Release notes for v2.40.57

Overview

This release introduces GPU acceleration to GFQL, Louie.AI security support, and standard security, performance, and bugfix patches.

Graphistry Versions

Server	2.40.57 (was 2.40.53)
JS React+Vanilla	4.6.6
Python PyGraphistry client	0.33.0 (was 0.31.3)

Third-Party Versions

Arrow	10.0.1
Caddy	2.6.4
CUDA (In-Docker)	11.5
Dask	2023.1.1
DGL	1.1.1
Django	3.2
Docker (CE)	23.0.1
Docker Compose	2.17.3
fsspec	2023.1.0
Gremlinpython	3.5.0
Jupyter	1.0.0
Neo4j-python-driver	4.4.5
NodeJS	14.21.3
Pandas	1.5.3
Postgres	14.7
Python	3.8.16
neo4j node driver	4.0, 4.1, 4.4
networkX	2.6.3
notebook	6.5.2
Nvidia driver (cloud)	470
RAPIDS	23.02
Redis	6.2.7
Splunk node SDK	1.9.0
Tornado	6.2
torch	1.12.1
torch-geometric	2.3.1
torch-god	0.3.1

Features

• PyGraphistry:
  • GFQL graph queries now support optional automatic GPU acceleration, which have been benchmarked for up to 100M edges on a 16GB GPUs

Infrastructure:

• Additional OpenTelemetry tracing in streamgl-viz

Admin

• .Louie.AI: See migration notes

Security

• Marketplace: Remove Python2.7, Python 3 from the host-level of all cloud image distributions 

Fixes & tweaks

• UI: Fix categorical encodings issues for large graphs with many unique numeric values
• UI: To maintain browser compatibility, we verified the 'TransformStream' API support (see Chrome 116 fixes); if unsupported, we optimized the response handling instead of using 'TransformStream'. This fixes misshaped streamed graph issues for some interim Firefox releases.

Docs

• PyGraphistry - GFQL: GPU acceleration mode demo notebook

Migration

• Louie.AI: For external redirects, e.g., with Louie servers, users must now explicitly define a comma separated list of trusted external URIs using the ALLOWED_EXTERNAL_URI setting in the admin panel (https://YOUR_HOST/admin/constance/config/)
• Note next release (2.41.X) will be CUDA 12.X
• See previous notes