News: 1626765906

  ARM Give a man a fire and he's warm for a day, but set fire to him and he's warm for the rest of his life (Terry Pratchett, Jingo)

Ad tech ruined the web – and PDF files are here to save it, allegedly

(2021/07/20)


In January, an online publisher launched a website called Lab 6 that serves its content as a PDF to protest the state of the modern web, and has caused quite a stir.

There's nothing novel about posting PDFs to the web, but doing so as a protest against web technology is akin to taking a stand in the [1]tabs-spaces debate – PDFs have long been [2]a polarizing topic . Since then, two more issues of [3]Lab 6 have been published, [4]prompting disagreement among other web developers about the wisdom of this approach.

The author, who asked to be identified simply as "James" in an email exchange with The Register , justified use of the PDF format as an attempt to push back against modern web technology.

[5]

"PDF has many shortcomings," he wrote in his initial issue. "But in its cold, immutable fixity, it stands in opposition to the mercenary, dynamic web of rubbish."

[6]

[7]

His post cites the toxic effect of advertising on the web, complaining about how search engine optimization has led to the generation of "megabytes of adverts, distractions, upsells, misdirections, invitations to sign up to newsletters, cookie warnings, [and] GDPR warnings."

It rails against the mutability of web pages, updated without notice in an attempt to keep the attention of search engines. It laments the tracking and monitoring scripts that litter websites. And it decries the capture of the web standards community by browser vendors.

[8]

James, in short, rejects the way the web has evolved from a document-centric platform to an application-centric system. His solution is to publish files as PDF/A, "which forbids interactive content (normal PDFs can contain JavaScript!) and ensures your PDFs are absolutely self-contained, even embedding the fonts."

'Ignorance and stupidity'

[9]Maximiliano Firtman , a mobile and web developer and author, dismissed the cri de coeur as noise. "PDF Is not a format suited to share in different formats and diverse devices," he told The Register . "It's a format created for printing. So it's like using a boat to drive across a street."

"There are much better choices, starting with just HTML with no external resources, epub, [or] markdown," he said. "PDF is not responsive, it's slow, it can't be translated, it's insecure, it's not accessible, [and] it can include embedded JavaScript proving it can be as harmful as any web page. It's just ignorance and stupidity."

In an email to The Register , James explained that Lab 6 is a personal project. "My employment is in an industry completely unrelated to web technology," he said. "My interest in the web is purely non-commercial, growing out of the Geocities pages I used to make in the 90s as a kid. Lab 6 is about more than shilling for Adobe (there’s a longitudinal capture-the-flag forensics exercise hidden in it, for example)."

He said he was surprised his site had received social media attention and noted that his most recent post is a hybrid PDF and [10]Gemini file , "which might go some way to addressing a lot of the hate that PDFs attract."

[11]

James said there wasn't a single moment where web technology went off the rails for him. It's not all bad, he said, but he cited four specific concerns.

First, he objects to the movement away from HTML as a finished specification.

"HTML is currently effectively just whatever the browser vendors say it is, written down in a standard of sorts, but with no end in sight," he said. "It will never be finished, so the bugs will never be fixed, and the complexity has grown so immense that nobody other than the incumbent browser vendors can realistically implement the web at all any more."

Second, he took issue with the way independent bloggers have moved to specific platforms rather than running their own services. He said he could understand that as a response to the hostility of the internet – who wants to run their own infrastructure when it's under constant attack? – but he said the cost has been the death of format experimentation as content gets squeezed into standard templates and distributed through a handful of aggregators.

Third, he bemoaned the monetization of everything. "It’s impossible to search for, say, recipes online now and find an actual recipe without wading through insane volumes of advertising and SEO’d blather," he explained. "The kind of recipes you find on page one of search results seem to exist solely for the sake of attracting eyeballs, not because someone genuinely loved a recipe and wanted to share it."

And fourth, he denounced the web's shift from a document-centric to an application-centric model.

"The web is now an application hosting environment," he said. "You can still write a document in very plain HTML but there’s no demarcation between static documents and web applications. Even techies who pride themselves on writing efficient lightweight markup usually can’t resist putting a comment feedback form on the page or hiding some tracking JavaScript in the background."

[12]The W3C steers the way the World Wide Web works. Yet it is reluctant to record crucial meetings – and its minutes are incomplete

[13]Good news: Google no longer requires publishers to use the AMP format. Bad news: What replaces it might be worse

[14]Googler demolishes one of Apple's monopoly defenses – that web apps are just as good as native iOS software

[15]Sir Tim Berners-Lee's World Wide Web NFT fetches $5.4m at auction while rest of us gaze upon source code for $0

James is not a fan of PDFs. They're ugly and inelegant, he explained. But he sees them as a way to mitigate the harm of the modern web ecosystem, one that's accessible to non-technical people through common tools like Microsoft Word.

Asked whether he felt the inability to alter PDFs through client-side intervention – something typically available to browser users viewing HTML documents – represented a loss, he disagreed.

