created Oct 7, 2016 - updated Aug 14, 2017
The "open web" and "silos", those terms are confusing. What do they mean?
Here's a 2010 post at tantek.com, titled What is the Open Web?.
tantek started or helped to start the Indieweb movement in 2010. In that 2010 post, he wrote about the open web:
In summary:
- open content and application publishing
- open ability to code and implement the standards that such content depends on
- open access to content, web-applications, web standards implementations (browsers), and the internet.
I like email, RSS, and http(s). The first and last are internet protocols. Throw in IRC, ssh, and sftp. More internet protocols. I like APIs and interoperability. IFTTT and brid.gy can act as conduits between the open web and silos.
To me, the open web means hosting content on one's own domain name, and the content is publicly available to all with an internet connection by using a web browser or command prompt tools, such as cURL.
And ideally, the website owners would fund their own serving hosting setups. This would be a server-hosted solution, instead of a CMS-hosted solution. The former provides more flexibility. But the server-hosted solution requires more computer tech skills than a CMS-hosted setup.
Websites that support the open web should use progressive enhancement to allow their content to be available when JavaScript is disabled or when their article pages are accessed via cURL.
The open web means publishers have the ability to change the look of their websites to be anything that they want. No limitations on theming.
Some CMS-hosted solutions provide enough customization options on the server and for the browser display to satisfy most authors. But the server-hosted solution permits maximum flexibility for choosing database servers, caching servers, web servers, configurations, and more.
The open web means that publishers can change the software that produces their website. This is an advantage of a server-hosted solution, which permits authors to use the same server while switching web publishing software. Or site owners can customize their publishing software. For example, authors can add support for the Textile markup language in addition to Markdown. Or they could create their markup language. Site owners could change their APIs. No limitations on programming. Authors can install nearly anything on the servers that they lease.
With a CMS-hosted solution, authors can move their content from one CMS-hosted solution to another. That generally means switching CMS businesses.
While I prefer the server-hosted solution, most authors don't want to be sys admins, which makes a CMS-hosted solution better, since the CMS-hosting company takes care of managing the operating system and the web publishing software for security updates and bug fixes.
The open web might mean that publishers only use a simple text editor to create static HTML pages, and they can markup their web pages anyway they desire. Authors can create and update their pages locally on their laptops, tablets, etc. and upload them to their web servers. Or they can log onto their servers to create and update their web pages.
The open web may mean that publishers have nearly unlimited flexibility. The authors rarely encounter a wall that cannot be knocked down.
That's my beginning definition of the open web. But the open web vs silo debate probably means different things to different people.
I read blogs that do not use a domain name, except for what Blogger provides, which would be yoursitename.blogspot.com. Is that the open web or a silo?
People who use Medium, WordPress, Tumblr, and Blogger without domain name mapping have carved out their own web spaces on the internet. That feels somewhat like the open web to me.
But those users have little to no access to modify the source code that creates their websites. Authors can make theme changes, but mostly, those users are walled in to what those services provide by default. And that may be good enough for those publishers.
- tweet by Indieweb developer:
Funny how "the open web" is now used to refer to closed silos like Blogger, Twitter, and Medium
It's comical that the above opinion is a Twitter post, but that Indieweb user stores his social media posts on his own website, which is how the Indieweb works. Any content posted elsewhere is also posted to the publishers' personal websites. The Indieweb mantra is post on your own site first and then syndicate elsewhere.
Anyway, if someone creates a blog at WordPress.com, which means using the WordPress or Automattic servers to host the content, then the publisher won't need to create a server hosting account with Digital Ocean, Dreamhost, Amazon Web Services, etc, which means the publisher won't need to download, install, configure, and manage the web publishing software. All of the server hosting and software management are conducted by WordPress.
The publisher is free to focus on a website design theme and creating content. The publisher's domain name would be yoursite.wordpress.com. Is that the open web or a silo? WordPress supports domain name mapping, which means publishers can buy domain names and point the names to their WordPress accounts, creating yoursite.com. Is the latter more open web than the former even though all hosting is the same?
This is what's confusing. In the above scenario, the setup meets most of my requirements of the open web. The author with the above setup cannnot make significant overhauls to the server-side source code. But most publishers don't care about changing the CMS source code. Besides, authors can install Wordpress plug-ins that satisfy their requirements.
When small publishers such as The Awl moved their content from their own server hosting accounts to Medium, did those publishers move from the open web to a silo even though their domain names remain unchanged? Confusing. At least when The Awl hosted on their own servers, they could change the CMS source code, and they could create a unique look on the front end. But many publishers want to focus on writing and not on programming, desisning, and sys-admin functions.
