3 ms·
I guess by modern standards this load time is acceptable, but when you argue that PDFs are a way to move forward, you're competing with HTML 4/5. And by that st
by danShumway 5y ago
I guess by modern standards this load time is acceptable, but when you argue that PDFs are a way to move forward, you're competing with HTML 4/5. And by that standard:
- Crud this website is so slow. Unacceptably slow. If your technology stack is spending 10 seconds just to fetch and render 13 pages of large-screen text, then either you're doing something wrong or it's a bad technology stack. That load time alone should kill this idea.
- There's no way for me to turn off images. This is the opposite of a client-respecting webpage, the only way you could make it worse is by rendering to Canvas or shipping me a PNG. My mobile browser doesn't fetch fonts by default. You're overriding my choice to do that.
- Mobile? Reflow? Responsive design? Adjustable font sizes? The author kind of offhandedly says that PDFs can do reflow right now, but how many clients actually support that. Does the PDF format handle this by default?
- Saying "you can technically make PDF accessible" is exactly the same as saying "you can technically use just a subset of HTML." It's the same argument. Nobody does it, PDFs are generally hostile to accessibility, and there's no way to signal that a PDF is accessible or enforce it as a community standard.
So, the much bigger question: what's wrong with Gemini[0]? I've been critical of Gemini in the past on multiple fronts, but if you are in this space where you want to burn everything down and make your blog static, Gemini really does seem to solve every problem that the author has, except better. It's also trivial to proxy Gemini documents or statically re-render them to HTML, which makes them accessible to people outside the community. And by default, they're both pretty accessible to screen readers, and much more efficient than what the author is proposing.
The author argues that using static HTML wouldn't be good enough because there's no standard that forces you to exclude Javascript. Then they point to PDF/A, which is not a standard that is enforced by most browser PDF viewers. To me, this argument isn't any different from telling website authors to choose not to use Javascript, what is going to force anyone to use PDF/A? Every web browser PDF reader supports Javascript. NoScript support in Firefox is better than the controls/extensions for disabling PDF scripting.
And Gemini is right there: for the most part it's actually working today. So I just don't get it. Why pick a technology that's tangibly worse than the web on (and I mean this quite literally) almost every single axis and every single metric, when you could instead switch to a markup language that actually does have use-cases, that does simplify deployment and blogging in some instances, that does have a real community, that does have some real advantages over HTML, that does have some real momentum behind it, and that doesn't disrespect my choices about what fonts/images I want to download?
[0]: https://gemini.circumlunar.space/ https://gemini.circumlunar.space/