At or more likely after the openmod in Zurich (2018) a discussion started, if most of the openmod still agree about the aims (or the method) of the Q&A category.
See here for the original mailing list discussion.
Up to now only oemof managed to move the discussion to the common forum and the question arose that this is a subject to the neutrality of the openmod because all the other open models are not that visible for openmod partners.
To further discuss how to handle it in the future we asked 2 questions that you can answer for your models/frameworks in this wiki-post and that can be used as a basis for further discussing the Q&A section.
-
why didn’t most of the models which were part of the openmod by that time open their Q&A to the community on the forum? (@OSSeMOSYS, Calliope, PYPSA, urbs,….)
-
would the newcoming openmodels open their Q&A to the community (on the openmod forum or, if necessary to change place, somewhere else)? (@FINE, GLAES, TSAM, PyOS, AixLib, TEASER, pyam, DPSA.jl, enlopy, MESSAGEix, JuMP (@Robby, thanks for the list!))
what would be your pros and cons?
Answers:
MESSAGEix
When preparing for the open-source MESSAGEix release a few months ago, we opted for a public Google group - the main reason being that it presents the lowest possible entry barrier to new modellers using this as a community forum and message board.
PYPSA
For PyPSA we also have a public Google group. I think one of the main advantages is that it also serves as a place to broadcast announcements to our users (new releases, bug fixes, new features, new publications, new examples - also for our users to broadcast these things to other users). So it’s dual use - technical support and broadcasting. More generally I guess I feel, like Mark put it nicely, that decentralisation is more robust and allows for variety and experimentation. Centralisation is convenient but comes at the price of uniformisation and potential risk when services go down (or get taken over by suspect multinationals :-)). With arXiv.org (for preprints), zenodo.org (for data) and orcid.org (for researcher profiles) I feel a bit more comfortable because they have big academic institutions/libraries backing them up. Here’s some thoughts on decentralised git by the way:
https://medium.com/@alexberegszaszi/mango-git-completely-decentralised-7aef8bcbcfe6
I would see openmod serving more Mark’s “watering hole” purpose, offering platforms for exchange, informing people about each other’s projects and fostering collaboration.
one oemof view
I am an oemof developer and I like the idea of having a central stackexcchange-like forum for open energy models and this was the reason to move the oemof support to the openmod forum.
- For me the border between the models is not so strict and people may combine different models or use different models in different projects.
A joint forum shows the variety but if we are the only framework using the openmod forum as a support forum it does not make sense.
- Sometimes the border between specific and general question is not clear. Installation support is specific, that’s no question but how to model a heat storage can be both.
In my opinion using the forum as a general forum and (!) a support forum will increase the number of users to the benefit of all. People will be aware of all the other models and may stumble over interesting questions besides their main topic.
Altogether I see some advantages in a central forum but most of them have no effect if it is just oemof using the openmod forum in this way.
And if other users feel bothered by the way oemof uses the forum we should move away because this shouldn’t be a reason to create bad mood.
second oemof view
as I initially talked to Robbie and started the oemof-internal discussion, my two cents as one of the oemof-developers:
I.) I see Robbies and Toms arguments regarding the difference between generic and specific questions and mostly agree with them. Having own mailing lists, forums, etc. is common practice in many projects and I would also rather expect a project-specific google group rather than some external forum I don’t know. Furthermore, if I used the software in another (not energy related) context, I would wonder why the support is provided on the openmod-platform.
II.) From my point of view, the questions on how to bring open models closer together and make it easier for users and developers to know about other models are answered by the model factsheets and periodical workshops. This would also be in line with Marks water[ing] hole analogy.