Hacker News Reader

Comment by antirez | Hacker News Reader

Regardless of my tweet, that is a personal opinion based on my feelings, background and experiences, the discussions in the replies are interesting from the POV of understanding what different people think in the programming community. Some will say that critiques are only from old people, in the style "when I was young we had only zeroes! You now have ones and zeroes". Others say that the web is slow because of ADs and trackers and modern frameworks would otherwise be fast. There is who believes that the problem comes from big companies imposing new standards in order to turn the web into the new Java applets (a full fledged application platform), and so forth.

View on HN

I read it and feel it's the same as how we perceive modern music. Everyone says music "used to be better" but that's just survival bias. Just like music, there was a LOT of trash web development back in the day as well. Sites built with tables in dreamweaver best viewed on netscape navigator at 800x600 with animated gifs bogging the download speed existed long ago.

What we have today are the same problems with a new wrapper, created by a not dissimilar group of people from the past.

>I read it and feel it's the same as how we perceive modern music. Everyone says music "used to be better" but that's just survival bias.

Well, if the top 100 tracks from, say, 1961 to 2021 progressively get less musically diverse, with simpler chords, less harmonies, less timbral variety, lesser melodies, more repeatition, less dynamics, less genre variety, more infantile lyrics (something that has been studied and measured several time, e.g: ), etc, then it's not some "survival bias".

https://www.researchgate.net/publication/346997561_Why_are_s...

https://newatlas.com/pop-music-trends/23535/

https://www.nature.com/articles/srep00521

https://www.res.org.uk/resources-page/economics-of-music-cha...

https://pudding.cool/2018/05/similarity/

To be fair, you have to adjust for the relevance of top 100. In the past the top 100 was the only thing most people were exposed to because that's all the radio played. Today Spotify and Youtube and their recommendation algorithms make the top 100 almost completely irrelevant.

The size too: there are far more songs released now than in the 1960s. So it makes sense that the "creme" on the top is more homogenised.

Why does that make sense? More available music should imply a more varied range of good music to rise to the top.

It happens in some simplified models. Suppose a thousand tracks get released every year, and these tracks are evenly divided among 50 different types of music, 20 tracks per type. Some of these types of music are very popular (4.95 stars), some are slightly less popular (4.90 stars, maybe because they offend some powerful group), and some are much less popular (4.00 stars). So the most popular 100 tracks will be the 20 tracks from each of the 5 most popular music types.

If we increase the number of tracks to ten thousand, there are a couple of ways we can go. We could increase the number of types of music to 500, and in that case, we'd see better music rising to the top. (Or at least more popular music, which may or may not coincide with being better.) Or we could increase the number of tracks per type to 200, in which case a random half of the tracks of the most popular 4.95-star music type will be the "top 100" for the year.† Or we could go for the middle of the road: maybe now we have 150 types of music and 67 tracks of each one. Or we could have musicians and record companies that respond to incentives by trying to produce more music of the popular types, somewhat handicapped by the fact that those popular types change every year, and however much Nickelback might try it, recording the same song under twenty different titles doesn't actually give you twenty top-ten hits.

Regardless, there are a lot of ways that more published music could both provide more variety and less varied top hits.

______

†Of course the Billboard Hot 100 is per week, not per year, but that's the least of the oversimplifications in this model!

Not necessarily. That might be true if the top X you're tracking is also expanding along with the size of the catalog, but isn't true if X is fixed, like a top 100 music chart.

More common variants of more popular genres could easily crowd out moderately popular genres. The long tail has been a long noted issue at Spotify with several attempts at fixing.

It reminds me of why Apple Pie is America’s favorite pie.

50-100 years ago when parents were deciding what pie to make or buy, parent’s favorite was X, eldest kids favorite was Y, youngest kid’s favorite was Z.

Parents instead bought Apple Pie, because it was in everyone’s top 5 favorite pies.

Top 50 is an average of all of the world wide listeners. No one’s favorite songs are in the Top 50, but they are the average top 50.

The articles you're linking are proven to be misrepresenting.

They for example look at a really small subset of music "Million song dataset" and only analysed basic metrics that could be automatically measured. To be honest I think the linked Spanish paper and the senationalist "Science proves modern music is bad" should be retracted since the methodology is flawed.

Don't take my word for it, take Tantacrul's - composer, video creator, Design Lead for MuseScore [0].

It's such a complex topic to study "music" - you have to narrow it down - genre, country of origin, purpose - it's a major simplification to say that "all music is worse now", there are a lot of types of music that didn't exist in 1961 (synths, electronic, hip-hop) - how can you compare it?

[0]: https://www.youtube.com/watch?v=VfNdps0daF8

>The articles you're linking are proven to be misrepresenting.

Citation needed. The articles are actual proof and measurements, whereas Tantacrul's take is just an opinion.

>They for example look at a really small subset of music "Million song dataset" and only analysed basic metrics that could be automatically measured.

The "small subset of music" is the most popular music of any year. That is what reveals music tastes over time, and this is the kind of music that permeates culture.

As for the "basic metrics" there's not basic as in trite/insignificant but basic as fundamental.

It's just that some people want so much to cling to "each period is the same, there are no ups and downs in cultural production" to not be seen as backwards oldsters, whereas milenia of history teaches that that there are absolutely ups and downs in cultural production (periods of stagnation, etc).

>it's a major simplification to say that "all music is worse now"

It's also a strawman. Nobody said that here.

Really love Tantacrul's video on this. I was actually about to look for it to reply to the grandparent post when I saw your post.

>(synths, electronic, hip-hop)

Synth began in the 70's, and Moroder/Jarre is far better than the 90% of today's crap.

If it's better than 90% of today's synth music, does that mean that 10% of today's music is better than Moroder/Jarre?

That's a lot!

No, just means it's equal but derivative...

The thing about music is that "more complex chords, more harmonies, more timbral variety, etc..." doesn't say anything about whether or not the music is better. It's a silly pursuit anyway given that music is a subjective experience.

That's neither here, nor there.

This is not about whether or not this or that particular song is better, is the top (most listened/streamed/talked about) music in general losing variety in all these aspects (harmony, melody, timbre, lyrics, genres, dynamics, etc).

That's not subjective, that's objective, and has been measured to get worse.

The "subjective experience" could be good, the same way people can prefer McDonalds over a wholesome meal by the best chef. Doesn't mean its also good by other metrics, and I for one don't believe individual taste is everything. There are people with shit taste, loads of them.

Again, “worse” is subjective.

Even saying someone has “shit taste” is subjective.

In the end, everything is subjective, even morality. There's no objective physical law that says killing is bad. Animals do it all the time and could not care less about it.

But to the degree that we have a culture, though, we also have a non-100%-exact but nonetheless existing hierarchy of artistic works. There might be disagreements, even strong ones, but there's also some general agreement, that not everything is a fuzzy blob of equal value, left for the individual taste to sort or not, and this just for itself.

Is the idea that the Beatles are better than The Monkeys or Oasis, that Aphex Twin is better than Skrillex, that Dua Lipa is better than Justin Bieber, that Michael Jackson is better than Milli Vanilly, in some non-measurable but tangible way really that difficult?

That this, once a common and well accepted idea (related to the idea of the "canon"), appears like beyond the pale for the 21st century solipsistic individual, where only the subjective taste matters, is not really the fault of the idea itself.

I wonder how deeply tinted your rose-colored glasses need to be to go to these lengths to prove that things were better back in the day.

Or how deep a fear of being called an "oldster" one has to have, to need to go to these other lengths to prove that things are always a fuzzy blog, and there are no periods (and historically even decades or centuries) with worse or better artistic output...

There's a simple metric for the state of modern music - where are the modern equivalents, in musical stature, of David Bowie, Pink Floyd, Roxy Music, Bob Dylan, Elton John, U2, Fleetwood Mac, Steely Dan & The Eagles? Answer - there isn't anything anywhere near the pantheon of talent that graced the late 60s to the mid-80s. It's just a phenomenon of history. It's not relativism due to ageing. It's a fact in the same sense that Shakespeare's work outlived that of his contemporaries.

I guess this depends on 'musically diverse' === 'better' though, which isn't necessarily the case when people think about what makes the top 100 tracks for a year.

The top 100 hasn't been a relevant indicator of modern music since Napster. The top 100 is a record industry owned entity with a tremendous amount of self-interest.

If you can't find more diverse modern music than the top 100 offers today you're not trying that hard.

That's neither here, nor there.

This is not about being able to find this or that niche musician that's great or even greater than any in the past, but about what the masses listen getting worse.

Music isn't just a solitary experience, but also a part of general culture.

As streamed music is getting cruder over time, the majority of the people are listening to increasingly shitty songs. That's chilling, regardless of whether someone can find 10000s of niche bands to listen themselves.

It's reductive to suggest that the top 100 represents "general culture" (whatever that ambiguous term means). There aren't solely niche musicians outside of the top 100. Theres plenty of musicians with millions of listens/views on streaming media that don't enter billboard lists that have extremely prodigious careers that have complex lyrics, use a wide assortment of instruments/equipments, evolve genres, etc.

Your data might be objective, but it's still a narrow slice of a much broader ecosystem.

