In writing this up - https://simonwillison.net/2024/Jul/30/aws-codecommit-quietly... - I found out about a much more significant deprecation: Amazon QLDB (Quantum Ledger Database - a blockchain-ish thing they launched in GA in 2019) is being hard-deprecated too! They're shutting it down completely on 31st July 2025, having announced the shutdown a few weeks ago (on July 18th).
Not surprised. CodeCommit was released alongside a stable of other mediocre tools for CI/CD (CodePipeline and CodeDeploy if I recall correctly) that reflected the pinnacle of AWS's mid-to-late 2010s attitude, which is to find something popular and offer an incredibly mediocre alternative to it that will still be used by those teams who want to move as much as possible to AWS. Seems like that stalled out a bit, mostly due to it being so insanely bad that even the most dedicated AWS fanatics didn't bite. I feel like a lot of the recent stuff is just drop-in replacement AWS alternatives to popular tools (like Kafka and Cassandra) with outrageous price tags.
>be used by those teams who want to move as much as possible to AWS
By teams that have been mandated to move as much as possible to AWS by their company's senior leadership, because it simplified accounts/they negotiated a "great contract" etc.
Doesn't it reflect more on the quality of the people involved? These tools came out at the same time as a bunch of other AWS tools that are still used. The platform itself, the company around these teams was the same, so why did Code Commit suck but other AWS products from the same vintage turned out great?
> find something popular and offer an incredibly mediocre alternative to it that will still be used by those teams who want to move as much as possible to AWS.
Any insights on _why_ they turned out to be mediocre? Is it because of them being released as MVP and staying that way for many years or just a lack of interest afterwards?
> If you would like to use an alternative to AWS CodeCommit given this news, we recommend using GitLab, GitHub, or another third party source provider of your choice. We have written a blog which describes how to migrate your repository to one of these other solutions.
I wonder how long AWS will keep Code Commit running for their customers who are already using it? I'm guessing many, many years.
Weird that there's no announcement anywhere (that I can find) about CodeCommit ceasing to onboard new customers. Apparently it happened on June 6th but this forum post from July 26th is the only thing that comes up in search.
The worst thing about CodeCommit is that repo URLs are not unique, Depending on the AWS account you log in to, you can have a different repo with the same url.
Second worst thing is that it basically refuses to show a diff for any file longer than a few lines.
Third worst thing is you need a login helper and special generated credential to create a login.
Fourth worst thing is the absolute slowness of it. Good riddance.
I once had high hopes Amazon would be able to build a real GitHub competitor. It's a real shame Amazon doesn't seem to be able to build anything that reaches beyond their walled garden.
I've tried all of these tools over the last few years as part of vendor due diligence and none of them impressed me. Code Deploy is probably the best of the bunch and it's really just a very opinionated deployment tool that only works on specific use cases well.
CodePipeline is another good example of just horrible user experience. You need a lot of hours of training to figure out all the different permissions required to get it all to work. It is a mess.
I use GitHub Codespaces now for much of my deployment automation.
Not surprising the ui is not very friendly and getting code in and out when you are not inside the AWS walled garden (need AWS keys) really restricted it
clones a repository named "foobar" in the namespace of the currently logged-in user. So it can give you completely different results with different credentials.
While this isn't surprising considering how bad it was, it is surprising considering that they recently announced the whole Amazon Q for software development thing. Who's going to trust them with Dev tooling now?
AWS CodePipeline + CodeDeploy are horrors¹ from what I saw at a client.
They had mandated .Net + Javascript for all development, GitHub for versioning, and AWS for all other cloud services.
The different projects they had going were not complex or huge.
Yet it took months to get it up and running.
The last month at least they paid for AWS specialists to come in and
set it up and even they spent weeks.
Throughout it all, any notion of trying a different CI/CD stack were
rudely dismissed.
Once it was up and running nobody dared touch the pipeline
again.
From all the AWS services that the customer used nothing
was ever comparable to the horrors of their CI/CD.
Setting it up with Azure's offerings would have been damned near trivial.
I have however not used those in production so I do not have the
experience to able to say it is a better solution over all.
¹ In fairness, AWS had just recently released the CI/CD offerings
and things may be a lot better now. I havent look at it again
since then.
