JuliaClimate / meta

For discussions about JuliaClimate implementations
MIT License
6 stars 1 forks source link

Priorities for discussion & setup for the organization #3

Open gaelforget opened 4 years ago

gaelforget commented 4 years ago

In addition to #1 & #2 there might be other obvious things that you should be discussed in priority or asap.

Candidates may include:

Happy to open issues for each one of of these but seems better to let everyone weigh in on a priority list first. Please consider giving +1 or -1 to individual items below and add more items?

gaelforget commented 4 years ago

Code of conduct

gaelforget commented 4 years ago

Contributing guidelines

gaelforget commented 4 years ago

Logo design (ok for now? compete proposals? volunteers? ideas?)

gaelforget commented 4 years ago

A registry repo (do we want one early on? can wait?)

gaelforget commented 4 years ago

Periodic telecons (how soon? how often? format?)

gaelforget commented 4 years ago

Discourse channels (now? later? how many? names / scope?)

Datseris commented 4 years ago

To justify my votes: I love logos, I've designed four of them by now, most in collaboration by cormullion. Have a look and tell me what you think: if you like them I can try to hack something for here.

https://juliadynamics.github.io/JuliaDynamics/ (6 logos here)

https://github.com/JuliaMusic (the org logo)

I think registries are extremely important, given how many things are now in a highly unstable manner and thus should not be used in serious scientific contexts.

I like the idea of sort, lets say monthly video calls. One to two hours per month seems to be a good medium ground.

For the code of conduit, I never found it important. I think we are welcoming enough as people that we don't have to write it in a piece of paper as if the lawyer of our institute forced us to... (I hope I don't raise too many eyebrows here :D )

For the rest I don/t have plus/minus.

visr commented 4 years ago

Regarding the Discourse channels, since there is already a Geo domain that has this topic description:

This subcategory is for geo-related discussions. This includes using Julia in geosciences, or dealing with geospatial data.

And this domain is not overly crowded, I'd say lets just use that?

In the About the Geo category wiki post I just added JuliaEarth and JuliaClimate as two other relevant organizations, feel free to edit it more.

natgeo-wong commented 4 years ago

Telecons would be good once in a while, I agree! At least we'll get to meet each other all at the same time instead of having all separate convos scattered around, and we'll be able to get everyone on the same page.

Registry repo/website/logos should wait imo until we've gotten more stable versions of the current repos in the organization, then we can go wild with publicity, etc.

hdrake commented 4 years ago

I would love to have some example notebooks that show how the various packages we already have can be used together. I still don't really have a good sense of what all of these packages people are referencing do and definitely have no idea how they fit together (if at all?) and if there are any redundancies (whether good or bad).

natgeo-wong commented 4 years ago

Good point about notebooks. I'll try to cobble together some examples for ClimateEasy and ClimateSatellite over the next week before school starts.

Balinus commented 4 years ago

The idea of notebooks is perhaps a good starting point. We know the package by name but not necessarily used them!

gaelforget commented 4 years ago

I would love to have some example notebooks that show how the various packages we already have can be used together.

I could move https://github.com/gaelforget/GlobalOceanNotebooks to the org -- these are meant to illustrate the use of MeshArrays (C-grids), NCTiles (tiled netcdf), and compute ocean transports to machine precision (OHT, streamfunctions). I am working on additions in the near future, which could include e.g.

The repo scope could also later (or now?) be extended to cover more of the Julia climate stack (see #2) if there is interest in doing that?

Balinus commented 4 years ago

I could move https://github.com/gaelforget/GlobalOceanNotebooks to the org

That we be quite relevant imho!

Alexander-Barth commented 4 years ago

A registry repo (do we want one early on? can wait?)

Do you mean our own JuliaCimate registry, separate from the general registry (https://julialang.github.io/Pkg.jl/v1.1/registries/#Adding-registries-1) ? If, yes, it is unclear what the advantages are. I think that the package there would loose significantly in visibility. The maintenance burden can be quite high too and it is unclear if the github bots also work on a separate registry.

Alexander-Barth commented 4 years ago

Logo design (ok for now? compete proposals? volunteers? ideas?)

What about showing the 3 main components of the climate system: atmosphere, ocean, land reusing the 3 colors of the julia logo?

Balinus commented 4 years ago

Logo design (ok for now? compete proposals? volunteers? ideas?)

What about showing the 3 main components of the climate system: atmosphere, ocean, land reusing the 3 colors of the julia logo?

Anyone with some artist skills?

visr commented 4 years ago

Anyone with some artist skills?

Dare I tag chief logo officer @cormullion to see if he would be up for it?

cormullion commented 4 years ago

I’m always happy to help! I’ll make a start soon. 😂

Balinus commented 4 years ago

I’m always happy to help! I’ll make a start soon. 😂

Awesome! 😄

natgeo-wong commented 4 years ago

15820356001841889792019113767548

Something like that? 😆😆😆 sorry it's horrible

cormullion commented 4 years ago

Thinking of the layers of the earth's climate sounds like a profitable line to take. I quite liked this first sketch:

Screenshot 2020-02-18 at 17 58 13

It might need some tweaks for brightness perhaps - those Julia colors aren't too bright...

