Disclaimer: This article is intended to convey more market information and does not constitute any investment advice. The article only represents the views of the author and does not represent the official position of Mars Finance.
Editor: Remember to pay attention
Source: Orange Book
Original author: Sinclair
原文标题：《The Rise and Demise of RSS》
Original compilation: Yao Changlin
preface:
The older generation of Internet people reading this article should be very touched, and the new generation of blockchain will also see a lot of familiar phenomena. Huo Ju, a player with both identities, feels more complicated, and he recommends this article like this:
"By the point in time of 2019, looking back at the history of the emergence and decline of RSS can more clearly see the gains and losses." This article provides a complete history of the rise of RSS in 1999 to the closure of Google Reader in 2013, when content reading and distribution were replaced by social networks.
I recommend this article because this history has a lot of similarities with blockchain, RSS is a more distributed, more evenly entitled, more open system than today's social networks, but it was ultimately defeated. One of the important reasons is because centralized products are easier to design and provide a better experience, easier for users to understand without having to learn complex concepts, and better ways to make money for running businesses.
The contrast between blockchain and centralized products today happens to be the same, this time, can decentralized products go a different way? ”
This article is a bit long, there are a lot of strange and ancient English technical vocabulary, but it doesn't matter if you don't understand the technical details. In addition, there are many English names in it, we have replaced several important names with Chinese, and when they first appear, they will be marked in red, remembering that these people will not be lost.
More than ten years of old netizens should be familiar with RSS. In fact, RSS has two different definitions, Really Simple Syndication and Rich Site Summary, but both are essentially a program-friendly way of subscribing to information. There are still apps and websites that use RSS technology today, but for most users, RSS has become a vague concept.
Looking back at the history of RSS, there are two stories worth telling: the first is a great vision for the future of the web, which has never been realized; the second story, the collaboration to develop common standards, is how it became the most controversial schism in the history of open source software.
In the late 1990s, a dreamy 10 years between Netscape's launch and the .com bubble, no one knew where the Internet would go, but everyone realized that the Internet would be more imaginative than it is now. There is speculation that the Internet will revolutionize with "converging networks." Originally, the Internet was a point-to-point transmission, passing information from the server side to the user in one direction, but the new model may break the original architecture and repackage and distribute the whole network information in a "channel" manner.
At that time, there was a influential Newsletter among investors, called Releas 1.0, and Werbach predicted in an article: "The aggregation network will evolve into the core model of the Internet ecology, and companies and individuals can control their online identities and enjoy the benefits of the vast network."
The future of RSS was once so bright, what happened then?
Weibach asks the reader to imagine a scenario in which a fencing enthusiast has two options when buying a epee, one is to log in to an e-commerce website or run to an offline store, and the other is to log in to the fencing website every day and click on the right side of the advertising bar to buy it. Similar to the radio and television network, the programs of large TV stations can be broadcast on local small TV stations, so that more people will watch. Aggregation networks can provide information to users through intermediary websites. This makes it easier for users to control their information interactions with the Internet.
RSS is one of the most promising standards for this convergent future. Weibach sees RSS as "the paragon of lightweight aggregation protocols." Another contemporaneous article argues that RSS was the first protocol to realize the potential of XML (Extensible Markup Language). RSS helps readers and content aggregators customize channels from the sites they want.
But today, 20 years later, with the rise of social networks, Google shut down Google Reader, RSS is only used in podcasts, technical podcasts and some news sources, becoming a technology that is slowly dying. Indeed, there are still many people who rely on RSS readers and stubbornly add RSS to blogs as a sentiment. These insistences turned into a protest against a centralized network, a protest against several large companies that controlled the entire network, against a network that was completely different from Weibach's imagination.
The future of RSS was once so bright, what happened then? Is the collapse of RSS inevitable? Is it really the result of infighting over standards?
Stuck in the mud
RSS was invented twice, which means it doesn't have a recognized inventor and gets bogged down in endless bickering. But it also shows that RSS is a good idea at the right time.
In 1998, Netscape was looking for the next user increment, and its flagship product, Netscape Browser, once held an 80 percent share, is now rapidly eroding its market share by Microsoft's Internet Explorer browser. So Netscape needed a new project. In May, Netscape convened a team to embark on a new project, Project 60, an RSS journalism tool based on Ben Hammersle's Atom project. Two months later Netscape released its new portal product, My Netscape, which competed with Yahoo, MSN, and Excite.
In March of the following year, Netscape added a new feature to the portal My Netscape, "My Netscape Network." Users can customize the My Netscape page, add channel functionality, and subscribe to the latest headlines from other sites. As long as the website publishes a "specified format" file, users can click "Add Channel" on their My Netscape home page to subscribe to the user's favorite website. A module containing website headlines will bring up the user's My Netscape page.
