I posted this as I think it’s fun to look back at how some “early” web stuff was done and a bit of a trip down memory lane on a Saturday morning.
When I was experimenting making websites as a teenager 20 (ouch!) years ago I would love to investigate how other sites worked. The BBC had this little “Text Only” link on all pages of the site that redirected to a version stripped of all images and tables (tables were used for layout, much like HN) with high contrast and large text for the visually impaired and screen readers. Brilliantly ahead of its time thinking about accessibility, which I suppose is the BBCs mandate.
The url to these pages were prefixed with something like /cgi-bin/betsie.pl?url= and the bottom of the pages liked to this page I have submitted (sadly now only in the Internet Archive), it’s affectively a proxy written in Perl that transforms each page on the fly. I remember setting Betsie up on my own site, which no one visited because who would visit the site of a 14yo.
I believe Betsy was still in use till about 2010. You can read more about it here:
Within 2 weeks we managed to get our first band signed. A Swedish band called Maas. I worked with a producer Andrew Barron, who takes a lot of credit for seeing the project past the finish line. Andrew sat with me over a weekend and deep into a Sunday night to get it finally launched.
The project was written in mod_perl and was designed to be extrememly high performance because Radio 1 at the time (and perhaps still is) the most popular radio station in the UK. And we expected a surge of traffic every time the DJ mentioned OneMusic, which is exactly what we got.
I was based in Oxford Circus and I worked with the ops team which - at the time, and I'm not sure if this is still the case - was based out in some mansion in the country somewhere. They had a kind of Bletchley Park vibe about them - this team of absolute gurus who you only got to speak with over the phone, and who would somehow make your application handle the massive amounts of traffic that would be thrown at it. They did a great job hosting that application on Apache/mod_perl in the dark days before Nginx was invented.
One of the absolute worst design decisions I made was to design the application so that it would support sessions for users without cookies enabled. Which is why that Archive.org link has the OMSID query string parameter in it. The app would dynamically modify all URLs on the page to embed that assigned session ID so it stuck, even if you didn't have cookies enabled. It would follow you around the site as a query string param. I stupidly suggested the idea to Andrew and he was like "Hell yes, lets do that!" and so I did. But it had big performance implications as the URLs had to be dynamically generated for every pageview, and it added significantly to the time it to for me to create the project. But hey, we did it, and got it launched, and it actually worked and provided sessions for users who had cookies disabled.
The project taught me a lot about high performance websites, delivering a critical project on-time to a big audience, and I went on to create WorkZoo (Time Mag top 50 website of 2005) which I sold, Feedjit (live traffic feed on over a million sites) and my current cybersecurity business Wordfence which employs 40 and is the best available security product for WordPress. I think creating all those projects set me up to create a WordPress plugin that is high performance, widely compatible, very stable, and delivers on user satisfaction.
When I left the BBC, the team bought me a bottle of Gran Centenario Reposado tequila out of their own money. I'll never forget that and I'll never forget them. Such a great group of people.
If you include the session ID in the url it can leak to other uses if they send a urls to someone, it also exposes the session id in intermediate proxy/router log files (if using plain http, not tls). So it’s basically a security risk.
Before cookies had widespread support it was basically the only way to implement sessions.
Disclaimers like this used to be a little more common on the web, as I recall.
But the BBC is a national broadcaster and news organisation. I think you have to keep in mind the volume, nature and often bad-faith unfairness of the kinds of complaints it receives. It's not difficult to imagine the "I was on the BBC website and two clicks later I was looking at bestiality material" complaints, tabloid news stories, extreme right wing/left wing/nationalist politicians etc.
A disclaimer like that shoots down that entire range of complaints and was probably constructed with Ofcom's agreement for that reason. (They are the relevant regulator.)
If you google "BBC disclaimer" you'll see there's a bunch of them they have had to construct over the years to fend off bad faith complaints from all sorts of angles.
(And one or two good-faith complaints that stem from their peculiar position, like the "Other listings magazines are available" disclaimer they have/had to make when they mention the commercial TV/Radio listings magazine, The Radio Times, that is produced under licence. That disclaimer has entered into popular culture and is the basis of a running joke among BBC presenters at times when a contributor inadvertently mentions or places a product. "Of course, other <insert product category> are available.")
