[gmurphy]

I designed Chrome, and was responsible for decisions around its UI, including the addition and removal of the RSS button.

Our design philosophy at the time, which was in reaction to the bloated I-need-engagement-for-my-team browser UIs of the time, was to only offer what people needed, and allow extensions to cover everything else.

I loved RSS. We all did. I still use Feedly every day and mourn the loss of Google Reader. But even back then, practically no-one else cared, even amongst our early adopter userbase. If we had lowered our usage bar to allow the RSS button, the bar would've been low enough that a thousand other features-you-don't-want-but-other-people-do would've been in there too (omg the arguments about having a "print" button).

Extensions was our "if you want it, you can add it" answer. It was imperfect, because it didn't allow ideas such as RSS to be advertised to the mainstream, but we had already clearly seen that that hadn't worked, and regardless, your daily tools should not be a place for pushing agendas unless you have total confidence that they will be valued.

I still think "following stuff" is an unsolved, undervalued problem and big opportunity space

[sirspacey]

This is something those of us who grew up with the Internet don’t want to face.

Whether by nature, nurture, or some other force - user agency lives in the very long tail of any software usage.

In every category, agency drives the birth of tech but dies in the scoreboards.

Every product I’ve loved, every founder I’ve believed in, was about human possibility & creativity.

But the ones lucky enough to scale to species level adoption discover that the ethos of creation is lost in the swarm of consumption.

Perhaps I’ve arrived at a poor conclusion, but I’ve simply decided to accept that I am a fairly unusual person for wanting to craft, make choices, and discover the weird, wonderful, or just discover what I didn’t know there was to discover.

I’ve weened myself off of consumption over the years. Here’s to making all my future decisions in tech about putting my attention, energy, and thought into “what do I want to create? learn?”

I don’t think the product universe I live in will ever be dominant. But I do believe it can exist if we accept that we can either build it or trade it, but we cannot do both.

[gmurphy]

I don't think you are alone here - there's a reason Chrome invested so heavily in dev tools and made them (and things like view-source) accessible in the core distribution even though they were also perfect candidiates for being addons. Many of us owed our careers to the path that view source opened, and we wanted to pay it forward.

I do regret not finding ways to do more - for example, an unrealized goal of mine was to find a way to add a 'create' button to Chrome to nudge everyone to be a creator - my favorite simple+cheeky idea was to have a way to enable contentEditable on the current site, then save the resulting changes as an extension you could give to your friends, allowing you to graffiti the internet in private small groups.

[DonHopkins]

A brilliant idea that would never make it through the legal department.

Somebody I worked with at TomTom had the brilliant idea of gamifying the driving experience with internet connected TomTom PNDs, by publishing a real-time Leaderboard on every single road on the map, so you could see the top ten speeds, and compete with other drivers for the high score! No matter where you were driving in the world, you could instantly see the top ten speeds of other TomTom users who drove down that same stretch of road, and put the pedal to the metal to claim or defend your own spot on the leaderboard!

My TomTomagotchi idea didn't go over well with TomTom legal either: they didn't want their PNDs getting depressed and dying if you didn't give them food, attention, and drive them around everywhere they wanted. But I was sure there's a revenue model having drive through Burger Kings and car washes pay for placements.

[AlexanderTheGr8]

Unrelated but you are a great writer! "user agency lives in the very long tail of any software usage" "agency drives the birth of tech but dies in the scoreboards" "scale to species level adoption discover that the ethos of creation is lost in the swarm of consumption"

You remind me of the writing of Antonio Garcia Martinez in the book Chaos Monkeys.

[leobg]

Quite a Nietzschean insight. Both in therms of its direction and the quality of its expression in language. I am bowing to you, Sir.

[greenie_beans]

> If we had lowered our usage bar to allow the RSS button, the bar would've been low enough that a thousand other features-you-don't-want-but-other-people-do would've been in there too (omg the arguments about having a "print" button).

this is when businesses should make decisions based on vibes rather than data.