This "specified format" file is the RSS file. But in the My Netscape Network's statement, Netscape defines RSS as "RDF Site Summary." This definition is not precise, because RDF (Resource Description Framework) is a syntax that describes specific properties of a specified resource.
In fact, the 1999 W3C also intended to draft RDF standards. Although RSS should theoretically be based on RDF, the RSS reference documentation given by Netscape does not use any RDF tags at all. In the Netscape RSS specification documentation, its author Dan Libby mentions that in the MNN version Netscape deliberately limits the complexity of RSS. The version number of this specification is set at 0.90, which means that subsequent versions will be more consistent with the W3C standard.
The original RSS standard was developed by Libby and Netscape colleagues Eckart Walther and Ramanathan Guha. Guha's email mentioned that much of the ideas Guha and Walther developed came from Libby's early ideas; after AOL acquired Netscape, the two of them left and Libby took on the main update. Guha has contributed a lot to the development of RDF, and he and Walther intend to apply RDF to RSS. Later, Libby mentioned in the email that the RDF version of RSS was eventually deleted for two reasons: the first was the time limit, and the second was that RDF was too complicated for ordinary users.
At a time when Netscape was caught up in a "portal war" trying to win page views, "web blogs" quietly became popular. Winer, CEO of UserLand Software, invented the first content management system that allowed ordinary people who didn't know how to build their own blogs. Wiener's blog Scripting News is one of the oldest blogs on the internet. More than a year before Netscape released my Netscape Network, on December 15, 199, Wiener announced that its blogging product would support both XML and HTML formats.
Wiener's XML format is known as the "ScriptEd News Format". Some argue that it is similar to Microsoft's channel definition format, but there is no documentary evidence. Like Netscape's RSS, Wiener defines the structure of the text that other programs can read and call. When Netscape released RSS version 0.90, UserLand announced support for both formats. But Wiener found Netscape's version "lamentably incomplete" and "missing the core parts that authors and readers need most." It can only reference a series of links, while the Scripted News Format can contain more: it contains more paragraphs, and each paragraph can have many links.
In June 1999, two months after Netscape released my Netscape Network, Wiener also released a new version of the "Scripting News Format", ScriptingNews 2.0b1. Wiener accelerated the promotion of its own standards, and users did not recognize the huge flaws of RSS 0.90. Wiener's version adds some new elements that are compatible with RSS. But there's a huge difference between the two standards: The format Wiener calls "fat" can contain an entire article, not just a few links.
In July, Netscape was about to release RSS 0.91, and one of the major problems was to update the text specification. RSS no longer stands for "RDF Site Summary", but "Rich Site Summary". The RDF elements have all been removed. Many scripted news tags have also been merged. In this text specification, Libby explains:
The RDF reference has been removed from the standard. RSS, which contains RDF, provides a metadata format for website summaries. There are two important considerations for deletion. First, data parties need to provide an aggregated data format, not a metadata format, and RDF files must be very precise to meet the standard. However, this will cause difficulty in reading, and it is difficult to manually develop RDF files directly. Second, there is no tool to generate RDF files. Based on the above two points, we decided to adopt a standard XML approach.
Wiener was so pleased with RSS 0.91 that it declared "better than I thought" as an alternative to the ScriptingNews 2.0b1 format. At one point, it was agreed that there would soon be a unified official version of RSS.
Great division
After 1 year, we found that RSS 0.91 has many shortcomings. There are a lot of things that people want to do with RSS that version 0.91 can't do, and there are a lot of superfluous restrictions, such as a maximum of 15 links per channel.
At this time, RSS technology has been widely used. Netscape has shown little interest in RSS 0.91, and most other companies are using Wiener's standards, such as O'Reilly Net's RSS aggregator Meerkat, news aggregation sites Moreover.com, and so on. Representatives of various stakeholders communicated by email on how to improve the 0.91 standard, but no consensus has been reached.
Disagreement over namespaces is disagreement over the nature of RSS
A collection of messages called Syndication, which records all messages discussed about the RSS standard, is still accessible today. Still a valuable history today, documenting how these deep divisions ultimately tore apart the entire RSS community.
The representative of the divided side is Wiener. He was keen to improve RSS, but iterated in a relatively conservative way. In June 2000, Wiener released version 0.91, but it was not much different from the version of Netscape. Wiener said in his blog that since Netscape no longer maintains updates, 0.91 serves as a starting point to show how RSS can be used in real-world scenarios.
At the same time, he also believes that the easy-to-use RSS has been successful enough; the complex new features mentioned in the Syndication email will not provide any value for content distribution. In particular, he was opposed to adding namespaces and to restoring deleted RDF forms (namespaces allow programmers to customize the subformat of RSS, meaning that new features require everyone's agreement. But namespaces also make it harder to read and write to RSS. In the Syndication email group, Wiener mentioned that these changes are very important and may lead to a split in the standard:
We're also thinking about how we can move forward with RSS. Of course, I also want to use ICE-like content in RSS2, and the publish and subscribe functions are also highly prioritized, but everything is premised on simplicity. I also want more room to expand, but definitely not to use "namespaces", "schemas" or repeat the old RDF path. I understand that there may still be people who need to keep these features, so there may be a new fork version. I have a lot of ideas for the forked version, which will be announced to you when the time comes.
