WebObjects Overview (2001) [pdf] | Hacker News Reader

WebObjects Overview (2001) [pdf]

developer.apple.com

I am not sure why this stuff is posted now, but as someone who still need to use WebObjects at work because of a legacy product, we use instead Project Wonder:

https://github.com/wocommunity/wonder

https://wiki.wocommunity.org/

Because Wonder is more or less patched & maintained while WebObjects is in abandonware state since 2008 (and officially since 2016).

What legacy product do you support that depends on WebObjects...?

It's a niche ERP. Sorry i will not say more to keep my privacy.

The editor is in the process to rewrite it with Spring & AngularJS frameworks, but it will still take years to do so.

My second job was working for a small WebObjects consultancy in 2008. They developed bespoke line-of-business applications for small companies. Having the complete stack - from the HTML templating through to the ORM - provided by one vendor and designed to work together made development faster than anything I've worked with since.

It's a shame the cost was too high for too long, and even when Apple made it free, they never open sourced it.

Steve loved WebObjects. Trawl around YouTube and you can find some of his enthusiastic demos from the late NeXT era. In one, IIRC, he builds an entire web store from scratch in about 15 minutes.

But his message on returning to Apple was all about focus. Apple was not going to try to be a web middleware company, so they used the tech where they could benefit from it internally (and did so for a loooong time) and that was that.

Late ‘90s Steve didn’t only put bullets in the heads of technologies (Newton, OpenDoc) that he didn’t like.

Powerbuilder was (I guess it still exists) sort of that way. Not as sophisticated as WebObjects but for throwing together CRUD screens for business functions it was pretty good. One way to do things, you didn't really have to stop and think "how am I going to build this." Sort of like framing a house with dimensional lumber.

> They developed bespoke line-of-business applications for small companies. Having the complete stack - from the HTML templating through to the ORM - provided by one vendor and designed to work together made development faster than anything I've worked with since.

How is that different then .Net?

It was the 90s.

(Not OPs story, but WebObjects availability)

I worked with .NET as well around the same time, and in my opinion it never quite achieved the same level of integration as WebObjects.

It also had quite a high price-of-entry until Visual Studio Express editions were released in 2005.

I don't have any experience with modern .NET/MVC specifically but modern approaches in general shy away from the decisions that made WebObjects so good because they do tend to hit nasty scalability limits.

It predates .NET for about 10 years.

WebObjects and Distributed Objects Everywhere, are two Objective-C projects that eventually influenced how JEE came to be, yet another connection between Java and its Objective-C influences.

Ah, WebObjects, one of the sad "dead ends" of web tech, not because of being technically obsolete, but because the world moved to different stacks…

I'd put ColdFusion, AolServer and Seaside in the same category.

WebObjects ran the website of the nhl for a long time

It’s a tad ironic to see todays modern tad rediscovering the use of presentation templates much like WebObjects, ColdFusion and others quietly did for a long time and were ostracized for.

I did WO/EO for several years until 2001 including the USPS Postmaster app, and it was incredibly productive. This was before JS took over everything, but I can't ever recall being able to do so many client/server web apps with so little effort. Sure the tooling was sometimes lame (ProjectBuilder was a precursor to Xcode) but Objective-C at the time was amazing compared to all other web programming languages. I think WO was far easier to develop in than today's complex web stack, but clearly serving the simpler need of that time. It wouldn't be useful today.

ColdFusion, may it remain buried for good. I was forced to use it in the 90's and was struck by how much more work it was to get something sensible up and running with it compared to using Perl. I finally got them to ditch it and life was good - or at least better.

Its spirit lives on Jakarta EE and Spring.

Some say it is still better than today's Ruby Rails / Laravel in terms of productivity. And if it was open sourced at the time the world would have been very different.

All we know is, it's called WebObject.

Fun fact: WebObjects was the most expensive product Apple has ever sold, at $50000. It managed to keep NeXT alive after the hardware business collapsed

Adjusted for inflation perhaps. But the last Intel Mac Pro definitely went above that.

At $53,799, not much, according to https://www.theverge.com/circuitbreaker/2019/12/10/21003636/...

My mind was blown when I've seen WebObjects in 2001. We were using XML/XSLT at that time to render web pages and plain servlets.

I worked at a university and we used WedObjects for a lot of project. I still believe that it could be around today and competing with the best "stacks" available. Limiting factors would be price and deployment options.

But alas, it wasn't viewed in the light by everyone and other "stacks" became more popular.

I wasn't even a programmer back in 2001, but flipping through this book shows how certain concepts endure, even as technology evolves. It's a bit like the snake eating its tail - even as we move on, we find ourselves revisiting tried approaches.

I was working on BBC News Online shortly before it launched in 1997. That was WebObjects-based, and it seemed really powerful and really nicely designed.

Page 39 has a simple code example. I wasn't expecting Java.

Reply on news.ycombinator.com