+++ date = "2021-09-30T02:39:44+00:00" publishdate = "2023-12-29T07:08:55+00:00" title = "A Few Links" slug = "a-few-links" author = "Thedro" tags = ["bookmarks"] type = "posts" summary = "Over the course of cross browser testing websites, my bookmarks have become scattered and out of sync." draft = "" syntax = "1" toc = "" updated = "2021-10-01" +++ {{< image source="/images/a-few-links.png" title="Image showing a few links from my list of bookmarks." >}} Image showing a few links from my list of bookmarks. {{< /image >}} Over the course of cross browser testing websites, my bookmarks have become scattered and out of sync. I've grown too accustomed to using multiple web browsers and their various styles of bookmark management. I've begin consolidating my bookmarks into one location in an attempt to reduce my Internet usage. The constant researching of websites and articles that are already bookmarked or known from memory isn't particularly fun. To make matters worse, website rediscovery through a search engine is abysmal nowadays with sites becoming increasingly ranked by comment recency on social media aggregators, {{< sidenote mark="SEO" set="right" >}} Search Engine Optimization. {{< /sidenote >}} spam, and bot generated content. Rediscovering an original website or source is {{< sidenote mark="more" set="left" >}} Proxy search by social media comments; The title of an article and website domain is not always enough to directly reach a source, without first having to hop through an aggregator or social media site. {{< /sidenote >}} work than necessary. If the origin website isn't mentioned recently by a top ten social media website or aggregator, then as far as search engines go it doesn't exist. Nothing yet in my opinion beats a list of links in a {{< sidenote mark="text" set="right" >}} Back to the basics. {{< /sidenote >}} file. The `awk` code below dumps a list of records from a `bookmarks.txt` file as partial `html`. I'm hoping to eventually integrate my public bookmarks into this website using partial `html` output. These bookmarks can be serialized into any format and it doesn't have to be `html`. ```awk #!/usr/bin/env -S awk -f # Record separator (RS) will split on blank lines. # Field separator (FS) will split on new lines. BEGIN { RS = "" FS = "\n" } { printf "\n", NR, $1, $2 printf "%s\n", $2 printf "

%s

\n", $1 printf "\n\n" } ``` I'm a lazy programmer. My requirements are simple; the website title is my tag, and the link is the additional context needed to find the exact site or source. The `bookmarks.txt` file that this `awk` code parses is a list of records containing a link and a title. A description snippet can also be appended to the record for extra fuzzy search context, but that's not a priority. ```text CSS Icons https://cssicon.space/ NixOS: the good, the bad, and the ugly http://www.willghatch.net/blog/2020/06/27/nixos-the-good-the-bad-and-the-ugly/ Learn X in Y minutes https://learnxinyminutes.com/ An Introduction to the World of ThinkPad https://mckinley.cc/blog/20200411.html All Abstractions Lie https://wiki.c2.com/?AllAbstractionsLie ``` Interestingly enough, many browsers incentivize the use of tabs as bookmarks, others, history as bookmarks, and some, tabs as history --- and vice versa. Bookmarks, tabs, and browser history eventually start to look the same and perhaps may end up unified in some way, or at least that's my impression after using many different web browsers for so long. Once my bookmarks are unified again, I'll return to my old Internet browsing style; no history or tab dependency, and no extension or plugin dependency --- just bookmarks. That provides the best case scenario for little dependency on my choice of browser. ## Conclusion It goes without saying that a list of links is not a particularly useful way for most people to find or store content, but with the trend of increasingly poor and siloed re--discoverability of original content, even search engines will eventually loose significant utility. That's not a bad trade--off though, perhaps it's not to long before someone implements a way for web browsers, given a specific prompt, to surf the Internet as an agent on the user's behalf, display the resulting websites, and skip the annoying parts of Internet browsing entirely.