"The fact that PDFs are not easily edited by users is a feature, not a loss," he replied. "It’s a terrible format for distributing data, but so is HTML (if you want to distribute data, it’s better to attach it as a separate tab-separated-values file than to rely on users being able to parse it out of an HTML table)."

"I don’t see PDFs replacing HTML by any means, but I do see them as one path to recovering the concept of the document," he elaborated. "A good outcome here would be for HTML to become the app platform of the web, and cede ground to PDF, Gemini, and others, as the document platform for the web."

"I would like to see more format experimentation, more diversity of thought and approach… a heterogeneous pluralistic web where people still try new things and challenge centralization and build out simpler, more sustainable, slower communities that ditch the like & subscribe culture." ®

Get our [16]Tech Resources



[1] https://www.jwz.org/doc/tabs-vs-spaces.html

[2] https://www.nngroup.com/articles/pdf-unfit-for-human-consumption/

[3] https://www.lab6.com/

[4] https://news.ycombinator.com/item?id=27880905

[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2YPaevgOP4du39rG87CTRuwAAAJI&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0

[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44YPaevgOP4du39rG87CTRuwAAAJI&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0

[7] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33YPaevgOP4du39rG87CTRuwAAAJI&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0

[8] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44YPaevgOP4du39rG87CTRuwAAAJI&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0

[9] https://firt.dev/

[10] https://gemini.circumlunar.space/

[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33YPaevgOP4du39rG87CTRuwAAAJI&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0

[12] https://www.theregister.com/2020/07/22/w3c_minutes/

[13] https://www.theregister.com/2021/06/28/google_amp_core_web_vitals/

[14] https://www.theregister.com/2021/05/04/one_of_apples_monopoly_defenses/

[15] https://www.theregister.com/2021/07/01/world_wide_web_nft_auction/

[16] https://whitepapers.theregister.com/



The Register print version next!

Anonymous Coward

Come on, you know you want to. Just think of all those Airport shelves you could dominate.

Phil O'Sophical

HTML is currently effectively just whatever the browser vendors say it is

Isn't that the idea of a markup language? It just says "emphasize this", "use large text" or "make a list", and it's up to the browser to do that in whatever way it feels appropriate. It was never intended to be a precise page description language, even though it is sometimes (ab)used for that.

His argument seems to be that HTML is too flexible, and he doesn't like the way people take advantage of that, so he wants to rigidly define how information is displayed. That's his (rather OCD?) choice, but it's more a personal choice of a particular publication mechanism than of a standard for the web. A bit like a film maker choosing to use black & white silent film, and decrying colour or sound, it's just one choice among many, and will have its fans & its detractors.

Steve Graham

I think you've misunderstood what he was saying. What he was getting at is that the browser vendors can add features willy-nilly to HTML and get them written into the standards.

Lon24

Whilst rummaging in the garage over the weekend I came across a backup CD from 2005.

First shock was just how many website I was able to stack uncompressed on less than 1GB of disk. Second shock was just how many ran on a modern Apache2 server without a hitch including some SSI and Perl scripting in the cgi-bin. They ran rather fast on my RaspberryPi.

They were all personally hand-chiselled HTML & Perl code. But then I came across one site that used php and mysql (An old phpBB version 1 forum). Impossible to hack into a working site on a modern system because of changes in all three 'packages' blew dependencies. That's where the rot started.

Simpler times, more dependable times when you did everything serverside and used flat databases. Oh and because you could assume no more than a 54k modem and a 640x480 screen they were light and compact.

I always argued that a good website was about content, not appearance. I made more money from ads that blended and were more relevant to the page (the original adwords) then when they tookover and created the ad-blocker market and a never to be resolved war of attrition.

So a pint James. Achieving a similar result by a another method.

Maybe

vtcodger

"His argument seems to be that HTML is too flexible"

I read it as something along the line of "HTML has standards. But nobody adheres to them. Because standards are so confining "

Seems to me like a pretty good analysis.

I'm not so sure about the proposed solution. For one thing, PDF supports scripting and once you allow scripting, every evil in the digital universe probably swarms out of the box if you open it. see -- https://nora.codes/post/pdf-embedding-attacks/

PDF - does nothing it says on the tin.

Tom 7

"Asked whether he felt the inability to alter PDFs through client-side intervention"

News to me.

James is not a fan of PDFs. They're ugly and inelegant, he explained. But

Anonymous Coward

despite his valiant effort to stir proper discussion - the world marches on :(

Dunning-Kruger

FF22

Firtman is right. This guy is so ignorant, that he's incapable of recognizing how stupid his idea is, for a multitudes of reasons also ADDITIONAL to not fulfilling the goal he supposedly set out for it.

Re: Dunning-Kruger

quxinot

If PDF is the answer, the question needs slapped.

Re: Dunning-Kruger

rg287

Firtman is also wrong.

PDF was originally designed to fix a layout for print, but now includes features like text reflow for viewing on arbitrary screen sizes. Like HTML it has evolved beyond it's original design goals. Is that a good thing? That can be debated ad nauseam , but if HTML is allowed to evolve to HTML 5 then PDF is surely allowed a crack at branching out.

