive. We'll see what happens. I am certainly not sending them any more of our money once we get our customers away from their clutches. They are OUR customers, not theirs.
I own and run several businesses, collectively they address pure software products (mobile, web, embedded, system), hardware (electronics, mechanical, optics) design and manufacturing and mainstream (nothing weird) retail consumer products.
Most of the FB experience I am describing applies to the consumer products.
