“Developers, fed up with cobbling together proprietary or open source solutions themselves, increasingly are comfortable letting AWS, Microsoft Azure, or Google Cloud remove that complexity for them.”
Well, there it is right there. You now know how to fight back. Do what they’re doing with open-source stuff that’s both self- and cloud-hosted. Boom. It’s what IBM did with AS/400’s that were integrated, all-you-need, fire-and-forget kind of machines that ran alone at branches or tied into their mainframes at centralized locations. Net Integrators did something similar for UNIX with their Nitix appliances with all-in-one FOSS behind pretty interface with high-availability at about $1,000-2,000 more than Windows alternative. Acquired by IBM & rolled into Lotus offerings. Smart defense from the AS/400 company… (rolls eyes) Later I saw small and large vendors encapsulating popular software in virtual appliances for VMware pre-configured or including configuration scripts. Then there’s the container craze that’s nothing like that but mentions similar advantages and disadvantages.
Anyway, it’s pretty simple. Those proprietary offers that made either proprietary or FOSS software seemless in a turnkey way worked. FOSS advocates need to build and sell more like that out of their stuff done [in a practical version of] their way. End of story. Or, rather, the beginning of a better one. Maybe. :)
Anyway, it’s pretty simple. Those proprietary offers that made either proprietary or FOSS software seemless in a turnkey way worked. FOSS advocates need to build and sell more like that out of their stuff done [in a practical version of] their way. End of story. Or, rather, the beginning of a better one. Maybe. :)
I don’t really understand your response. No matter how easy one makes a FOSS database, they cannot integrate it into the Azure or Amazon interface, so there will always be a bit of friction there. So your response of “it’s simple, just make it simpler!” doesn’t seem to understand the problem, as far as I can tell.
“Companies like Amazon offer a way around traditional procurement vehicles, minus the headache of assembly. As RedMonk analyst Stephen O'Grady has asked: "If you select Nginx, Kubernetes, Docker, Chef, and MongoDB, for example, what assurances do you have that these all work reliably together?” On their own, the answer is “none.” But in the cloud, the answer changes. "
The article claims that the cloud offerings are combinations of tech that pre-proven to work, easy to acquire, and very convenient to use. There’s basically nothing to setup, deploy, and manage. That, not the “interface” or being in cloud itself, is the actual selling point. The interface is just how they use that convenient service. The same selling point has been going on for decades with AS/400’s where they’re an integrated bunch of components you set-up once & then just use. Admins sometimes forget they’re even there they’re so seemless & reliable. There were also appliances, physical & virtual, that were similar. Plug them in (copy if virtual), turn them on, quick setup, and you’re good to go. The databases listed in the article don’t default on that kind of convenience, though. The burden is all on the buyer. They eliminate that burden like cloud offers do and the uptake might increase dramatically.
Quick example: I remember Sentinel’s HYDRA firewall in high-security field also aimed for usability. Rare for such stuff. It was a web & web-app server with protections against about every web application attack built-in all running on an embedded RTOS to reduce attack surface or bugs. Quite a few security appliances for web protection but this came on a PCI card that took 5 minutes to get running through easy GUI after plugging it in. How long does a similar stack of FOSS components take to install, initialize, and secure? And does it involve easy GUI dialogs or arcane commands you hunt for with Google? If it’s not 5 minutes by average admin, then there’s a market opportunity there, too. :)
In my experience (and as far as I can tell the author of this post has not done any research per se, either), the idea that just making solutions easier to setup by hand will allow it to compete with cloud offers just isn’t true as far as I’ve seen. Many people I meet just want to click a button in an interface and get their system. A lot of stuff in the cloud is worse than you could do by hand. DynamoDB, for example, you can easily iops yourself out of uptime. But the convenience is what many people I interact with want, even if the solution isn’t as good. Heck, they are going to be bought by Google or Apple anyways so that’s a tomorrow problem :)
What you say sounds right. Im thinking more they’d compete with stuff cloud is competing with. They might aldo get traction in groups that like cloud convenience but not 3rd parties having the data. There’s also customers that prefer tried and true stuff that changes slowly. Lastly, it’s sometimes cheaper or higher performance to go with dedicated hardware for a certain function.