I have a different opinion. When we moved from the old way to semantical HTML5 generated still server side but which the layout was defined by the finally mature CSS, with the dynamic parts handed by JavaScript and RPCs, everybody agreed it was better. Such web was faster, more compatible, simpler to parse, and so forth. I think that, if in the 20 years that followed, frontend development would progress in the same direction of improvements, everybody would agree again now.

And indeed everybody agrees that JavaScript itself is now better as it is the modern web APIs. If the rest is so controversial there must be a reason.

I’ve heard and read stories from seasoned engineers, here on HN and elsewhere, that say that FE dev was as prone to tire fires back then as it is now, so I don’t think that point is as incontrovertible as you claim it to be.

Some of those seasoned old-school engineers have, believe it or not, embraced modern FE frameworks and architectures and wouldn’t go back to the old thing if you paid them to do so. I’ve seen testimonials here on HN and surely it must be a factor in the world moving in this direction - it was not the juniors who made these decisions after all.

My point is different: I said that true evolution from tables to create the layout, continuos refreshes of the page to update the state, to a better development environment was universally recognized. And if it was again a case of good design, everybody would see it. If many voices are talking about over engineering now I bet there is a reason.

We had functional applications for forms and the like, in the 90s, running on machines with 8MB or 16MB of memory. They weren't as pretty, but they had simple development paradigms, VB and Delphi, easy to get started.

HTML has been the wrong place to make creating UIs (as opposed to marked up text) for such a long time. Things are getting better from multiple angles, but it's still very uneven.

It's always easy to solve a problem by taking away a number of it's requirements. VB was good at forms but visually unacceptable for branding. It was difficult and unsafe to deliver over the internet. And again, there were a LOT of really bad VB applications that hurt usability and performance.

If all you're doing is building forms for an application then it's acceptable, but that's an easy problem solved with elegant solutions today as well. HTML can immediately deliver the styled application without installation to every device regardless of OS. It can propagate changes without reinstallation. It is accessible for every kind of user. And it does this while looking miles better than any VB application.

>And it does this while looking miles better than any VB application.

Far less usable too.

> It was difficult and unsafe to deliver over the internet

No one cared, we had physical media.

> And again, there were a LOT of really bad VB applications that hurt usability and performance.

Most of them were fast enough.

OTOH, i think you can actually measure the quality and accessibility of tooling based on how abundant the long tail of mediocrity/low-quality output there is.

in music’s case, one of the replies to this comment describes decline in musical complexity/sophistication, which i’d personally attribute to democratization of the tools (which are also much more powerful, allowing kids w computers to do what took whole teams and studios full of equipment before).

so i think only seeing high quality UIs in the wild is more of a mixed bag than is intuitive to us — a world absent of shitty soundcloud rap is a world with worse music tooling.

I definitely agree with the music survivor bias thing. This is also very noticeable in the “computer generated imagery in movies looks artificial” meme: you just didn’t notice all the extremely effective and convincing CGI.

But I’m curious, what are the examples of great websites “back in the day” that have stood the test of time and would be considered good web development today?

Wikipedia and Zombo.com haven’t changed much.

You're on one right now. Although it'd benefit from some mobile compatibility...

Craigslist

> Everyone says music "used to be better" but that's just survival bias.

Not "just" survival bias.

I think it's quite reasonable to take a position that there was more innovation and creativity in pop music in 1950-2000.

As the genre has matured, popular/commercially successful music has depended more and more on fewer and fewer producers.

Indeed, there are fewer commercially successful artists: the US Billboard Hot 100 Top 20 this week contains 3 tracks by Justin Bieber, 2 by the Weeknd, and 2 by Drake. That would have been unheard of.

> it's quite reasonable to take a position that there was more innovation and creativity in pop music in 1950-2000.

A lot of people said the same in 1990, except their cutoff was 1980. All that synth is just rubbish, man, what about Led Zeppelin, that was new! Except they were just raiding blues and laying some electric guitar on top, of course, so really the cutoff is 1960; but to be honest them bluesmen were just recording stuff that had been sung in the cotton fields forever, so really the cutoff is the beginning of the slave trade; but really those melodies were just brought all the way from Africa, so really... etc etc etc

I think this essentially agrees with my point (that there was more creativity in the broad space of 'pop music' when pop music was a newer phenomenon).

I appreciate the reminder that a lot of pop music has a longer history as it developed from Black music styles. Thanks.

I think you've got that completely backwards. 2000 is about the time that the "long tail" exploded, where Youtube and Spotify made non-top 40 music a lot easier to find. In consequence there is a lot more really good music available today than there was pre-internet and it's a lot easier to find. It's not on the radio, but that's always been the case.

