Apple’s first online store played a crucial role in the company’s resurgence | Hacker News Reader

Apple’s first online store played a crucial role in the company’s resurgence

appleinsider.com

95 pointsspking70 comments

One of the interesting things is that Dell was really the only retailer with significant online sales at the time, and their online store was built on NextStep's WebObjects, just as Apple's store would be.

Also interesting that Jobs had Apple pay to use Amazon's one click ordering patent at a time when other online retailers didn't see the point of removing sales friction.

> Why Amazon’s ‘1-Click’ Ordering Was a Game Changer

https://knowledge.wharton.upenn.edu/podcast/knowledge-at-wha...

> One of the interesting things is that Dell was really the only retailer with significant online sales at the time,

I thought gateway 2000 was a big online retailer around that time. In fact I thought it was them more than Dell that lead the online push.

I guess I’m misremembering as it was 25 years ago.

Both started their online stores in 1996. Dell's whole business model, though, was cutting down on inventory (JIT/Lean style) and intermediaries. Dell didn't sell through big box stores and others like Gateway and the rest did.

Gateway tried to expand by opening its own stores, but ultimately closed them all because they lost money.

Thus when Apple opened its own store it was seen as a desperate gamble (perhaps it was? I no longer remember) and destined to fail.

I was thinking the same thing, as we got a 486 dx2 66 from gateway 2000, complete with a faulty quad speed cd writer, ms office and encarta. And of course… the cow box. But this was ordered over the telephone from adverts I kept up to date with in computer magazines and the moment the dx2 66 was the best price (the 100 was the silly price one) I got my parents to pull the trigger and buy it. Happy times circa 1993.

Wow, a CD writer at the time a 486 was still a reasonable proposition must have cost more than the machine itself!? We had one of the earliest CD writers in the UK, paired with a Pentium with MMX, and the writer price was still comparable to the rest of the machine put together!

Yes! I remember a lot of friends getting the same cow box we did ;) I conflate things and timelines though.

Gateway was everyone’s computer at that time. All my friends and family had Gateway. Dell blossomed later.

Another interesting thing, is that for all the hate Java EE got, it was originally an Objective-C framework, Distributed Objects Everywhere, developed at Sun, during their Sun/NeXT OpenSTEP collaboration, inspired by Web Objects.

I remember being really furious that "1 click ordering" was "patentable" when I first learned about it as a young geek.

The longer I stay in this industry, and the more I learn about compliance, customer surprise, and the difficulty in balancing usability/simplicity with flexibility/customizability, the more I realize it absolutely was.

I recall an interview with someone about this - I think it was Tim O'Reilly - where it turned out that Bezos had a hard time getting the developers to even see the value in it. They kept putting in confirmation dialogs and he'd have to go back and tell them to take the confirmation out, and that was what persuaded O'Reilly it really was innovative, the fact that it was difficult to get the programmers of the feature to even understand the requirement.

Edit: maybe not O'Reilly, since he penned this [1]. Maybe it was Kevin Kelly?

[1] https://www.oreilly.com/pub/pr/537

I’m still furious that this is considered patentable.

Can you elaborate on the difficulty?

> the difficulty in balancing usability/simplicity with flexibility/customizability

Not sure what the OP had in mind, but my current favorite example is Expedia. I have used Expedia consistently for over 15 years. Have bought many flights for the family etc. over the years. They know who I travel with. Yet if I open Expedia now to book a flight, I will have to enter the ages of my kids again, as if Expedia didn't know their birthdays. Apparently it is difficult to get the balance correct.

It's not a question about difficulty, it's that nobody really gives a shit.

I'm sure everyone on this site knows that it would not be difficult at all to implement this.

Perhaps there is some valid legal concern over remembering minors' data, but if so, you can pretty easily come up with another dozen examples of sites elsewhere having similar issues with non-sensitive data.

The problem is that you usually have to have someone who is actually passionate about the product to drive all these little things home. Maybe this was a bug that was the next item on their backlog, but then someone in the chain of command said "it's good enough now, ship it.", and then everyone got moved on to the next thing.

Almost everyone at most companies, at all levels, is just doing a job. Very few people actually care about the product.

Was the patent technically nonobvious? The concerns in this seem to be saying it can be hard to overcome organizational and legacy friction instead, as far as I can tell.

I never heard of anything nonobvious at the time either.

Should storing people's ages also be patentable?

To the extent that business models are patentable, maybe so? We're 25+ years into online travel, and nobody has implemented it. Maybe it's one of those things that looks simple but isn't?

For technical difficulty, it's certainly no less difficult than storing credit card details for reuse, and that passed muster at the PTO.

