|
xzzy posted:The red hat model doesn't even work for redhat, that's why they're killing off free access to RHEL rpm source. I am generally not a fan of calling a big exit a “success”, but I think $38 billion is big enough to count regardless of what happens afterwards.
|
# ? Aug 18, 2023 00:14 |
|
|
# ? Jun 8, 2024 06:49 |
|
My company is cncf so we can never change our license.
|
# ? Aug 19, 2023 21:34 |
|
12 rats tied together posted:If you have some system by which you manage conditions outside of CloudFormation (any programming lanuage with a json library, terraform, ansible) the only thing StackSets really solves for you is "for account in accounts", which was already the easiest part of this problem.
|
# ? Aug 20, 2023 19:10 |
|
Also, re: OSS-ish companies trying to make money chat. The elephant in the room is that most of these companies are essentially all competing against hyper scalers that basically print money and capture markets via rent mechanisms, and if they are screwed on their stock for whatever reason they potentially become takeover targets that destroy these OSS ecosystems permanently. See what happened to Puppetlabs, Chef, Saltstack, and even Docker for all these projects-turned-companies that just can't seem to stay stable and force industry trends. In fact, one of the few stable-ish for 8+ years projects in this space now is Kubernetes, which is basically subsidized by the massive profit margins of rent-seeking type companies and also fuels Peter Thiel's assertions in support of monopolies for tech, which is sadly becoming more and more true than I'd like it to be. To put it another way, K8s has been basically on top for longer than Puppet and Chef ever even existed possibly, which is kind of a scary thought and also reminds me way too much of how old I'm getting and just makes me depressed. Of the various OSS projects that are still hobbling along and doing solid technical work without an encumbrance of tons of corporate fiefdom-seeking measures off the top of my head there's the Linux kernel and Postgres. Everything else is basically a ward of a hyperscaler, absolute dickwad of a corporation, or a desperately fighting for survival formerly Apache-licensed company (MongoDB, Hashicorp, Elastic, etc.).
|
# ? Aug 20, 2023 19:20 |
|
Docker was just exceptionally poorly managed, I can't think of a worse managed high profile company in recent years. They tanked before the #metoo movement but had they survived they probably would have been hit with some of that as well
|
# ? Aug 20, 2023 20:46 |
|
Hadlock posted:Docker was just exceptionally poorly managed, I can't think of a worse managed high profile company in recent years. They tanked before the #metoo movement but had they survived they probably would have been hit with some of that as well Hashicorp. Trust me. They’re worse. E: I interviewed there. First off they wanted someone in pst and I said I love working those hours so I’m good. They interviewed a bunch of other people and came back to me. Then they wanted me to list the reasons I left all my other jobs. None of your loving business. Then quiet for 2 weeks until I get a email asking me salary requirements. At that point I just told them to gently caress off. jaegerx fucked around with this message at 21:09 on Aug 20, 2023 |
# ? Aug 20, 2023 20:58 |
|
jaegerx posted:Hashicorp. Trust me. They’re worse. FWIW, the last time I got one of those "reasons you left all your other jobs" things from a different company, I filled each one with an identical response of "offered better opportunity elsewhere" and they did not ask any other follow-ups before getting to offer stage Vulture Culture fucked around with this message at 16:02 on Aug 21, 2023 |
# ? Aug 21, 2023 15:59 |
|
xzzy posted:The red hat model doesn't even work for redhat, that's why they're killing off free access to RHEL rpm source.
|
# ? Aug 21, 2023 16:14 |
|
re: hashicorp, I just got out of a meeting where I got to yell at the azurerm product manager for lack of qa
|
# ? Aug 21, 2023 17:16 |
|
The Fool posted:re: hashicorp, I just got out of a meeting where I got to yell at the azurerm product manager for lack of qa When is your next meeting, continue yelling at them
|
# ? Aug 22, 2023 02:21 |
|
I, too, shall join in a video conference and yell at them free of charge if it pleases anyone working on a SOW to do just that.
