[Idea] Polls of Deceased Regions

[La Xinga]

When a region ceases to exist, all polls posted in the region, possibly numbering in the hundreds, are lost forever, seemingly. It seemed that they were simply unrecoverable.
After poll date was restored from May, though, many polls have resurfaced which are from regions long gone. Examples include:
https://www.nationstates.net/page=poll/p=199734
https://www.nationstates.net/page=poll/p=199864
https://www.nationstates.net/page=poll/p=199799
https://www.nationstates.net/page=poll/p=198262
https://www.nationstates.net/page=poll/p=198393
https://www.nationstates.net/page=poll/p=197979

Does this mean it's possible for poll history of a deceased region to be saved without too much work, or were the examples shown above only an exception due to the hardware failure?

A weird thing which I noticed is that there are a very small amount of deceased regions which do retain some poll information unrelated to the hardware failure. Examples include:
https://www.nationstates.net/page=poll/p=150778
https://www.nationstates.net/page=poll/p=197843
https://www.nationstates.net/page=poll/p=145414
https://www.nationstates.net/page=poll/p=148035

These are definitely exceptions, though, with the vast majority of polls from deceased regions disappearing.

[Roavin]

This is a side effect of restoring polls from an older backup. Normally, when a region ceases to exist, everything related to it (RMB, polls, etc) are wiped. The polls you mentioned will probably be cleaned up at some point.

[La Xinga]

This is a side effect of restoring polls from an older backup. Normally, when a region ceases to exist, everything related to it (RMB, polls, etc) are wiped. The polls you mentioned will probably be cleaned up at some point.

Thank you for the response. I suppose this means there's no hope of deceased regions' polls being saved in the future?

[Sedgistan]

It's not on the agenda at present, but people can make the case for it if they think it's a worthwhile development.

[All Wild Things]

It would make a fascinating dataset. For example we could research what the most common poll topic was (I suspect "pineapple on pizza" could be the most common, but it could be "is a hotdog a sandwich?"). We could aggregate the results over the years, and come up with a definitive all-time answer to the "pineapple on pizza" question. We could also look at the trends over time, and see if pineapple came in and out of fashion.

If, like me, your insomniac moments are consumed by considering these important questions, then you know that recording poll history is the right thing to do.

[La Xinga]

It's not on the agenda at present, but people can make the case for it if they think it's a worthwhile development.

I hope people think believe it is; thank you for the response.

It would make a fascinating dataset. For example we could research what the most common poll topic was (I suspect "pineapple on pizza" could be the most common, but it could be "is a hotdog a sandwich?"). We could aggregate the results over the years, and come up with a definitive all-time answer to the "pineapple on pizza" question. We could also look at the trends over time, and see if pineapple came in and out of fashion.

If, like me, your insomniac moments are consumed by considering these important questions, then you know that recording poll history is the right thing to do.

Perhaps a search feature for all the polls could be added (assuming it's possible), with criteria for poll creator, voter, region, etc, for all polls ever posted. Maybe other nations' poll history too, as you can only currently view your own, though perhaps that was intentional.

[Frisbeeteria]

This, I believe, the key question:

If, like me, your insomniac moments are consumed by considering these important questions,

I'm thinking the population who are "like you" is likely in the single, or at most, double digits. Do we really want to ask the extremely busy developers to spend time working on a system that does not actually improve the game, but exists solely to "make a fascinating dataset"?

Not seeing it.

[TalAkMaChen]

.... "pineapple on pizza" could be the most common [question]

There is only one answer to that question and it is a clear "no way, this is disgusting".
Else, you can still collect the data on your own, so why should admins/coders collect hundreds if not thousands of polls in some "historic record"? Sure, keeping record is a thing for historians but to what cost? It takes work to implement and test that feature, it takes up even more data tables that will for 99% be unused anyhow - and all for the amusement of a few "nerds"? Nah, I agree, there are better things to code first.

[Trotterdam]

We already have 58151 archived polls, many of which are in strong regions that are unlikely to expire (or even in game-created regions that can't expire), and will therefore be retained permanently. Retaining an extra few polls from regions that happened to cease to exist would take up a trivially insignificant amount of additional disk space. Polls in general, being mostly text with stringent character limits, take up very little space compared to other things like images, factbooks, or even RMB posts (a poll probably takes up a little more space than the average single RMB post, but people make RMB posts a lot more often than they make polls).

The only real difficulty is if then someone recreates a new region by the same name, and you might get confused about whether the poll came from before or after the refounding. Which doesn't seem like a big deal, especially since the same can happen with a region getting raided and being placed under new management without ever ceasing to exist.

[Bormiar]

A lot is lost every time a region expires, most notably RMB posts. We have ways of mitigating that damage, including through offsite resources which store old RMB posts and WFE changes, and the F/S update which makes refounding a region less appealing.

If you want to make a dataset like the kind All Wild Things describes, and you're not satisfied by the 58151 existing archived polls, you could always start keeping track of that data yourself. I'm not saying this update shouldn't happen. Just that there are other solutions which place no strain at all on the very few developers trusted to access NS' source code.

By the way, AWT, that information sounds sort of tricky to extract from a limited amount of unfiltered natural language data, and of little use. If you're interested in and have the time for this sort of thing, you can try out https://www.kaggle.com/ instead.

[All Wild Things]

.... "pineapple on pizza" could be the most common [question]

There is only one answer to that question and it is a clear "no way, this is disgusting".

Um, no. You're trying to impose your evil pizza-fascist view on others. Despite 61% of Europeans also being evil pizza-fascists who are willing to ban pineapple on pizza, this wider poll reveals that approximately 40% of voters say "Yay" to pineapple on pizza. Research conducted by Osiris about 4.5 years ago revealed that pineapple was the third favourite pizza topping, behind pepperoni and "just plain cheese". Further research conducted by TEP a few months later had similar results, but with pineapple and "just plain cheese" tied for second place behind pepperoni. And you have the cheek to call me a nerd...

BTW, the world still needs to know: is a hotdog a sandwich?

If you're interested in and have the time for this sort of thing, you can try out https://www.kaggle.com/ instead.

I'll take a look at that, thanks!