I'm not very convinced by the results of this "Auto Linker". I find most of the links in that article are rather useless, linking only to have a link to something.
IMHO links should be chosen "wisely", with the intent of helping the reader to find more information on important parts of the article or referencing other work etc.
Linking to the definition of rabbit holes is not really useful, if I'm not talking about varying sizes and depths of rabbit holes due to different breeds of rabbits and their prefered style of rabbit hole digging.
I used to link to reduce redundancy in what I had to write. Now, I tend to duplicate large parts of documentation to remove links, because I found that people actively avoid links. I'm guessing this aversion is related to a unconscious awareness of the "doorway effect", which exist even in virtual environments, and I assume, to some degree, on web pages:
> The doorway effect is a known psychological event where a person's short-term memory declines when passing through a doorway moving from one location to another when it would not if they had remained in the same place.
You know how Visual Studio can bring up in a mini window a "peek" into another function? I think that's excellent and I think we need something similar for regular documents.
Something which could expand incrementally. It would be harder to write, but it would be excellent for documentation. Maybe an LLM would be good for such a thing. Say, instead of just linking to the documentation of another function, you could "zoom" into a description of it, tailored to screen estate or number of words:
Links have two problems: (1) they carry you away into a different context and (2) they rot.
So, to bring something to attention, you need to both quote enough and provide a link.
For one's own archival purposes, it makes sense to locally store a copy if any important enough web page. No, that text or image is not guaranteed to be there tomorrow.
How much of this is down to the examples chosen? It seems to me that in practice you would generally use it with a specific link in mind (like the "Rust prehistory" or "notes on a smaller rust") ones; although I can see that those might be a bit more cumbersome to show in the post, given that they might use longer queries.
> It’s not that I don’t know what I want to link to. Most of the posts I write are responses to things I’ve read. I know what I want to link to, but I don’t want to lose flow to hunt down a link while writing. I could go back at the end and insert links retrospectively, but it takes more time than I’d like to hop between browser, editor, and what I was thinking about at that point. It would be nice if I could mark links as I write, while still being able to refine each link in-editor as I edit and revise later.
> I wish there was a faster way to link the posts I write.
I agree about the importance of links. In fact, when I come across an article about something that many others have written about, and it has no links or references, I often just don’t read it.
But the links in this article, the ones generated by a script, are worse than useless. Meaning, they’re a waste of time that makes the text worse.
First of all, they are set to open a new tab. Rude. Second of all, a link to a wikipedia page for a term, for example, is just stupid noise. If I want to know what wikipedia has plagiarized about a certain topic I can go there and look it up.
Links are not an afterthought, something that can be usefully tacked on to the finished text. Each link should be genuinely helpful, or back up an assertion. Ideally, each one should be non-obvious.
Thank you for taking the time to read the post and write a thoughtful reply. You make a good point: it is hard to find good links. These links suck. I apologize for this poor first-order approximation: I want to link to stuff interesting enough to merit a post or two in their own right.
To that end, I keep a list of every interesting thing that I read. I plan to index that list, and make it searchable. Reliance on a search engine is a stop-gap to keep scope sensible at this point. There's nothing stopping other people from taking this idea and running with it.
Given your personal involvement in this area, I understand your apprehension. Like you, I want to see a denser, more interesting web. But without tools to overcome the static friction of linking things together, I don't think we'll see density and utility of links increase any more than it already has. That's what I hoped to highlight in this piece.
Well, I appreciate that you’re thinking about the problem of linking in general and writing about it. If finding and inserting appropriate links is laborious—well, so is writing. Linking is part of the process—part of the content. You can’t automate it away.
Like most HN readers, I've written my own content/information management system, and this makes keeping track of sources and inserting/formatting links much easier. Almost a pleasure, in fact. But there are plenty of open-source and other products out there designed to help with this task.
Could you elaborate a bit on your workflow that makes it almost enjoyable? My personal workflow can always use an upgrade and I'd love to hear what works for you!
Very briefly, it depends on Pandoc with citeproc, and vim (lately nvim) with the vim-markdown plugin, which adds hiding to citeproc links, making the text easier to read in the editor.
Next I have bibliographic information on every source that I use stored in a postgres database. I have a collection of scripts (in lua, python, bash, ...) so I can do what I want within the editor: supply a URL and have some information scraped from the page and stored in a note, one file per resource, with bibliographic information in YAML blocks and anything else that I want to add. Another script stores the important information from the YAML blocks into the database. I have a script that builds a bibtex file from the citeproc links in my document, for formatting bibliographies in the final version. I can generate HTML, PDF, DOCX, or plain text from the same source.
The pleasure comes from having a set of keyboard shortcuts that let me jump instantly from a citeproc tag to my notes file, a local copy of the resource, or the copy on the web; also to search for notes with keywords, and a few more things. I can select part of my text and hit two keys to transform it into plain text or HTML with properly formatted links and references.
For each notes file corresponding to a resource another keyboard shortcut brings up the list of articles referring to it, and I have a script to draw a graphviz graph of this reference network.
I’ve built this up over the years; it’s a mess and no one else should use it. Maybe I’ll clean it up some day. It’s really paying off now for me though: I’m writing a book with about 50 references for each chapter; it would be unmanageable without some kind of system.
I know that there’s Zotero and other systems out there. But I prefer that this work in the peculiar way that I like; these systems tend to become brain extensions and are somewhat personal.
> I plan to index that list, and make it searchable.
You could do that, or you could use one of the reference managers that already exist and do this for you. At the very least, try out a few and see what features they offer. If none of them work for you out-of-the-box, perhaps try out what extensions and customizations exist, or create your own. EndNote, Mendeley and Zotero are popular. EasyBib.com and RefWorks also come to mind. Most of these work with browsers and website to make adding and cataloging references easy.
I wonder if you could change the tool to search your browser history, rather than the web. That way, it's more likely to pick up the things you actually read before writing the post.
Why is it rude to open a new tab? I truly hate when I’m trying to view a link and it navigates me away from what I’m reading. IMO new tabs should always be the default.
