Run test version of Etherpad-based IMPS

[joereddington]

We are outgrowing Google Drive at the same time that it is making itself harder for us to use. Increasingly groups are finding it harder and harder to log in.

It's time for us to have an alternative, either as a backup, or as a complete replacement.

We're going to base our client on etherpad-lite. I created this issue: https://github.com/ether/etherpad-lite/issues/3599, which has been fufillied.

We previously attempted to set something up on Digital Occean in #77 but it felt down over config and access issues. Now that we have moved the WWW stack to AWS (in #139) we can do it properly.

Plan

• [x] Prove everypad works in AWS (commands and links in #139)
• [x] Update the url - 'etherpad' is a bit of a mouthful.
• 'pad' is standard
• 'imps' maybe? That is the place
• 'orcle' would be taken...
• 'Joe is awesome'
• imps it is.
• [x] Make everpad pretty by cutting down the unecessary icons
• [x] Write script that reads 'seed' file and generates cards, ordering file, and info.
• [ ] Alter IMPS to read the ordering file version. (it gets the ordering file, and then it gets the rest of the files
• [x] Add the multiple dashboards, one for each character.

Notes

[joereddington]

• Testing the old python scripts with the API key Planning out how I want Oracle to be.

• The main insight was this: there needs to be a central 'ordering' file that writers can use to change the order of their cards (Possible problem with this is if they delete some of them)

• The second thing I want is to be running from one repo - NO more cloning a new one each time.

• I want changes to happen on a button press. No 'new version every ten minutes'

• I want the creation of cards to be automatic from a form like this: https://whitewaterwriters.com/imps/p/SampleSeed

[joereddington]

There is a file called germinate.py on the server that works. It correctly generates the files.

Next actions:

• Produce the php that runs the file.
• Add the errors if the files already exist
• Write a JSON file which saves the names of the characters
• Also upload the info file and the ordering file
• Put under version control properly.
• Proper if main
• Refactor massively

Then, and only then.

• Write the code to bring the files back down, based on the slug (and the ordering file)
  • Write one (1) large latex file out of them.
  • No, because more than one character will be interested in the same file.
  • Create Json file that includes down to the bullets and which is included in the dashboard.
  • I mean, how much of a problem is it if it includes right down to the full text? The pdf are only half a mb? It's 2022.
  • Put some \typeout{message}. messages into the code so that we can see what is happening properly.

[joereddington]

11/02/22 08:02 to 08:20, https://github.com/eQualityTime/Public/issues/150 +EQT

11/02/22 08:46 to 10:22, https://github.com/eQualityTime/Public/issues/150 +EQT

Lots of files now working.

• Now can give a slug and get a wordcount. That's cool.
• Need to be able to save it in a way that works for the php.

11/02/22 10:30 to 11:20, Grant writer call. +EQT

Lots of good notes about

• Redoing the website
• Proofreading the website for WWW
• Accounts formatting
• Stop answering questions that weren't asked for.

11/02/22 11:22 to 13:17, https://github.com/eQualityTime/Public/issues/150 +EQT

• I've now got a json file with one book in it.
• I've built a very basic dashboard in php.
• I've done a brainstorm about what bits of the code need to happen in what order.
• I considered how much of the work I should do in Php to save the downloads and of course that's a stupid question because everything is on the server anyway. I'll do it all in python.

11/02/22 13:17 to 13:34, Need to rewrite this file and catch my breath https://github.com/eQualityTime/Public/issues/150 +EQT

11/02/22 13:34 to 14:03, Family break

11/02/22 14:03 to 15:54, https://github.com/eQualityTime/Public/issues/150 +EQT

Update_books script first.

• Really noticeable how much better things are when they are object oriented
• Requires me to turn off the direct 'overwrite'
• Setup vim tags
• Wrote 'export code' and script for tar
• Setup crontab, now need to debug it.
• Redid Germinate so it uses a synopsis and correctly creates the info and ordering files
• Edited dashboard to review the newly generate samples
• Added the wordcount and synopsis extraction to the python
• Dashboard now shows wordcount and synopsis (
  • And recovers okay when it doesn't
• Characters:
  • It's done! Time for a bit of a commit
• Total word counts were done in php.