Provide firm date on multicore/thread support

Started by Seimus, June 27, 2024, 06:37:46 PM

Previous topic - Next topic
Honestly you can not compare Oracle to ZA, Oracle is literary keeping companies hostage due to their own fault.

Anyway,

If its really true, that multi-core support is behind a paywall than I am personally done with ZA and not going to renew my licenses/subs.

When I discussed the payed subscriptions with ZA sales few years ago, it was told that the multi-core support will come that year (I was already skeptic then), it didn't. There was no mention that such a base feature will be behind a higher tier paywall.

Honestly,
1. This is just anti-consumer as much as possible
2. Not only that ZA already creates a bottleneck for Intranet communication, now there is a possibility that you have to pay extra even if your HW is totally capable without ZA
3. You are literary alienating home-labers and selfhosters, that mostly help you to report, identify and tshoot BUGs


Last but not least, ZA, if you want to go the corporate way, by all means feel free. BUT DO NOT backstab the community that partially helped you and is helping you to create better product.


I hope ZA team will provide an official statement, because this is by all means a real concern.

Regards,
S.
Networking is love. You may hate it, but in the end, you always come back to it.

OPNSense HW
APU2D2 - deceased
N5105 - i226-V | Patriot 2x8G 3200 DDR4 | L 790 512G - VM HA(SOON)
N100   - i226-V | Crucial 16G  4800 DDR5 | S 980 500G - PROD

Comparing Oracle to a packet filter? Sounds like an ZA executive. Multicore support in this day and age should be a base feature. Oracle is relic and has been around for age. I'm fairly confident that there license model was crafted during a time where multicore was a commodity. In this and ages it isn't. Smart watches have multiple cores.

Well I can see the comparison to Oracle. They and others moved a while ago to per-core licensing to extract more from customers as the core counts started multiplying and they realised customers were getting more processing from each socket. I get the reasoning.
BUT, I see it just like Seimus. Alienates the prosumers that gave you community beta testing.
Want to get big enterprises pay more for your product, of course. Want to keep your lowly homelabbers and prosumers, you should.
And I reckon you can do both. Similar to your pricing structure for free and home, make it free up to X threads/cores.

So, Oracle example was was from "A Company" and "Multicore" perspective regardless of what product performs.

Now back to ZA, I agree with you @Seimus on all counts but to be the devil's advocate here...
what's it for a company to buy a DECISO's BEEFY box and throw in the ZA and run with it for free... Yeah, I do understand the Mom'n'Pops shops should be considered home users as the restrictions in budgets'n all... 

BUT! ZA has to make moola! So, from that perspective they could Cap the multi to a session rate or what not so that the grandma in Norway with 400Gbps pipe would have to cough up USD to satisfy her 8K TV. ROFL!

We may be witnessing a Broadcom moment! (vmware).

Or an IBM moment (CentOS).

Or an Oracle moment (Open Solaris).

But all of those are big, they can float through angry people dropping their licenses...

Just another example, Deciso made the OPNsense Business level cheap by comparison, their hardware competes with similar enterprise level hardware (at least what I'm looking at). And yet they still have the free open source version with really no crippling to the features like other "freeware". It might be worth making ZA cheaper to see if sales volume increases and surpasses current levels.

In the end, make your choices, my fourth example seems to be working and no body complaining. But if you decide to be like the other three, grab it and own it and be prepared for the backlash. Broadcom is taking a beating right now by all the people that used to buy a yearly license (myself included), now it's all locked up and a mess. They've lost a lot of young talent who will not be willing to use the product when they get big jobs, the old talent retires and no one is left to handle some of these massive systems. And when most companies figure out how to move away, they are gone for good (the Broadcom "squeeze all you can from it, and dump it" plan).