The opposing side of Wiener is dominated by 3 people, Rael Dornfest of O'Reilly, Ian Davis, CEO of Search Field startup Calab, and Aaron Swartz, 14. Swartz is the co-founder of Reddit, a well-known hackathon, whom we are familiar with. In an email Davis sent me, in 2000, Swartz's father often accompanied him to technical conferences.
All 3 of them agree that RSS needs namespace capabilities to meet the different needs of users. In several other emails, Davis suggested building a namespace-based module that would make RSS more extensible and less complex. The pro-namespace camp believes that RSS will soon be used not only for synchronized blogs, but also for many other use cases. Namespaces are the only scenario without adding complexity.
The controversy over namespaces is only a symptom, and the core of the controversy is "what exactly should RSS be used for?" Wiener first made his standards to synchronize his blog. Netscape released the RSS standard to build a miniature website in the portal. Some argue that Netscape's original intent should be respected. In an email to Syndication, Davis mentioned that RSS was originally designed to build "mini-sitemaps," but with a lot of new needs over the past year, RSS should now expand to support more information types, not just simple news headlines.
This expands Netscape's planning for RSS, and Libby mentioned in an email to me that the debate over the development of RSS centers on" vs. "Making it easier for people to publish their own work."
In his reply to Davis' email, Wiener mentioned a completely different logic: Scripting News was the first RSS network, and the purpose was completely different from Netscape. The community is divided over the inventor of RSS and his goals, and divisions are inevitable.
The version split occurred after Dorne announced the proposed RSS 1.0 specification and formed the RSS-DEV Working Group (which included Davis, Swartz et al., but none Wiener). In this version 1.0, RSS is again defined as "RDF Site Summary" and RDF elements are added again. Considering Wiener's historical contribution to the spread of RSS, version 1.0 did not delete Wiener's name. But version 1.0 also mentions that RSS will not follow the path planned by Wiener. Simply adding some elements to RSS without considering scalability, RSS will lose a lot of application scenarios. Version 1.0 also defines the module system based on XML namespaces.
RSS-DEV arbitrarily named it "RSS 1.0", and Wiener was outraged. In a separate email, he mentioned that he had a huge work stolen, presumably referring to O'Reilly and its RSS-DEV working group.
Others in the mailing group also believe that the RSS-DEV team should not use the name RSS until it has community consent. But the Working Group still insisted on using it. Dan Brickley, a member of the working group, argued that RSS 1.0 is based on the earliest vision of RSS and can be traced back to MCR (the predecessor of RDF) and CDF, among others. He also believes that RDF is originally part of RSS, and that the 1.0 version contributes far more to RSS than Wiener and is more worthy of the name RSS.
The RSS-DEV Working Group released its final version in December. Around the same time, Wiener released its own upgrade to RSS 0.91, RSS 0.92, and several changes to the upgraded version were quickly adopted by various podcasts. At this point, RSS officially split.
If the RSS-DEV working group had seriously invited Wiener to join, the split might have been avoided. Wiener is clearly important, and the working group acknowledges that he is the lead author of Syndication, making an important contribution to the popularization of RSS. But Davis's email also mentioned that Wiener wanted to control RSS and wanted to turn RSS into private property, so he didn't want to work with us. Wiener declined the working group's invitation. Tim O'Reilly, CEO of O'Reilly, explained at a Symposium at UserLand in September 2000:
Everyone gathered to discuss the evolution of RSS, and Wiener was also present. When the opinions of those present turned in a direction he did not support, Wiener withdrew, saying O'Reilly wanted to take his place through discussion, even though O'Reilly's Dorne was only one of a dozen authors, and Dorn had gone through the entire history of RSS.
Wiener replied in writing to Tim O'Reilly:
I had just met Dale two weeks before the conference and he didn't mention RSS 1.0 at all. Before the launch, I was on the phone with Dorne on Friday and there was no news either. The first time I heard about RSS1.0 was in the official announcement.
Let me ask you an impolite question, what would you do if "RSS 1.0" had been conspired out without any votes and discussions, without letting the committee come together to make decisions?
UserLand has done a lot of work to develop and popularize RSS. Now you kicked away and took the name. This is very excessive. If I want to keep developing, I'll have to use a new name. Tim, you tell me, why did this all happen, and how did it happen?
I didn't find any discussion of using the RSS 1.0 name in the history email. Wiener said in the email that he was not trying to control RSS, but just wanted to use it in the product.