Middle mouse button (or control+click) in any browser opens a new tab (as long as you're using actual links, don't get me started on the morons who think "span with a onclick handler" is an acceptable substitute)
Forcing to open on a new tab breaks conventional behaviour, it should be left as the user's decision (either explicit every time, or a setting in their browser. This setting already exists btw, if you like more tabs you can enable it in your browser).
(Maybe a case could be made if the website is actually an application and there would be data loss if not for the new tab, but I would argue design your application so as data loss is not possible in that case)
---
As for myself, the browsing history is something very useful to have, I like keeping the amount of tabs down to near one per activity, so I regularly use the history to go back to what I was reading before. Forcing new tabs means I have to reach for more advanced tools like tab trees, which are often not worth the cognitive effort.
The reader should decide. It’s usually trivial to hold down a modifier key to open a link in a new tab or window, but if the `target` attribute is set in the link it’s not as easy to override it.
This debate raged in the early days of the web: what is the better practice? A consensus emerged among web designers that leaving it to the reader makes more sense. I’m proud that I convinced J. Zeldman that his early practice of using `target='_blank'` hurt the user’s navigation, and so played a minuscule part in nudging designers in general away from this.
Unfortunately, with the advent of SPAs, Web3.0 and the takeover of the web by Angular and friends, we can no longer rely on "links" behaving that way. For the time-being, we've lost that battle and the web is worse-off as a result.
I.e. A lot of the time, links are actually various DOM elements that have onclick handlers and various JS-fuckery meant to obfuscate true behavior from the user and make it more difficult for the user to control the browser. Worse still, how about the user-hostile behavior of browsers that now allow JS to takeover and disable right-click, select and copy functions.
Someone defended SPAs by pointing out that frameworks like React use Javascript and browser APIs to fake paths and history, and another noted that reimplementing basic browser functionality was madness.
Worse than that, even if you somehow try to force onclick handlers to not obfuscate true behavior, most truly bad SPAs don't even have valid intermediary states accessible thru URLs; the only way to navigate the "application" is sometimes by changing the state with interactive SPA use.
Because if the link is a normal one, the user can choose whether to open it in same tab (regular click) or a new tab (lots of UX shortcuts for this like Ctrl+click).
But if the link is hardcoded as a new tab, the user has no choice. There's no way to open a new-tab link in the same tab. :( Not to mention that there's generally zero indication it's a new-tab link anyways, so not only does it take away choice, but it does so non-predictably. Ugh.
Personally I have set `browser.link.open_newwindow = 1` in Firefox to force all links to open in the same tab, then I can middle-click to open in a new tab. This way I can do what I want without worrying about what the website author thinks I wanted.
It does have some downsides though. Some sites don't use <a> tags but instead <div> with JS to open the link, then you can't choose to open in a new tab (on the other hand they can't force you to open a new tab so it is sort of even on average). Some sites break because they are trying to launch a new window that communicates with the parent. But overall I think it is a big improvement to the web.
It overrides the user's browser settings. As a user, technically you can override the override, but it's only accessible sometimes through hidden settings or user scripts.
I don’t mind new tabs but what really annoys me is when things are set to open in a new window, and you have to fiddle with it to get it to merge back into the window you already have open
Firstly, when mousing over a link which will open in a new tab, show a different pointer. I get a hand with the index finger extended when hovering over a normal link; it could be that but with a diagonal arrow, like the common "open in new window" badge, for a link which opens in a new tab.
Secondly, in the context menu for a link which will open in a new tab, replace the "open link in new tab" option with "open link in this tab".
As long as links are real links, and not JavaScript devilry, this would 85% fix it for me.
I didn't downvote, but links opening a new browser window, except for the occasional auth pop-up, hasn't been a thing since... 2008? What browser still doesn't use tabs by default?
> Second of all, a link to a wikipedia page for a term, for example, is just stupid noise.
I sometimes link to the Wikipedia article when there are many things by a certain term, or when I'm using terminology that doesn't show up in search results (but is common in colloquial use). Should I do this a different way?
You’re using links to disambiguate your terms; it makes sense. My unpopular opinion would be to keep linking as you’re doing, but find better sources than wikipedia. Something reliable and, if possible, authoritative, and that’s more likely to have stable content—where you can be a bit more sure that what your readers will see when they follow the link is what you saw.
I think it's a cool idea. What would be even cooler would be if it would do a content based search of your browser history, and suggest the most relevant pages for the thing you want to link. So, ideally it would be finding the pages you actually looked at. (Although, in practice it becomes difficult when you use all kinds of different devices to browse with, plus whether there's enough information to search with to retrieve the page. But I imagine something could be thrown together that does work a decent percentage of the time.)
(defun my/org-insert-wikipedia-link ()
(interactive)
(apply-to-region (lambda (string)
"Convert a string to a link to English Wikipedia"
(concat "[[https://en.wikipedia.org/wiki/" (subst-char-in-string ? ?_ string) "][" string "]]"))))
I imagine you could do something similar for the other examples in the post. Makes adding these links pretty painless.
Wow. I thought this was going to be a big rant piece on why we should all go back to dynamic linking again and I was so ready to start my Monday off with a spicy take.
But yes we should all also be adding links in our writing and publishing on our own platforms.
This is how I was using Twitter as well, as a network of people sharing interesting and often niche content (which is why algorithmic timelines are useless to me because the human curation was the entire value).
>I have so many linkless posts waiting to be published: I could go ahead and publish them as-is, but by doing so I feel as though I’d be treating you, dear reader, unjustly.
How about allowing the readers to participate? Posts could be in a pre-publish state and the community could add link suggestions and annotations.
I am a blogger, and I'll stay a blogger even with GPT stuff.
I agree. As bloggers continue to link only to good stuff between them, those blogs will become infinitely more valuable than the flood of trash that people will post from GPT.
It seems hopeless that people will find that good stuff, but word of mouth will do it, I think.
