PS: The idea that freedom of speech is an absolute right that should have (almost) no limitations is not a universal one. While that right is granted to people living in all democratic countries, outside of North America it’s accepted that such right comes with limitations, and usually that has roots in the history of those places.
For example, in Italy where I grew up, the same constitution that grants freedom of expression (speech, press, etc) also criminalizes “apology of fascism”, or propagating the ideas of fascism; it also sets other limits on speech that is hateful or discriminatory. Other European countries have similar laws, such as the outlawing of Nazi rhetoric and symbology in Germany.
Update June 2020: This blog is not served through IPFS anymore.
Fun fact: if you’re reading this article, you’re using the distributed web. Since mid-February 2019, this blog, With Blue Ink, has been served through IPFS and the Cloudflare Distributed Web Gateway.
Last November I blogged about how to run a static website from IPFS. I was already running a couple of apps in that way used by myself and my family, and I felt it was time to migrate my blog too. This took a bit longer than expected as I dealt with some issues, some of them explained below, but around a month ago I flipped the (DNS) switch and definitively turned off the single-instance VM that was hosting the blog.
That decision made me anxious at the moment, but a month in things look almost entirely good.
For a blog that used to average less than 3,000 page views per month, this happened:
On Wednesday, March 13, traffic was 5,060% higher than on the same day a week earlier. In a single day, With Blue Ink got almost double the page views than in an typical month before that.
Despite the significant increase on traffic, here’s what happened to the CPU usage of one of the main IPFS nodes serving the website:
Nothing.
Thanks to distributing the content via IPFS, and serving it via the Cloudflare CDN, With Blue Ink saw virtually no impact on performance and availability following a 5,000% traffic spike.
Not just that: tests show the website has been incredibly fast for users all around the world, consistently.
Let’s see the architecture
The architecture of the blog is relatively simple:
Pushing some new code to the master branch on GitHub triggers an automatic build in Azure Pipelines, which clones the source code and runs Hugo to build the website (it’s all free!). You can see the configuration in the azure-pipelines.yaml file inside the repo.
After the build is done, Azure Pipelines triggers an automatic release job. The release pipeline has two stages (you can read how I configured them in the other IPFS article):
Copy the files into one of the IPFS VMs, then via SSH invoking the ipfs-cluster-ctl pin add command to add the documents to the cluster and replicate them across all nodes.
Make a REST call to the Cloudflare APIs to update the DNSLink, which is a TXT DNS record _dnslink.withblue.ink containing the IPFS hash of the website.
What I learnt / could go better
HTTP turned thirty years old this month, while IPFS is still a new technology. With IPFS, some things work differently than we’re used to, and others just don’t work at all.
IPFS isn’t serverless; it’s also definitely not free. You do need at least one server seeding your data. The good news is that you don’t need large servers. A burstable 1-core VM offers more than enough CPU; if you’re running IPFS Cluster too, however, you do need 2 GB of memory. Adding three nodes like I did was probably an overkill (but a great learning experience–and really fun).
I spent a large part of 2019 working with the distributed and decentralized web, especially IPFS, also known as the “Inter-Planetary File System”. I’ve written a few articles on the topic, on how you can host a web app on IPFS, one of which even ended up on the front page of HackerNews.
For about a year, I hosted my blog and other apps through an IPFS cluster. I wrote a utility for making pinning files easier on Pinata, a third-party cloud service for IPFS. I made some small contributions to the IPFS core projects. I built some projects with it, including one that I never released–nor fully completed–that used both IPFS and Ethereum. And I even gave a talk about hosting static web apps on IPFS at Node+JS Interactive last December in Montreal.
That all changed in the Spring of 2020. I called myself out of the distributed web.
My blog and other apps I built aren’t hosted on IPFS anymore. I don’t participate in those online communities anymore. I’ve stopped writing about the distributed and researching about it. I’ve shelved all my projects that were using those technologies.
I also updated my blog posts about IPFS adding a note that my blog isn’t hosted that way anymore. More than a few people asked me why, and I always gave the same answer: a mix of technical issues and mostly personal reasons. So, I think it’s time I explain the personal reasons.
First, I need to explain why I got involved with IPFS in the first place.
When I first read about IPFS, my mind immediately saw it as an exciting new platform I could build my apps for. The premise of a fully-decentralized platform included unlimited scalability, ultra-high availability and resiliency, no single points of failure, and resistance against attacks like DDoS.
Coming from a background in which I am always thinking about SLAs, number of nines of uptime, disaster recovery, etc, IPFS sounded like a dream platform that would magically solve all my concerns. And, aside from some performance issues at times, it did. Plus, the small engineer inside me was really excited about being able to play with a new, shiny toy, that had lots of hype around it!
