Hacker News new | past | comments | ask | show | jobs | submit login
How the BBC News website has changed over the past 20 years (bbc.co.uk)
89 points by dberhane on Nov 18, 2017 | hide | past | favorite | 30 comments

How wonderful to see the BBC General Election site from 2010. Me and one other guy built the whole output because the World Cup was on and all the main staff were working on that. The other engineer even had to make the results service work on Ceefax [1]!

We had over a billion hits in 24 hours to the statically published JSON file that updated the results every 15 seconds and we could even control the client poll rate from that file just in case.

Over a weekend we built the first version of the BBCs live page amongst other things. It was terrifying watching constituencies declare for the first time; about as close to a live performance of web development as you can get.

[1] https://en.wikipedia.org/wiki/Ceefax

[2] http://news.bbc.co.uk/1/shared/election2010/results/ (slightly broken now...)

One thing I've always appreciated about the BBC website is that the old articles are still in the old style of the time. For example: http://news.bbc.co.uk/1/hi/business/3559050.stm or http://news.bbc.co.uk/1/hi/business/574132.stm (there might be a way to reliably find more examples, I'm not sure).

I'm not sure whether this is by design or if the content was too tightly coupled to the style to migrate, but I like it. I think in the relevant RFCs it says that a URL should always point to the same thing, so it's good that they don't try to update them.

I wouldn't say it's by design - more of a fun consequence of the way the BBC's web platforms have been retired so far. Most of the retired platforms generated static content (some XML, some HTML) so when the platform needed to be switched off it was relatively trivial to archive all of the content. In about 2013 the website was transferred piece by piece to a dynamic platform, which it (mostly) still runs on.

The current platform is already slated for decommission some time in the next couple of years and there's still no cohesive plan for how pages on that platform will be archived. My guess is that they will be rendered on the new platform(s) indefinitely, since everything is driven by a CMS API now. I think it's kind of a shame - there's something nice about being able to look back in time without the pages being distorted through the lens of the wayback machine.

I really hope they will archive statically rendered pages. It matters that you are able to see old websites in their original form.

EDIT: does anyone know whom we might contact at the BBC to encourage them to do this when they make the transition?

I'm happy as long as there is an archive. The way front end dev is going these days it seems the best hope is a json api stays alive.

I think those two pages are the fastest-loading web pages I've viewed in the last 6 months. The second is just 109kB with all assets.

That is what the web was like if you had a fast connection back in the 90s. Those were the days...

I was a web producer (front end developer) and then web developer (CMS template developer) for BBC News online 2000-2004. It was my first job in tech at 18 and arguably still my fondest time in the industry. I led some of the redesign from single column to the more modern design we see today.

I'm now startup founder and a VC - v far from the non-profit public service nature of the BBC.

Happy to AMA if anyone had any questions.

An interesting part about BBC News's tech stack, which was touched on in another comment here, is that we didn't have the budget to serve dynamic pages and so everything SPG'd (statically page generated) as we called it so it could be served statically from the production web servers.

It meant an incredibly novel use of Apple WebObjects - which was set up to render all of the versions of the story (flavours, I think we called them) via the WOA and then FTP them to production. We used this system to produce not only web versions but content for interactive TV output and I believe even Ceefax at one point.

Another curious fact now: most of the BBC News website was coded in Objective C in the early 2000's, to much chagrin as it meant us developers having to learn this 'dead' language when really we wanted to use the Java version of WebObjects instead. Many of the devs went on to have leading roles in the UK iPhone development industry when apps were enabled for iPhone given the amount of experienced we all unknowingly had received in what suddenly became a much-sought after language.

I wish I had a specific question to ask, but honestly I'd love it if you just told more stories about this period in your life!

How did you make your VCing money?

Co-founding a successful startup (WP Engine) and before that some angel investment funded by consultancy earnings in the late 2000's. I also joined Uber just around the 1000 employee mark, and I remain bullish that equity will increase further in value.

Hardwired widths, inaccesibility on small screens etc. were always annoying, but in many respects, I'm retrograde enough to actually prefer the clunkier designs and layouts of the paleolithic web-era. No-nonsense, unambiguous, clearly delineated sections, and a healthy emphasis on content over form.