This is one of the main reasons I think advertising is harmful and all ads should be blocked. We don't have a problem with shortage of content, we have a problem with filtering out all the bullshit where they've put all the effort into search engine optimization and ad monetization, and no effort into creating content.
Ad apologists will claim "but how will we fund content creation!" and the answer is, if the content is actually good, people will be willing to pay for it--probably even if not forced to do so: lots of content creation is being funded by donations right now. "But that doesn't scale!" cries Hacker News. But what you mean is, it doesn't scale for a centralized entity that wants to collect rent on other people's work, and frankly, fuck those centralized entities. They're the ones creating the filtering problem. Paid, quality content creation inherently doesn't scale because it's inherently serial: a person/team creates one piece of content at a time. You can scale distribution, but you can't scale quality creation.
And the fact is, people will create because they want to create, even if they aren't paid to do so. Maybe they are paid, but that's secondary to the human need to create. As Quincy Jones said, "When you chase music for money, God walks out of the room."
If I'm being honest, most "content" I've come across on the web doesn't deserve to be funded and its creators are better served creating something else if they're looking to be compensated.
people will create because they want to create
What great content I have come across is usually created within a similar ethos to FOSS. People know they've learned a great deal from others in the past and want to pay it forward, or they want to share something with the world to see how others might iterate upon it.
if the content is actually good, people will be willing to pay for it--probably even if not forced to do so: lots of content creation is being funded by donations right now
I've also seen great content from people connected to a patreon, so there is definitely room for compensation, if the content is good enough.
There is a class of content that is not interesting enough and requires too much work to be maintained by hobbyists. Perhaps you will always find people willing to write interesting content about vintage computers, but you won't find many people who want to write about the intricacies of local bureaucracy, the tax system, and other boring-but-necessary topics.
Donations just don't work, and people should stop pretending they do. I make literally 100 times more from affiliate marketing than I do from donations. If I had no integrity, that multiple would be even bigger.
Let me put this another way: the bare minimum level of affiliate marketing covers all my expenses. Donations don't ever buy groceries. I could not put this much effort into my work if I had to get another job to put food on the table.
> Perhaps you will always find people willing to write interesting content about vintage computers, but you won't find many people who want to write about the intricacies of local bureaucracy, the tax system, and other boring-but-necessary topics.
> Donations just don't work, and people should stop pretending they do.
ProPublica and Mother Jones are doing just fine on donations.
But more to the point, these are absolutely not topics we want advertisers to have control over. I'll admit that donations leave many important forms of content underfunded, but the alternative you're proposing is a well-funded fox in charge of the henhouse.
> I make literally 100 times more from affiliate marketing than I do from donations. If I had no integrity, that multiple would be even bigger.
You might have integrity but you're a stopped clock: even when it's right it's never useful. The only means your customers without expert knowledge have of distinguishing your motivations is by looking at your funding, and as long as you accept incentives to lie, you can't be trusted.
> Let me put this another way: the bare minimum level of affiliate marketing covers all my expenses. Donations don't ever buy groceries. I could not put this much effort into my work if I had to get another job to put food on the table.
That's not evidence that donations don't work. That's just evidence you've built your business around advertising and not donations.
I have put a lot of effort into getting donations, but they just don’t come. It’s a dead end. Even if I annoyed my readers in the manner or Wikipedia, I still would not make rent, but I would make my website significantly worse.
In any case, it’s futile to argue with you. You have the luxury of judging from a distance what I’m constrained to describe from first hand experience. You bring up Propublica and Mother Jones as if these rare successes mattered to people with humble readerships. It goes to show that you don’t have a tenuous grasp a reality I’m familiar with. You express desires, I express facts.
In any case, why should anyone care about what you want? You get a free product and complain that it’s not free enough. The cheapest customers are truly the most demanding. It behoves any self-respecting person to ignore them.
> I have put a lot of effort into getting donations, but they just don’t come.
What have you done, exactly?
If by "I've put a lot of effort into getting donations", you mean that you've put a lot of effort into soliciting donations, it's quite predictable that that wouldn't work.
If by "I've put a lot of effort into getting donations", you mean that you've put a lot of effort into finding out the needs of your audience and producing content that meets those needs, it would be very surprising to me that this didn't get any response.
It's certainly important to make it easy for people to donate, but it's arguably more important to produce something people want to donate for. And, I don't want to underplay the fact that producing quality content is hard. But it's not impossible.
> In any case, it’s futile to argue with you. You have the luxury of judging from a distance what I’m constrained to describe from first hand experience.
Not the case--don't make assumptions about people you don't know. Perhaps you should read some of my comments where I gave numbers on what kinds of subscription numbers I was able to get when I was running a substack.
> In any case, why should anyone care about what you want? You get a free product and complain that it’s not free enough. The cheapest customers are truly the most demanding. It behoves any self-respecting person to ignore them.
1. I donate a significant amount to content producers, and pay for a few different subscriptions, as well as being willing to pay to purchase content. My point all along has been that people are willing to pay for and donate to the production of quality content. Maybe if you cared more about what content consumers thought, you'd get more of them to donate.
2. I don't think you can reasonably disagree that quality content is content that serves the content consumer. You want to produce quality content, right? If content consumers tell you what they want, it behooves you to listen.
3. If you're a sociopathic content producer, sure, you have no reason to care about people except as potential sources of income. But as a content consumer, and more importantly, as a decent human being, you probably do care about other people as more than just sources of income. You've said you could make more money on ads if you didn't have integrity, so you clearly have other motivations besides money. You probably aren't a sociopath. So if you set aside your emotions about this disagreement I think you probably do care about people even if they don't directly contribute to your bottom line.
I think donations would work, if there was not all the spam content to sift through. Imagine if all that spam was not financed or pushed by ads and marketting, how many more users would feel the need to find good content, rather than sheeply accepting whatever is pushed into their faces. If there simply was not so much of artificially pushed meaningless or generic content, then their need would move to other content. If that need was not fulfilled, then the demand would outweigh the offer and more than enough money would be available for donations.
