mugnyte's Journal: Reading as a Transaction

There's been a bit of flurry in marketing and advertising circles ever since Danny Carlton posted a rant asking for detection of Firefox's Adblocking plugins in web sites and for them to stop delivering content if advertising could not be sent as well. This went through the usual wash and wear cycle here on slashdot and digg with the expected results: Silly boy! Many good arguments were made, here's an expanded summary from Digg, Adblock's own forum and Slashdot:

Is it illegal to block ads?

• There's little legal ground to claim a web site is a copyrighted work in total with its ads. The true intellectual content of a web site is always in the new information presented. Otherwise, a conglomeration of ads in a specific configuration (with no original content at all) could be considered copyrighted and of value, when it is commonly understood today that such pages are "empty".
• There are no laws requiring any content gained from web sites to be strictly bundled with their advertising in all mediums. For example, one may copy/paste text from a web site, store the sent HTML file, or simply re-write the concept into one's own web without original site without breaking laws, save for the copyright restrictions the source has outlined (noncommercial/personal use is always open).
• Radio and TV consumers who skip/record around commercials are not stealing, a legal decision descending from the tape recorder, VCR and CD, with DVD muddying the waters with the prevention of consumer breaking protection schemes. Technically, the equipment makers have been "incentivized" to not provide automatic commercial skipping by content generators, but nothing forces one to consume all ads on a recording - you can skip them without infringement, albeit perhaps manually. There are no ad delivery mechanisms online that can be construed as using "protection scheme" that ad blocking software would be "cracking". Indeed, the requirements for having a proven and powerful protection mechanism have been tested legally several times, and trivial cases have been discarded.

Is it technically feasible to prevent ad blocking?

• Browsers are under no obligation to display any portion of an HTML file they deem incompatible with their parser, display system, etc. Audio, print and text browsers fall under these scenarios already, as do older browsers. In total, there is no lasting way to discern one browser from another; if a Firebox release provides an "Exact IE mode" (identification, behavior matching per javascript idiosyncrasy, etc) the web site could not make a distinction between Firefox and IE.
• If a web server forces its own content to be delivered only after all advertisements are sent to a web browser (instead of sending content inline as is typically the practice now), a totally new paradigm must be built into many content delivery systems. In these systems, advertisements source from other servers and can be requested in any order. Additionally, even if all content is delivered, it need not be displayed, as outlined above.
• If all advertising is delivered from the source content web server, instead of various other ad-syndication servers, ad blocking must discern between content and commercial. However, ad blocking is not typically based on the semantic meaning of the content, but the format (image, animation, etc). Thus, unless a server revolves the names of the files it serves as advertising, it will have those files blocked after their first impression, and other file (content) will be kept. One could argue that this is "post impression", but as modern blockers share information from client to client, a browser could find that all of its revolving image names are blocked on all clients after one impression anywhere, which goes back to blocking pre-impression.
• Ad blocking is implemented in some fashion on all major browsers (IE, Opera, Safari). Singling out Firefox is simply from a lack of knowledge about the browser marketplace. This is a feature born of the grassroots request for browser extensibility. There is no technical way a browser itself could regulate the actions any of its extensions, save for the API to influence the root browser behavior. The HTML stream and request stack is usually under the control the extensions - this is the core method for providing extension functionality.
• Ads that use Flash (or now, Silverlight) must ask that browsers have this extension enabled. Nothing requires a browser to have this besides the site itself. Some sites deliver a non-flash version, some do not. As Silverlight and Flash, and possibly any other graphics package compete for market share, web sites cannot expect all browsers to have these extensions present. Thus they must accept that only a subset of total web users will use their site if they employ such delivery mediums. This is the status quo and is well understood.

Is it unethical to block (some) content from web sites?

• Every real-world metaphor applied to Mr. Carlton's request opinion fails to prove his point. Customers who discard newspaper inserts without reading them are not stealing. Passengers that close their eyes to roadside billboards (revenue from which can pay for maintenance of the road they are using) are not stealing. Removing bundled software with a package software install is common and in most cases prudent.
• Web sites that use advertising as primary income without selling a product themselves are highest impacted from ad blocking software. However, the information they provide is essentially free and replicated once online (the original content may not be replicated exactly, but the information is paraphrased, summarized and editorialized). Thus, one cannot expect such information to be a constant source of income via ads as that information becomes "stale". Only through constant updates, living up to the "content is king" euphemism, will a site continue to draw viewers. Such a provider must choose between a subscription model and an ad-support model, and each will have different advantages. Neither can be "policy" for online server behaviors.
• Mr. Carlton explains that the "monetization of the internet" is the fuel behinds its growth. Historically, this is patently untrue and for a large swatch of the WWW's lifetime, there was a struggle to provide commercial services. Advertising is not the primary method for information dissemination on the medium, since text-based searching remains king for information retrieval. The bundling of ads with content can be a source of income via Google AdSense service, but this is not necessary for all web sites to exist and profit.

  Overall, Mr. Carlton's comments express an opinion that suffers from a short-term knowledge of the technology of the internet. Since its inception decades ago, the continuing design decisions are around freedom of the client side. Information can be gather and collected from almost any source server in an automated fashion. Advertising attempts to link and defend the cost of internet participation with the transfer of commercial information. This has never been true, and today as savvy users remove unwanted information from web sites the core design features become present again: The client can do anything it wants with the information your server provides. Indeed, this freedom is how innovation and creativity caused the web to explode in popularity entirely; the web grew from having an immense data source of users, references, content, and techniques available for manual or scripted use.

  The rise of ad blocking can be attributed to the disconnect of a web site's content to the relevance and design of its ads. Most unintelligent ad servers do not screen their delivered content for offensive or annoying presentations. Web sites have traditionally had little say in which ads could be rotated through their system, save for the largest categories of pornography, violence, etc. The top provider in the market at the moment, Google's AdSense, does allow for the strongest intelligent delivery and ad filtering concepts. It also works to remain relevant to not just the site, but the content itself, as well as any behavior about the client that can be tracked. This results in the least annoying, and subsequently, least blocked ads. One could infer from this relationship that if web sites took more personal responsibility for the ads they delivered, users would accept/reject the site in total, rather than blocking/partitioning the content.

  This author expects that, like all things online, this meme of Danny Carlton's will not die fully but live on in a limbo in perpetuity. Functionally it will not change much of the function of the web. If a large percentage of the online users implement file/script blocking, we may see a movement for advertisers to provide full-page ads that require an impression before moving on to ad-free pages. As an example: Salon.com experiments with this format with some success. It cookies the impression and one can use the site without (full-page) ads afterwards, and also provides a 2nd tier subscription model for more content. For Salon this is a good balance, since the site seems to take responsibility for it's advertising, much like the models for TV broadcasters.