A new wiki for the project

Avatar

By Sanglorian 29 Jul 2016 23:46

Champion · 501 comments

Thank you ShadowApex! The import looks very good overall, and will have saved us a whole heap of effort.

There are only a few problems I'm noticing.

1. The files seem to have all of their revisions, but then have their original version duplicated as the most recent one. So if a file was created as [[Category:Dog]] [[Category:Brown]] and then edited to [[Category:Dog]] [[Category:Reddy Brown]], the most recent version is [[Category:Dog]] [[Category:Brown]] again. (EDIT: Or maybe there's just no revision history, and the most recent version is whatever it was originally imported with, see https://wiki.tuxemon.org/index.php?titl … uana_2.png )

I tried fixing this by re-importing some files data from tuxemon.referata.com, but this doesn't work because the newer revisions still exist on wiki.tuxemon.org, they just are followed by a revision containing the oldest version.

We can edit this manually if there's no easy automatic fix.

2. The Referata wiki will from time to time run through and "refresh" each of its pages. That doesn't seem to happen here.

3. The Monster template is set up along with the CSS file to make an animation out of the two face sprites by making one of them alternate appearing and disappearing. That doesn't seem to work here, although I'm not sure why.

4. The "set" function doesn't seem to work (see for example the bottom of the Memnomnom entry: https://wiki.tuxemon.org/index.php?title=Memnomnom )

Finally, not an error, but is there a way that we could allow for the entire wiki (pages, images, database, etc.) to be downloaded/forked as a whole? I feel like this would be a good way of making sure it's all backed up, as well as being in the spirit of open source (other Tuxemon-like projects could just grab the whole thing and hit the ground running).

--

Thanks again! This made my day yesterday when I saw you'd done it.

Last edited by Sanglorian (30 Jul 2016 00:39)

Avatar

By tamashihoshi 1 Aug 2016 17:26

Champion · 251 comments
ShadowApex wrote

I finally finished a working script to migrate all the images (and image history) over from referata to wiki.tuxemon.org! big_smile

thats so cool!!  \ (,,° 0 °,,) /


We'll meet again, don't know where, don't know when. But I know we'll meet again some sunny day!

Avatar

By ShadowApex 5 Aug 2016 00:35

Lead Developer · 374 comments
Sanglorian wrote

1. The files seem to have all of their revisions, but then have their original version duplicated as the most recent one. So if a file was created as [[Category:Dog]] [[Category:Brown]] and then edited to [[Category:Dog]] [[Category:Reddy Brown]], the most recent version is [[Category:Dog]] [[Category:Brown]] again. (EDIT: Or maybe there's just no revision history, and the most recent version is whatever it was originally imported with, see https://wiki.tuxemon.org/index.php?titl … uana_2.png )

I tried fixing this by re-importing some files data from tuxemon.referata.com, but this doesn't work because the newer revisions still exist on wiki.tuxemon.org, they just are followed by a revision containing the oldest version.

We can edit this manually if there's no easy automatic fix.

Unfortunately there's no easy fix for this. sad

Sanglorian wrote

2. The Referata wiki will from time to time run through and "refresh" each of its pages. That doesn't seem to happen here.

What kind of refresh is triggered? Is it specific to Semantic MediaWiki? I can add a cron job to implement this if needed.

Sanglorian wrote

3. The Monster template is set up along with the CSS file to make an animation out of the two face sprites by making one of them alternate appearing and disappearing. That doesn't seem to work here, although I'm not sure why.

I can upload the CSS if you know what was used.

Sanglorian wrote

4. The "set" function doesn't seem to work (see for example the bottom of the Memnomnom entry: https://wiki.tuxemon.org/index.php?title=Memnomnom )

I just installed the "SemanticInternalObjects" extension, so this should work now.

Sanglorian wrote

Finally, not an error, but is there a way that we could allow for the entire wiki (pages, images, database, etc.) to be downloaded/forked as a whole? I feel like this would be a good way of making sure it's all backed up, as well as being in the spirit of open source (other Tuxemon-like projects could just grab the whole thing and hit the ground running).

Because the database contains user data/passwords, I would not want this to be available to everyone. I might look into making the pages and images downloadable though. It's also possible that I could dump everything in the database except the user table, but that may break things depending on how MediaWiki handles user ids. In any event, I do maintain off-site backups of the website and wiki.

EDIT: I added the cron jobs listed here: https://www.semantic-mediawiki.org/wiki … #Cron_jobs


Avatar

By Sanglorian 8 Aug 2016 07:20

Champion · 501 comments

Hey ShadowApex,

1. No worries - I did a manual read-through of all the uploaded assets and most should be working now.

2.

What kind of refresh is triggered? Is it specific to Semantic MediaWiki? I can add a cron job to implement this if needed.

2. I'm not sure sorry - but you can see for example on the SMWAdmin page that the "Data repair and upgrade" isn't running despite being turned on for a few days. I think on most SMW wikis, every page visit by any person makes a few of these jobs run in the background.

Sanglorian wrote

3. The Monster template is set up along with the CSS file to make an animation out of the two face sprites by making one of them alternate appearing and disappearing. That doesn't seem to work here, although I'm not sure why.

I can upload the CSS if you know what was used.

Thanks, I've actually already uploaded the CSS here. For some reason that does result in an animation on the old wiki, but not on the new one. 

I just installed the "SemanticInternalObjects" extension, so this should work now.

Thanks!

Sanglorian wrote

Finally, not an error, but is there a way that we could allow for the entire wiki (pages, images, database, etc.) to be downloaded/forked as a whole? I feel like this would be a good way of making sure it's all backed up, as well as being in the spirit of open source (other Tuxemon-like projects could just grab the whole thing and hit the ground running).

Because the database contains user data/passwords, I would not want this to be available to everyone. I might look into making the pages and images downloadable though. It's also possible that I could dump everything in the database except the user table, but that may break things depending on how MediaWiki handles user ids. In any event, I do maintain off-site backups of the website and wiki.

No worries. I would like some sort of dump to happen, just for peace of mind and to make sure the assets are readily available (it's so sad to see so many other community projects where links are broken and assets are lost. One of the great things about Github is there's basically no chance of that happening for our code!)