Perhaps an example, not sure if that one works well: I have my personal music collection. I choose music to listen to based on mood. I do not own a spotify account and I do not want to own one. I go out to my music collection and choose music for myself or I go online and search for more music of some genre or fitting my mood. I actively seek that out. On spotify people let stations play some music, based on algorithms. The content is pushed into their faces (ears). If there was not so much of music "to fill the gaps", then more people would go on an active search, instead of relying on it being pushed to them. They would maybe discover some music from some composer and seek out their website, shop, bandcamp, whatever, and buy their music there. I have done so at least and found great music this way. Basically cutting out the content pusher middleman, that has nothing to do with the actual creation of content, but merely acts as a publisher. More people being actively seeking content themselves, less people relying on publishers.
I guess it is a continuous fight of donation financed content creators against the generic artificially marketted stuff, that is financed in different ways. We can however see, that there are content creators out there, who do make due on donation basis. Personally, I hope that we will see more of that and less affiliate marketting pushed stuff.
People have repeatedly shown that they are not willing to pay for the content they consume. They might pay 10€ a month for Spotify or Netflix, but small independent creators are lucky to get an occasional donation.
Affiliate marketing is a deal with the devil, but it's still a deal worth making. They let me work full time on building a useful resource and offering it for free. I can work with dignity, without having to beg for money, and still without selling out.
They similarly finance many people whose content saved me time, money and stress.
They also finance lazy people, some of which steal my content. However killing their income stream kills mine, and it still leaves you with far more prevalent first-party blogspam.
In my opinion, the solution lies in better content curation tools. None of this would be a problem if search engines surfaced quality content once in a while.
> In my opinion, the solution lies in better content curation tools. None of this would be a problem if search engines surfaced quality content once in a while.
Agreed!
Unsurprisingly, search engines funded by advertisers surface content which is effective at serving ads, rather than quality content. Do you see how this might be a problem?
> Affiliate marketing is a deal with the devil, but it's still a deal worth making. They let me work full time on building a useful resource and offering it for free. I can work with dignity, without having to beg for money, and still without selling out.
> You can scale distribution, but you can't scale quality creation
I'm going to have to start quoting this; it's a much more succinct argument than I've heard before, and it captures the essence far better than I would be able to do even with many more words.
It's not my first try: I've had years of having this discussion to improve its communication. I'm glad I've improved it enough that somebody thinks I'm doing it well!
> I can feel with your sentiment, but don’t see a practical way to do that.
> Define “ad”.
"How to define 'ad'" is a question worth asking, because it is important that we block ads without blocking valuable content. And how to block harder-to-detect forms of ads is a real, open problem.
But it's just not an argument that blocking ads is impractical. The vast majority of ads are ones that no one is confused about, and which can be detected fairly easily. Imperfect != impractical.
What you're saying is an example of a perfect solution fallacy. Obviously adblocking is never going to be perfect, but it works pretty well.
> And for true ads, would you want to block a starting blog author from promoting their work online?
Not in a void, the way you're presenting it. I'm proposing blocking all ads, not just newcomers' ads. Ads provide the greatest advantage to the best-funded, and that's usually not newcomers. Advertising benefits entrenched interests, not newcomers. Uniformly blocking all ads would be a boon to newcomers, not a harm.
At a more fundamental level, ads fundamentally prevent capitalism from providing the greatest value for the least cost. Ads allow advertisers to provide lower value at greater cost by spending more money on advertising.
An absence of advertising means that people find out about products from higher-signal-to-noise ratio sources of information, such as independent review sites, word-of-mouth, and subject experts. Consumer reports is an excellent ROI: buy nice or buy twice. My friends are better at recommending books and movies to me than any advertiser. And it should be the responsibility of the FDA to disseminate information about new treatments to doctors, and the responsibility of doctors/prescribers to keep up-to-date: drug ads and prescriber kickbacks from pharmaceuticals should be unequivocally illegal.
> "But that doesn't scale!" cries Hacker News. But what you mean is, it doesn't scale for a centralized entity that wants to collect rent on other people's work, and frankly, fuck those centralized entities.
The problem is, payment is hard.
For authors:
- Micropayments are still ridiculously expensive
- Users want it to be as simple and hassle-free as possible, which means you're stuck with Stripe, Ko-Fi, Patreon or PayPal as that is what people have, and they don't want to trust Joe Random Blogger with their credit card information or their real name
- PayPal has been known to randomly close down accounts and hold your money hostage for months and years, especially if you're dealing with "morally grey" content (e.g. if you write about sex or anything related to crime)
- Accounting is annoying because each micro-donator needs a formal invoice, you have to deal with dozens to hundreds of micro invoices at tax time
- You're virtually stuck with credit cards and their limitations (again, see above, most of PayPal's ridiculous policies comes from the big CC networks) if your audience is international, but guess what, less than half of Germans have one [1], and outside of Europe the situation is even worse.
For users:
- Fuck no I don't want to fill out a form with CC number, CVC code, my real name and my address just to give a dollar for Joe Random Blogger's coffee cash
- Fuck no where did I just put my phone for 2FA
- Why is Joe Random Blogger now blasting my email address with newsletters?
- Uh, I did not consent Joe Random Blogger to suddenly draw in 20$ a month instead of 5$
The centralized entities (Paypal, Stripe, ko-fi, Patreon, whatever) take care of a ton of the associated bullshit, but they are expensive and, as I wrote, prone to randomly ban you without recourse - and not just ban your commercial entity but also your personal account - for life.
In contrast, ads are easy: you embed some piece of code on your website (or set some identifiers in your WordPress theme/plugin) and get a check or bank transfer every month. That's it.
Ehhh, it's not hard, it's just centralized, with the associated problems of a few entities having an oligopoly. I can have Stripe and/or Patreon integrated with a site in a few hours of work. PayPal isn't hard to integrate either, they just suck. (I have no experience with Ko-Fi).
