[Discuss] A Web Server is No Longer a Web Server
Kent Borg
kentborg at borg.org
Sat Oct 25 10:23:03 EDT 2025
On 10/24/25 5:01 PM, Rich Pieri wrote:
> Took me a couple of re-reads to figure they mean the complexity is
> what's unnecessary for serving static content. Ambiguous wording, and
> maybe misstating what Heroku does or has become?
Sorry I was unclear.
They are moving from paying money to have Heroku generate pages on the
fly, to static pages that are free to host wherever they are parking
them. And, more secure, etc.
So they no longer need Heroku. Which, to the writer of the thing I read,
translated into they no longer need a web server.
I was being snarky about how the original thing that web servers used to
do (serve static web pages), has become so trivial and obscure that the
software needed to serve them appears to vanish.
A system designed for hypertext has turned out to be the architecture
for so much of modern life, but it wasn't designed for that. And, if one
were to design something for user interfaces over a large network,
hypertext would have been a strange way to go about it. It has turned
into an OS, kind of.
I do grant that, if one were to design something for the task, if one
were to build something like a better version of X11, the result would
probably be a way to build distinct programs. But because of the
hypertext origins of the web we the result is much more integrated, it
naturally wants to be a single thing, and there are advantages to that.
But there are disadvantages, too. Building a secure web site (e.g., for
a bank) is hard. And the integrated nature means URLs can show up in so
many places, including places that haven't yet been invented, making
phishing a serious ongoing problem that it would not have been without
the hypertext origins of the web.
-kb
More information about the Discuss
mailing list