|
# ? Aug 22, 2023 02:32 |
|
SeaborneClink posted:When is your next meeting, continue yelling at them Wednesday, and I intend to
|
# ? Aug 22, 2023 02:36 |
|
Wednesday will be our third meeting on this issue in 4 business days and I am so tired but am also so annoyed that I'm not going to let it drop
|
# ? Aug 22, 2023 02:38 |
|
necrobobsledder posted:Also, re: OSS-ish companies trying to make money chat. What do you mean by hyperscalers? What's Thiel's support for tech company monopolies? jaegerx posted:Hashicorp. Trust me. They’re worse. I could see something like someone applying for a senior executive at a bank or something getting asked why he left their lost job for but typical engineers? That's really weird.
|
# ? Aug 22, 2023 10:15 |
|
The Fool posted:re: hashicorp, I just got out of a meeting where I got to yell at the azurerm product manager for lack of qa Keep yelling. You have my support.
|
# ? Aug 22, 2023 10:15 |
|
hashicorp has a lot of faults, but azure being poo poo to work with isn't theirs
|
# ? Aug 22, 2023 10:24 |
|
Crosby B. Alfred posted:I could see something like someone applying for a senior executive at a bank or something getting asked why he left their lost job for but typical engineers? That's really weird. It's an rear end in a top hat filtering question. Most of the time, you just need to be able to to say anything at all without starting a rant about your old company or the people you worked with.
|
# ? Aug 22, 2023 11:01 |
Crosby B. Alfred posted:What do you mean by hyperscalers? What's Thiel's support for tech company monopolies? Hyperscalers is a bizarre term people who are obsessed with running their own datacenters or work with too many Expedients use to describe the three big public cloud providers
|
|
# ? Aug 22, 2023 12:16 |
|
management uses that term exclusively and i hate it
|
# ? Aug 22, 2023 14:11 |
The first time I heard it was Expedient marketing materials circa 2017. Not sure if they're really responsible for the term, but I absolutely blame them and other datacenters trying to position their bullshit "IaaS" services against the cloud providers as though VPN access to a VMWare tenant of some kind is an equivalent service (not to mention when they go whole hog and use VMWare Cloud Director and pitch it as a "cloud-equivalent" solution because the barely functional API can be used ). Once I hear someone say hyperscaler I tune them out entirely. Quit drinking the koolaid and use the right terms ffs
|
|
# ? Aug 22, 2023 14:33 |
|
hyperscalers are the 'real' clouds. they used to just be the clouds but then linode and every dipshit with vmware started calling their poo poo a cloud so there needed to be a new word for the ones that operate at 100 - 1000x the scale of everyone else.
|
# ? Aug 22, 2023 15:50 |
I just say cloud, no one I work with mistakes that for Digital Ocean. I do literally only work on public cloud poo poo so might be some… I dunno confirmation bias on my part
|
|
# ? Aug 22, 2023 15:52 |
|
vanity slug posted:hashicorp has a lot of faults, but azure being poo poo to work with isn't theirs The CLI was this ridiculously heavyweight Node.js thing too, so if you tried to parallelize it more than a couple dozen times your whole system would fall over Absolutely hostile experience, I'm sure it's gotten a lot better
|
# ? Aug 22, 2023 15:55 |
When was that, 2016?
|
|
# ? Aug 22, 2023 15:56 |
|
i am a moron posted:When was that, 2016?
|
# ? Aug 22, 2023 17:44 |
|
Hadlock posted:Docker was just exceptionally poorly managed, I can't think of a worse managed high profile company in recent years. They tanked before the #metoo movement but had they survived they probably would have been hit with some of that as well its pretty incredible the amount of money and opportunity docker had and they just... did nothing with it.
|
# ? Aug 22, 2023 17:59 |
|
Mr. Crow posted:its pretty incredible the amount of money and opportunity docker had and they just... did nothing with it. I wonder if they could have built the dev ecosystem in a more productive direction if they had bought a company like Parallels and owned the VM tech they needed, but I absolutely don't see those company cultures getting along at that time.