[smusamashah]

Unlike other browsers at the time, why was it decided not to even read RSS xml files and show them like a nice page like other browsers? Even now RSS shows up as raw xml.

Edit: reread your comment. I guess it was because this could be done using an extension. I believe this to be the main reason for lowering adoption / use of RSS with rise of chrome. To this date chrome shows raw xml when you open RSS xml while other browsers give you something interesting and readable to look at and even let you subscribe to that interesting page.

[gmurphy]

Unfortunately I don't remember the exact reason, but with some squinting at the past I believe it would've come down to two things - resources and how we thought about the role of developers:

Chrome was a no-crunch 50-person project, which sounds like a lot until you think about how we were porting webkit to windows, adding a sandbox, creating a new javascript engine, a new windows UI toolkit, etc and we were ruthless about prioritization - any engineer who could've spent on an RSS renderer would've been taken away from valuable webkit windows compatibility work (omg ACID tests), and weeks mattered.

Second, we had an extremely strong belief in letting developers own as much as they could - our impossible ideal was that the Chrome UI wouldn't exist and developers would just create content you could access emphemerally through your operating system. At the time, we had seen that it was possible for developers to use various trickery (XSLT or CSS?) to style RSS feeds so they would show up somewhat like how you described (IIRC Feedburner did this)

So if you take those two things together (and again I may be misremembering) you see it being a "developers can do it so if they care they can do it, we should not spend time making something that overrules them"

It's not possible for me to know whether that was the right idea in hindsight because I don't know what we would've sacrificed (I think I would've preferred using that time to make the standalone image renderer have better zooming/panning controls)

[rossjudson]

This is the reference answer, direct from the source. Much appreciated.

Sorry that you have to put up with speculative nonsense from the willfully misinformed.

[mattparmett]

Your comment made me curious, were there any other features or ideas that you (& team) loved but didn’t have the user base or buy-in to include by default?

[gmurphy]

Thank you for taking me down memory lane with this great question. It's been almost twenty years, so I may misremember the internal popularity of many things, but one that stands out was browsing history - almost everyone that joined the team for the first ten years wanted to redesign history and felt it could've been a huge differentiator. We had mockups that looked like subway maps, implemented floating overlays showing your recent tasks, and we designed the history database to store everything forever.

In reality, Google search was good enough to find what you wanted 95% of the time, and most people would not bother discovering and learning a separate pathway to find the remaining 5% - especially when combined with our poor human memories and local search capabilities at the time.

[MichaelZuo]

> We had mockups that looked like subway maps, implemented floating overlays showing your recent tasks, and we designed the history database to store everything forever.

That would still be a real innovation over the current 90 day retention maximum.

[yencabulator]

Unfortunately, today's version of that is "Web & App Activity", sent to Google servers with the Terms of Service allowing it to be used for ad targeting and what not.

(And I think the real innovation would be showing the branching, not mangling everything into a linear history. But it's harder to show and use well, so it doesn't happen. Very much the same story as editor undo/redo.)

[makeitdouble]

> But even back then, practically no-one else cared, even amongst our early adopter userbase

This reminds me of chapter support in podcast players. For a very long time the every player dev was arguing that nobody cares. And it took years for the tune to become "nobody cares except germans". Then it was implemented for a "vocal ninority", and now everyone gets chapter support and I don't think any user is lamenting about the evolution.

This is one of the worst aspect of the "nobody I know cares about it" mindset (I don't think we can really work around it, but it should be acknowledged and kept in mind).

The other path is to push for many more places building their own tools and services, so they don;t even have to ask for these changes and do it themselves, with everyone else getting to enjoy the result.

Basically, while Google was the reason Chrome existed in the first place, I think it's also the main liability to have a product designed with a global audience in mind.

PS: the issue with chrome extensions for core behaviors is they'll always be second class citizens. I'm still happy there's an escape hatch though.

And also thank you for your work on the browser, I don't want the tone of this to be overly negative when it's really about side points.

[fomine3]

Perhaps such users tend to disable telemetry

[martinsnow]

I just want to say thank you for killing IE