All that said, I am excited to see what's (hopefully?) coming. A travel site should be able to intuit what kind of trip I'm planning, or ask me if it's not obvious. If I do a search for a city that I fly to every other week, perhaps it's okay to just go ahead and sort to the top the same flights and hotel I used last time. If the flight is in the summer and to a popular vacation area, perhaps guess that my family will also be traveling.

These things have all been technically possible for decades, but we don't see them in the wild. Maybe they are less obvious or more difficult than they appear from the outside (I have never built a travel agent).

OK, I think the difficulty isn't the first level of implementation, but all the edge cases.

I think that's true about a lot of innovation.

So yes, the elevator pitch is just "remember the shipping and billing details, and use it by default".

But the question is the edge cases:

First, you have to remember that this came at a time when people were still nervous about even giving your credit card to a website. Cross-web identity wasn't a thing, password managers weren't a thing.

Second, you need intelligence for what are the default options to select? It's easy for iTunes when the delivery is digital, or for a Prime member when 99% of the time the shipping is free.

But what about for everyone else - should 1 click choose the fastest shipping or the cheapest? Is there a 3rd option? OK, you might say the customer should be able to customize that in the profile. But A) what if they don't, and B) what are the extreme edge cases? If someone wants it the fastest, is it worth it if it costs 100$? Or is it a % of the cost of the good? Is faster shipping by 1 day worth it if the shipping costs twice as much?

And so on.

Wow, I never knew that Apple actually paid Amazon to use the patent. Very insightful!

> Michael Dell's comment came after Apple had started to develop its online store, but it definitely smarted — as you can see in video of Steve Jobs launching the Apple online store. You can see it, but you can't really hear it. Video survives of the presentation, but it is close to inaudible.

Even though this article was published today, the video link is dead due to the posting account ("SteveJobsArchive") being terminated. I found a mirror here, though, and the audio seems fine: https://vimeo.com/30998833

That is the presentation I was looking for.

I didn’t realize SteveJobsArchive was dead though. Dammit all.

After Steve makes a $16,000 purchase from the store, and he says "You can go the website ... so you don't take up the time of the salespeople at CompUSA" I bet Hal Compton (the CompUSA CEO who had just left the stage) was in the wings hopping mad at Steve.

Said around the 41 minute mark of the above linked video.

Audio cuts out around 25:42?

Looks like there's an issue where the audio is actually just mono, but is split into stereo channels, but after that point in the video, one of the channels is inverted. So if you're listening to them through a single speaker or otherwise mixed down to mono, the channels will cancel out.

Try using headphones or stereo speakers, or download the video and play it back in a program like VLC that lets you select just one of the audio channels to play.

I never really thought about this and took the website for granted. It's really brilliant how we can go back in time with the wayback machine: https://web.archive.org/web/19991008055013/http://www.apple....

This is pretty precious.

> The new online Apple Store brought in $12 million of revenue in its first 30 days, for an average of $730,000 per day. That's three-quarters of where Dell's daily revenue had reached after its first six months.

Uhh...

12,000,000 / 30 = 400,000

If I wanted to be charitable, I could guess that they might have gotten their numbers mixed up talking about net and gross revenue, but... This feels like a pretty big oversight.

Hmm, so 730k * 30 = 21.9M... with $21M and $12M figures in play, there may have been a digit transposition mistake somewhere, plus maybe truncation instead of rounding... Not that those are lesser mistakes, of course.

WebObjects was so vastly ahead of everything else in the space at the time, it's ridiculous and hard to overstate.

It did the MVC-ORM-Persistance-Templates so early, and so well.

I was fortunate enough to do one small project with it back in the day. Really nice to work with.

They ruined it when they moved it to Java. I get why (no one wanted to learn Objective-C before the iPhone), but it decimated so much of what was great.

And how rewriting of history saved Steve multiple times ?

Never seen a guy save a company so many times. Begs the question why it needed saving so often...

To be fair, Jobs-interregnum Apple was such a hilariously mismanaged mess that it did actually need saving multiple times over. Michael Dell's quip about "shutting Apple down and giving the money back to the shareholders" was not entirely as misguided as it sounds today.

I'm not sure I follow.

It's pretty widely recognized that Apple was circling the drain in the '90s, and things didn't turn around until they bought NeXT and put Steve in the CEO chair. Since that point, they haven't needed "saving". All these different stories you see are about the changes Steve made to the company that probably contributed to the turnaround.

I believe the common thought is that Apple got down to as little as 90 days before having to declare bankruptcy.

If Steve hadn’t made drastic (successful!) changes fast, there would be no Apple. They really were right on the edge.