Balinus commented 4 years ago

I like the general intention! I'm wondering how possible it is to have the "flow" more dependent on the system? For instance, air is pretty turbulent, ocean less and land is "almost" static. I don't know if this idea is adequate once executed artistically though!

cormullion commented 4 years ago

:) It's always worth remembering that icons and logos are rarely accurate depictions of anything in reality - their job is to be a simple, distinctive, memorable, and eye-catching image that's generally in harmony with the project There's often no literal depiction of any reality - Julia's dots, Apple's apple, Google's er, whatever... At least this is symmetrical at the moment, which is nice. I can play with the parameters if you think it's worth pursuing.

hdrake commented 4 years ago

Could you try rendering it with a few less lines (maybe half)? Looks great!

cormullion commented 4 years ago

No problem::

Screenshot 2020-02-18 at 18 53 04

I'll do more versions tomorrow - it's now dinner time for me...

natgeo-wong commented 4 years ago

Maybe the background could be white (or close to white) instead of dark grey? But loving it!

Edit: I took a vectorstock of a hurricane and recolored it just for fun. Not sure if it's allowed, but I kinda like the end result so just dropping it here.

haha

Original image was taken from vectorstock.

Edit 2: I'm enjoying this too much for my own good, so I made a spinner.

hurricanespinner

briochemc commented 4 years ago

No problem::

Screenshot 2020-02-18 at 18 53 04

I'll do more versions tomorrow - it's now dinner time for me...

What about the same with a white inside background color, but a circle/outline for the edge?

cormullion commented 4 years ago

The colors are slightly less impactful, perhaps... Screenshot 2020-02-19 at 15 59 11

cormullion commented 4 years ago

This one turned out a bit messy - perhaps adding "clouds" wasn't a brilliant idea... Screenshot 2020-02-19 at 16 37 13

cormullion commented 4 years ago

This one's a bit dull - especially compared to a hurricane... :) But simple is good sometimes...?

Screenshot 2020-02-19 at 16 40 00
briochemc commented 4 years ago

This one's a bit dull - especially compared to a hurricane... :) But simple is good sometimes...?

Screenshot 2020-02-19 at 16 40 00

I think this one but with the wavy flow would look really nice!

cormullion commented 4 years ago

some interesting little geometry puzzles, but it's promising...

juliclimate-4

natgeo-wong commented 4 years ago

some interesting little geometry puzzles, but it's promising...

juliclimate-4

This is awesome!!

Balinus commented 4 years ago

some interesting little geometry puzzles, but it's promising...

juliclimate-4

Indeed, nice one!

briochemc commented 4 years ago

@cormullion Awesome job! I'd love to see the same as the last one but with the less organized flow that you had first, with thicker white lines too. I liked it because it hints at some convergence/divergence while the last one looks more organized and IMHO too laminar (not sure this is the correct term here). I apologize if asking all this is a pain for you, I just find them really cool! BTW do you generate these with Julia only? If yes would it be possible to see the code, too?

cormullion commented 4 years ago

No problem. Remember, these aren't diagrams! 😂🤣😅

Screenshot 2020-02-20 at 08 52 33

source: https://gist.github.com/cormullion/b212622d4e92d26596ea800ce7753272

gaelforget commented 4 years ago

Thanks @cormullion for these awesome logo drafts. I kept scrolling down in the discussion thread and getting more and more excited by this! Might just suggest a couple tweaks to the last one above: (1) rm the outer ring and (2) reorder the color. Like this:

juliaclimate-gf1

https://gist.github.com/gaelforget/4744a63b86f74362a71a89d27de44c90

About the colors ordering: I feel this one a bit more natural & loosely representative of the stacking of climate system components. (to me red could thus be thought to represent e.g. earth / geology / mantle, green as land or ocean ecosystems, blue as either ocean or atmosphere, purple as the night sky / space)

About the outer ring: I find the logo more aestetically pleasing without it, and I like to think of earth as an open system (there is no lid between atmosphere and space; eg material or radiation can go in and out)

cormullion commented 4 years ago

Looks good to me!

Balinus commented 4 years ago

Awesome!

gaelforget commented 4 years ago

I would love to have some example notebooks that show how the various packages we already have can be used together. I still don't really have a good sense of what all of these packages people are referencing do and definitely have no idea how they fit together (if at all?) and if there are any redundancies (whether good or bad).

Please take a look at GlobalOceanNotebooks. For now it's centered on packages that I know best but hopefully this is already a useful start. Should work in binder except for the last two in OceanTransports/. All work fine with Julia 1.3 on my Mac but would be great if someone could try rerunning the notebooks. Please submit issues to the notebook repo for potential feedback, questions, or maybe even to offer to contribute additional notebooks.

I hope to keep moving this notebook folder forward over coming weeks probably with a focus on stuff I have been doing with databasis frameworks (e.g. erdapp, Pangeo), analyses frameworks (e.g. climate tools , esdl), and numerical models (e.g. climate tasks, mitgcmtools)

gaelforget commented 4 years ago

Since it seems that we have broad agreement on the logo design I will swap it in place of the placeholder enso plot. Thanks to all for such efficiency!

Later today I will try to send a PR to the meta repo incl. png, gist link, & credit to @cormullion + @juliaclimate community for the design (in meta/readme.md)

cormullion commented 4 years ago

definitely a team effort - #teamjuliaclimate!

gaelforget commented 4 years ago

Please take a look at GlobalOceanNotebooks

Quick update. Did more work on this recently to streamline the following notebooks which I just ran without issue on Binder :

and

The rest of DataStructures/ is expected to change more soon...