NPR is in a similar situation, I imagine -- having to put disclaimers into their content to protect themselves from the kind of bad faith criticism that will get brought up the next time their funding is being discussed.
I'm sure non-British HNers have nothing but a positive impression of the BBC, are aware that it showcases Britain in a positive light, and therefore find it beyond baffling that it is the target of so much vitriol at home from, ahem, certain quarters.
Even GNU recommends liberal licenses for snippets with less than 300 lines (and allows for commercial use anyway!):
https://www.gnu.org/licenses/gpl-faq.html
"
What if the work is not very long? (#WhatIfWorkIsShort)
If a whole software package contains very little code—less than 300 lines is the benchmark we use—you may as well use a lax permissive license for it, rather than a copyleft license like the GNU GPL. (Unless, that is, the code is specially important.) We recommend the Apache License 2.0 for such cases.
"
While LOCs poorly reflect the effort made, in this case a less anal-retentive approach would have opened up a long-term contribution to societies at large. Considering that the BBC was mainly finances by quasi-public money it's a shame that it was opted for a license with an unnecessary special case (NC).
> Considering that the BBC was mainly finances by quasi-public money it's a shame that it was opted for a license with an unnecessary special case (NC).
Unfortunatly this is why it has to be licensed in that way. Having the BBC release code for free to commercial use is often deemed to be a "market distortion", and thus not allowed to happen, as such the lawyers will often err on the side of caution.
Also remember this was 1999, 5 years before Apache License 2.0 was approved.
> Having the BBC release code for free to commercial use is often deemed to be a "market distortion", and thus not allowed to happen, as such the lawyers will often err on the side of caution.
People outside the UK: this isn't just theoretical, please look at the history of BBC Micro. That's how the corporation was structured at the time, and it should be read with this in mind. There are a lot of very fine pieces of BBC software that is stricter than that and can't be legally released to the public, for example most broadcasting software like its MPEG2 encoder aren't available because it'll disadvantage broadcast equipment manufacturers (although for the MPEG2 encoder FFMPEG's is now better). This is also not specific to the BBC: before its disbandment, the Independent Broadcasting Authority (which owned half of the transmission network in the UK, the other half was previously BBC but both were privatized and now owned by Arqiva) required to do product development and research by law but can't commercialise those research by themselves.
Very true. I've mentioned in another comment the "Other <product category> are available" disclaimer, that used to be tacked onto any content mentioning The Radio Times or any of the BBC Gardening/Good Food/Sport products from BBC Enterprises or licensees.
It's now a running joke, but the disclaimer about The Radio Times really did emerge out of a market distortion legal complaint by a competitor, TV Times.
These market distortion complaints were often timed around BBC Charter renewals to give the BBC's critics in Parliament (generally on the Tory side) ammunition. But at the same time, it's also the era of the Radio One Roadshow, the BBC Micro as you say, the emergence of the VHS box set, etc., etc., and the BBC did need to be required not to distort markets accidentally as it expanded commercially (which ironically it felt encouraged to do in order to find future non-licence-fee revenue)
It's a little bit complicated but before 1991-1994 (where there are large-scale changes outside of BBC), Radio Times were exclusively for the BBC while TV Times were exclusively for the ITV companies (yes, ITV wasn't just a single company those days!), Channel 4, and S4C (technically there is a third one, Channel Times, but that's for the Channel Islands which has a very interesting ITV setup at that time).
Yeah -- I remember more than a full decade of two listings magazines per week!
To break it down a little for non-British HNers (it is very confusing and messy):
The BBC does not run adverts, but it does have a carve-out from that to mention its own associated products in limited circumstances. Most significantly the listings magazine, The Radio Times.
We start out in the situation zinekeller mentions where the BBC listed BBC TV/radio channels in The Radio Times, and TV Times listed programmes from the ITA (Independent Television Association - ITV/C4/S4C/London Weekend Television etc.)