What happened next for me was a reckoning with the reality of what many people behind the IPFS core project and the community around it saw: the dream of a radically open, unfiltered, and by-design un-censorable platform.
I have recently opened up about my experience, over a decade ago, with building an app with good intentions but that was then misused (That time I accidentally built a spying app). I learned early on in my life and (pre-)professional career about the importance of ethics in software development, and I am now a proponent of the idea that just because something can be built, it doesn’t mean it should be built.
And that brings me back to why, after spending some time in the world of the decentralized web, I have called myself out, and why I think that should things like IPFS actually become mainstream, they might cause more harm than good in the world.
I have seen, and I am seeing every day, the dangers of completely unrestricted speech, and I don’t want to be the one enabling that.
I know that last sentence is a strong ideological statement; some might call it a political statement, but for me it’s more than just political, which is often used to describe extemporary beliefs.
Many of you reading this will not agree with me, and that’s fine. I’m not going to try and change your beliefs with this blog post. Rather, I’m looking to explain why, while I respect that others might have differing opinions, I stopped doing anything that would actively advance a technology whose ethics I question. To put it in other terms: your freedom of speech isn’t my obligation to enable you and give you a platform.
In short, I think that while the Internet has helped the world in countless of ways, it has also brought out the worst in people.
I do believe we need some filters on the Internet. It’s not just about stopping criminal activities, terrorism and child pornography: while I am obviously unsupportive of all them, I also think they’re not the biggest dangers coming from the Internet (yet they’re a very convenient pretext for politicians).
Instead, I think that regular people’s writings on the Internet is hurting the world on a bigger scale. And the collective sentiment is often manipulated by some “agitators” that are exploiting anonymous online speech for their own agendas: that includes online militias–for example sponsored by foreign governments–whose goal is to destabilize a society.
In the last few years, completely unregulated online speech has given rise to fake news and conspiracy theories that have actually killed people. It’s offered a megaphone to those promoting dangerous ideas like white supremacy, Islamophobia, anti-Semitism, homophobia and other anti-LGBTQ positions, and sometimes outright Nazism. It has tilted many democracies towards right-wing populism and fascism.
All these extreme ideas have divided societies and increased social tensions. And they’re responsible for a number of acts of terrorism which caused the death of too many people.
Given our experiences so far, there’s no sign that indicates that a fully decentralized and unrestrained web would be anything but a dangerous wild west.
In fact, despite being tightly centralized and controlled, social media companies are facing significant challenges regulating what people write on their platforms, and in fact they are usually at the center of every scandal of these years. Decentralization and less control won’t be the solution to this issue, but rather the opposite.
If you believe that I’m overthinking this, and that it’s not going to be bad this time, I urge you to think twice.
First, there’s no indication that a new Web would be better than the previous one just on virtue of being decentralized. The same actors that are using today’s Internet to wreak havoc around the world would not disappear in the new Internet, and actually, they could be even more unrestrained.
Second, while almost everyone in the communities supporting a distributed web are good people, with good intentions, seeing some names in there is concerning to me. Regarding IPFS, advocates (at least for a while) included people like Nick Lim of BitMitigate and VanwaNet, companies responsible for rescuing, among others, pro-nazi website The Daily Stormer and the platform 8chan, a cesspool full of Nazi propaganda, child pornography, and other hate speech. Gatherings on 8chan have been blamed for at least three mass shootings in 2019 alone, including the one in the mosque in Christchurch, all of them motivated by racial hatred.
The first real examples of the distributed web aren’t particularly encouraging either. Among some of the most popular apps (“popular” in relative terms, of course) for the distributed web is DTube, a sort of YouTube that is built on top of IPFS. As you can expect, the website is full of questionable content, including conspiracy theories, cryptocurrency scams, weapons, RT International’s Russian propaganda… and of course, porn.
In essence, if it’s true that a good beginning makes a good ending… with such a mixed beginning, the outlook isn’t too rosy.
I understand that my opinion is somehow a minority one, and people will continue to build IPFS and other technologies part of the distributed web. There’s also a chance they might become successful and potentially get mainstream adoption–although at this stage the barrier to entry is too high for the average user.
However, I feel that it’s my responsibility to not be helping to advance this technology and the beliefs of at least some advocates in the world of the distributed web hold. If the advancement occurs, it won’t be because of my help.
PS: The idea that freedom of speech is an absolute right that should have (almost) no limitations is not a universal one. While that right is granted to people living in all democratic countries, outside of North America it’s accepted that such right comes with limitations, and usually that has roots in the history of those places.
