Replies: 8 comments 29 replies
-
|
Dear @nojaf , Thanks for the issues raised, let me try to address them
As a side note - the fork haven't happened out of the thin air. I'm starting a new project at work, so I didn't want to spend months at finding a way to approach Dustin and more months to agree on solution, and life is short! Also, there are particular things that I wanted to fix in Giraffe from it's inception, but which will be a breaking change for it, so a fork definitely gives me such a chance. |
Beta Was this translation helpful? Give feedback.
-
|
I appreciate the offer but I was trying to go thru similar steps of transition to a maintainer that could commit more fully but I was never able to get there. I'm glad this kind of transition is finally happening. If you just want a backup "still has a key just in case" kind of maintainer, I can do that. |
Beta Was this translation helpful? Give feedback.
-
|
From a SAFE Stack point of view (and the wider F# community), I think that there is value in trying to keep Giraffe going as a brand - the same as Saturn - because it's a known thing. Forking will unfortunately just create more confusion and split the community, so personally I would rather have that as a last resort. Moving to Giraffe / Saturn from Suave caused enough headaches and I really don't want to repeat anything like that again. We really aren't in trouble with Giraffe, I chat with Dustin regularly, but simultaneously I think he would probably appreciate having some extra maintainers / contributors to the project. |
Beta Was this translation helpful? Give feedback.
-
|
Hey, I actually only saw now your original tweet and blog post and had a read. I admit I have no interest in maintaining Giraffe any longer. Not just Giraffe but really any open source project. I just don't have the time for this. I still love programming and I work on a lot of side projects in my spare time, but nothing which is open source or visible to the public. Open source work is a lot of work where you have to constantly lend your brain capacity to the desires, questions and wishes of random people around the world for very little to almost no financial reward. It's something I don't enjoy doing anymore. I rather work on my own issues or on products which people pay money for as I find this line of work more satisfying whilst also being more in control of where and how I shape the software. This is basically a long explanation where I am mentally and why I've let Giraffe "rot". That is my mistake and when people reached out to me to help I often put it on the backburner thinking I need to take some time to reply to the kind people who want to help but then I never found the time to respond, which meant that I've actually done the opposite of what I wanted to do: I appeared to be ignoring people who reach out when in fact I just delayed it because I waited to devote them more attention than I was currently able to offer. So long story short, if you want to can take full ownership of Giraffe. I can transfer everything to fsprojects today or make you the most powerful Admin of the giraffe-fsharp GitHub org or do it via another way if you have a preference. If however you want to start something new now then I also understand that and you have my full blessing! The F# community deserves a good functional web framework and I've not been able to live up to that over the last few years. |
Beta Was this translation helpful? Give feedback.
-
|
Oh I misunderstood that. My understanding was that @Lanayx wanted to focus on Oxpecker now or have I got this wrong as well? |
Beta Was this translation helpful? Give feedback.
-
|
Dustin, you are right. My interest now is in building a new framework that will be competitive wth other modern frameworks by improving Giraffe codebase as much as I can and not looking at backward compatibility with Giraffe api. Keeping compatibility and supporting existing Giraffe codebases is not something I want to work on, so I can't lead Giraffe project itself. |
Beta Was this translation helpful? Give feedback.
-
Should I send the e-mail as well? I'm in btw. |
Beta Was this translation helpful? Give feedback.
-
|
@64J0 Yes please if you send me an email so I got your email address and we find a slot either this or next week, that'd be great. I'm also following up with a few other people who showed interest because I think having 2 core maintainers would be preferable. |
Beta Was this translation helpful? Give feedback.
-
|
Sent right now 🤝 |
Beta Was this translation helpful? Give feedback.
-
|
My strong, strong preference would be to keep the Giraffe brand going and make additions / breaking changes to that rather than a brand new project. I've looked through the blog post and there are some ideas in there that Dustin and I have already been chucking around in the past few months so I think we're all thinking along the same lines. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you for the opinion! As I said in other messages, I really feel the need in core breaking changes, and I'm not sure if keeping just the same Giraffe repo and package will be user friendly, since they will look for a way to "upgrade" one version to another and will still remember old maintenance problems. And they will also complain about breaking changes. It reminds me the story with Angular and Angular2, where second version was really a brand new project, so it was no different - to switch to Angular2 or Aurelia. As for new users, it will be no difference for them to use Giraffe, Oxpecker or anything else, they will go to the place where documentation is better and there are enough stars and maintenance, also if someone or something will guide them, say a big message in giraffe's readme - you better go to another repository. |
Beta Was this translation helpful? Give feedback.
-
|
Hi all, Cross linking this to keep the discussion in sync: fsprojects/.github#49 (comment) Whether "fsprojects" or "giraffe-fsharp" is used as the containing org, the main first thing is to identify 2 possible initial active maintainers. These are usually people who have recently contributed PRs and are motiviated to work on the project as a whole, including making releases. If the org "giraffe-fsharp" is used, then there should also be 1 or 2 backup maintainers who have ownership rights on org and nuget packages. The fsprojects policies may be helpful here: https://github.com/fsprojects/.github/blob/master/profile/README.md Don |
Beta Was this translation helpful? Give feedback.