> Micropayments are still ridiculously expensive
So don't do micropayments. Payments in the range of $5-$10 are less expensive, and it makes you more dependent on loyal fans rather than generous drive-bys. It's harder to get loyal fans--you have to produce quality content--but when you get them it's a more stable source of income. And from the perspective of content consumers, that's exactly what I want: higher quality content. Larger payments also means your account doesn't get used as often by hackers testing out credit card numbers they bought off the darkweb.
> - PayPal has been known to randomly close down accounts and hold your money hostage for months and years, especially if you're dealing with "morally grey" content (e.g. if you write about sex or anything related to crime)
Yeah, that's a real problem, and one of the reasons we need legally-mandated payment-neutrality if we're going to give so much power to credit card companies.
You'll rarely hear me say anything positive about cryptocurrency, but this is a rare exception: cryptocurrency is a viable workaround for content creators that need a workaround (which isn't most content creators). But to be clear, I'm not talking about crypto-bro flavor of the week: USDC is pretty ideal for this, or the two popular zk-Snark/ring signature privacy coins[1] if privacy is desirable. Variable value is a problem, not a feature, so I don't see a reason to accept that unless you have a need for privacy: paying with Bitcoin is dumb.
> - Why is Joe Random Blogger now blasting my email address with newsletters?
> - Uh, I did not consent Joe Random Blogger to suddenly draw in 20$ a month instead of 5$
Again, these tend to be problems with lower-quality content. If you're trying to build a loyal fanbase you do that by providing quality content, not by spamming or stealing.
[1] I'm not naming these coins by name because that tends to get flagged.
> Ehhh, it's not hard, it's just centralized, with the associated problems of a few entities having an oligopoly.
It's hard to pull it off without relying on the oligopoly entities. Like, I can publish my SEPA banking information on my blog, but that would just invite scammers. And it's not real-time which means I have to manually process payments.
And that's just dealing with the payment itself and not solving the problem of accounting and taxes. It's a difference if I go to my tax preparer and hand him twelve advertising revenue bills or if I'm handing him a wad of receipts by individual people.
> Payments in the range of $5-$10 are less expensive
They are, but good luck getting people used to free content to pay that range. Even 99.9% of Onlyfxns models rarely make a single subscriber above that. Cheap mass content is what the market wants (no matter if in porn or blogs), you got to fulfill a very specific niche if you want to lure in some whales.
> You'll rarely hear me say anything positive about cryptocurrency, but this is a rare exception
Cryptocurrencies bring in their own host of problems:
- people will automatically discard your opinion simply because you offer crypto - the NFT and shitcoin crazes as well as the environmental issues crypto has (e-waste, CO2 footprint, ...) have burnt so many people that offering crypto acceptance these days outside of anonymity context (e.g. VPNs, piracy) is just as negative on your brand as paying for Twitter Blue is (which is why Musk decided to not show any more if an account is legacy-verified or Twitter Blue subscriber).
- can't do KYC any more so you have a harder time doing taxes (may be more of an European problem, given the IRS seems to even allow you to enter crime proceedings - if I would do that on my taxes here in Germany, I'd get the cops knocking on my door)
- your whole infrastructure will now be targeted by sometimes highly sophisticated attackers going after your coins
> Again, these tend to be problems with lower-quality content. If you're trying to build a loyal fanbase you do that by providing quality content, not by spamming or stealing.
As a user, all I have at a moment is to look how the content currently is. I have no recourse when the author decides to sell out, similarly to what happened to MANY Chrome extensions that suddenly went and embedded malware (e.g. [1]), or when they / their database get inevitably hacked.
> It's hard to pull it off without relying on the oligopoly entities.
Right--as I said, that's a problem, and I'd actually say it's hard enough to be effectively impossible.
Then again, I'm not sure how this is an objection to donation-based monetization of content creation as opposed to ads, given the average content creator isn't running their own ad network.
> It's a difference if I go to my tax preparer and hand him twelve advertising revenue bills or if I'm handing him a wad of receipts by individual people.
I'm not sure I understand this. It's not a wad of receipts, it's a single spreadsheet of receipts with a total at the bottom if you've set it up at all reasonably. I'm not convinced this is a real problem.
> They are, but good luck getting people used to free content to pay that range. Even 99.9% of Onlyfxns models rarely make a single subscriber above that. Cheap mass content is what the market wants (no matter if in porn or blogs), you got to fulfill a very specific niche if you want to lure in some whales.
I'm not sure where you're getting your data here. I had a Substack and was able to get a number of $100/year or $10/month subscribers within only a few posts, by providing well-thought-out educational content for an audience with disposable income. A commitment to posting weekly ended up being stressful so I ditched the Substack, refunded where it made sense, and am slowly working toward publishing it as a book. This is probably a less-profitable route, but it's more conducive to my mental health.
I'm not saying it's easy, but it's not that hard.
> people will automatically discard your opinion simply because you offer crypto
The sorts of content which I would recommend using crypto for as a workaround for centralized payment processors tends to not be reputation based. If you're writing a blog about permaculture, sure, people might discard your opinion because you accept crypto, but you also probably aren't likely to have your blog payments canceled. If you're writing a smut blog that might have credit card processing revoked nobody was reading it because they cared about your opinion.
> - the NFT and shitcoin crazes as well as the environmental issues crypto has (e-waste, CO2 footprint, ...) have burnt so many people that offering crypto acceptance these days outside of anonymity context (e.g. VPNs, piracy) is just as negative on your brand as paying for Twitter Blue is (which is why Musk decided to not show any more if an account is legacy-verified or Twitter Blue subscriber).
If you respond to what is actually in my post, you'll note that privacy is exactly the context I was suggesting cryptocurrency for.
> - can't do KYC any more so you have a harder time doing taxes (may be more of an European problem, given the IRS seems to even allow you to enter crime proceedings - if I would do that on my taxes here in Germany, I'd get the cops knocking on my door)
There's nothing stopping you from doing KYC in addition to accepting USDC if accepting cryptocurrency is merely a workaround for having your payment processor canceled. Accepting anonymous payments is obviously off the table where KYC is required.