For example, in Italy where I grew up, the same constitution that grants freedom of expression (speech, press, etc) also criminalizes “apology of fascism”, or propagating the ideas of fascism; it also sets other limits on speech that is hateful or discriminatory. Other European countries have similar laws, such as the outlawing of Nazi rhetoric and symbology in Germany.
If you got this far the spoiler alert should tell you that I was quite shocked by Alessandro’s decision to withdraw from Web 3 development. What follows is my Journey to the end of 2019 when I did mother-of-all bumper stickers just to file my development work for the Grub Street Journal before going back into the Suit world and relaunching my Property Development business. Here we are 5 years later.
October 27, 2019
mother-of-all bumper stickers
#MOABS MOTHER OF ALL BUMPER STICKERS. PUTNAM ON THROWING THE BABY OUT WITH THE BATHWATER. WEB 3 ITS REALLY A P2P END TO END ENCRYPTION THANG!
I am still waiting for the nod from Zero Hedge after several ‘applications’ , what prompted my current interest was this interesting article appearing there today,
I started storyboarding the Grub Street eco system yesterday with a legacy Wix site, a Beezer App connecting to you, John Ward and David Malone to start looking at ways to extract data from the web on AUthors as they register on the platform, The App should extract data as served by Google and Mozilla and then optimise metadata to secure better legacy search results for registered Writers/Filmmakers on the platform,
This Mother of all bumper stickers is attempting to capture the breadth of functionality condensed into the proposed Grub Street Journal App the App is a Librarian and Curator for multimedia creators and a Curator/Librarian Sensai teacher devils advocate reading group and search engine for the Journal Subscribers. Every Subscriber is an Editor. critic and is the completer of the Grey space.
PLAYING WITH FUNCTIONALITY, GRUB STREET JOURNAL IS AN APP IN DEVELOPMENT WHICH IS TOKEN AGNOSTIC, WE WISH TO PLACE OUR ARTISTS, WRITER AND MULTIMEDIA WORKS ON WEB 3 ACROSS THE 3 WEB NETWORKS AND ALL PLATFORMS LEGACY, CLOUD AND WEB3, PROPAGATION OF THE ARTISTS WORK AND ITS AVAILABILITY FOR COLLABORATIVE/LICENSED USE IN WHOLE OR PART WITH CRYPTOGRAPHIC ROYALTY COLLECTION AND ATTRIBUTIONS WILL BE MANAGED BY OUR BACK END LIBRARIAN/BROWSER (“EXEGESIS, HERMENEUTICS FLUX CAPACITOR”). IMMUTABILITY AND SECURITY FOR ALL CONTENT WILL BE COURTESY OF THE WEB 3 DISTRIBUTED NETWORKS
Social Networks have been all the rage years ago when they reached their peak and destroyed a couple democracies around the world (including mine) but this is not a text about that. The lens I want to use for us to observe social networks is the content creation and manipulation lens, so grab your lens and follow me through some underexplained reasoning.
Social Networks try to present a minimal UI, just enough so that the core features are usable. They are target at common, easy idioms that foster passive engagement in the form of content consumption and sharing. Content sharing is not the same as creation or manipulation, you’re supposed to bring content from outside.
Social Networks have been optimized for mobile-first UIs which provide a better experience in detriment of a more powerful and flexible experience.
Social Networks prioritize artificial rituals such as liking which gives the trappings of interactivity without generating new content (while also helping fine tune their profiling and tracking data for advertisement purposes). In essence they are metric factories where most features are present to provide measurable forms of engagement to feed ML powered advertising engines. The content sharing is just the fuel used to burn through these metric rituals.
So, my case is, Social Networks limit the amount of interaction
WITH START PAGE AND ONE OF THE PRIVACY BROWSERS THE HYBRID WEB BROWSER EXPERIENCE IS WITHIN REACH. #GRUBSTREETJOURNAL AND #HYPATIASEYEBROWSER COMINGSOON.
via @therealslog #IABATO#TheSlog its all bollocks and that’s official, Follow The Slog
mentions
#IABATO#TheSlog its all bollocks and that’s official, Follow The Slog. The slog makes the calls on the hard plays this from 2014. Right again @JoeBlob20
mentions
#IABATO#TheSlog its all bollocks and that’s official, Follow The Slog. The slog makes the calls on the hard plays this from 2014. Right again @JoeBlob20 from 12 Sep 2013 Still getting it right 6 years later!
mentions Stop #PaedoBrino follow the Slog today he brought you, @wiki_ballot and gave you #IABATO follow & retweet, Read today’s explosive revelations.
