Feat: create `dart` version of `cid` package
nelsonic opened this issue · 13 comments
At present there is no package on pub.dev matching the keyword "cid": https://pub.dev/packages?q=cid
Todo
- Port the
Elixircode from https://github.com/dwyl/cid to this repo indart-lang♻️ - Create a Pull Request to get early feedback on code. 🆕
- Publish the package to
pub.devthis will be our firstdartpackage! 📦- Document the process for publishing a package: dwyl/learn-dart#22
Note: for biz continuity purposes, the package will be published under the
dwylnamespace/ownership.
There appear to be a few options available for Base 58 in dart: https://pub.dev/packages?q=base58
So at least we don't need to implement that part before being able to build cid in dart. 👌
@LuchoTurtle as discussed on Standup this morning,
If you want to fork the https://github.com/C0MM4ND/dart_fast_base58 for security reasons
and include it as a git dependency in our cid implementation, go for it:
dependencies:
flutter:
sdk: flutter
dart_fast_base58:
git:
url: https://github.com/dwyl/dart_fast_base58.git
ref: main # branch nameHaving read:
https://sovereign-individual.xyz/posts/future-proof-crypto-hashes/
https://sovereign-individual.xyz/posts/ipfs-content-identifiers/
https://docs.ipfs.tech/concepts/content-addressing/#cid-conversion
https://multiformats.io/multihash/
and perusing dwyl's CID implementation at https://github.com/dwyl/cid/blob/master/lib/cid.ex,
I tried to search for implementations of Multihash I could use in Dart and found there to be none. The closest I found was multibase, which really isn't the same thing and it's not how IPFS does things.
So should I try to implement my own version of Multihash in Dart? Never done this nor done anything crypto-related 😅 .
What do you think?
@LuchoTurtle Good research and follow-up question. 👍
We can do the Multihash later. It's not a prerequisite for cid. 💭
Once we have a basic cid implementation in Dart, 
enough to do what we need in our Flutter App, 📱
very happy to do any deep-dives into more advanced features. ⏳
Please create an issue in the backlog for Multihash now 🆕
so we know what is next and can link to it from the README.md -> Roadmap Section. 🔗
Stoked you're getting a chance to dive into Crypto. 🔐
It's a real differentiator on your CV to have a deep technical understanding of the topic. 🤓 👌
If along your journey to understanding all this you learn new things, 🔍
please add a section to: https://github.com/dwyl/learn-cryptography 📝 🙏
We can do the Multihash later. It's not a prerequisite for cid. 💭
You are right it is not a prerequisite for CID. For CID v0 it is not used but I feel like going for v1 would be better not only for us but for anyone who wants a cid in Dart.
After doing research, I noticed that there wasn't a Dart implementation of Multihash so I went along and have an mvp working (encoding and decoding) -> #2.
I still want to iron out a few features and then get it featured on the canonical repo of multiformats https://github.com/multiformats/multihash#implementations. There's more info on the PR #2.
This will change the scope of the repo to Multihash. I will use this one and work on the CID making use of Multihash on a different one. Is that reasonable? 😄
@LuchoTurtle great that you proactively implemented something.
Figuring out how all of this works under the hood is a non-trivial task.
And you are growing leaps and bounds as an engineer by exploring it. 🎉
Remember that:
"Writing code? That's the easy part.
Getting your application in the hands of users,
and creating applications that people actually want to use - now that's the hard stuff." ~ Jeff Atwood
Now for the "not nice" part:
When I asked you to create an issue for Multihash it wasn't a suggestion.
Issues are how we keep track of the work that needs to be done. Without them it's chaos.
I don't like Chaos and I don't know any leader/entrepreneur who does.
If they say they do; they're either joking or lying. Chaos is the enemy of progress.
I say/write "Please" to be polite. But please don't confuse direct instructions for optional requests again.
It's not difficult to open an issue [5 mins] before spending a day [8 hours] on a tangental task. 💭
Create an issue before you start the work
Here's the issue: dwyl/dart_multihash#1 it took me T5m to create it.
My time is worth 20x so 5 minutes of my time is 100 minutes of yours ... You're welcome.
As far as "bosses" go, I try my best not to be an A55H0L3
but I've worked a few places where doing the exact opposite
of what the "boss" asks gets people fired on-the-spot. 🔥
Remote Work === Communication.
If you want to work in micro-team in an office on closed source software
where written communication isn't required, that's fine.
But if you want to work remotely, then communicating early and often is a MUST, not optional.
I've worked on 200+ people remote teams and the engineers who ignore communication
get labelled as "Difficult to Work With" and get "let go" very fast. 🙅
I'm not saying this to be "mean" or "difficult", in fact the exact opposite is true!
As you know my goal is to help you grow and flourish as a person and engineer. 🚀
I expect you to go on to do great things and work in super interesting places.
But the number of those places in PT/Porto is vanishingly small.
Which means you either need to move away from all your family, friends & creature comforts
to a big international city where you know nobody, have zero friends or family and will be miserable.
OR you have the "best of all worlds" and work remotely.
Your Time < My Time
Your "bosses" time will always be more valuable than yours.
In some cases it's "only" 5x but if you work somewhere "big" like Google, it's 100x.
Never waste your bosses time. Ever.
When you [eventually] become the "boss", e.g. when you start your own company.
You will want the people working for you to respect your time.
I sincerely hope you reach that point.
Because the alternative is mediocrity and nobody wants that. 😢
In the process of adding property-tests, akin to what's happening here -> https://github.com/dwyl/cid/blob/194d5e10d8713a6bce186d74021153ca43d5782a/test/cid_test.exs#L114
It's taking a wee of time to get it working the same, processes work a bit differently in Dart. But I should get it working today.
I kept README fairly simple because I didn't want to duplicate information from https://github.com/dwyl/cid
Good plan. thank you for being thorough. 👌
Finished IPFS tests based on what was done with dwyl/cid -> #3 (comment)
Regarding documenting the process of publishing a package dwyl/learn-dart#22, I'm waiting for dwyl/dart_multihash#2 to be reviewed so I can publish it, properly use it on this repo instead of mocking.
@LuchoTurtle dwyl/dart_multihash#2 merged.
Please proceed. 🙏 (Thanks again!)
Just pushed the necessary changes with #3
Marked your comment as outdated by mistake, wanted to hide it to make scrolling easier but ended up making a mistake. Sorry about that.
Published
https://pub.dev/packages/dart_cid