> - your whole infrastructure will now be targeted by sometimes highly sophisticated attackers going after your coins
There's very little reason to actually ever hold coins for a significant amount of time, so losses if they occur should be minimal.
"Infrastructure" in this case can be generating a memo ID to associate transactions with KYC before they are able to view the payment address, which can be the address of a paper wallet. This doesn't require a sophisticated knowledge of crypto.
> As a user, all I have at a moment is to look how the content currently is. I have no recourse when the author decides to sell out, similarly to what happened to MANY Chrome extensions that suddenly went and embedded malware (e.g. [1]), or when they / their database get inevitably hacked.
Sure. When you buy a book, you're worried that the author might write a bad book in the future? If so, I don't suspect your concern is shared by most content consumers.
I'm not sure the risks of running code that can auto-update itself are really relevant to content. So far I haven't heard of anyone figuring out how to go into my brain and auto-update my memory of blog posts I've read, and I'm quite happy to pay for things if I've already learned from them.
A lot of these objections are a bit odd. What's your reason for preferring ad-supported content?
> A lot of these objections are a bit odd. What's your reason for preferring ad-supported content?
I don't like ad-supported content and hate tracking with a passion. However, it's reasonable to say that, at least at the moment, advertising is
- easier (and not just technically, but also from the bureaucratic side!) to implement than donations or "real" payments, and note that some countries like Germany make a legal distinction between these when it comes to taxes, and mis-classifying income as donations comes with severe penalties
- guaranteed, predictable income for authors
- zero effort for the reader
- zero direct, financial risks for the reader
The result of this is the current, mostly ad supported crap infrastructure we currently have on the Internet.
Personally, I'd advocate for drastic banking regulation and tax code changes to make paying for online content easier so that regular people can take advantage too, without going through middlemen extracting rent everywhere:
- for clearly non-commercial content which most personal blogs/vlogs/podcasts fall into, completely exempt donations from tax and other bureaucratic (AML, KYC, invoices) requirements
- create a globally usable (!) financial network with no censorship other than what's illegal in the recipient's country, low caps on transfer fees (0.1%), real-time transfers and "deposit only" accounts, so that transferring money across countries actually gets realistic, and I can offer my bank account number without having to fear someone draining my account. SEPA gets pretty close to that, but it's not joined with the US, Australia or Asia so at the moment there is no alternative for cross-continental donations other than to rely on PayPal and friends.
You bring up a lot of real problems and good solutions, but ultimately I don't see an argument for the existence of ads in your post. If anything, advertising is one of the reasons a lot of the problems you mention exist. If it weren't so easy to monetize garbage with ads, there would be greater demand for easier payment solutions.
I'll also add, about this supposed benefit of ads:
> - zero direct, financial risks for the reader
"Direct" is doing a lot of work here. Ads absolutely are a financial risk to readers. "You can just not buy the product" puts a lot more faith in human agency than is warranted. Everyone buys goods and services they don't need, or which are worse than less-advertised alternatives, because of advertising. People are manipulable and that includes you (and me). No one is immune. Even if you never see an ad, your friends pass bad information from ads to you. Even if you aren't easily fooled by lies, advertisers have a great deal of control over what truth gets placed in front of you. You are being manipulated into spending money by ads--it's practically unavoidable in today's society--and if you aren't taking steps to protect yourself, you are being manipulated even more.
Arguably financial harm isn't even the most fundamental harm done by ads--I'd argue that the psychological harm done by ads is actually worse.
Leaving all of this aside, people just don't donate. They'll watch 100 hours of a YouTube channel and eagerly await new videos, and still not donate a cent. Donors are a very small fraction of the audience.
I get 1 donation per 14,000 visitors. Even among the people who email me with complex questions, and get detailed answers that no one else offers, only a few donate. The link is not hidden or anything; people just expect content to be free.
Just consider how much content you voluntarily consume, and how much of it you paid for.
> Leaving all of this aside, people just don't donate. They'll watch 100 hours of a YouTube channel and eagerly await new videos, and still not donate a cent. Donors are a very small fraction of the audience.
The thing you seem to not understand is that ads and donations are not compatible business models. If you are making money off ads, that is why you are not making money off donations.
The content you make to sell ads is attractive up front, with clickbaity headlines and overwrought thumbnails. If the platform you're on rewards watch/read time, you then just have to provoke some sort of addictive emotion like outrage or fear to keep people there for a bit. On platforms that don't reward watch/read time, you don't even have to do that.
The content you make to attract donations is useful, educational, or otherwise beneficial to your audience. This is incompatible with pinning people down with brain chemicals so you can ram ads down their throat.
If you interrupt your content with a NordVPN or Athletic Greens commercial, people are less likely to donate. Why would I donate to someone who is trying to manipulate me? You're not doing me a service, you're doing advertisers a service.
Every example you have given of people not donating, is easily explained by the fact that the content is ad supported.
> I get 1 donation per 14,000 visitors. Even among the people who email me with complex questions, and get detailed answers that no one else offers, only a few donate. The link is not hidden or anything; people just expect content to be free.
Alternative hypothesis: people just expect that you're already getting paid by advertisers, which by your own admission is true.
You've set up this exchange where you use your audience to sell ads. You chose to make that relationship with your audience.
It should not be surprising to you that the audience uses you as they see fit. Those are the terms you signed up for.
- Apple Pay, Google Pay and Stripe Link will remember your CC details - so no need to fill them in again and again. This is as smooth as can be.
- Micropayments are expensive. Forget about it. Creators need to come together under common umbrellas to get paid. That's what traditional publications were in the end. If people could put their egos aside a little, they could start making some good bundles.
- You are dead wrong on accounting. Nobody has to make formal invoices for all their subscribers.
- It's easy to accept SEPA payments and many other payment forms without credit cards through Stripe. Those germans who don't have credit cards have debit cards and they work the same.
