5 ms·

One of the most terrifying parts of 1984 (the Orwell novel, not the year) for me was the memory hole - a physical hole in every office where paper containing il

by clucas 8y ago

One of the most terrifying parts of 1984 (the Orwell novel, not the year) for me was the memory hole - a physical hole in every office where paper containing illegal ideas or "wrong" information was dumped to be immediately incinerated. So much of what we know about the past and who we are today comes from the written record and interpretations thereof - it was basically erasing history as it happened. For a long time, we didn't really have to worry about this sort of thing, because the physical infrastructure required for such a system didn't exist, would be prohibitively expensive to create, and would likely lead to massive protests if someone tried to create it. But now we are in the digital age, and we have slowly been creating our own memory hole infrastructure, bit by bit, in the form of centralized cloud services. This deletion is an example. Now, these service providers should be (have to be) allowed to remove whatever they want from their platform, but the fact that they can do so should be front and center in everyone's mind when they post content to them, be it code or text or video. We should be patronizing alternatives to github, youtube, facebook, and all other large centralized data stores as much as possible.

yters 8y ago
We also have an accidental memory hold once we've lost the reading method for previous digital records.
- thaumasiotes 8y ago
  That's nothing new. We've been forgetting how to read old records since at least 1200 BC (when the Greeks forgot how to read and write). Interestingly, while Greek history from that period is lost, Mesopotamian/Assyrian/Persian/etc. records continue with cultural continuity from ~3000 BC right up to the Arab conquests, at which point they forgot how to read their old records too. This strongly suggests that that loss was intentional.
  - aasasd 8y ago
    The most recent example in relation to written records is probably the switch of Germany, under the Nazi rule, from the traditional blackletter Fraktur script and its cursives Kurrentschrift and Sütterlin to the ‘Latin’ script. Soon enough, people didn't know how to read their grandparents' handwriting. Try for yourself: https://en.wikipedia.org/wiki/Sütterlin https://en.wikipedia.org/wiki/Sütterlin But it seems that Hitler's rationalization of “if we're to trade with the rest of Europe, those nations must understand German writing” came to be fulfilled.
    - thaumasiotes 8y ago
      This example is a little bit softer than the ones I mentioned. 'Latin' German is, I assume, letter-for-letter identical to blackletter German, just with noticeably different glyphs. Given, say, a box of correspondence, a German who'd never been trained to read blackletter could probably puzzle it out. Akkadian and Sumerian were thousands of years old (and, obviously, long dead) when they got wiped out. And while Mycenaean Greek is not much different from classical Greek (with a gap of only a few centuries), it is written using a writing system entirely unrelated to that of classical Greek, which we assume was originally developed for some other unknown language because it fits Greek so poorly. > “if we're to trade with the rest of Europe, those nations must understand German writing” I don't see why this would make any difference. We trade with Japan (and had been doing so for centuries before Hitler). Recognizing the letters in a German contract isn't much good if you can't understand the words.
brlewis 8y ago
Since git users keep complete copies as a matter of course, the effect of github deletions is less dramatic. I realize this doesn't help for related information such as issue discussion.
- crooked-v 8y ago
  On this note, I'm surprised there's still no de facto standard for having web UI issues associated with a repo actually be powered off text files committed in a subfolder.
  - trickstra 8y ago
    bugseverywhere.org is (was? seems a bit dead) a project doing exactly that - saving issues as part of a git repo
  - coldacid 8y ago
    There have been a few different projects for tracking issues as objects in Git storage, but they all do things in slightly different (incompatible) ways and none have gained enough mindshare to be commonly used. My guess is that none will ever gain enough mindshare without being packaged in with git itself, either, so we got ourselves a chicken and egg problem.
TimJRobinson 8y ago
This technology exists, it just needs to be more widespread. Scuttlebutt[1] is a system for building peer to peer applications. It's main use currently is social networking[2] but someone has already implemented a git store /GitHub alternative on top of it[3]. Nobody controls the network so it can't be censored. If we can get more developers building on top of it we could have a much more decentralized internet. [1] https://scuttlebutt.nz https://scuttlebutt.nz [2]https://github.com/ssbc/patchwork https://github.com/ssbc/patchwork [3]https://github.com/noffle/git-ssb-intro https://github.com/noffle/git-ssb-intro
tick_tock_tick 8y ago
The EU's whole right to be forgotten is the first step in this direction. Honestly the whole EU has been kind of scaring me lately in terms of what they view as acceptable limits on freedoms, especially for speech.

by clucas 8y ago

yters 8y ago

We also have an accidental memory hold once we've lost the reading method for previous digital records.

thaumasiotes 8y ago
That's nothing new. We've been forgetting how to read old records since at least 1200 BC (when the Greeks forgot how to read and write). Interestingly, while Greek history from that period is lost, Mesopotamian/Assyrian/Persian/etc. records continue with cultural continuity from ~3000 BC right up to the Arab conquests, at which point they forgot how to read their old records too. This strongly suggests that that loss was intentional.
- aasasd 8y ago
  The most recent example in relation to written records is probably the switch of Germany, under the Nazi rule, from the traditional blackletter Fraktur script and its cursives Kurrentschrift and Sütterlin to the ‘Latin’ script. Soon enough, people didn't know how to read their grandparents' handwriting. Try for yourself: https://en.wikipedia.org/wiki/Sütterlin https://en.wikipedia.org/wiki/Sütterlin But it seems that Hitler's rationalization of “if we're to trade with the rest of Europe, those nations must understand German writing” came to be fulfilled.
  - thaumasiotes 8y ago
    This example is a little bit softer than the ones I mentioned. 'Latin' German is, I assume, letter-for-letter identical to blackletter German, just with noticeably different glyphs. Given, say, a box of correspondence, a German who'd never been trained to read blackletter could probably puzzle it out. Akkadian and Sumerian were thousands of years old (and, obviously, long dead) when they got wiped out. And while Mycenaean Greek is not much different from classical Greek (with a gap of only a few centuries), it is written using a writing system entirely unrelated to that of classical Greek, which we assume was originally developed for some other unknown language because it fits Greek so poorly. > “if we're to trade with the rest of Europe, those nations must understand German writing” I don't see why this would make any difference. We trade with Japan (and had been doing so for centuries before Hitler). Recognizing the letters in a German contract isn't much good if you can't understand the words.

brlewis 8y ago

Since git users keep complete copies as a matter of course, the effect of github deletions is less dramatic. I realize this doesn't help for related information such as issue discussion.

crooked-v 8y ago
On this note, I'm surprised there's still no de facto standard for having web UI issues associated with a repo actually be powered off text files committed in a subfolder.
- trickstra 8y ago
  bugseverywhere.org is (was? seems a bit dead) a project doing exactly that - saving issues as part of a git repo
- coldacid 8y ago
  There have been a few different projects for tracking issues as objects in Git storage, but they all do things in slightly different (incompatible) ways and none have gained enough mindshare to be commonly used. My guess is that none will ever gain enough mindshare without being packaged in with git itself, either, so we got ourselves a chicken and egg problem.

TimJRobinson 8y ago

This technology exists, it just needs to be more widespread. Scuttlebutt[1] is a system for building peer to peer applications. It's main use currently is social networking[2] but someone has already implemented a git store /GitHub alternative on top of it[3]. Nobody controls the network so it can't be censored. If we can get more developers building on top of it we could have a much more decentralized internet. [1] https://scuttlebutt.nz https://scuttlebutt.nz [2]https://github.com/ssbc/patchwork https://github.com/ssbc/patchwork [3]https://github.com/noffle/git-ssb-intro https://github.com/noffle/git-ssb-intro

tick_tock_tick 8y ago

The EU's whole right to be forgotten is the first step in this direction. Honestly the whole EU has been kind of scaring me lately in terms of what they view as acceptable limits on freedoms, especially for speech.