> In consequence there is a lot more really good music available today than there was pre-internet and it's a lot easier to find.

Tangentially, if (like me) you find this "long tail" interesting but don't know where to start, I've enjoyed going through Ted Gioia's "100 Best Albums of [previous year]" [1] each year. It's obviously subjective and subject to whatever biases he has, but the sheer diversity of the list is quite cool. Listening to a few albums per week is a pretty easy way to sample -- most of the albums are on YouTube or Bandcamp for free.

[1] http://tedgioia.com/bestalbumsof2020.html

Exactly. One only has to look at the myriad of tv shows, video games, movies, etc. to see that we have more quality options than ever and are seeing even more niche, but commercially successful content. There’s more static too, but there’s a lot of good out there.

Everybody thinks we are in a golden age of television series, don't they?

I don't claim that these examples have the same history as music.

I don't think so at all - there's a lot of pastiche, some of it loving, but there's only so many ways to do a pop song. It's alright for forms to die. The real "long tail" that the internet introduced is that we now had access to all of the good music we missed before the internet. There's no reason to bitch about modern music anymore because there's no reason to be exposed to it other than aggressive fandoms and marketing.

I honestly don't even understand the impulse to listen to new music. The quality of "newness" is not a desirable one for me in music; I prefer qualities relating to sound. To me it seems like a distorted version of the impulse to be cool in high school.

The real problem is that commercial, recorded music crowds out local, live music. For local music I understand wanting to know what's new - you may miss out because bands aren't live forever.

And before, soulseek.

> Indeed, there are fewer commercially successful artists

Source?

I reckon every decade had a handful of artists dominate the charts, so it wasn’t unheard of at all to see multiple chart toppers by the same artists at any given time. In fact, given the gatekeepers of popular music largely were concentrated around radio DJs and the “industry machine” of record labels, I’d hypothesize there’s a larger on average number of unique artists in the top 100 since the 2000s than before.

>Source?

https://www.billboard.com/charts/hot-100

You can change the week yourself.

If you pick a few random weeks before 2000, you likely won't find this pattern that I identified, with the same artist in the Top 20 multiple times.

If that's not enough evidence then there is a Python web scraper: https://github.com/jocelyn-ong/data-science-projects/blob/ma...

Good, and commercially successful, don't always go together.

Amazon is commercially successful but few would argue their UX is good.

Do we know what the UX goals are for the site or what requirements they had? It's really difficult to gauge these things if we don't even know what they're going for.

Personally, there are some things I have found difficult to do (getting a chat going with a customer service agent) and some things that are very easy (checkout process, making a return). It's no surprise to me that they would make it slightly more difficult to find a live person to chat with since they probably want me to exhaust the other automated methods of solving my problem before doing that. Is that bad UX? I don't think so.

Amazon's user experience is fantastic. Why else do you think they are so commercially successful?

Amazon pages are a cluttered mess, but they are a -familiar- cluttered mess because everyone has used Amazon for so long. Amazon succeeds in spite of its design, because everyone has become familiar with how it works. This approach would not work outside of Amazon, which Nielsen Norman discussed some 15 years ago (https://www.nngroup.com/articles/amazon-no-e-commerce-role-m...)

I wonder how this might read if written now, rather than 16 years ago.

NNG often say to follow conventions and Amazon has been leader for so long now that I can't help but think they have established many conventions now that might have been viewed negatively 16 years ago.

> Amazon pages are a cluttered mess

Ahh, you want "clean" design where maybe there is just one box with beautiful font and not all these options "cluttering" things. You want those options buried in an option tree maybe 8 levels deep.

Well, I'm pretty sick of this minimalist design philosophy that requires 10x more clicks to accomplish anything because a designer decided the original was just too easy to use and wasn't "engaging enough" to generate enough clicks. They have made the web a much worse place for getting stuff done online, and it's no wonder that Amazon, the home of one-click purchasing, has continued optimizing for having the UI get out of your way and accomplish the most with the least effort rather than building attention-seeking UI "experiences" that require lots of interactivity to get simple things done.

The big one that always gets me is sort by price doesn't actually sort by price. Or at least, not the price I care about (which is price for condition new, seller Amazon).

Consumer logistics - which rely on not paying drivers and warehouse people much, and forcing them to work insane hours.

And AWS.

The web store UX has become appallingly bad - unreliable/fake reviews, tens/hundreds of sellers all roboselling the same identical items from China, poor quality items, unreliable/poor product search ("10TB external hard drive" shows... not many 10TB external hard drives). And so on.

