Thank you for documenting and sharing your position on the WASM matter. To me, this post conveys that ~“Dash aims to serve the enterprise user with data-intensive applications.” If that is the case, then I’d argue that there are better ways of doing so, and that there are easier/ better aims in the first place.
Dashboards shouldn’t be data intensive
Sure, data-intensive computation should take place on a server. However, dashboards (any interface really) should fundamentally be serving lightweight (summary data/ results/ 100-row previews) to end users while offloading heavy work to remote systems wherever possible via callbacks. That 1GB example you mention should be pointing to an S3 bucket via a pre-signed URL, not passing through a Dash app. Which brings me to my next point.
Plotly doesn’t have a competitive advantage in data capabilities
Even if data-intensive analysis is the correct strategy, where are Dash Enterprise’s cloud service/ datastore connectors and batch processing capabilities? If you Google “dash enterprise aws batch” or “dash plotly aws lambda” - nothing comes up. The celery-based (very webapp-centric way of thinking) k8s jobs implementation just reiterates the VPC/ on-prem enterprise theme.
Embrace your competitive advantage
Plotly has a unique core competence in that it deeply understands front-end development and is able to delivery intuitive APIs that enable developers to build things they wouldn’t otherwise be capable of. Plotly is better positioned that anyone else in the world to attack the WASM space. To me, you guys are front-end people, not data people. Why? (a) you chose celery, (b) you’re developing low-code features instead of data-focused features, (c) the data table components would be way better if you were, (d) there would be way more data-focused components, (e) you made Plotly+Dash and they are amazing!
Serve & monetize the community you’ve built
People (the massive, non-enterprise Plotly community) want Plotly to provide an easy, reasonably priced way to serve their dashboards into production. Why the hesitance to build a platform to enable the community you have worked so hard to build? Don’t let it go to waste! Streamlit’s acquisition means they will be dragged into enterprise work, which will open up the SaaS dashboard space again. Marketplace; let developers sell subscriptions to apps that are served on your platform and take a cut. Tap into assets of production that you don’t own and stop getting bogged down in 1-1 linear growth enterprise engagements that must be staffed.
WASM should be a way to get Dash apps in the hands of everyone for every use case (without running up Plotly’s cloud costs). Charge app developers based on the amount of static content served or some other API metering. It should be Plotly’s chance to definitely codify/ monopolize what the next 2 decades of realtime app development and the future of the internet looks like.