|
# ? Aug 22, 2023 18:23 |
Vulture Culture posted:Oh yeah. It was real rusty, the console would randomly bounce you between Service Manager and Resource Manager, and I'm pretty sure cloud-init wasn't even supported on ARM templates yet. I couldn't believe they released it like that, but it would have been way worse not to. It’s come a long way. Azure powershell and Az CLI aren’t perfect, but they are very robust and easy to use regardless of end user OS. Az CLI even has a badass terminal program that can autocomplete resource names and makes on the spot use of CLI easy. Any years where Azure Classic still appreciably existed were uh… rough
|
|
# ? Aug 22, 2023 18:26 |
|
Vulture Culture posted:In the early years, they had an absolutely impossible market position if they wanted to be anything other than a services company. They were clearly sitting on something that was able to be the hub of multiple entire ecosystems, there was a huge amount of motion in the container space, and it wasn't clear where they were supposed to provide leadership vs. where they were supposed to let other software fill in the gaps. Their play at developer desktops relied on immature virtualization tech they didn't own and couldn't control. The cool hip startups where they gained traction had no money to pay them. Their aim at enterprise services was constrained by a dependency on Linux at a time when many companies were only adopting new management tools with cross-platform support, and middle-of-the-curve businesses had just virtualized their datacenters. That left HPC and Big Data, which they went after aggressively enough to hire the CEO of Gluster, but that space was too caught up in revolutionary changes in the data landscape to invest much time into swinging their grid computing effort into containers. i mean, google offered them borg and they said they weren't interested and docker swarm was the way and the light so google kicked it to cncf, right? that's how i recall it (hilariously) going down
|
# ? Aug 25, 2023 06:48 |
|
it is always going to astound me that they had a technology which was adopted more rapidly than fire and they could not figure out how to monetize it
|
# ? Aug 25, 2023 06:50 |
|
Junkiebev posted:i mean, google offered them borg and they said they weren't interested and docker swarm was the way and the light so google kicked it to cncf, right? Kubernetes was a complete disaster to operate at this point—I think etcd hadn't even hit 2.0 yet. It was the development of the partner ecosystem that made it viable for businesses that aren't trying to run at Google scale. poo poo, remember Kubespray? Vulture Culture fucked around with this message at 16:20 on Aug 25, 2023 |
# ? Aug 25, 2023 16:16 |
|
The Fool posted:re: hashicorp, I just got out of a meeting where I got to yell at the azurerm product manager for lack of qa direct result of my rabble rousing the azurerm changelog posted:The field win32_status has been replaced by win32_status_code the fact that this is the solution they went with is hilarious to me
|
# ? Aug 25, 2023 16:23 |
|
Vulture Culture posted:I don't think so. There was less than a year between dotCloud renaming to Docker (October 2013) and Google welcoming Docker to the Kubernetes community (July 2014), itself barely a month after the first GitHub commit for Kubernetes i mean, the entire container orchestration landscape was crazy in that era - remember Mesos/DCOS? Cattle? Cluster Suite? it was the wild west!
|
# ? Aug 25, 2023 18:06 |
|
The Fool posted:direct result of my rabble rousing Did they change the name of an existing field in a non-major release or something here? You mentioned it was a field that you didn’t even use which implies this is an output, and still requires state surgery on hundreds of statefiles. What on earth happened here?
|
# ? Aug 25, 2023 18:13 |
|
The original issue was this breaking change in a non-major release:quote:the win32_status property of the status_code block in auto_heal has changed from string to int win32_status is an optional property, but if you use auto_heal and don't set it, it will get set to "" by default If you had a web_app resource deployed prior to this version, when you tried to use this new version your plan would fail because state validation would see the "" on a value that was supposed to be an int The only solutions they initially offered us were to manually touch state for ~600 workspaces (some have multiple web_apps in them) or to redeploy all of our affected web_app instances.
|
# ? Aug 25, 2023 18:39 |
|
Junkiebev posted:i mean, the entire container orchestration landscape was crazy in that era - remember Mesos/DCOS? Cattle? Cluster Suite? Mesos is the only thing keeping Twitter running right now.
|
# ? Aug 25, 2023 21:56 |
|
Crosby B. Alfred posted:What's Thiel's support for tech company monopolies?
|
# ? Aug 26, 2023 06:10 |
|
what's the current docker alternative? podman?
|
# ? Aug 26, 2023 16:08 |
|
bobmarleysghost posted:what's the current docker alternative? podman?
|
# ? Aug 26, 2023 17:12 |
|
|
# ? Jun 8, 2024 06:49 |
|
bobmarleysghost posted:what's the current docker alternative? podman? The alternative that is pretty much a 1:1 to the docker client is nerdctl https://github.com/containerd/nerdctl
|
# ? Aug 26, 2023 18:17 |