If I host at WordPress.com, Blogger.com, Medium.com, Tumblr.com, DigitalOcean.com, or Amazon.com/aws, how are some hosted solutions considered a silo while others are considered the open web? Again, self-hosting at DO or AWS enables full control over how the site is built and how it looks. Maybe degrees of open webness exist. That would make things even more confusing.
In my opinion, public websites that are meant to be read by browsing-only users and are built as a JavaScript single page applications (SPA) do not support the open web. That's an easy one, regardless of where the site is hosted. When JavaScript is disabled, these sites display no article content.
When SPA pages are curled with a command prompt utility, the downloaded content consists mainly of JavaScript code or a link to the JavaScript that will fetch the content and display it dynamically in a web browser that can execute JavaScript. SPA pages fail in the text-based Lynx web browser. When a web page is considered not curlable, it means that the page contains no content.
2015 tantek.com post: js;dr = JavaScript required; Didn’t Read
Pages that are empty without JS: dead to history (archive-org), unreliable for search results (despite any search engine claims of JS support, check it yourself), and thus ignorable. No need to waste time reading or responding. Also known as, if it’s not curlable, it’s not on the web.
Excerpt from a 2015 Hacker News thread about curl:
"If it doesn't load through curl, it's broken." --someone
... requiring code execution in order to read data [text] is madness. I wasn't saying not to do the fancy stuff but rather to start with something which degrades well and then have your JavaScript enhance that basic experience.
That last sentence touches on the concept of Progressive Enhancement.
If the SPA or JavaScript-heavy website, however, is a "web app" that requires a user to login to perform functions, such as using email, preparing tax returns, building a market research survey, banking, shoppping, etc., then I can understand the use of JavaScript, and I have no problem with it. But if JavaScript is required simply to read a website, then that's anti-open web.
If an app other than a web browser is required to be downloaded and installed simply to read a website, then that's also anti-open web. It doesn't matter if the app uses web technology, such as REST, JSON, http/https.
But users may prefer apps over the open web, even when acting as browsingly-only users. Designers and developers respond to user preferences.
An "old" 2010 Wired article with the sensationalistic title The Web is Dead. Long Live the Internet
As much as we love the open, unfettered Web, we’re abandoning it for simpler, sleeker services that just work. You’ve spent the day on the Internet — but not on the Web. Over the past few years, one of the most important shifts in the digital world has been the move from the wide-open Web to semiclosed platforms that use the Internet for transport but not the browser for display.
It’s driven primarily by the rise of the iPhone model of mobile computing, and it’s a world Google can’t crawl, one where HTML doesn’t rule. And it’s the world that consumers are increasingly choosing, not because they’re rejecting the idea of the Web but because these dedicated platforms often just work better or fit better into their lives (the screen comes to them, they don’t have to go to the screen).
I store most of my images at Flickr to be embedded into my web posts. That's using a silo to support my open web sites. If Flickr disappeared without warning, then my pages would contain missing images. But I backup my images to DVDs, or I used to, which permits me to retrieve the images to be uploaded to either my own site or to another image hosting service.
I also store some of my images on a server that I pay to use, and I use my own image uploading web app: http://waxwing.soupmode.com. My Waxwing JavaScript code works within the browser to lower the quality and size of the images, which reduces the upload time and reduces the amount of space consumed on the server.
I use servers that are hosted at Hurricane Electric, Digital Ocean, and Amazon. I'm not using a server in my own home, except for my Tor site http://zwdqwr2p2xwkpbyv.onion.
When I host at DO or AWS, is that supporting or opposing the open web? I could lose my websites if for some reason the server hosting company deleted my accounts, went out of business, or my credit cart expired.
When a Facebook page is only accessible when logged into Facebook, then that's a silo or closed web. It's odd that some businesses set their Facebook pages to a private setting that excludes people like me who do not have an active Facebook account.
But if the business's Facebook page is available to everyone, is that the open web or still a silo? Based upon my requirements above, it's a silo. The business owner has little to no access to change how the site looks and functions.
The publisher chose to host content at Facebook, instead of at WordPress or Medium. Maybe the publisher is unconcerned about owning a domain name. Maybe the publisher does not want to install and manage software on a server. And small business owners have more important things to be concerned about. Unless some guidelines are violated, the Facebook page will probably last as long as the business.
Quote by someone:
"If you aren't paying for something, then you aren't the customer. You're the product being sold."
One aspect of the open web could be that the author's content and actions are not re-purposed by the hosting service and used for targeted advertising. That eliminates Facebook from open web contention.
Some people think that in the future, independent websites won't exist. People, businesses, and orgs will publish on platforms, like Medium.com, Facebook, Instagram, and Snapchat. It's already that way for some small, locally-owned businesses in the Toledo area, which do not have a website with their own domain name.
