User Details
- User Since
- Oct 12 2014, 9:48 AM (532 w, 5 d)
- Availability
- Available
- IRC Nick
- Melos
- LDAP User
- Melos
- MediaWiki User
- Melos [ Global Accounts ]
Oct 23 2024
Oct 1 2024
Sep 29 2024
Aug 26 2024
Jun 28 2024
May 4 2024
Analizzando i tool già presenti abbiamo Xtools che permette di estrarre una determinata categoria di edit automatici:
- https://xtools.wmcloud.org/api/user/automated_editcount/{project}/{username}/{namespace}/{start}/{end}/{tools} (Doc)
Settando il parametro {tools} su true restituisce il numero di edit raggruppati per i tool presenti qui, quindi facilmente aggregabili in base alle nostre esigenze.
La lista global include i principali tool automatici (lista parziale non esaustiva):
- AutoWikiBrowser
- JWB
- PAWS
- WPCleaner
Occorre individuare gli altri eventuali tool non conteggiati (in attesa di feedback dalla comunità).
May 2 2024
L'etichetta porta benefici di verifica postuma degli edit. Sarei per il supporto locale essendo un tool il cui sviluppatore è quasi inattivo (vedi per esempio questo).
Apr 30 2024
Mar 18 2024
Feb 28 2024
Jan 31 2024
It works fine. Thank you!
Jan 30 2024
You got a new talk message! https://it.wikipedia.org/wiki/Discussioni_utente:DLynch_(WMF)#Test
Oct 7 2023
Sep 13 2023
Sep 7 2023
Aug 22 2023
Thank you for your fix! I think we can wait the normal deployment train (message can be easily overwritten locally).
Jul 24 2023
May 29 2023
May 27 2023
May 20 2023
Mar 31 2020
Mar 19 2020
I have seen this bug on my local installation when the core version is updated to 1.35.0-wmf.24 and vector version remains to 1.35.0-wmf.23.
Mar 1 2020
Feb 27 2020
Feb 24 2020
Feb 23 2020
Feb 17 2020
Feb 10 2020
Feb 7 2020
Fixed in py2 to py3 migration
Feb 6 2020
Feb 13 2019
Feb 10 2019
I've created a virtualenv with:
*pip install irc
*git clone https://github.com/sixohsix/python-irclib.git (http://python-irclib.sourceforge.net/)
*pip install MySQL-python (https://pypi.org/project/MySQL-python/)
Now bots are running (via "source venv/bin/activate"+restart script) in the virtualenv but we need to update crontab (or bash scripts) to start using it
Feb 9 2019
Dec 23 2018
This feature comes with https://gerrit.wikimedia.org/r/#/c/mediawiki/core/+/405015/ since 14/06/2018, see https://phabricator.wikimedia.org/source/mediawiki/browse/master/includes/Storage/PageUpdater.php$1159
and T196400
Nov 16 2018
Nov 15 2018
I have updated my php version from 7.0.10 to 7.0.32 and everything works fine.
Nov 14 2018
Also REL1_32 branch is affected by the same error.
Nov 13 2018
master version.
After a manual installation there are other errors while attempting to save like Fatal error: Cannot use MediaWiki\Revision\RevisionSlots as RevisionSlots because the name is already in use in includes\Storage\RevisionSlotsUpdate.php on line 28
I don't know if they came from https://gerrit.wikimedia.org/r/#/c/mediawiki/core/+/461714/
Apr 6 2018
Mar 31 2018
Mar 7 2018
- In some places, such as the screenshot you included from Special:AbuleFilter, the field is treated plain text.
Correct, it is only a description so plain text is correct IMHO.
- In other places, such as Special:AbuseLog and the "preview selected message" feature when editing a filter, it's treated as wikitext and gets parsed to HTML.
Here there is the first problem, we need to change it in plain text (see for ex. T141670).
- And in some places, such as the warning produced on the edit page when the filter actually fires, it's treated as wikitext and parsed to HTML and then that HTML is treated as wikitext and parsed a second time, which breaks links in the original "wikitext" and leads to display as shown in this task's description.
What really needs to happen is that one interpretation, plain text or wikitext, needs to be chosen and used everywhere. IMO it should just be plain text. And then either way the edit page warning needs to be fixed to not double-parse.
I agree, plain text is the better choose and so we can also remove parseinline.
Nov 19 2017
Nov 18 2017
Nov 16 2017
Nov 12 2017
As I have explained in gerrit imho the patch have to check $wgBlockAllowsUTEdit. If it's set to false you don't have to show the checkbox as in Special:Block
Nov 11 2017
Nov 7 2017
Oct 27 2017
When 'afl_ip' => '' should we add a message to inform user that Originating IP address was dropped/too old or leave the user to see an empty "Originating IP address" in the results? In every case the try to access private data must be logged if $wgAbuseFilterLogAccessingPrivateDetails = true.
Oct 26 2017
We could check for global $wgCUDMaxAge but if wgAbuseFilterPrivateDetailsMaxAge is setted it override $wgCUDMaxAge so site admins can have more flexibility in configuring AF.
Oct 24 2017
Another point: In Wikimedia sites we could decide to view data only from a X period (for ex 3 months). Since IP are stored in afl_ip field and we can't saefly drop it from the database we should add another time variable for it to compare with afl_timestamp. Then, if filter log timestamp is older than our max time limit, user will be warning with a "too old" message.
While I'm testing the patch above I was thinking to create another configuration variable to require a valid reason. Since are private data site admins could decide to require from users a valid reason to see them. So setting a "AbuseFilterLogAccessingPrivateRequestReason": false in extension.json and checking for it in the code should be a good point.
Oct 22 2017
What about to separate rights for logs and for view information as in the CU extension? This will make it more configurable.
For example. A trusted user group can be able to see only the log without any risk.
Oct 19 2017
Oct 16 2017
While I'm working on something that show links for CentralAuth and GlobalBlocking I've sent a patch for "checkuser-userlink" message, so we can add the equivalent of "checkuser-userlinks-ip" also for registered users. Than we can close this bug and open another one as new feature request.
Oct 15 2017
I'm working on it and I'll send a patch for code review. I'll try to add new links as integration with global extensions.
Oct 11 2017
Oct 10 2017
I think it don't pass the part ( IP::isIPv6( $ip ) && $range < $wgCheckUserCIDRLimit['IPv6'] ) ) since 29 < 32 (range is too wide). Is the bug in the block stuffs instead?