Are they shutting down CodeCatalyst as well? Or are they just forcing new customers over to CodeCatalyst? It’s weird (for Amazon) they recommend Github and Gitlab over an AWS alternative if they are shifting people over to CodeCatalyst
This is very bad, and the silent deprecation of Amazon QLDB even more shocking. It looks like at AWS the MBA's are fully in charge...
Many in this thread commenting on CodeCommit not being GitHub or GitLab are missing the point.
I don't want to have my companies code with GitHub/Microsoft and being used to train their AI. Also don't want to have to rely on a third party like GitLab, that is a company who makes no money, and whose losses are $55 million dollars a quarter,
and has shaky internal technical governance. Did not forget about their Prod database one man setup....
I don't care about the lack of features of CodeCommit. It's usefulness was essentially providing a managed git server. Did not need more, but needed it as a managed server in the AWS Cloud.
The alternatives now will be the third parties, or the additional effort of running EC2 instances and managing the resilience and architecture.
This does not predict a good future for tools like Amazon CodeGuru Reviewer for Java and Python, CodeArtifact, CodePipeline and CodeBuild. And even if these plus QLDB were probably money losers for AWS, the MBA's are missing the point.
I would not be surprised if these silent decisions, are not reversed shortly, or if the service is just kept forever as is, but not deprecated.
To me it looks like a huge hack (download your code as a zip file just to upload it to CodePipeline) rather than a solution - I'll not be surprised to discover other migtations to look similarly
AWS had and probably still has the least interesting and immature way of investing in new products of any organization with the resources it has. Apart from the early wave of AWS products virtually everything has been a dud. There are so many random AWS products that are live but basically dead because Amazon/AWS never invested enough into them. Totally predictable outcome. I’m not sure why Bezos made Andy Jassy CEO.
Their business models has always been basically MS's strategy: Build out a massive sales and consulting organization, and then build products so that sales people can generate feature matrices to compare themselves with the competition. I'm certain they're keenly aware that they build shit products -- but they know that the majority of their customers are willing to pay a premium for a worse product if it means avoiding talking to finance and ensuring compliance for any external vendor.
This link doesn’t appear to have that information. It’s just an unanswered question about being unable to create CodeCommit repos. Is there an actual announcement somewhere?
It was never great anyway, although I wonder how they will handle some of the AWS stuff that would use it by default (automatically deploying private Lambdas, etc. IIRC).
Very disappointing. CodeCommit had value in its APIs for managing and accessing Git repositories programmatically e.g. from Lambda functions. I've used it to implement simple, Git-compatible version control in some projects by calling the PutFile API. Now it will be much more complicated to achieve something similar.
Every AWS feature requires some stupid aws cli call that gives you 12 hours of usage. So every day depending on how many products you're using, you have stupid shit to run.
No. You could either set it up with SSH or HTTPS credentials (managed via IAM Console or CLI) or use the AWS CLI's credential helper that managed everything automatically. The credential helper was a little weird to set up the first time I used it, but once I wrote up the instructions it was easy enough.
Honestly I've never come across the '12 hours of usage' thing for AWS. But I only use CodeCommit, EC2, S3, SNS, SES, and other low-level services.
simonw|1 year ago
QLDB shutdown announcement in the release notes: https://docs.aws.amazon.com/qldb/latest/developerguide/docum...
Their blog post about how to rewrite QLDB apps to use Aurora PostgreSQL instead: https://aws.amazon.com/blogs/database/migrate-an-amazon-qldb...
Hacker News discussion from when QLDB was first announced in 2018: https://news.ycombinator.com/item?id=18553387
I also found this handy community GitHub repo which tracks these breaking AWS changes and lets you subscribe to them via an Atom feed: https://github.com/SummitRoute/aws_breaking_changes
formerly_proven|1 year ago
Straight callback to all the "you don't need a blockchain, you just need a database".
jamesfinlayson|1 year ago
KingOfCoders|1 year ago
schoen|1 year ago
andrewstuart|1 year ago
viccis|1 year ago
rgblambda|1 year ago
By teams that have been mandated to move as much as possible to AWS by their company's senior leadership, because it simplified accounts/they negotiated a "great contract" etc.
vasco|1 year ago
thr0w|1 year ago
rkwz|1 year ago
Any insights on _why_ they turned out to be mediocre? Is it because of them being released as MVP and staying that way for many years or just a lack of interest afterwards?
simonw|1 year ago
I found that blog post: "How to migrate your AWS CodeCommit repository to another Git provider" from 25th July 2024 https://aws.amazon.com/blogs/devops/how-to-migrate-your-aws-...
