Page 2 of 2

PostPosted: Thu Jul 12, 2018 4:09 pm
by Imperium Anglorum
All of it's a great plan, Fris! I hope it happens.

PostPosted: Tue Jul 24, 2018 5:31 pm
by Imperium Anglorum
Bump.

PostPosted: Fri Aug 17, 2018 12:55 pm
by Altmoras
Bump

PostPosted: Fri Aug 17, 2018 1:36 pm
by Eluvatar
I'm unfortunately busy right now. I'll get to this when I can, and keep you posted.

PostPosted: Fri Dec 21, 2018 11:09 pm
by Wallenburg
Bump. Again.

PostPosted: Mon Jan 07, 2019 4:54 pm
by Wallenburg
Bump.

PostPosted: Tue Jan 08, 2019 6:07 pm
by Imperium Anglorum
Definitely still support this idea.

PostPosted: Wed Jan 09, 2019 8:25 pm
by The Northern Light
While it'd be great if admins implemented this idea, as a temporary solution, I fixed a long running bug in my WA Vote & Proposal archive, so that proposal text is now stored correctly:

http://cerulean.nsr3n.info/wa_vote_records/proposals/

You can look at the GA entries for "Repeal: “WA General Fund”" and SC entries for "Repeal: "Commend Lyras"" and "Commend Paffnia" to see what new proposal entries will look like. They should have all the information they used to have, plus the full proposal text.

Unfortunately, I can't fix older entries. For those, there's still the other pieces of information (author, title, approvals, etc.), in case those are of use.

PostPosted: Sat Jan 12, 2019 9:47 am
by Nagatar Karumuttu Chettiar
I know my opinion isn't particularly significant, but I do support the idea, for the record. Old proposals could be a wellspring of ideas for aspiring authors (cough cough, like myself).

PostPosted: Sat Jan 12, 2019 1:15 pm
by Borovan3
I don't support indexing every submitted proposal when some are embarrassing only if they have reached queue or are good

PostPosted: Sat Jan 12, 2019 5:29 pm
by Imperium Anglorum
Borovan3 wrote:I don't support indexing every submitted proposal when some are embarrassing only if they have reached queue or are good

If they're already archived by TNP (and now, Europe) then how is that not non-unique?

PostPosted: Sat Jan 12, 2019 7:28 pm
by Lord Dominator
Borovan3 wrote:I don't support indexing every submitted proposal when some are embarrassing only if they have reached queue or are good

'Good' is an extremely nebulous

PostPosted: Fri Mar 15, 2019 7:30 pm
by Altmoras
Bump

PostPosted: Fri Mar 15, 2019 8:15 pm
by Wallenburg
IA is maintaining an archive HERE at this time.

PostPosted: Fri Mar 12, 2021 11:01 pm
by United Calanworie

PostPosted: Sat Mar 13, 2021 7:07 am
by Comfed
This is a good idea.

PostPosted: Sat Mar 13, 2021 10:54 am
by Wallenburg
Comfed wrote:This is a good idea.

It is, and it would be very easy to implement. Fris has indicated as much earlier in this thread.

PostPosted: Sat Mar 13, 2021 3:13 pm
by Sedgistan
I've brought it to Violet's attention.

PostPosted: Sat Mar 13, 2021 8:21 pm
by [violet]
I'm not sure what Elu has/had in mind, but imo this isn't practical with our current system architecture. The entire WA has to remain resident in memory and is very slow to load and save itself--it does not use a database like dispatches do. Storing a thousand or so ex-proposals would increase the size of this WA object by an order of magnitude and create a significant performance drag.

So implementation requires a pretty fundamental rewrite of how the WA stores data.

PostPosted: Sat Mar 13, 2021 8:59 pm
by Merni
[violet] wrote:I'm not sure what Elu has/had in mind, but imo this isn't practical with our current system architecture. The entire WA has to remain resident in memory and is very slow to load and save itself--it does not use a database like dispatches do. Storing a thousand or so ex-proposals would increase the size of this WA object by an order of magnitude and create a significant performance drag.

So implementation requires a pretty fundamental rewrite of how the WA stores data.

So, even the 1000 or so (GA+SC+historical) passed resolutions stay in memory? If those are somehow stored, couldn't proposals be stored the same way?

PostPosted: Mon Mar 15, 2021 2:20 am
by [violet]
Merni wrote:So, even the 1000 or so (GA+SC+historical) passed resolutions stay in memory? If those are somehow stored, couldn't proposals be stored the same way?

It all stays in memory, yep, but the bigger it gets, the worse it performs.

PostPosted: Mon Dec 11, 2023 11:24 am
by Eluvatar
[violet] wrote:I'm not sure what Elu has/had in mind, but imo this isn't practical with our current system architecture. The entire WA has to remain resident in memory and is very slow to load and save itself--it does not use a database like dispatches do. Storing a thousand or so ex-proposals would increase the size of this WA object by an order of magnitude and create a significant performance drag.

So implementation requires a pretty fundamental rewrite of how the WA stores data.

That is indeed why it hasn't been done prior. Some of the (smaller) changes made recently to support frontiers may be a promising direction to take here, but proposals are more complicated than membership.

PostPosted: Sun Mar 17, 2024 5:36 am
by Fachumonn
I don't know if I'm still allowed to post here, but how about just storing resolutions that got to queue, whether legal or illegal. This could surely cut down on a bunch of the spam proposals?

PostPosted: Sun Mar 17, 2024 12:30 pm
by Unibot III
Is it possible to have a different part of the site infrastructure scrap the relevant pages, filter for the proposal text or failed resolution, then index by proposal ID in a different, more flexible database? Essentially the whole operation, included the recorded proposals, would live in a database separate from the WA?

PostPosted: Sun Mar 17, 2024 3:01 pm
by Imperium Anglorum
Unibot III wrote:Is it possible to have a different part of the site infrastructure scrap the relevant pages, filter for the proposal text or failed resolution, then index by proposal ID in a different, more flexible database? Essentially the whole operation, included the recorded proposals, would live in a database separate from the WA?

Well, the possibility is obvious given that it's already been done that way (by me). The difficulty I think for UC is implementation. That part I can't say I am at all privy to.