Oscar Wilde defined a cynic in Lady Windemere’s fan as; ” a person who knows the price of everything and the value of nothing”.
Big Data is the new oil, the Black and White gold or, the pixelated fools gold perhaps? Huge valuations are placed upon companies who can however incongruously or implausibly graft on the pre-fix or Suffix “- Tech, Tech – TECH”.
An old computing maxim is Garbage in Garbage out, that’s self-evident of course, “pigs ears and Silk purses” and all of that. But let’s remain in the porcine metaphor is Big Data really just “pearls before Swine”.
Perhaps I should explain? Military commanders know all too well that The Map is not the territory, for the purposes of “Prop-Tech” that’s an apt starting point. Of course, the Map is not the territory and the Model is not the thing, just as the name of anything is not the thing in itself. These are far more than just semantic games and when we look at Data, data processing, databases and the hugely powerful new generation of Graph Databases, we have to be very careful to remember that any description of a Thing, a phenomena, a market a moment in time or a reporting quarter is just that, a description of something that has happened in the past at a fixed point or over a period of time. Statically for a snapshot, we can describe say’ ? an unexpired lease term, the passing rental, the repairing obligations the asking price, the previous list of tenants and so forth. Databases are very good at this sort of description and comparisons over time and in parallel can give a sense of the dynamics of individual variables within what becomes a computer model based upon the Data across a whole album of individual pictures.
At this point, one usually swoons with eyes glazed over and we will all want to say something indicating understanding, something indicating the conversation has been helpful and interesting and signalling that we agree, Big Data is full of goodness and very very useful.
“You must have a very very clever Algorithm to go with that”
Perhaps a warning label should come with each Big Data truth claim.
“Two Algorithms and some snake oil to go”.
Valuation is an art and not a science, there is also art in Mathematics both pure Maths and Applied maths always have more than one way to get to a particular solution. That’s why they are Solutions and not answers, the word Solution implies the inevitability of there being more than one way to wiggle the Elephants Trunk.
Based on the paper:
“Drawing an elephant with four complex parameters”, by
Jurgen Mayer, Khaled Khairy, and Jonathon Howard,
Am. J. Phys. 78, 648 (2010), DOI:10.1119/1.3254017
“The company will now focus on investing in technology, both in terms of its people and its assets. SEGRO plans to bulk up the Proptech and digital teams that now sit in the new division. “We had already created this team as we had been working on this for a year to 18 months,” Sleath says. “We have got a small group and will be recruiting into this.
“We are looking for data scientists. We want to use data better when we install centres and pilot a dozen or so warehouses and we want to put in sensors so we can capture data.”
There are some interesting developments in the US property world which are off Radar but happening on the OIP platform in which I have been involved. Overstocks Patrick Byrne has an operation called Medici Land, Why is that relevant, let me tell you. Web 3 will revolutionise Database hosting, Data Centres will not be required for the bulk of Data transfer which will be encrypted peer to peer. This is the key to understanding where to focus ones DATA base ambitions in the future. ( outside of the Surveillance Capitalism State, at least. )
Anyway, on such things, you will find that Medici Land ( Overstocks, Patrick Byrne) is way ahead of the curve and if the BIG DATA seduction is happening at SEGRO what is happening in Web 3 is highly relevant.
The centralised proprietary API accessed DataBase model is so Last Millenium, Money for old rope, I Say. The future is Distributed and individually devolved to each terminal input node. This has profound implications for Publishing and customer outreach and advertising. BIG Tech, Big Data and Prop Tech are both overrated and overvalued on the outmoded and outdated proprietary centralised database model, be warned! There are UBERS around every corner, and We Work’ lurking in every Balance sheet.
I give you,” Skills Matter, but so does Cashflow”.
Here is the latest publically available Balance Sheet of the very useful, but for me neither fish nor foul Costar UK Limited. As with Uber one wonders whose lunch they, Costar Uk Limited, has designs upon. The US operation is quite distinct, Brokerage in the US is a rather different business to how things work in the UK Market. In the UK it seems that the US directors of Co Star think things will change, EG and Property Weeks are clearly taking different paths but it’s not only their Lunch that is at risk.
That’s a selection of some of the Work on Distributed platforms we have been doing things with at Tychastic Investments Group over the past 5 years.
We are an Internet of Things, Think & Do Tank. The kind of Tank that does not invade Poland, does not Steal your Data and sell it back to you at an inflated price. More of a Giant Aquarium than a Goldfish Bowl and certainly with a good attention span.