My appliance concept could get market share in these areas. People wanting a point and click solution with no concerns about 3rd partg risk past a ailability would go with cloud over my recommendation in a heartbeat. Still leaves a good chunk of market, though.
“Developers, fed up with cobbling together proprietary or open source solutions themselves, increasingly are comfortable letting AWS, Microsoft Azure, or Google Cloud remove that complexity for them.”
Well, there it is right there. You now know how to fight back. Do what they’re doing with open-source stuff that’s both self- and cloud-hosted. Boom. It’s what IBM did with AS/400’s that were integrated, all-you-need, fire-and-forget kind of machines that ran alone at branches or tied into their mainframes at centralized locations. Net Integrators did something similar for UNIX with their Nitix appliances with all-in-one FOSS behind pretty interface with high-availability at about $1,000-2,000 more than Windows alternative. Acquired by IBM & rolled into Lotus offerings. Smart defense from the AS/400 company… (rolls eyes) Later I saw small and large vendors encapsulating popular software in virtual appliances for VMware pre-configured or including configuration scripts. Then there’s the container craze that’s nothing like that but mentions similar advantages and disadvantages.
Anyway, it’s pretty simple. Those proprietary offers that made either proprietary or FOSS software seemless in a turnkey way worked. FOSS advocates need to build and sell more like that out of their stuff done [in a practical version of] their way. End of story. Or, rather, the beginning of a better one. Maybe. :)
I don’t really understand your response. No matter how easy one makes a FOSS database, they cannot integrate it into the Azure or Amazon interface, so there will always be a bit of friction there. So your response of “it’s simple, just make it simpler!” doesn’t seem to understand the problem, as far as I can tell.
It’s in the article:
“Companies like Amazon offer a way around traditional procurement vehicles, minus the headache of assembly. As RedMonk analyst Stephen O'Grady has asked: "If you select Nginx, Kubernetes, Docker, Chef, and MongoDB, for example, what assurances do you have that these all work reliably together?” On their own, the answer is “none.” But in the cloud, the answer changes. "
The article claims that the cloud offerings are combinations of tech that pre-proven to work, easy to acquire, and very convenient to use. There’s basically nothing to setup, deploy, and manage. That, not the “interface” or being in cloud itself, is the actual selling point. The interface is just how they use that convenient service. The same selling point has been going on for decades with AS/400’s where they’re an integrated bunch of components you set-up once & then just use. Admins sometimes forget they’re even there they’re so seemless & reliable. There were also appliances, physical & virtual, that were similar. Plug them in (copy if virtual), turn them on, quick setup, and you’re good to go. The databases listed in the article don’t default on that kind of convenience, though. The burden is all on the buyer. They eliminate that burden like cloud offers do and the uptake might increase dramatically.
Quick example: I remember Sentinel’s HYDRA firewall in high-security field also aimed for usability. Rare for such stuff. It was a web & web-app server with protections against about every web application attack built-in all running on an embedded RTOS to reduce attack surface or bugs. Quite a few security appliances for web protection but this came on a PCI card that took 5 minutes to get running through easy GUI after plugging it in. How long does a similar stack of FOSS components take to install, initialize, and secure? And does it involve easy GUI dialogs or arcane commands you hunt for with Google? If it’s not 5 minutes by average admin, then there’s a market opportunity there, too. :)
In my experience (and as far as I can tell the author of this post has not done any research per se, either), the idea that just making solutions easier to setup by hand will allow it to compete with cloud offers just isn’t true as far as I’ve seen. Many people I meet just want to click a button in an interface and get their system. A lot of stuff in the cloud is worse than you could do by hand. DynamoDB, for example, you can easily iops yourself out of uptime. But the convenience is what many people I interact with want, even if the solution isn’t as good. Heck, they are going to be bought by Google or Apple anyways so that’s a tomorrow problem :)
What you say sounds right. Im thinking more they’d compete with stuff cloud is competing with. They might aldo get traction in groups that like cloud convenience but not 3rd parties having the data. There’s also customers that prefer tried and true stuff that changes slowly. Lastly, it’s sometimes cheaper or higher performance to go with dedicated hardware for a certain function.
My appliance concept could get market share in these areas. People wanting a point and click solution with no concerns about 3rd partg risk past a ailability would go with cloud over my recommendation in a heartbeat. Still leaves a good chunk of market, though.