Because business is hard without leveraging IBM's monopoly power. Most companies fail.

Something something John Sculley?

Who was basically brought in to Apple by Steve Jobs AFAIK

“Do you want to sell sugar water for the rest of your life or come with me and change the world?”

(Sculley worked for Pepsi at the time.)

There's a neat interview with Marc Benioff where about 15 mins in he mentions talking to Steve Jobs about an App Store, and I think later in the interview he also tells how he gave Steve the domain AppStore.com and it all is just a very sweet story really :)

The first paragraph implies that Apple takes down the website to promote new product launches - is that... correct?

Seems silly to take their website down for new products.

Yup, try http://store.apple.com during the next WWDC keynote: "we'll be right back!". And it doesn't come back until the keynote is over.

I assume that in ye olde days they were actually switching the website over with new products, or had some other reasonable technical explanation. These days I have to believe it's just for nostalgia or something.

Back when they used a third party CDN, I have to assume they were paranoid about uploading the details of new products to servers controlled by a third party until the keynote was already in progress.

They still use a third party CDN.

They rolled their own CDN a decade ago.

> Apple’s multi-terabit, $100M CDN is live

https://arstechnica.com/information-technology/2014/07/apple...

We're talking about the website in this thread, and the website still uses Akamai. An nslookup on their domain will show you as much.

It's showmanship. Drama, anticipation, and a reveal.

Also, if you’re not someone who follows Apple how would you like to buy something during the announcement of its replacement and find out an hour later?

I agree with another comment here that it’s more of a showmanship thing/making it more of an event these days.

I have to assume it's to avoid the headache of cancelled orders, and ensure strong sell-through of the brand new high margin product.

I’d never considered that scenario but I believe you’re spot on. I agree that it probably began as a true technical issue, but your scenario is something less obvious that would enrage a customer who just placed an expensive order.

It does take time for the changes to the site to propagate to all the servers in your Content Delivery Network.

https://en.wikipedia.org/wiki/Content_delivery_network

It’s a hype building exercise.

Somewhere between Macintosh & iPod + iTunes and MacOS (was: OSX (Unix, Bash,)) I think Apple was saved.

And the white cabling with the silhouettes

This is exactly how I remember and have experienced it, in that order.

MacOS became big because of the included Unix. Developers flocked. And al ballman once said... developers, developers, developers!

Much as it pains me to say, as a die hard Mac fan, Apple's success really largely rides on other products.

The iPod (which was not at all developer friendly) saved Apple commercially for several years. Yes, the Mac business had turned around a few years before, but e.g. the year before the iPod, Apple had the disastrous Power Mac G4 cube launch, which led to some unprofitable quarters. The iPod ensured that Apple's fortunes were no longer tied to new Macs (even though there were no G4 cube repeats).

And the big influx of developers came with iPhone. Not sure that Unix was that important a selling argument in that segment — it's certainly there behind the scenes, but does not play much of a role in developing.

Getting on Intel was also really big. That allowed a developer to have 1 computer where the primary OS ran Unix, but could also run commercial apps like Photoshop, while at the same time allowing them to dual boot or run a Windows VM. It was the jack of all trades.

iTunes on Windows was also key, as that let the iPod blow up with Windows users (using Rhapsody before that was awful). Those iPod sales really allowed them to build up their cash hoard.

After the iPod, this was what really helped. As someone on a college campus in 2006, the influx of people with MacBooks (not necessarily Pro’s but the regular one in black and white cases) was astonishing to watch, as they suddenly could dual boot Windows if they absolutely had to for their coursework. Combined with a much better version of iTunes as the iPod rose to prominence (the Video one came out the year before) and it became an easy sell to any 18 or 19 year old and their parents.

Yeah, I remember that explosion of MacBooks + iPods appearing everywhere in ~2006. It really stood out to me - suddenly almost everyone I knew was getting a MacBook (particularly people who had never owned anything from Apple before). The platform was really good then (IMO). :)

Sorry to tame down those rosy memories, but nothing like that was happening in Europe, or rest of the world for that matter. Apple was seen as vastly overpriced to PC (something students and their parents react very much to), without any added value if you ignore massive marketing (not going into other Apple products since they really stood above the crowd, but ecosystem wasn't existing much back then).

Basically a very expensive PC built together with custom OS, without any possibility to extend or modify it, very un-hacker if I may say so. If you used it as Windows machine it was much smarter to go to similarly-priced enterprise laptops, which offered things like NBD on-site warranty repairs (I had whole keyboard block replaced once on Dell, from manifestation to fix it took maybe 1 hour and I just took a long coffee break), with which Apple simply couldn't compete in business segments.