But then you get a situation where TV viewers are buying two increasingly competitive magazines each week just to get the full listings. As satellite and cable TV appears, this becomes unfair on the consumer (to the benefit of the BBC) and it has to change.
The disclaimer came in part from those publications seeking and being granted access to the listings data that the others were publishing. TV Times got access to BBC listing data, and Radio Times published the ITA channel listings.
But the BBC was and still is able to mention the Radio Times, whereas TV Times cannot advertise on the BBC. A market distortion.
I suppose there must have been calls to ask the BBC to get out of the listings magazine market (I don't remember anything specific) but ultimately the disclaimer ("Other listings magazines are available") was imposed on the BBC.
I should correct myself above because the group wasn't the ITA, it was the ITCA -- the Independent Television Companies Association. Which is now UKIB:
Yeah, sadly. But good you say it explicitly and it a fair spin(?) -- I just wanted to point out the IMO damage that license does, not blame the authors/BBC.
In Germany it's the same with the "Mediathek" nonsense. Shows paid with public money can only shown for a ridiculously short time -- no freely accessible archive.
Wow this brings back memories. While it was tedious it's something which has humbled me over the years.
Due to various laws we had to guarantee accessibility of most if not all 'official communication' to native speakers of several languages and various (dis)abilities. With some minor structural tweaks and guidance from an 'accessibility coordinator' who worked at a college we were able to realistically make it very accessible to the 1% (or lower) of the population who rely on it for their own independence, but the changes and influence permeated from the ground up and IMO made everything better for everybody.
Remember when websites let you choose the font size? Actually sitting down with people who have difficulty seeing was insightful, most of them found the font size adjuster widgets were useless, poorly located, too small or had low contrast (so as to not annoy the sighted visitors, the irony) - instead preferring to switch over to the text-only version (if available) and using keyboard shortcuts or browser settings to adjust font size.
Likewise with screen readers, we spent a lot of time annotating images and structuring links, content & context so that it made more sense and was generally predictable. We had to use them ourselves in the same way that you'd check a design across various tablet / mobile devices and browsers.
Unfortunately this kind of effort isn't made very often outside of government and charity websites these days, but accessibility software has also gotten much better over time (I was absolutely blown away when a blind guy showed me how he used is iPhone, it's much more efficient than normal visual use for core functionality).
The key thing is with ear buds in, which works even if you're fully or partially sighted you can keep it in your lap and quickly respond to messages or lookup facts without it looking like you're breaking attention, e.g. in a meeting or even during a full blown conversation much in the same way you'd read an article and listen to TV at the same time.
33 comments
[ 2.9 ms ] story [ 85.3 ms ] threadWhen I was experimenting making websites as a teenager 20 (ouch!) years ago I would love to investigate how other sites worked. The BBC had this little “Text Only” link on all pages of the site that redirected to a version stripped of all images and tables (tables were used for layout, much like HN) with high contrast and large text for the visually impaired and screen readers. Brilliantly ahead of its time thinking about accessibility, which I suppose is the BBCs mandate.
The url to these pages were prefixed with something like /cgi-bin/betsie.pl?url= and the bottom of the pages liked to this page I have submitted (sadly now only in the Internet Archive), it’s affectively a proxy written in Perl that transforms each page on the fly. I remember setting Betsie up on my own site, which no one visited because who would visit the site of a 14yo.
I believe Betsy was still in use till about 2010. You can read more about it here:
https://www.bbc.co.uk/blogs/bbcinternet/2012/02/accessibilit...
Source code:
https://web.archive.org/web/20000824120750/http://www.bbc.co...
It was quite useful when you were stuck with baroque HTML generation code you were unable to modify.
The BBC produce a bunch of interesting tools for web development, another being the layout comparison tool, Wraith:
https://bbc.github.io/wraith/
https://web.archive.org/web/20050416110703/http://www.bbc.co...
Within 2 weeks we managed to get our first band signed. A Swedish band called Maas. I worked with a producer Andrew Barron, who takes a lot of credit for seeing the project past the finish line. Andrew sat with me over a weekend and deep into a Sunday night to get it finally launched.
