Disscussing and reorganizing our Roadmap

[Thespica]

Since our roadmap haven't been updated for a long time, it will be better to reorganize that.

Any suggestion is wellcome!!! We can discuss here even if it's a draft or you have a cool idea but you do not have enough time to implement it.

Potential proposal from my perspective:

1. support loading GraphAr to HugeGraph through its loader (first), and export to GraphAr (later). I can help with this task but have not time to implement that.
2. support loading GraphAr to cuGraph to make GPU-accelerated graph processing more convenient.

cc @yangxk1 @SemyonSinchenko @SYaoJun

[SemyonSinchenko]

Hi @Thespica !
Let's add the feature @Iskander14yo is working on: sampling API for GCNN training on data, stored in the GAR format.

• Discussion
• Proposal

As well let's add #917, it looks like there is no scalable ways to convert to GAR now.