I wonder how long AWS will keep Code Commit running for their customers who are already using it? I'm guessing many, many years.
Weird that there's no announcement anywhere (that I can find) about CodeCommit ceasing to onboard new customers. Apparently it happened on June 6th but this forum post from July 26th is the only thing that comes up in search.
frikkie444|1 year ago
Second worst thing is that it basically refuses to show a diff for any file longer than a few lines.
Third worst thing is you need a login helper and special generated credential to create a login.
Fourth worst thing is the absolute slowness of it. Good riddance.
upon_drumhead|1 year ago
paulddraper|1 year ago
AWS can build low level solutions -- EC2, EBS, S3. High level stuff is garbage.
jeffwask|1 year ago
paulmendoza|1 year ago
I use GitHub Codespaces now for much of my deployment automation.
Codebuild is also a real pain.
whazor|1 year ago
nunez|1 year ago
duttonw|1 year ago
teraflop|1 year ago
rohansood15|1 year ago
ThinkBeat|1 year ago
The different projects they had going were not complex or huge.
Yet it took months to get it up and running. The last month at least they paid for AWS specialists to come in and set it up and even they spent weeks.
Throughout it all, any notion of trying a different CI/CD stack were rudely dismissed.
Once it was up and running nobody dared touch the pipeline again.
From all the AWS services that the customer used nothing was ever comparable to the horrors of their CI/CD.
Setting it up with Azure's offerings would have been damned near trivial. I have however not used those in production so I do not have the experience to able to say it is a better solution over all.
¹ In fairness, AWS had just recently released the CI/CD offerings and things may be a lot better now. I havent look at it again since then.
kjfarm|1 year ago
donw|1 year ago
belter|1 year ago
Many in this thread commenting on CodeCommit not being GitHub or GitLab are missing the point.
I don't want to have my companies code with GitHub/Microsoft and being used to train their AI. Also don't want to have to rely on a third party like GitLab, that is a company who makes no money, and whose losses are $55 million dollars a quarter, and has shaky internal technical governance. Did not forget about their Prod database one man setup....
I don't care about the lack of features of CodeCommit. It's usefulness was essentially providing a managed git server. Did not need more, but needed it as a managed server in the AWS Cloud.
The alternatives now will be the third parties, or the additional effort of running EC2 instances and managing the resilience and architecture.
This does not predict a good future for tools like Amazon CodeGuru Reviewer for Java and Python, CodeArtifact, CodePipeline and CodeBuild. And even if these plus QLDB were probably money losers for AWS, the MBA's are missing the point.
I would not be surprised if these silent decisions, are not reversed shortly, or if the service is just kept forever as is, but not deprecated.
Edit: Just found this
"AWS breaking changes and price increases" - https://github.com/SummitRoute/aws_breaking_changes?tab=read...
Don't know what is going on but now also Cloud9 seems to have a shaky future? Not on-boarding new customers?
AMZN Earnings Release is next Thursday after market close. https://www.nasdaq.com/market-activity/stocks/amzn/earnings
I am going to listen to that earnings call, very carefully....
bugsense|1 year ago
sluongng|1 year ago
carlsborg|1 year ago
denysvitali|1 year ago
To me it looks like a huge hack (download your code as a zip file just to upload it to CodePipeline) rather than a solution - I'll not be surprised to discover other migtations to look similarly
blackeyeblitzar|1 year ago
ChadyWady|1 year ago
irjustin|1 year ago
They better not touch CodePipeline and CodeDeploy though.
skywhopper|1 year ago
GardenLetter27|1 year ago
genericacct|1 year ago
kennu|1 year ago
nyarlathotep_|1 year ago
coding123|1 year ago
Was codecommit like that
camtarn|1 year ago
Honestly I've never come across the '12 hours of usage' thing for AWS. But I only use CodeCommit, EC2, S3, SNS, SES, and other low-level services.
__alexs|1 year ago