The project was written in mod_perl and was designed to be extrememly high performance because Radio 1 at the time (and perhaps still is) the most popular radio station in the UK. And we expected a surge of traffic every time the DJ mentioned OneMusic, which is exactly what we got.
I was based in Oxford Circus and I worked with the ops team which - at the time, and I'm not sure if this is still the case - was based out in some mansion in the country somewhere. They had a kind of Bletchley Park vibe about them - this team of absolute gurus who you only got to speak with over the phone, and who would somehow make your application handle the massive amounts of traffic that would be thrown at it. They did a great job hosting that application on Apache/mod_perl in the dark days before Nginx was invented.
One of the absolute worst design decisions I made was to design the application so that it would support sessions for users without cookies enabled. Which is why that Archive.org link has the OMSID query string parameter in it. The app would dynamically modify all URLs on the page to embed that assigned session ID so it stuck, even if you didn't have cookies enabled. It would follow you around the site as a query string param. I stupidly suggested the idea to Andrew and he was like "Hell yes, lets do that!" and so I did. But it had big performance implications as the URLs had to be dynamically generated for every pageview, and it added significantly to the time it to for me to create the project. But hey, we did it, and got it launched, and it actually worked and provided sessions for users who had cookies disabled.
The project taught me a lot about high performance websites, delivering a critical project on-time to a big audience, and I went on to create WorkZoo (Time Mag top 50 website of 2005) which I sold, Feedjit (live traffic feed on over a million sites) and my current cybersecurity business Wordfence which employs 40 and is the best available security product for WordPress. I think creating all those projects set me up to create a WordPress plugin that is high performance, widely compatible, very stable, and delivers on user satisfaction.
When I left the BBC, the team bought me a bottle of Gran Centenario Reposado tequila out of their own money. I'll never forget that and I'll never forget them. Such a great group of people.
https://www.bbc.com/historyofthebbc/buildings/kingswood-warr...
I think we all did this at least once back then and all regret it for the same reasons...
Before cookies had widespread support it was basically the only way to implement sessions.
Wow, they still use this exact disclaimer 23 years later, to the letter.
(I wonder why they need it...)
Disclaimers like this used to be a little more common on the web, as I recall.
But the BBC is a national broadcaster and news organisation. I think you have to keep in mind the volume, nature and often bad-faith unfairness of the kinds of complaints it receives. It's not difficult to imagine the "I was on the BBC website and two clicks later I was looking at bestiality material" complaints, tabloid news stories, extreme right wing/left wing/nationalist politicians etc.
A disclaimer like that shoots down that entire range of complaints and was probably constructed with Ofcom's agreement for that reason. (They are the relevant regulator.)
If you google "BBC disclaimer" you'll see there's a bunch of them they have had to construct over the years to fend off bad faith complaints from all sorts of angles.
(And one or two good-faith complaints that stem from their peculiar position, like the "Other listings magazines are available" disclaimer they have/had to make when they mention the commercial TV/Radio listings magazine, The Radio Times, that is produced under licence. That disclaimer has entered into popular culture and is the basis of a running joke among BBC presenters at times when a contributor inadvertently mentions or places a product. "Of course, other <insert product category> are available.")
NPR is in a similar situation, I imagine -- having to put disclaimers into their content to protect themselves from the kind of bad faith criticism that will get brought up the next time their funding is being discussed.
Even GNU recommends liberal licenses for snippets with less than 300 lines (and allows for commercial use anyway!): https://www.gnu.org/licenses/gpl-faq.html " What if the work is not very long? (#WhatIfWorkIsShort)
"While LOCs poorly reflect the effort made, in this case a less anal-retentive approach would have opened up a long-term contribution to societies at large. Considering that the BBC was mainly finances by quasi-public money it's a shame that it was opted for a license with an unnecessary special case (NC).
Unfortunatly this is why it has to be licensed in that way. Having the BBC release code for free to commercial use is often deemed to be a "market distortion", and thus not allowed to happen, as such the lawyers will often err on the side of caution.
Also remember this was 1999, 5 years before Apache License 2.0 was approved.
