gravypod

79 ms·

811.
by gravypod 6y ago
I hope this is expanded to other package repositories (Maven, Docker, npm, etc). I'd really like to make it easier for people at edge locations to build my code.
812.
by gravypod 6y ago
> Why not use > file? I'm assuming sponge buffers everything in memory so you don't get concurrent modification bugs.
813.
by gravypod 6y ago
Thanks! I hope to generalize this to other backends as well.
814.
by gravypod 6y ago
Hey everyone! We've started making heavy use of both gRPC and MongoDB at Caper. When we first started using MongoDB we manually created pojo classes and mapped our protos into Java classes before saving/loading from our DB. This w
815.
Show HN: Pronto – Java, Proto, Mongo Connector(github.com)
3 points by gravypod 6y ago|3 comments
816.
by gravypod 6y ago
Shopify has traditionally been an example people have pointed to for scaling a monolith with a large growth factor in all areas: team size, features, user base size, general "scale" of the company. Does anyone on here, who has wor
817.
by gravypod 6y ago
Do you have a breakdown of "Common Reasons For Startup Failure" by total funding amount bracket? Is there a trend?
818.
by gravypod 6y ago
Bazel isn't really designed for dynamic linking applications. You can include system libraries if you need: https://groups.google.com/g/bazel-discuss/c/Ndd820uaq2U You'd also need to make sure every
819.
by gravypod 6y ago
Bazel generates reproducible builds by default and is very cache friendly. It forces hermetic build and test environments; each build/test gets it's own sandbox that only contains and preserves deps and outputs. It's not magi
820.
by gravypod 6y ago
Came here to say just this. It's a marvel of engineering and the slack for the community is really helpful. I hope more people join in the future.
821.
by gravypod 6y ago
If you're using libraries for everything, nothing. If you need to implement network IO or something, you no longer need to use NIO to get NIO performance. You can just do blocking reads and writes in your threadlet and get the same per
822.
by gravypod 6y ago
I don't think this is the case. Last I read up you are personally liable for HIPAA violations. If you do something, knew it wasn't something you should do, and refused to fix it, stuff gets really bad. To stay safe with medical
823.
by gravypod 6y ago
You can setup wildcard DNS names or use `{STAR}.{localhost,local}` for local development. Just use a reverse proxy on `:80` that allows you to define wildcard subdomains like `container_name.{STAR}` or setup the reverse proxy to automatical
824.
by gravypod 6y ago
The main reasons are: 1. My applications have no dynamic "linking" 2. The containers are distroless 3. Services can be built independently 4. I run the same exact sha in dev/staging/prod To do what y
825.
by gravypod 6y ago
Looks like a nice tool. If it could watch for changes, rebuild, repush, and restart pods it would be the perfect dev environment. Even still this looks like a great start. Have you thought about polishing it a little more and doing a Show H
826.
by gravypod 6y ago
Multiple approaches I've seen in the past. My most recent approach is to rely heavily on unit tests and do 99% of your testing there. Before opening a PR/MR/CL into `master` I test my feature out end-to-end using `docker-comp
827.
by gravypod 6y ago
I think it's nuts too. I'd rather work from home and live in the middle of nowhere. 3k/month is a mansion in most parts of the world.
828.
by gravypod 6y ago
You can rent-to-own a 6br house on a 6k+ sqft lot with a detached garage in NJ with a ~1hr commute to Manhattan for $3,500/month in NJ. $3k/month on an apartment gets you a luxury huge 2 to 3 bedroom with in unit w/d and bowl
829.
by gravypod 6y ago
> What I am seeing here is very clear also. The SIP order has removed many of the greatest parts of SF, the food, the bars, the culture, the density to be able to go anywhere and see your friends easily, and whats left are many of the pr
830.
by gravypod 6y ago
Yea, exactly, this. I see all sorts of people making build systems and tools and they say you just do "my_command build my_target && docker-compose up -d" and I can't help but wonder: are these people ONLY developing
831.
by gravypod 6y ago
There's a great talk from Chick-Fil-A about running kube clusters on bare metal [0]. I'm also going to be taking up a similar problem soon-ish and I'm also looking into Container Linux/derivations for doing a lot of the
832.
by gravypod 6y ago
`kind` and `k3d` are great but I don't think their aimed at being a developer friendly cluster from what I understand. They both require a lot of tweaking and setup to make them look like their a managed kube cluster (LB, storage, ingr
833.
by gravypod 6y ago
The solution here is to choose a build system that is language agnostic. The most common of these is docker & docker-compose. There's also the Blaze-likes (Pants, Buck, Bazel, Please) that let you hermetically include tools like no
834.
by gravypod 6y ago
No, other tools do not suffer from what makes make complex. Make is "hard" because of a few things: 1. The syntax is very complex 2. It's just a wrapper around shell command execution 3. There's no single s
835.
by gravypod 6y ago
Makefiles are nice in theory but don't scale well to team members who don't know how to use make. If your team is full of experienced C/C++ devs then this will probably work great for you. If you are working with a bunch of a
836.
by gravypod 6y ago
At my current company, since I'm using gRPC, the build contexts for my containers are annoyingly large and I'm definitely feeling this. It's a massive pain and really annoys me but If I had to choose between 40 minutes of str
837.
by gravypod 6y ago
Depends on what cloud services you use. I am very opposed to using any cloud service that I can't self host in some way or another. That's a sure fire way to end up screwed. The main cloud services I use at work are: 1. S3&#
838.
by gravypod 6y ago
If you are using a managed kubernetes cluster from a cloud provider you mostly don't need to worry about these sorts of things. If you're not, and deploying to bare metal, the main things you need to worry about are: load balancer
839.
by gravypod 6y ago
> Run your applications in Docker from day 1 (with docker-compose it’s as valuable for dev as it is for production) and think carefully before letting your applications store local state. I think this is the key take away for many startu
840.
by gravypod 6y ago
Are there any companies that help you incorporate all across the world with a single button press and manage the employee pay & PTO stuff in accordance with local laws?

More ›