In all honesty, PDF often "just works", even if it's not as elegant as other options in various scenarios. You can archive it, checksum it. So many sites these days basically break the Wayback Machine because the Web Archive simply can't keep up with the dependencies and dynamic structure of what could (and often should) be a static page.

In the meantime: Behold, [1]the lightest, most compatible and responsive site on the web . An example to us all.

[1] https://www.sepehrlaal.com/

re-sizing and re-wrapping text

Dr Paul Taylor

I agree with most of the sentiments of this project, but surely going back to sanitised HTML would be better.

I am finding it increasingly difficult to read small text, so a basic thing I do with web pages is to enlarge them. In basic HTML, the text gets re-wrapped. "Clever" web pages often have fixed line lengths, making this impossible. (I do also resize PDFs.)

Re: re-sizing and re-wrapping text

Pascal Monett

I agree with you. The sentiment is most assuredly well-placed, but plain HTML would have been a much more valid demonstration.

What is killing the Web's usefulness is JavaScript. The over-reliance on this demonstrably dangerous technology is not just a nuisance, it is responsible for 99% of all malware infestations and, at the best of times, just transforms a web site into something you can't even bookmark properly.

Long live NoScript !

Re: re-sizing and re-wrapping text

vtcodger

You're correct of course. For one thing, I doubt that it will ever be possible to secure an internet that permits Javascript or any other form of scripting. You'll surely end up playing Whack-a-mole with an boundless sea of exploitable OS and application vulnerabilities. BUT, if you don't allow scripting, how are you going to do slippy maps (Google Maps, OSM, etc)? Editable forms? Spell checking?

For example, composing and posting this message looks to require Javascript.

Re: re-sizing and re-wrapping text

veti

The thing about sanitised HTML - at a casual glance, it's very hard to differentiate from common, iQuery-infested HTML.

What he says is true: even good devs with sound instincts find it very hard to resist the lure of "diagnostics" or "feedback". I've done it myself. PDF avoids temptation.

There's a lot of truth in what he says....

LDS

Of course he know PDFs are a provocation, but what he says about turning a simple way to share information using "hypetexts" into the ugliest way to develop the ugliest applications is true. And the only working business mode is really to litter them with tracking to sell ads - leading straight in the "social networks" issue.

mark l 2

"PDF Is not a format suited to share in different formats and diverse devices," he told The Register. "It's a format created for printing. So it's like using a boat to drive across a street."

PDF's used to be a format for printing, but those Fscktards at Adobe had to try and cram in a load of other functionality which wasn't needed and made the bloated insecure mess that is Acrobat reader and ruined it for everyone.

Sure Lab 6 are publishing them as PDF/A without the ability to run JS but how do you know it a PDF/A and not a different version which does allow JS without opening it first?

I personally use the built in Linux Mint PDF reader Xreader which after trying a couple of PDF which contain javascript appears not to support JS so at least that should make it a bit safer to open PDF files from unknown sources.

Flash!

Sceptic Tank

Will save every one of us!

Lab 6

Howard Sway

It had to happen : ladies and gentlemen we have found the real Nathan Barley.

For instance, [1]you could start here, with Issue 2 : A Gemini-PDF polyglot and The Tilde Quilt where "If you enjoyed last month's PDF-MP3 polyglot doc-cast, you're going to love this hybrid Gemini-PDF file".

You will be dumbfounded by The Tilde Quilt, which "combines the exclusivity of private pubnix membership, the value-hoarding artificial scarcity of sequential integers, the inspiration of spatial constraint, the north star of the unbounded cartesian grid, and the immediacy of named pipes" and is a "brand new media paradigm".

It is a crappy little piece of ASCII art.

[1] https://www.lab6.com/2

Couple of things

Flywheel

I was quite excited when I read the article on my laptop - everything looked good and seemed to make sense. Tried it on my mobile and had to remind myself I'm not a youth, nor a hawk.

Also, PDFs are largely benevolent, right? But here's an example I found earlier....

unzip -v pocorgtfo19.pdf

Archive: pocorgtfo19.pdf

Length Method Size Cmpr Date Time CRC-32 Name

- -------- ------ ------- ---- ---------- ----- -------- ----

15905 Defl:X 14500 9% 2019-01-04 16:19 838d0657 defender.zip

90291 Defl:X 23139 74% 2019-03-04 19:23 8da8fe88 phrack6612.txt

181069 Defl:X 181080 0% 2019-01-04 16:19 a00486ec jonesforth.tar.bz2

......

19676 Defl:X 17862 9% 2019-03-09 19:41 d8324af6 polyocamlbyte.zip

83900 Defl:X 77184 8% 2019-01-04 16:19 c1f51574 PDFRick.zip

- -------- ------- --- -------

33897715 30472096 10% 33 files

.. and yes, it acts exactly like you'd a PDF to.

Alan E. Davis: Some files at llug.sep.bnl.gov/pub/debian/Incoming are
stamped on 10 January 1998. As I write, nowhere on Earth is it now 10 January.

Craig Sanders: That just proves how advanced debian is, doesn't it :-)
-- debian-devel