Yes I know, but why release a product you don’t want people to use??
Would Apple ever do that?
“Hey don’t update to IOS16 because it doesn’t work correctly.”
No they would not. Why does IOHK then do it? It simply makes no sense to me.
Anyway I am now using Docker v. 1.33 and it works fine.
You got it wrong here. IOG wants pool operators to use 1.34.1 because it has some nice enhancements for them. They just could not get it to work on Windows. They messaged that clearly. And - as you stated you don’t want to run a pool and 1.33.0 works for you I don’t understand your complaining.
And - please don’t compare IOG with Apple - IOG is a serious company that doesn’t just produce toys for the stupid…
I don’t get it wrong. Why do you change the discussion? If IOHK don’t have a working Windows version then they should not release it. It’s fine they version for the operating system where it works.
Apparently you don’t understand that a node for windows is primarily used for development. Production and stakepools are running on Linux (majority)
Regarding Apple then I can compare IOHK to whatever I want. I don’t use Apple products but saying they produce toys for the stupid is below any standard for a serious conversation.
I have to keep my opinion that IOHK should NOT release anything not tested.
Do they “release”? The Release Notes do very specifically say that you should not install on Windows. (I tend to agree that the reason given does not really explain, why it manifests as missing DLL on your machine.)
They don’t directly link to a Windows binary, but only to this Hydra build cluster. (…, which could be criticised in itself. I’d much prefer an official list of binary releases than this Hydra labyrinth.)
Removing one platform from the build cluster completely, because it has one showstopper, sounds also wrong.
What do you mean by “random Joe can commit”?
In the end, I don’t think the quality of IOG’s software (and documentation) is beyond doubt, but it is also not horribly unusable. And the same holds for Apple and Microsoft. They make sense neither as a shiny example of how it should be done (“Would Apple ever do that?”) nor as an example of how it should not be done (“toys for the stupid”).
I mean for community software the quality of the software will vary a lot. It’s fair enough since people are using there free time. In some cases it is very useful in other cases it’s not. Having too many different approaches to solved the same problem makes it hard to figure out what to use and get an overview unless you only deal with Cardano and dedicate nearly all your time for that.
I just say that Apple would not release a software package saying don’t install this because it does not work. That’s all. I don’t talk about documentation like you do.
But since you mention documentation then that’s the usual issue with almost all software packages. Not that many are able to provide clear documentation with good examples.
If you look at eth web3 for sending a transaction it is clear what you should do. The same goes for the rest of their documentation web3.eth.personal — web3.js 1.0.0 documentation.
So let that be an inspiration for us all.
…, but IOG’s cardano-node is not community software. They have quite a bit of funding for putting paid and professional developers on it.
I know.
The reason for posting here is as a dev wanting to get started with Cardano development the process is painful if you compare to eth, sol or even bitcoin. The lack of official tools like a simple official NPM package for wallet interfacing is not available. Either you have to do yourself (I do that) or use a community package that may or may not work as expected.
So since IOHK has around 404 devs then why don’t we see better tools development? It’s been years now. Solana did it in almost no time.
Now there will be a horde of Cardano soldiers defending IOHK. Fine but I don’t understand why it is not allowed in this forum to be critical. Missing dev tools is harmful for Cardano and the mismatch between what Charles say in his videos and what is done by IOHK should be clear to every dev. Also those I would call Cardano maximalists.
I tend to be with you on that one.
Personally, I’d need an “official” Python lib much more than an NPM package, but I’d also be happy with much clearer documentation on how to do it. There are the academic papers – good but not concrete. There is the Haskell source code – not extensively documented and only accessible to the initiated. And there is documentation of the various command-line tools – scattered, often out-of-date, widely varying quality. And that’s basically it.
Your concrete example in this thread – specific version not working on Windows – was not a good example in my opinion, though. They said that it will only work on the other platforms. I would not count the link to https://hydra.iohk.io/build/12997302#tabs-constituents a “release” (re your “Apple would never do that.” point). But I also wonder, why the Windows build has a green check mark. I don’t really understand this Hydra thing, but shouldn’t the fail from https://hydra.iohk.io/build/12970899 be inherited, so that people are warned?
Please, be critical!