On Sat, Jul 8, 2023 at 3:08 AM Joseph M Hellerstein <hellerst...@berkeley.edu> wrote: > > On Fri, Jul 7, 2023 at 5:23 AM Roman Shaposhnik <ro...@shaposhnik.org> > wrote: > > > On Thu, Jul 6, 2023 at 10:37 PM Joseph M Hellerstein > > <hellerst...@berkeley.edu> wrote: > > > > > > My 2 cents from the peanut gallery would be to prioritize efforts that > > > help grow the community for MADlib on PostgreSQL. From a marketing > > > perspective, the "top of the funnel" is PostgreSQL users, and there are > > > lots more of them to try and reach than any other platform. I worry that > > > excitement about MADlib could languish because of a historical focus on > > > Greenplum. > > > > Huge +1 to that! > > > > > On a related note of growing adoption, it would be great to see more > > > interop for other PostgreSQL-compliant backends, e.g. some of the cloud > > > offerings. This might require finding interested contributors at vendors. > > > If I can be helpful making connections, let me know -- I'm in touch with > > > relevant folks at AWS, Google, Azure, etc. if somebody in the open source > > > community is interested in pursuing those conversations. > > > > Joe, from your vantage point, what are the exciting > > PostgreSQL-compliant backends that you're seeing (especially when it > > comes to startup companies). Things like TimescaleDB and Neon, etc. > > > > In terms of reach, none of the startups compare to the cloud vendors' > in-house offerings. I'd focus energy there.
Agreed! That said, I'm still curious about things like Neon, simply b/c they do seem to care about ML use case as one of their differentiator for thin-cloning: https://neon.tech/docs/introduction/branching#machine-learning But I guess I just need to ask Heikki ;-) > The challenge is that, to my knowledge, none of them support arbitrary C > UDFs. So I'd explore both business and technical solutions to that. On the > business side, one could try to convince vendors to anoint MADlib as an > approved extension. That would likely require some customer advocacy. On > the technical side, the goal would be to get MADlib running using only > approved UDF frameworks. I can see various approaches there that seem > feasible. That's a good point actually! A bit of BD activity would really help... > Again, I'm very much in the peanut gallery here, not having kept a close > eye on MADlib usage or the PostgreSQL ecosystem in recent years. So take my > advice as worth about the cost you paid for it :-) ...not that I personally can help with any of that ;-) so it really echoes your comments here. Thanks, Roman.