Yeah, literally not one of the people I'm thinking of that bought a MacBook in those days cared about extending/modifying or hacking it. Not even adding RAM. They were legitimately good machines with a good operating system, that worked really well for these people. I was "the designated tech guy" for basically everyone I knew in those days, and I got almost zero queries about these Macs in those days. BTW, this was in Vancouver, Canada, not in US. Either way in NA the Apple platform had a serious resurgence then. The MacBooks were actually seriously solid in comparison to equivalently-spec'ed Dell or HP laptops, and those manufacturers didn't even come close in terms of hardware quality (especially the trackpad). Combined with the skyrocketing popularity of the iPod, there's really no doubt it was a good time for Apple and its users.

Not sure why this was down voted. I for one appreciate a perspective outside the US bubble. Perhaps someone else from Europe can weigh in?

Right. Intel, and bootcamp which fully supported windows with drivers made it a very good choice for basically anyone.

I had one of the original iPods, with the physical scroll wheel. The software it used on Windows was “Musicmatch Jukebox”; I don’t remember Rhapsody. It was definitely better when iTunes came out for Windows.

You're right. I was misremembering, or thought Rhapsody owned Musicmatch.

They partnered very heavily with them, providing their music store and services

It also looked absolutely amazing. We came from blocky ugly 256 color icons, bland shades of grey, and lumpy fonts to aqua. Each icon looked "real", either as a real thing you could hold (hard drive, CD, printer, trash) or a glossy photo realistic artists rendition (the network icon and favorites heart in particular were gorgeous). The faint pinstripes and transparency matched apple's hardware at the time, and it just looked absolutely incredible. It's contemporary oses, windows xp mainly, improved the resolution of their interfaces, but still had garish cartoonish icons, and Linux DEs didn't really shake their lumpy pixelart icons till Nuvola and Crystal came about

The Developer story is key, I think, in addition to content CDNs and SAST/DAST for apps.

iPod Linux: https://en.wikipedia.org/wiki/IPodLinux

Rockbox Firmware for an Archos and then an Nano w/ color and no WiFi: https://en.wikipedia.org/wiki/Rockbox

  podman run --rm -it docker.io/busybox

  xcode-select -p
  # installing homebrew installs the xcode CLI tools:
  /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

  brew install podman 
  brew install --cask podman-desktop

https://mac.install.guide/commandlinetools/3.html

https://brew.sh/

How to add a Software repository to an OS. Software repository; SLSA, Sigstore, DevSecOps: https://en.wikipedia.org/wiki/Software_repository

W/ Ansible:

osx_defaults_module: https://docs.ansible.com/ansible/latest/collections/communit...

honebrew_tap_module: https://docs.ansible.com/ansible/latest/collections/communit...

honebrew_module: https://docs.ansible.com/ansible/latest/collections/communit...

homebrew_cask_module https://docs.ansible.com/ansible/latest/collections/communit...

From my upgrade_mac.sh: https://github.com/westurner/dotfiles/blob/develop/scripts/u... :

  upgrade_macos() {
    softwareupdate --list
    softwareupdate --download
    softwareupdate --install --all --restart
  }

From https://twitter.com/mitsuhiko/status/1720410479141487099 :

> GitHub Actions currently charges $0.16 *per minute* for the macOS M1 Runners. That comes out to $84,096 for 1 machine year

GitHub Runner is written in Go; it fetches tasks from GitHub Actions and posts the results back to the Pull Request that spawned the build.

nektos/act is how Gitea Actions builds GitHub Actions workflow YAML build definition documents. https://github.com/nektos/act

From https://twitter.com/MatthewCroughan/status/17200423527675700... :

> This is the macOS Ventura installer running in 30 VMs, in 30 #nix derivations at once. It gets the installer from Apple, automates the installation using Tesseract OCR and TCL Expect scripts. This is to test the repeatability. A single function call `makeDarwinImage`.

With a Multi-Stage Dockerfile/Containerfild, you can have a dev environment like xcode or gcc+make in the first stage that builds the package, and then the second stage the package is installed and tested, and then the package is signed and published to a package repo / app store / OCI container image repository.

Continuous integration: https://en.wikipedia.org/wiki/Continuous_integration

Is there a good way to do automated testing like pytest+Hypothesis+tox w/ e.g. the Swift programming language for computers? CloudFuzz is built upon OSS-Fuzz.

SLSA now specifies builders for signing things correctly in CI builds with keys in RAM on the build workers.

"Build your own SLSA 3+ provenance builder on GitHub Actions" https://slsa.dev/blog/2023/08/bring-your-own-builder-github

Reply on news.ycombinator.com