An easy example of publishing to a silo is Facebook's Instant Articles, which only functions within Facebook's native app for the phone. I don't think that IA works in the company's native app for tablets. It's phone-app-only. Obviously, IA does not work for the web on any device.
Facebook's Instant Articles relies on a Facebook-flavored RSS file. RSS is an open web example, but the default RSS feed of websites won't work. CMSs need updated to produce a different RSS file for IA.
Instant Articles is an interesting publishing mechanism, since updates on the publishers' websites get accepted by IA. But again, it's available on Facebook's native phone app only.
Phone-app-only is not the open web. I don't want to install nor use native apps on the phone except for web and gopher browsers. A theory: The Open Web is the Answer to App Censorship.
The debate will continue, regarding the open web and silos. Most users, publishers, businesses, and other orgs don't care about such trivia. They're too busy with other facets of life. They want to use the easiest tools available that provide the most reach to their readers, fans, customers, and contributors.
And if the preferred publishing and communicating platform for these users is Faceook, then that is NOT Facebook's fault. It's the fault of the open web geeks for failing to create easier systems.
The Blogosphere of the late 1990s to the mid-aughts was a holistic, organic, decentralized social network of sorts that relied on personal websites, comments, RSS, pingbacks, trackbacks, and the blog search engine Technorati.
Facebook launched in February 2004, but it remained closed to certain email addresses or users until September 2006 when Facebook opened up to everyone. Facebook to the world is only 10-years-old. The internet began in the late 1960s or early 1970s. The web began around 1990. The web became popular to the masses by the mid 1990s.
The open web fans had a 10- to 15-year headstart on Facebook. And now 10 more years have passed.
Automattic/WordPress permits users to download, install, and customize their software. That supports the open web. WordPress is the most popular CMS. Numerous other CMS, blog, and wiki systems exist that can be freely downloaded by publishers.
People who have built open source products have contributed to the open web. Examples include publishing systems, web servers, database servers, caching servers, programming languages, and frameworks.
I think that the message boards, such as Hacker News and the Stack Exchange sites contribute to the open and closed web, since discussions help inspire people to build things, and meaty discussions can alter the opinions of makers.
In my opinion, the Indiweb people have done the most to support the open web. The Indieweb attracts new converts every year. Currently, the Indieweb ideas are probably too technical for mass adoption. But the Indieweb is taking a long, slow approach to change.
We don't know what the next big thing will be. We don't know for certain how people will be creating, sharing, and responding to content five to ten years from now. Hopefully, some of the Indieweb ideas will creep into the mainstream.
Update with info below
https://www.theverge.com/2017/5/24/15681958/what-is-web-definition
http://boghop.com/2017/05/23/the-internet-is-not-broken.html
http://boghop.com/2017/05/23/antiamp-rhetoric-may-2017.html
http://boghop.com/2017/06/01/antifacebook-rhetoric-from-oldschool-bloggers.html
http://boghop.com/2017/06/11/june-2017-hn-discussion-about-the-open-web.html
http://boghop.com/2017/06/02/indieweb-links-jun-2-2017.html
http://boghop.com/2017/06/10/reading-without-javascript.html
http://altplatform.org/2017/06/09/feed-reader-revolution/
q.
Making matters worse, these social silo readers have typically, if not uniformly, turned off all external access to their own RSS feeds long ago. If you want to read content in Facebook, you have to log in and have an account and participate there directly, you cannot just subscribe to five peoples’ content via RSS and read it anywhere you want. This monopolistic behavior is exactly the reason we call them silos. Content goes in, but doesn’t come back out.
q..
http://jothut.com/cgi-bin/junco.pl/replies/92639
from https://chat.indieweb.org/2017-07-11
here's a problem with using a cms-hosted solution for maintaining a personal website. tumblr may deactivate or recycle a username (name.tumblr.com) if the owner has not posted in a while. #indieweb #hosting - better to choose a server-hosted solution #blogging
From: JR's : micro blog - Jul 11, 2017 - reply
3 replies
JR: apparently a tumblr email to an author: "It's been a while since you've been on Tumblr, and we wanted to make sure that you're still interested in using the username [username]. If so, just hit this button:"
- 33 mins ago - # - reply
JR: "If not, you don't have to do anything. If we don't hear from you within two weeks, we'll just give you a temporary username and release your old one back into the wild."
- 33 mins ago - # - reply
JR: "You can come back and change your temporary name to whatever you want, whenever you’re ready. Your content will all still be here when you get back."
- 32 mins ago - # - reply
algorithm controls what readers see. anti-open web.
http://boffosocko.com/2017/07/11/the-facebook-algorithm-mom-problem/
is it anti-open web if authors cannot custom their websites and individual article pages? this is a problem with a cms-hosted solution.
http://practicaltypography.com/billionaires-typewriter.html
https://indieweb.org/admin_tax