People outside the UK: this isn't just theoretical, please look at the history of BBC Micro. That's how the corporation was structured at the time, and it should be read with this in mind. There are a lot of very fine pieces of BBC software that is stricter than that and can't be legally released to the public, for example most broadcasting software like its MPEG2 encoder aren't available because it'll disadvantage broadcast equipment manufacturers (although for the MPEG2 encoder FFMPEG's is now better). This is also not specific to the BBC: before its disbandment, the Independent Broadcasting Authority (which owned half of the transmission network in the UK, the other half was previously BBC but both were privatized and now owned by Arqiva) required to do product development and research by law but can't commercialise those research by themselves.
It's now a running joke, but the disclaimer about The Radio Times really did emerge out of a market distortion legal complaint by a competitor, TV Times.
These market distortion complaints were often timed around BBC Charter renewals to give the BBC's critics in Parliament (generally on the Tory side) ammunition. But at the same time, it's also the era of the Radio One Roadshow, the BBC Micro as you say, the emergence of the VHS box set, etc., etc., and the BBC did need to be required not to distort markets accidentally as it expanded commercially (which ironically it felt encouraged to do in order to find future non-licence-fee revenue)
To break it down a little for non-British HNers (it is very confusing and messy):
The BBC does not run adverts, but it does have a carve-out from that to mention its own associated products in limited circumstances. Most significantly the listings magazine, The Radio Times.
We start out in the situation zinekeller mentions where the BBC listed BBC TV/radio channels in The Radio Times, and TV Times listed programmes from the ITA (Independent Television Association - ITV/C4/S4C/London Weekend Television etc.)
But then you get a situation where TV viewers are buying two increasingly competitive magazines each week just to get the full listings. As satellite and cable TV appears, this becomes unfair on the consumer (to the benefit of the BBC) and it has to change.
The disclaimer came in part from those publications seeking and being granted access to the listings data that the others were publishing. TV Times got access to BBC listing data, and Radio Times published the ITA channel listings.
But the BBC was and still is able to mention the Radio Times, whereas TV Times cannot advertise on the BBC. A market distortion.
I suppose there must have been calls to ask the BBC to get out of the listings magazine market (I don't remember anything specific) but ultimately the disclaimer ("Other listings magazines are available") was imposed on the BBC.
https://en.wikipedia.org/wiki/United_Kingdom_Independent_Bro...
And the tape over product logos and labels to hide them.
In Germany it's the same with the "Mediathek" nonsense. Shows paid with public money can only shown for a ridiculously short time -- no freely accessible archive.
You could make this a bit simpler nowadays with some libs like Mojo and LWP.
It was a joke, and one based on nearly 27 years of writing Perl at that.
Due to various laws we had to guarantee accessibility of most if not all 'official communication' to native speakers of several languages and various (dis)abilities. With some minor structural tweaks and guidance from an 'accessibility coordinator' who worked at a college we were able to realistically make it very accessible to the 1% (or lower) of the population who rely on it for their own independence, but the changes and influence permeated from the ground up and IMO made everything better for everybody.
Remember when websites let you choose the font size? Actually sitting down with people who have difficulty seeing was insightful, most of them found the font size adjuster widgets were useless, poorly located, too small or had low contrast (so as to not annoy the sighted visitors, the irony) - instead preferring to switch over to the text-only version (if available) and using keyboard shortcuts or browser settings to adjust font size.
Likewise with screen readers, we spent a lot of time annotating images and structuring links, content & context so that it made more sense and was generally predictable. We had to use them ourselves in the same way that you'd check a design across various tablet / mobile devices and browsers.
Unfortunately this kind of effort isn't made very often outside of government and charity websites these days, but accessibility software has also gotten much better over time (I was absolutely blown away when a blind guy showed me how he used is iPhone, it's much more efficient than normal visual use for core functionality).
The key thing is with ear buds in, which works even if you're fully or partially sighted you can keep it in your lap and quickly respond to messages or lookup facts without it looking like you're breaking attention, e.g. in a meeting or even during a full blown conversation much in the same way you'd read an article and listen to TV at the same time.
code recycling detected