Project for personal tracking tasks of NDKilla.
Aug 2 2018
Jul 22 2018
Jun 27 2018
May 15 2018
Since the backend work is all done here, I am therefore resolving this task.
Looking at implementing variables, could use json_encode() to store in MySQL then using json_decode(), manipulate and retrieve as needed. Though how to manage it with respect to all.dblist will be interesting.
May 11 2018
Extensions can now be managed (in theory) by using Special:ManageWiki.
Apr 2 2018
Going to mark this as resolved as I have just published a base extension which will begin to replace this extension.
Apr 1 2018
Mar 19 2018
Mar 17 2018
That’s the sole purpose of this task/project.
Yes. I don't mean raw text. I mean checkboxes which are directly applied to the LS. Give full power of LS to wiki admin, but with safety layer.
Direct editing of LS would be difficult to do properly. The most I'd say could be done would be a series of checkboxes, instead of direct code editing. As otherwise, incorrect code or data leaks are likely to occur. The way we have in mind would be both user-friendly and safe.
my suggestion was merged to this task. But maybe different.
Mar 16 2018
Jan 18 2018
https://phabricator.wikimedia.org/T140941 is a task for making SiteMAtrix even less WMF specific.
Jan 1 2018
Dec 31 2017
Dec 30 2017
Dec 23 2017
Dec 6 2017
Nov 27 2017
Oct 29 2017
@labster Any updates? Do you think you'll be able to work on this sometime soon?
Sep 14 2017
Sep 1 2017
This is for Semantic MW, but it is an interesting example for what we could do: https://www.mediawiki.org/wiki/Extension:Semantic_Drilldown
Aug 29 2017
Puppet only manages stuff in puppet anyway so this can't be in puppet.
@labster I didn't mean "easy", but I meant if, you would do one setting, like "wgLogo" adding all the rest would be trivial (just copy it and replace the values) but of course programming the initial code would not be close to easy.
Aug 28 2017
On thinking about it, we can't have a single CDB for all of the wikis, because CDB files are immutable. I keep thinking it's like BerkeleyDB or something. I still feel like serialized blobs should be faster, because there's less overhead for something we'll just slurp into memory. (CDB is still a better option for LC, because you only need a few values for each request, not the whole thing.)
Aug 22 2017
@labster that is probably. It's better to have all values for every request, but we need to be sure there is no negative performance impact.
Aug 20 2017
@Southparkfan this still needs an answer.
Aug 13 2017
Tasks that go on for ever and ever are bad.
Jul 24 2017
It was unintentional. But anyway, doubt it'll be started in this goal period.
Why remove the goal?
https://meta.miraheze.org/wiki/Tech:Projects/Miraheze_Wiki_List I wrote this up for GSOC a while ago.
It would be nice to have a place where wikis are sorted into topics, or something. It's sad to see many wikis with good content, but only one editor.. I'm sure if other users interested in the topic knew about it, they would contribute
Jul 21 2017
More wikis (also in deleted.dblist) can now be dropped.
dragonwiki dsiguidewiki elo131wiki englishlanguagewikiwiki evlerwiki fcswiki filowiki fluxwiki h2751903wiki hendrickswiki hivemindwiki hrodrwiki humanumwiki jakeperswiki joongs14wiki juliesnoteswiki lgproduktsupportwiki lucilekbtwiki minecraftaddonswiki mylogicwiki mywiki ndnwiki oneclanwiki orbiswikiwiki rimworldwiki modelusgovwiki tufuswiki cosmoswiki
Jul 18 2017
libertywiki dump is on /home/revi. It can be dropped, though it can wait until I provide a new list.
@revi needs to act on libertywiki.
Jul 16 2017
teachingsecondswiki telegramvtwiki tennisdbwiki testrbcmdoewiki thamilwiki theseabeneathwikki tiddlywikiwiki tumblenetwiki videogameswiki vifakswiki vvwiki warbinwiki wcwiki weeglifewiki wethepediawiki wijoumedwiki wikislandswiki wikixrlianwiki xyzwiki yourovisionwiki zaffwltestwiki zonwiki zuersnetwiki zvtwiki wiki1776wiki 17huberkwiki 2l1v3swiki 5awiki adwiki aevumkathwiki anatoruswiki anet3dwiki asellwiki ashishwiki boss429wiki buswikiwiki casualtywiki cats11thstreetwiki cerhisdevwiki chgkirkwiki citywiki criptanawiki csstestwiki datawiki daudtwikiwiki dh8257wiki dndwiki nationstateswiki conworldwiki
I have a question about using CDB files. Assuming that we'll have to load every value in the configuration to process a request, is there a reason to use a key-value store at all? If we need to iterate through all of the keys in the file, why not just store a serialized array and just load that file? Or have a CDB for all the wikis, mapped as ( wikiwiki => [ serialized settings ] )?
Jul 10 2017
already dropped (before I tried to):
already dropped (before I tried to):