If ECMAScript, HTML, and the DOM didn't exist and you were asked to create a specification for applications where the client UI is remote, possibly very resource constrained with a connection to the back end that may be slow and only mostly reliable, what would you invent? Is there a better model already out there that isn't used because Javascript + HTML has sucked all of the oxygen out of the room?

Even Flex and ActionScript were better in many ways than the DOM.

My big problem with web application (in a true application sense - not things that are just hypertext and shouldn't be applications at all!) development is that the DOM+CSS model is not made for rich UI experiences. Basic paradigms from desktop applications like spreadsheet cell selection, draggable cards (think Vizio/UML), modals, and MDI / multi-document interfaces are non-standard and brutally challenging to construct in a reasonable way using the DOM.

What I'd invent would pretty much be Silverlight without the Microsoft, honestly - a typed UI framework built on a widget and widget-binding model which would allow a smooth mixture between OS-level widgets (and the accessibility affordances they provide) and hand-controlled rendering/drawing, with a stripped-down runtime enabling resource constrained clients to execute client-side code which would hide / paper over the resource constrained backend connection.

Anyway, I also think this is orthogonal to the argument in this thread, because I think that most of the conversation and the sentiment of the original tweet is to call out applications that SHOULD be hypertext, not applications. For applications that need to be applications, I think things have gotten better, not worse, although they're still pretty bad.

As a language, ActionScript 3 was actually pretty good.

The reason it's so hard to come up with a feasible alternative at this point is that we have done ~15 years of browser wars with perhaps an average of 3k "core" contributors (browser company employees etc) since "we" moved into this direction... pretty much by chance.

My understanding from working at Opera at the time (2004 and onwards) is that the "senior" (experienced) people were busy implementing stuff in the browser engines and various GUI platforms. We hired very young (often like 17-18) and very smart people who had experience actually writing HTML/CSS/JS to work on developing web standards. They naturally typically had very little commercial software development experience.

After a while it kinda became a competition - which browser company's web standards people would be leading in terms of ideas/innovations. How many web APIs could browser company A do, vs browser company B. That's when the complexity really started accelerating. Then Safari and Chrome happened.

I wish we had spent more time working with these web standards people (we had so much experience building GUIs, for instance). They were really friendly and approachable, but we were all so busy with actually building the browsers, during these browser war times. It feels like a missed opportunity, in retrospect.

Java Web Start was a thing, and had some nice aspects to it. JVM startup time has been a problem, and the UI toolkits maybe had some issues? And of course the early JVMs were notoriously full of security holes, which is something browsers somehow managed to avoid.

But nevertheless I think Java/JAWS got basic stuff still right: run applications directly from network (with auto-updates), have security controlled sandbox, be fully cross-platform and portable, and have useful stuff built-in, and even had the PWA-like thing that you could have desktop shortcuts to JAWS applications.

Regarding startup time, I do wonder what would be the startup time of your typical Electron app on a late 90s-early 00s PC hardware. Somehow I'm imaging JVM might not be that bad in comparison...

As far as I know, security issues were mostly due to security not being too important at the time, but could have been trivially solved. Startup time is similarly a question of what should be optimized for.

I would have liked to see a web built on Java tech.

Depends on use case, but for general purpose apps MS XAML platforms are IMO better. They were designed for rich GUI from the start and it shows.

The main downside, they only work on Windows, MS once had Silverlight for Mac but it wasn't particularly good.

X Windows was an option back in the 90's, and with frameworks like Motif++ it was kind of ok.

Does X scale? Can you have thousands or even millions of simultaneous clients hitting a server?

So, X flips the nouns, and the server is the part where the user sits, and the client is the part where the data processing happens.

It used to be possible (and done) to have a thin client X server that would have the whole desktop running elsewhere. You could run a computer lab with a bunch of thin clients and a handful of desktop servers for lack of a better word. That was probably hundreds, maybe thousands.

Since then, though, X moved a lot more work from the Xserver to the clients; text is almost exclusively rendered on the client and passed as an image to the server to display.

If you did a classical experience, it might work ok to millions, but the user experience would have to be pretty meager. There's also the issue that there's not really a good way to sandbox different X clients; so nobody in their right mind would let a random internet server connect to their Xserver. Connectivity issues abound as well.

> Others say that the web is slow because of ADs

This is pretty much objective truth in my mind. I might agree with people who hate the web if I always had to browse with ads enabled.

My worst experiences in the web come from mobile media (news) sites where I can’t block ads. They are absolute dumpster fires.

Everything else, including the aforementioned sites when ads are blocked, is pretty decent in my experience.

I think this is more about the limitations of our perceptions given we live such short focused lives.

Reply on news.ycombinator.com