> - Apple Pay, Google Pay and Stripe Link will remember your CC details - so no need to fill them in again and again. This is as smooth as can be.
Yes, yet another middle man, doing exactly the same as PayPal - and at least Google is known for banning people for having too many refunds or tripping some anti-fraud ML model. Not something I'd trust either as an user or as an author.
> Creators need to come together under common umbrellas to get paid. That's what traditional publications were in the end. If people could put their egos aside a little, they could start making some good bundles.
Again, middle men. Middle men that are going to censor you for whatever reason - no matter your political direction, by the way.
> You are dead wrong on accounting. Nobody has to make formal invoices for all their subscribers.
I'm German and I'd prefer to not have my donations re-classified as taxable commercial income. Our government's interpretation on what defines "gewerblich" can be pretty insane - you have to provide a full imprint with name and address on your stupid cat photo blog if you're reaching more people than your family. There have been a number of people in my Twitter feed who got into serious trouble for anything from running donation pools for their pets, beer money-style donations, and most recently Onlyfxns.
Our government is braindead when it comes to realities of digital life, and so are many others. Too many people just rely on never appearing on the government's radar.
> It's easy to accept SEPA payments and many other payment forms without credit cards through Stripe. Those germans who don't have credit cards have debit cards and they work the same.
As said in the first point, you're again at the mercy of a middle man that's hard to hold accountable. My local bank I can at least file a complaint at the banking regulator, good luck with any of the US institutions.
Yes, of course you are at the mercy of a middle man. That's been the reality since credit cards where invented. Literally every business online or offline has to accept this. The price is cheap considering what these middle men provide.
You don't have to use a US card processor, there are European options. The fact is that these processors have millions of companies using them, handling probably billions of transaction every year. They are much more reliable than European banks, who will close the bank accounts of sites and journalists writing things the banks don't agree with.
As for donations, fair enough. Why not get your donations to an account outside of Germany to escape that headache?
> Again, middle men. Middle men that are going to censor you for whatever reason - no matter your political direction, by the way.
I'm aware of the censorship concern. Still it's worth a shot to try to bundle your content with like minded, you don't necessarily need middle men for this.
All in all, something will have to change with the way online content is consumed and produced - at least if we want to take full advantage of the blessings of the information age instead of rotting at the mercy of a media oligopoly.
> You don't have to use a US card processor, there are European options.
You're still at the mercy of what the US networks deem acceptable. Even if you are German residing in Germany, have a German bank, your customer is German residing in Germany and has a German bank... say, you're a sex worker which is completely legal in Germany. And yet, you'll get booted off of anything where the US card networks are involved.
> As for donations, fair enough. Why not get your donations to an account outside of Germany to escape that headache?
Still tax evasion. Yes, I could create something like a Maltese shell company, but that is even more of an effort to set up and maintain.
Either pay the taxes your lords demand or don't pay them. You can't have it both ways. You don't need to set up a shell company, you can set up a bank account and never declare.
As for Visa/Mastercard - yes you are at their mercy. They are not perfect, but what are your options right now, realistically? Except for manual verification of payments?
The internet seems ripe for the return of a service like StumbleUpon. I used to find all kinds of great niche content through that service. A much better way to kill time than endlessly refreshing a feed or shitposting -- it feels more like exploring the internet.
I am waiting for the day that GPT can show it's work so to speak. i.e. "Debug last results" and get a list of all the trained input data including any URL's, books, text input and who performed the input, where the data came from. I am told this will never happen but I think courts will disagree once enough big mistakes and/or potential malfeasance or fraud occurs and people try to blame GPT. Or when GPT repeats the unsavory behavior we have all witnessed with social media algorithms leading to highly divisive outcomes.
I think this is a good idea. When we disagree about things in conversation, we as humans ask "where did you hear that?" and from there, we dig into the source of the thought process.
Unfortunately, a lot of what folks are saying to you is true. It's very hard for GPT to cite its results because the way it is trained it is by design made to garble the inputs down into a simplified concept of guessing the next words given some input.
It essentially just has a procedural generator for words that is seeded by your prompt. If you give "alfalfa" it will head off in a direction towards "farm", "hay", and "grazing" along with other connective words to form sentences. Because its concept of data is all around just word spaces, it can't really go "oh, I read about alfalfa on a Wikipedia article with x sources". It just knows "alfalfa is like the word grazing". I am simplifying to make a point, but this is in essence how these algorithms work, directions of traversal, guided by probability, towards word clouds floating in a grouped space.
This is sort of changing though. Bing and Google (as well as many other researchers) are using specialized databases to provide further context that is fed into your prompts that come from real search results. Theoretically, they could get this tuned enough that GPT and other LLM can have the right data to provide a connection to cited facts alongside the hallucinated glue language.
I feel like what you're asking for is valuable, but might take a bit before we really get it relatively accurate.
Basically what people who went against the “we won’t be able to tell what’s real and not” deepfake narrative even 5 years ago predicted. Ie a web of trust will still keep up paths of believable high quality content etc
Blog post curation services like Thinking About Things are great for finding the best of the internet without the hassle of wading through the cruft. Anyone know of any others?
Do most people write stuff and then go back and search for links to relevant content?
As a writer, I'm continually inspired by things I read or watch, and when I find something, I add it to my references. I use Zotero[1], but it's not the only solution. When I start writing, I already have my notes and the links to sources and related content. I don't add in links as an afterthought, the writing is informed by the related material.
Maybe, once in while, when I've done a couple of drafts and am getting close to the final version, I'll go back and add in links, but rarely do I spend time poking around on the web. I go to my references. Granted, the reference database takes effort to create and maintain, but the end result is far better than what I'd get just doing a web search for a term.
I 100% agree with this, and I try to stick by it for just about every article I write now. Does someone I'm interviewing talk about a game or movie they enjoyed as a kid? It'll get linked to the Wikipedia page for said work, or if possible a place the user can (legally) buy or experience it for themselves. Reference something I made in the past? That'll get linked where relevant, and the same goes for anything mentioned by a source too. And I'll always link to the original source if possible too, not the random reprint/paraphrased version from whatever popular news site or service covered it this time.
Love the idea of using a script to help with this too, though it's usually the harder to find stuff that's more valuable to link to, not things you ca immediately find with a quick Google search (like the Google home page).
I want to implement this in my Hugo blog build process. Initially, I thought it was ironic that there was not a single link to Linkoln, until I realized that code is on the article and there's no code in github.
This is where the article lost me. If it feels like a lack of discipline, it probably is.
By placing links in an article, the author asserts that those were intentional choices just as much as were any of the other writing decisions made while making the post. The same thing goes for decisions surrounding the theme of the blog and what DNS name the blog is found under. The author is presenting an idea; more or less effort can be put into this, but it will show if effort isn't put into it.
Not spending time on links is the equivalent of not spending time on citations, arguably one of the most important parts of an academic paper. It is almost as important a part of the writing as the writing itself.
I love the concept of getting links "out of the way," though I did it slightly differently. Instead of having links in the body of the text, links in the text of my docs tend to be links to citations in a reference section on the same page.
I find it quite odd and funny that the author specifically mentions Google, and his script has a function `def google_it(query):` but then in the function he uses DuckDuckGo. Is the author ashamed, or does he expect people to not trust DDG?
As a political blogger and upcoming author, links to sources are damn important for me, for a multitude of reasons:
- they allow me to back up claims with cold, hard facts instead of "pure hearsay"
- they defend me from getting held liable... say someone files a libel suit/C&D order against me. If I have a link to a proper, accepted medium (e.g. a newspaper or a TV station), I can defend my claim by pointing to the original source(s).
However, it is becoming increasingly difficult to link to stuff:
- publishers decide to put interesting content behind paywalls, leading to a constant noise of "can't read, paywall"
- publishers decide to re-launch their website, but not set proper redirects so all my archives are dead now
- German public TV/radio has to take web articles and TV/radio archives offline after a few days or months, because private media got a court decision and then a legal provision forcing them to take down content [1]
- some media absolutely LOVE live tickers, but you can't directly link to posts in these for posteriority
Here's a better approach - never, ever include a link unless it is important and cogent to the topic. Humorous or casual linking is just noise, and it can shroud actually useful links. We all can find information, so it isn't valuable linking to the definition of phrases or words, etc.
If you're referencing a study, link to the study. If you're citing a tweet, link to the tweet. Otherwise save spurious links as they add zero value and are a distraction and an unnecessary decoration making text less readable.
As an aside, I chuckled seeing the link to Atwood's "be a bigshot blogger" post where he recommended that people blog constantly about everything. For those who haven't kept track, that was a failure model. It made people basically give up on "blogs" because there was so much low value content, with people writing on a schedule rather than because they had actually interesting content. Now everyone just hopes that the rare useful post appears on a social news or media site.
Creating CRUDs has been so degraded that entry level to work on them is none. Unaware yet candidates don't even have to create a PDF. Companies recruit CRUD devs from Instagram and Twitter.
Just a word of writing advice, if you link content inside the text it should be first or last word, never in the middle. It just takes a lot of effort to know what the link in the middle of the text is about.
I don't agree with that at all. For example, if I refer to violinists like [Oscar Shumsky] and [Yehudi Menuhin], and I create links from their names to relevant articles about them, I don't think you'd have any trouble figuring out what it's about. I think this article did a good job of that, linking from relevant text that clearly indicates the topic of the destination, except maybe for "Writing consistently…".
Personally it constrains my eyes if the link is in the middle of article. It simply takes an extra effort.
However, its not just me. You can find online a bunch of research on the topic.
Can you share the research on this topic? If I understand you correctly what you’re saying is very different to my experience (and how popular sites like Wikipedia etc work). So I’d like to read the research. I’ve tried to google a bit but can’t find anything yet and not 100% sure what keywords to search for.
sorry it was weeks ago, And by spending 5 minutes googling i couldn't find it.
However I can summarise the point of that article. It basically says since links stand out, first thing you read from paragraph are links. If the link is on the beginning of the article, you naturally read the link and continue reading it. If it's on the end of the article, you'll first read the link and after it go back to begging of reading paragraph. If it's in the middle, you'll wonder what the paragraph is about, so your eyes will naturally read the first few words of paragraph and after it a couple of end words of paragraph. Before continuing to read article.
The author that wrote that article was obviously an expert in UI/UX and has spent a great deal thinking/doing articles.
An example that you've mentioned, Wikipedia, it doesn't have such huge contrast between link and text, and I'm expecting it to be full of links.I suppose thats the reason why but I don't do the described behaviour on it. However I often find myself doing the described behaviour when reading blogs, making me ignore/stop reading the blogs that do a lot of linking.
I'm still having trouble picturing how to actually implement the suggestion.
Could you, as an example, rewrite the first paragraph of the fine article, so that the 3 links are either at the start or end of the paragraph (as suggested), in a way that makes the contents of the paragraph and/or the link context clearer?
Looks like you're misunderstanding. There is a need to downvote because you're projecting your opinion (that some people strongly disagree with) as "advice", as if it's a fact. It's correct to downvote imho, which is what I did.
Similar to Wikipedia, it's ok, if not preferable, to link words in the middle of a sentence. I prefer it, some people I know prefer it. It's ok to give advice, but when the jury is still out, it's crucial to be humble.
> It just takes a lot of effort to know what the link in the middle of the text is about.
How does specifically linking the first or last word (of ...the sentence? paragraph?) make things any clearer?
Wouldn't it be best to just link the words that most describe what the linked-to page is about (maybe even the linked-to page title?), no matter where they appear?
IMHO links should be chosen "wisely", with the intent of helping the reader to find more information on important parts of the article or referencing other work etc.
Linking to the definition of rabbit holes is not really useful, if I'm not talking about varying sizes and depths of rabbit holes due to different breeds of rabbits and their prefered style of rabbit hole digging.