Many developers, tired of the endless debate in the community, decided to build a new version, and in 2013, the fork happened again. The developers developed a new version of Atom — eliminating RDF but embedding the XML namespace. Atom was submitted as the final version to the Internet Engineering Task Force (an organization responsible for building and promoting Internet standards).
Since then, there have been 3 different versions of RSS on the market: Wiener's 0.92 (updated to RSS 2.0 in 2002 and renamed "Really Simple Syndication"), RSS-DEV Working Group's RSS 1.0, and Atom. Still active today are RSS 2.0 and Atom.
decline
Different RSS standards did hinder the spread of RSS, but they did not prevent RSS from becoming popular in the 2000s. In 2004, the New York Times began using RSS to provide headlines, and began to popularize RSS and usage methods to ordinary users. Google Reader, which later had millions of users, was also released in 2005. By 2013, RSS was popular enough that the New York Times even declared in Swartz's obituary that RSS was "everywhere." Before a third of the people on the planet signed up for Facebook, RSS was the only link that many people had to internet news.
The New York Times published Swartz's obituary in January 2013. At this time, RSS has reached an inflection point and has gradually become a product that everyone is not familiar with. Google Reader closed in July 2013, ostensibly due to years of declining users. Many online reviews also claim that RSS is dead. But before Google Reader shut down, fewer and fewer users were using RSS. In May 2009, Steve Gillmor wrote on TechCrunch: "It's time to shut down RSS once and for all and switch to Twitter. RSS couldn't be worse. He noted that Twitter is a better information subscription tool because Twitter can provide different people's perspectives in addition to providing articles.
Today, RSS isn't dead, but it's far less popular than it used to be. Many people have tried to explain the current state of RSS, but the most convincing explanation comes from Gilllmor's proposal in 2009: social networks have replaced RSS, providing users with updated information and bringing benefits to the businesses that run social networks. It's like Google shutting down Google Reader to promote Google+. Because Google does get money from Google+, and Google Reader can't make a profit. In 2013, Marco Arment, founder of Instapaper, wrote on the podcast:
The shutdown of Google Reader may seem accidental, but it is actually the latest victim of the war between Facebook and Google on the Internet. It seems that Google Reader still has a large number of users in use, but it actually conflicts with the Google+ strategy: Google needs people to use Google+ to read and share in order to compete with Facebook for users' time, advertising data, advertising revenue, growth, and others.
As can be seen, users and tech companies believe that social networks are more efficient than RSS.
Another theory about the demise of RSS is also interesting. The New York Times, which has always wanted to recommend RSS to users, complained in 2016 that RSS is not friendly enough to ordinary users and is too geeky to use. In 2004, until the RSS icon was updated, the New York Times had been using orange boxes to link to RSS data sources, clicking on a web page with XML links all over the screen, daunting ordinary users. This wonderful tweet points out the essence of the demise of RSS:
Ordinary users do not find RSS easy to use, because RSS is not designed for ordinary users, there are too many technical thresholds. Once there is a better product, users will abandon RSS.
If iterates efficiently, RSS may be better used, perhaps RSS can connect people who subscribe to the same channel and share ideas with each other; perhaps the adaptation of browsers can be improved and the user experience is better. But while members of the RSS community are trying to create consensus, large companies like Facebook are rapidly upgrading their products and breaking the mold. While the community is still on the same page, efforts to improve the product are wasted on repetitive work.
Davis told me that if the community could compromise with each other and quickly form a consensus, then Atom would not exist, and the time spent arguing could be used to improve the product. So, when we ask ourselves why RSS is in decline, the first answer is that social networks replaced RSS; but ask deeper, why did social networks replace RSS? The answer, perhaps, is that RSS developers face more difficulties than building Facebook. As Dorne wrote to the Commission, "The political problems are far more serious now than the iterations of continuous development.".
So we're still drowning in information silos. Even so, the aggregation network that Weibach predicted in 1999 has been realized, but not in the way originally envisioned. After all, The Onion News was disseminated through aggregate networks like Facebook and Twitter, as did Seinfeld.
I consulted Weibach, who agreed with me. He sees RSS as a failed technology because it doesn't integrate the blog world, the content world, or different resources. But the revolution of social networks also lies in the ability to aggregate different content and resources, which is actually the original idea of RSS and aggregation networks.
Unfortunately, aggregating information on the modern web exists on only a handful of websites, meaning that no one can manage their online reach as much as Weibach imagined. One reason is that RSS doesn't give tech companies the opportunity to control access and sell ads, so tech companies won't support RSS.
One reason for this is that the uniform standards of a centralized company are much simpler than the standards that require community voting to form a consensus. Consensus is difficult to build and a waste of time, and without uniform standards, developers will propose multiple competing standards. If this article offers any lesson, it's that for a better and more open web, the world needs better collaboration.