Correct me if I'm wrong, but I always understood that if you write basic, un-styled HTML, you get flexible widths and small screen support for free. You have to deliberately decide to make it worse by hardwiring widths and element sizes. Is this not true?

Oh, absolutely. And all kinds of horror were perpetrated against html in the nineties.

My point was that even so, I found - and find - lots to like in the websites of yore.

You are correct. On the other hand, readability would be reduced on larger screens since the lines would be longer. (This could have been address by both users and web browser developers, but website designers took a different approach.)

Wayback Machine allows browsing website history by year. Press year, instead of arrows:

— The New York Times: http://web.archive.org/web/19961112181513/http://www.nytimes...

— WaPo: http://web.archive.org/web/19961220172326/http://www.washing...

— Financial Times: http://web.archive.org/web/19970607125328/http://www.ft.com:...

Interesting, although I wish they'd shown the whole of the historical pages. I'd like to work out when the front page dropped below 50% actual news stories (as opposed to 'The X that Y', 'Must See' and other buzzfeed type stuff). It's clear that in 2003 it was still nearly 100% news, but the later ones are cropped so you can't tell.

I am surprised how professional the site design looks even in 1997, also with rich media link like audio and video.

I interviewed to work as a web dev for the bbc back in tye very early 2000’s. In the interview I said I was better at going away and thinking about things rather than making decisions off the top of my head. I think that’s why I didn’t get the job.

As far as I know, it’s the only tech interview I didn’t get offered the job.

Tough crowd!

And still no https

Huh? It works for me, with a 301 redirect. Truncated output:

    $ curl -Lv http://bbc.co.uk
    > GET / HTTP/1.1
    > Host: bbc.co.uk
    > User-Agent: curl/7.56.1
    > Accept: */*
    < HTTP/1.1 301 Moved Permanently
    < Location: http://www.bbc.co.uk/
    > GET / HTTP/1.1
    > Host: www.bbc.co.uk
    > User-Agent: curl/7.56.1
    > Accept: */*
    < HTTP/1.1 301 Moved Permanently
    < Location: https://www.bbc.co.uk/
    > GET / HTTP/2
    > Host: www.bbc.co.uk
    > User-Agent: curl/7.56.1
    > Accept: */*
    < HTTP/2 200 
    < content-type: text/html; charset=utf-8
    < content-length: 270866

With a site like BBC News integrating with lots of different internal services, it can be more difficult than it seems on the face it to enable HTTPs, but I'm sure it's coming soon, other BBC sites have enabled HTTPs.

There's actually not much to stop the rollout of HTTPS on the BBC News site now. In fact, the major hurdles were overcome nearly a year ago. The final hurdle is convincing the product and editorial teams that enabling HTTPS is more important than whatever features they want to build.

Ah - fair enough. I made some assumptions based on some of the hurdles seen by some of the more recent moves to HTTPs on smaller BBC sites, not of any real knowledge of the current situation.

(FWIW, your article from Dec 2016 played a part in me joining the BBC recently as a developer, and I know of at least one other developer currently at the BBC who also was motivated to apply after your article :))

There are hurdles for sure. I guess I'm just publicly venting my frustration at having HTTPS on one of the world's biggest news websites be under-prioritised for years and years :)

I'm really glad to hear that btw. I wish I could have stuck around for longer!

I keep reading that - but I just don't buy it - you can offload SSL to edge devices - why do the servers even need to care?

TL;DR Chrome lit a fire on their collective arses. They’d never have switched otherwise.

> Earlier in 2016, the Chromium development team decided to implement a change to Google Chrome, preventing access to certain in-browser features on ‘insecure’ (non-HTTPS) web pages. In practice, this meant that key features of certain products, such as the location-finding feature within the Homepage, Travel News and Weather sites, would stop working if we didn’t enable HTTPS for those services.

Thanks Obama, I mean Google.

You can read BBC in countries that blocks it via an iframe proxy provider. Maybe that's why there is no https. But they can make http not redirect to https and still have https

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact