In order to make sure our codes stay tidy going forward (and we don’t have to do another epic week of 2-3 hour meetings) I am putting into place a bi-weekly coding meeting for POPREBEL. As long as there is at least one representative from each coding team per meeting, we are good.
The meeting structure will be straightforward – we will go through all the newly assigned codes since the last meeting (as in, all codes that have been newly created, not ones that have been reused) and make sure they aren’t duplicates of those which came before, as well as making sure we all know how to use them.
This will technically involve us all opening the ethno-poprebel codes page and sorting by “newest”. The meetings will be an hour, less if we make it through all the codes. We will also of course open the floor to other questions, comments, and discussions.
Let’s set a recurring day and time for this. We know from experience that 2pm UK to 3PM UK generally works for everyone. I propose Thursdays but am flexible. @Jan@Richard@Wojt@SZdenek@Jirka_Kocian please let us know what works for you, and we’ll get it set in the calendar as an ongoing event. Thank you!
I prefer Fridays or Wednesdays (Thursdays, I will be walking/travelling with my daughter from her school to her athletics training 15:00-15:30. Once she will be there, I will be available for our call, for one hour, 15:30-16:30.)
The same with Jan, once the semester begins I will be having a biweekly blast of teaching from 11 till 19:00 on Wednesdays with sime breaks in-between, so Friday seems like the best solution for me
OK, thanks all-- please set in your calendars that we will have a biweekly call from 2-3 UK, 3-4 Europe on Fridays, starting October the 9th.
We will not be shifting or moving around these calls from week to week — too much time has already been spent on scheduling — so make every effort you can to make it, please.
At the beginning of the call, please also be prepared to give a short update on the content activity on your part of the platform: what people are talking about, what you’re finding interesting, etc. Then we’ll dive into the code review.
Excellent. In my calendar. @Wojt and I just finished our next session and will meet again tomorrow. We are now well into the Socio-economic category and have collected by now a number of proposals to discuss, in addition to executing a quite a few more obvious mergers. All best, J
Just a note for our Friday meeting, not to forget it:
Shouldn’t we rearrange our shared categories in the spreadsheet according to what Jan outlined (to our new analytical categories)?
We would have to add new codes (as e.g. Jan’s point 3.9.x shows) we have not used yet (and recode our texts after that) in order to catch for example the extent or level of activity (category 3.1)
@amelia : We had just call with @Jirka_Kocian and till the next meeting on Friday we will work on:
(A) cleaning and merging of codes in “our” (old) categories (Education, Institutions, Housing, Health, Environment) - if it still makes sense since we are going to have a new system of analytical categories (see Jan’s post from Sept. 12). (If needed [and still relevant], I can merge Labour or Arts&Media during the next week. I could find some time for that.)
(B) creating and uploading Czech codenames (in the back-end) to as many codes as possible.
We are going to kill the use of the spreadsheet from here on out, as this was just to do this big push of organisation and merging. Now that all the codes and definitions are in the backend, there’s no reason to use Google Sheets — it’s a lot of duplication work since we’d have to manually add all new codes into it. We’ll instead do much more regular code review and use hierarchies in the backend. How to represent the top-level categories that won’t be part of formal hierarchies is a question we’ll need to answer, and we’ll discuss it on Friday.
Can one child code have more parents? @Jirka_Kocian wasn’t sure, if by making a hierarchy he would perhaps break some already existing hierarchy. I was not sure.
Should we do simple hierarchies like: “sexuality” as parent, and “LGBT”, “pedophilia”, “pornography”, etc. as children. And e.g. “LGBT activism” as child of “LGBT”? (Or a child of “activism”? - At best of both, right?!)
Please refer back to the hierarchies portion of the Open Ethnographer manual (and my prior posts on the subject, including notes from calls where we all discussed this). If you and @Jirka_Kocian need further clarification after that, we can talk about it in the call again.
@amelia Hope you had a great vacation. @Wojt and I are working on hierarchies. We understand that we should move our work from the spreadsheet to the backend, but we need a bit more time to finish our review (it is easier to review in the spreadsheet). Second, we have questions concerning several “mergers” and I suppose we should discuss them before we “lock” them in the backend. THIRD: we would love to have access to poprebel graphryder, so we can experiment with it in order to get a better/deeper “feel” for categories and hierarchies. We cannot find a link to it. Could you please provide it? All best, J
Remember, it does not update in real time. @hugi any chance we can do what you did for NGI, so the sliders and comment/posts appear? I know we need to have a bigger conversation about this.
I support staying in the spreadsheet until our meeting as it seems there is still some work to do on learning the category hierarchy distinction and the nature of hierarchies. Better to be sure across everyone in the project, otherwise we create a lot more work in undoing things.
@alberto@hugi@amelia@wojt@SZdenek@Jirka_Kocian Dear All: It is REALLY exciting to see the preliminary and early results. Many thanks! It will help enormously to sharpen the coding scheme and its categories, as well as to start developing interpretations of what is happening in those “mental maps.” I will play with it a lot before we meet next Friday, so I can have a precise list of comments and questions. Have a great weekend! Yours, Jan
That would be up to @matthias. I could set the API and dashboard up on another server, but that would mean some extra monthly costs. My estimate is about 50 USD per month in server costs to host both NGI and POPREBEL. If given the green light, I could set it up.
Writing this, I actually realize now that I’ve never invoiced for the server that’s been hosting the OpenVillage data, and it’s just been sitting there. That’s the server that I’ve now co-opted to host the NGI API and dashboard. It’s not much, but I should probably invoice ER Core for that soon…
I am very annoyed by this piece of … software that came without instructions and still has eluded me as to installing it on our main server so that everything works. It’s probably just an incompatible version of some other piece of software and I could figure it out in 1-2 days of debugging but honestly don’t think this would be a good investment of time because we want to replace this implementation of Graphryder anyway.
So: Hugi, you’re welcome to install the other two dashboards on the type of server where you got it working.
But I don’t think we have to spend 50 USD extra, as you could try first to install the Neo4J databases (and then also the Graphryder API and dashboard) on the same server. I was happy with ineo for this purpose, allowing to install many Neo4J databases on one host. On the edgeryders.eu host, the commands I used to create two Neo4J databases were as follows: