4 ms·
This makes much more sense if you think of Docker as two distinct companies that were awkwardly sharing the same name and corporate entity, and are now being se
by zapita 7y ago
This makes much more sense if you think of Docker as two distinct companies that were awkwardly sharing the same name and corporate entity, and are now being separated.
One makes developer tools, has a huge developer brand and community. It does not generate revenue except for Docker Hub which probably barely pays for itself.
The other sells enterprise products competing directly with Red Hat and Vmware, and indirectly with the big cloud providers. It generates meaningful revenue, but probably flat growth, which considering the huge amounts of VC money invested, makes it a failed business.
The investors probably decided that 1) Docker developer tools and brand still have potential, but 2) the enterprise business failed to deliver, so 3) they are jettisoning the latter and recapitalizing the former- essentially starting over.
- partisancat 7y agoI agree that docker never got branding right. but the whole thing makes even less sense when you see this nugget: "Mirantis will keep the Docker Enterprise brand alive, though, which will surely not create any confusion."
- zapita 7y agoSince the author of the article was too busy trying to be funny, I will do his job for him and explain why this makes sense. Docker needs to keep its brand, obviously, or they have nothing to build a new business from. But Mirantis is buying a business line called “Docker Enterprise”. That name can’t disappear overnight, it would be immensely confusing to customers, and impossible to pull off operationally. What you do in these situations is make an exception to the trademark exclusivity. Usually with a time limit. So, for example, Mirantis would have 2 years until they have to stop using the name “Docker enterprise”. On the one hand that is confusing to analysts and journalists; on the other hand, developers don’t care about the enterprise product anyway, so to them it won’t be confusing. And this was probably the least messy option.
- tpetry 7y agoDocker naming was all the time confusing. Think on Docker swarm which are 2 (3?) products.
- Patrick_Devine 7y agoTwo products. "Classic" Swarm, and SwarmKit. SwamKit is the thing which we think of as trying to compete with Kubernetes with a declarative model and is what people think of as being "Swarm". Very few people still use "Classic" Swarm, although it does still work.
- zapita 7y agoYes, Docker naming has often been confusing. No argument there.
- rolltiide 7y ago> that were awkwardly sharing the same name ah like Python 2 and Python 3!
- rhacker 7y agoDocker hub probably makes no money. We pay monthly for it, but could easily switch to AWS ECR at this point with a cheaper monthly price. The Hub is nice in that it essentially has unlimited storage, but I suspect this might be an internal struggle for the company. We have no reason to go in and clean up 2 year old images that are 1GB each. We generate images on each commit - trying to do as much as DIFF reductions as possible, but if we change package.json - boom, puppeteer + all over again. And that's quite often.
- mprovost 7y agoBut if they're using a deduplicating storage backend, they can shrink down the actual storage on disk across all customers to something quite reasonable.
- dharmab 7y agoThat only works if Dockerfiles are layered in such a way to maximize caching (many are not and naively pack many changes into a single layer). Yes there are block-level dedupes but that's a transparent feature of most storage systems and not a competitive distinguisher of Hub vs ECR/ACR/Quay.
- grosswait 7y agoI believe the person you replied to was commenting on the advantages to Docker, not the customer, and how (blocklevel is what I inferred) dedupe makes the quantity digestible to Docker's pocketbook.
- perlgeek 7y agoI'm glad somebody invested in Docker again, but I'd like to understand why. What's their business model, once you remove the parts that have been making money in the past?
- zapita 7y agoI don’t think they are betting on an existing source of revenue. They are betting that there are enough users of the free product, and a strong enough brand, that it’s worth trying to build a new business on top of those assets.