-
|
@Lanayx If you were to become a primary maintainer on Giraffe, I think it would be feasbile for you to do a Giraffe vNext (with breaking changes) alongside sustaining the current Giraffe. That's effectively what happened to Fabulous, which is a similar framework-sized thing. What do you think? It feels like there could be a solution along these lines, maybe with @TheAngryByrd as initial co-maintainer? I understand it's a different prospect for you though. |
Beta Was this translation helpful? Give feedback.
-
|
My opinion that is has to be a new thing. Giraffe's current endpoint routing was already a new thing and was really a breaking change, kinda repo inside repo. So a new project should be either Giraffe2 or GiraffeVNext or similar. It just can't stay the same Giraffe anymore. I think that even going forward with Oxpecker would be a good move, the brand value that @isaacabraham is talking about is a overestimated, it has lost a lot of it's value during the years of poor maintenance. For example the query "fsharp web framework" doesn't return it on the first page in Google, only the projects lists (which can be easily updated) |
Beta Was this translation helpful? Give feedback.
-
|
I don't think that the work done on Oxpecker will be lost in any case. After that, we're volunteers, you're free to do what you want, and in any case I'd like to thank everyone for their investment. I'm clearly not available to be the main maintainer on Giraffe, but I can participate. |
Beta Was this translation helpful? Give feedback.
-
|
@64J0 writes:
That is great! We'll collect all the replies and start to decide the way forward, but I'm pretty confident we can have you as one of the co-maintainers. |
Beta Was this translation helpful? Give feedback.
-
|
@Lanayx As far as breaking changes are concerned, Rich Hickey would say "you might as well change the name", but going the Giraffe vNext route certainly has advantages:
|
Beta Was this translation helpful? Give feedback.
-
|
@otto-gebb De-facto, the decision has already been made :) Giraffe will be supported as is and Oxpecker will continue it's own life. As for community fragmentation - I would imagine that Giraffe should specialize on original implementation and Oxpecker should specialize on endpoint routing, so users would have a clear understanding of what to choose. |
Beta Was this translation helpful? Give feedback.
-
|
"I'm clearly not available to be the main maintainer on Giraffe, but I can participate." Likewise, I'm willing to help in any way I can (documentation, translations, etc.) to push F# forward! |
Beta Was this translation helpful? Give feedback.
-
|
Everyone is welcome to throw in their hat. My plan (with the help from Don Syme and the F# Foundation) is to find at least two core maintainers who take ownership of Giraffe and facilitate the continuation and maintenance of the project. Then these two core maintainers can add more maintainers to the organisation as they see fit. The only requirement I have is that the people who volunteer to step in must be serious in wanting to maintain the project as this is the main issue we are trying to fix here. It will require more work than just updating the documentation once in a while. So long as people who throw in their hat are realistic about this and willing/keen to take over then you are definitely a candidate. I am basically waiting for more people who are keen to step up so we can pick two new maintainers. |
Beta Was this translation helpful? Give feedback.
-
|
Outside of reporting bugs and supplying a work-around or two, I’ve never contributed to an open source project before. However, I’d be willing to help out this project on a part-time basis, as I enjoy using F#. Having experienced breaking changes many times (e.g. Microsoft’s ASP.NET 6.x release broke one of my F# Giraffe applications, ongoing churn with various client-side Javascript frameworks), here’s my 2cents on how to do this transition:
|
Beta Was this translation helpful? Give feedback.
-
|
@psantoro This is really helpful, thank you! |
Beta Was this translation helpful? Give feedback.
-
Hello, when reading your tweet I couldn't help but wonder about a few things here.
These are all rather meta aspects of this repository and there won't be anything actionable. So if you don't feel like responding to this, I respect that and feel free to close this at any point.
Here goes:
I can't help but think that the original problem statement here is that the maintainer of Giraffe not being as active as people would like. This is a genuine concern, but I wonder if communication has really failed here or wasn't attempted enough. Did you get Dustin's perspective first-hand via a virtual meeting? Not that you need it, but do you have his blessing for this altered fork? I think it would help this story if that is the case. The way the blog post was written, I cannot shake the sense that the two of you never had a serious conversation about this. Maybe things could have been resolved without going the fork route. If you both didn't see eye-to-eye, that is fine and this outcome is still okay, but did that conversation even happen?
How are you not going to fall for the same trap? One of the issues with Giraffe seems to be that only a single person can release new NuGet packages. The reason this repository seems to exist is to prevent such a state from happening, but why did you not solve this issue first and announce this project when there were three or more maintainers on board in this quest? Because right now there is a lot of work left to do to get on a similar level as Giraffe. Getting started guides, documentation, starter templates, contribution guides, etc... These all take a lot of work and when it is just you, how far will you get before you get burned out as well?
Did you check anything with the SAFE stack before this move happened? This is again not something you are required to do, but as far as I understand this, they also have a stake in the health of Giraffe. My point is, did you really check with enough sources in the community beforehand? You could have found some co-maintainers first before diving into the code.
and move it to fsprojects once the project gets 200 stars, not a question here but some feedback: 200 stars is a poor metric to decide if something should go to fsprojects. What you are looking for is people you can trust to work with you on this. People you can give the publish keys to and can operate the project without your intervention.To summarize my thoughts on this, I fear that good communication could have been lacking in this story. Did all parties really hear each other? Does everyone understand why there is friction?
The F# community is a small one, most popular projects only have one maintainer and that is problematic.
People do not talk enough with each other and I cannot help but feel that all of this is the direct result of a human problem.
Beta Was this translation helpful? Give feedback.
All reactions