5 ms·
How do you know what the internet will look like in 500 years? Sites from 20 years ago are broken, you can expect that unless you’re using plain text and html t
by tekknik 5y ago
How do you know what the internet will look like in 500 years? Sites from 20 years ago are broken, you can expect that unless you’re using plain text and html that standards will change in 500 years and people will not be able to access your site.
Then there’s issues with domains. You’d have to setup a trust and again assume we will still be using domains in 500 years. If you use something like S3 then you’ll have to ensure they’re around for 500 years.
My perspective, this is entirely unrealistic.
- mywittyname 5y agoEven plain text is no guarantee. This assumes that future people will read/understand/use latin characters. ASCII could very well be replaced in the medium-term future
- theandrewbailey 5y agoI thought Unicode was already replacing ASCII.
- mywittyname 5y agoSure, but that is backwards compatible / a superset of ASCII. What I meant to say, is that it's totally possible that, in the coming few centuries, even basic ASCII won't be readily understood. As in the character mapping in modern systems will break down, i.e., int 97 is no longer 'a', but some glyph from a language not yet conceived. We take for granted backwards comparability. Just because ASCII has been readable for the past 60 or so years, doesn't mean it will continue to be for the next 60.
- necovek 5y agoWe do not take it for granted. Instead, we design systems to be such because it makes many things much simpler. This is the reason why UTF-8 has basically "won" over UTF-16 or UCS-4 when it comes to encoding Unicode characters. If anything, with the amount of data we have today, unless there is a big reason (probably political, but even they exist today for eg. China not to want to use an Unicode transformation based on American Standard Code for Information Interchange) to re-encode all historical data, backwards compatibility will be maintained with the computers of the future (if they still exist). Yes, even if we move their bytes to be 13-qubit qubytes :D To elaborate on the cost: re-encoding all data from 2050 is probably not going to be too expensive in 2400, but by then you'll need to re-encode data from up to 2400. To me this seems like a reason that backwards compatibility will make sense to be kept because there is not much to be gained. Eg. UTF-8 approach has shown us the best way forward. The trickiest is going to be to keep all video/audio encoding algorithms, especially as they are patent encumbered.
- arp242 5y agoUnicode is compatible with ASCII: any ASCII text will render fine (when UTF-8 is used anyway; some other Unicode encodings: not so much).
- trinovantes 5y agoThere's also the fact that human language will also change over the course of 500 years 500 year old English is nothing like English today
- blfr 5y agoLanguages change slower once they're written down. Today's English should be readily comprehensible.
- FeteCommuniste 5y agoFive hundred years ago is Shakespeare, roughly. Still readily understandable for the most part. Now if we go back a thousand years…
- kevinmchugh 5y agoShakespeare and the KJV are both closer to 400 years old. Shakespeare wanted to be accessible, though modern readers still need annotations to understand some of the words and most of the references. The KJV was intentionally somewhat archaic. There's a big difference in how accessible KJV is vs Shakespeare But otherwise yeah, the Norman Conquest did a number on the language
- paxys 5y agoShakespeare wrote in what is referred to as "Early Modern English". It is no secret that his writings were a big influence in the evolution of the language itself. If you look at some of his contemporaries from the same period, however, their language will be very different and much harder to understand.
- dhosek 5y agoMy specialty in undergrad was 15th and 16th century poetry. His contemporaries' writing (and writings from the early 15th c.) was no harder to read than Shakespeare.¹ The biggest challenge would be the irregularities of spelling—before the printing press and for at least a century afterwards, English spelling was inconsistent and flexible (and frequently was left up to the compositor for printed materials which led to different spellings for a word in the same document to make line breaks work better). ⸻⸻⸻ 1. A notable exception would be Edmund Spenser who wrote in a style that was archaic even to his contemporaries.
- mytailorisrich 5y agoIndeed. As I commented separately I think the only practical solution is to use a simple medium (USB stick) and simple, standard file formats, then to copy from medium to medium, and to convert obsolete formats as time passes. This requires descendants to keep at it over time, and it not really a "web site", but IMHO is the only way to keep the date accessible and useable over time.
- teawrecks 5y agoBroken but not illegible. I feel like you're arguing that because the world may forget how to read some data 500 from now, storing the data is worthless. But that's just not true. Archeologists find meaning in writings they've never seen before all the time.
- forgotmypw17 5y agoSites which were carefully designed are not broken, and I think that is a good starting point when designing with longevity in mind. Five hundred years is a long time, but I think it's reasonable to try to design a site that could last, for example, 25 years, because you can already write something which COULD HAVE worked for the PREVIOUS 25 years by testing with older browsers. You'll want to restrict yourself to a subset of HTML which is supported by all of them, perhaps with some progressive enhancement.