Call for additional Mapster maintainers

[andrerav]

Hi Mapster community,

2024 was rough year. In february, I unexpectedly lost a close family member. I have to admit that this has been tough to deal with on a personal level. Unfortunately, this indirectly affected Mapster, and for that I apologize. As the sole maintainer, I simply haven't been able to keep up with incoming issues, pull requests and getting the contributions pushed out to new releases. In short -- being a single maintainer simply doesn't scale well, and especially when times get tough. So, to ensure that Mapster remains a reliable and up-to-date tool for the .NET community, I’m now looking for additional maintainers to join the project.

Typical responsibilities may include:

• Pull Requests: Reviewing contributions, providing constructive feedback, and merging approved changes.
• Issues: Triaging new issues and, to the extent possible, resolving existing issues.
• Releases: Maintain a reasonable release cycle that meets community expectations.
• Code quality: Continuously improve the code quality.
• New features: Nominate and implement new ideas and features.
• Documentation: Improving clarity, fixing typos, and keeping the wiki up to date.
• Community: Engaging in discussions and questions and fostering collaboration and contributions.

If you have the time and interest, I’d love to hear from you. This invitation is open to anyone passionate about Mapster and open source collaboration. Selection criteria will prioritize relevant skills and past contributions to the project. I am hoping we can onboard at least 2 additional maintainers.

I will of course remain as a maintainer myself and continue to help out as much as I can :)

Please reach out via email at andreas.ravnestad@ramn.no to express your interest.

[andrerav]

I hope that @DocSvartz and @stagep will read this and give it a think :)

[devgopher]

Please accept my condolences, @andrerav

[iAmBipinPaul]

sorry for your loss @andrerav 🙏
if possible, please post this on dotnet reddit.
https://www.reddit.com/r/dotnet/

[DocSvartz]

I’m sorry for your loss @andrerav .

I can join Mapster maintainers. But at about the current level as I do it now.

In these areas I most likely cannot help:

Releases: Maintain a reasonable release cycle that meets community expectations.
Code quality: Continuously improve the code quality.

[andrerav]

That's great to hear @DocSvartz :) @stagep has also confirmed that he will step up. I have added you both to the organization and assigned appropriate roles. Should we keep all discussion on this board, or would you prefer a public Discord server to discuss next steps?

[stagep]

Let's keep the discussion here in the repo. We can start a new discussion that deals with the overall direction or just continue it here in this discussion. I spent the last couple of days looking at a range of issues. Some are old and may already be addressed and / or no longer relevant. I would like to close a lot of issues with a comment that if the issue is still active and unresolved then reopen the issue. Alternatively, we do the opposite and comment that this issue will be closed in 7 days unless there is a response back that it is still an issue. Do you have a preference @andrerav ?

[DocSvartz]

@stagep OK

[DocSvartz]

I plan to work on #755 and #723 please don't close them.)

[stagep]

Anything with recent activity will not be closed. I will focus on dormant issues to start.

[stagep]

Do you like the following comment on closing issues:

In order to identify issues that are still active, we are closing issues that we believe are either resolved or are dormant. If your issue is still active then please reopen. Thanks.

[andrerav]

Yep, looks good to me :)

[DocSvartz]

@andrerav @stagep I would like to agree on the format of communication between Maintainers
For example:
In case of agreement, is it enough to put a reaction 👍 or should always give a text response?

[andrerav]

Both are fine with me, just keep in mind that thumbs up won't give the other party a notification in the app or by email :)

[stagep]

I wanted to give @andrerav and @DocSvartz an update. After closing the issues that were old or appeared to be resolved, I decided to look at a recent issue #737. I was overwhelmed trying to understand the code and it was just too much cognitive load on top of the projects that I am currently working on for my job. As a solo full stack developer on a business application, I wear many hats and Mapster was one hat too many. I continue to monitor issues to do with configuration and usage of the library, although (ironically) there have not been any simpler issues recently. I would like to thank both of you for your continued work on this library.

[DocSvartz]

@stagep you inspired me by dealing with issues so quickly 👍👍👍
Regarding #737, at first I thought it shouldn't work )
But as far as I understand there is no particular problem there. I wrote my thoughts on this matter under isuue.

If you worked on implementing caching as part of the issue, I don't think there is an easy task.

[stagep]

@andrerav There are 3 issues that I believe are all discussing the same idea. #377 #451 and #622 . If they are the same issue then I would like to close #377 and #451 as duplicates. Please let me know.

[andrerav]

@stagep I agree, feel free to close them as duplicates of #622 :)

[reichigo]

I'm very sorry for your loss. If you'd like, I can join and try to help.

[DocSvartz]

@andrerav @stagep

Perhaps we should define a list of priority issues for the next release?

Just developing a solution and testing them at the same time (including developing the tests themselves) takes a lot of time.

Leave them here and I'll give you feedback. Will I be able to solve them before the release or not :)

[stagep]

@andrerav @DocSvartz

I was working on an issue with Mapster.Tools not working with .Net 8/9 in the development branch and so tested the master branch using .Net8/9 but forgot I was not on a fork and pushed my changes to the repo. I just reverted to the prior commit so master is back to how it was.

[DocSvartz]

Maybe the next prerelease should be released, at least with some new PRs.
@andrerav @stagep what do you think about it?

[stagep]

Yes on a prerelease. Perhaps one prerelease on all the simpler fixes that are very likely to be non-breaking, and one prerelease on all the null / mapping issues. We have got out of a steady prerelease package and need to at least provide packages. Even I am using my own fork and would like to move to a prerelease.

[DocSvartz]

@stagep Maybe I've already forgotten :) But in terms of null mapping issues there seem to be no significant changes.
The most significant change is related to the use of default values ​​and PRs based on it.

I can mark PRs as something like NextPre and NextPre+

[stagep]

@DocSvartz perhaps just all on the same prerelease then just to get a prerelease out there. It's a bit of a catch 22. Not prereleasing frequently makes each prerelease have more changes, which could have more breaking changes.

[andrerav]

@DocSvartz Sure, ping when everything is merged into development and I'll publish the packages.

[DocSvartz]

I merged my commits, as I wrote earlier, into two separate branches.
All that's left is to apply the community PRs to the second branch.

Can you publish from this branch? https://github.com/MapsterMapper/Mapster/tree/v-9-0-0
Or is it better to merge everything at once into the development branch?

[DocSvartz]

I checked, it all merge

[DocSvartz]

@andrerav all PRs merging to development branch

[andrerav]

@DocSvartz Packages have been published :) Might take some time before the nuget index is updated properly.

[DocSvartz]

@andrerav Please answer something about this issue #802.

[DocSvartz]

@andrerav I can do pre-releases if you agree.

[garfbradazEQ]

Is this Project still alive?

[stagep]

Let's see what @DocSvartz answers. @DocSvartz has a number of PRs that he wanted to publish as pre-releases but he never got an answer from @andrerav (the primary maintainer) to move ahead on. I wouldn't say it's dead. It's a mature package that for someone like myself, that just wants code generated mapping with no edge cases, it gets the job done.

[DocSvartz]

I'm currently busy with my own projects, but I'm ready to continue maintain Mapster.
If you have any problems or questions, create an issue and I'll see what I can do about it :)

[DocSvartz]

Hi @andrerav, let's do a new pre-release version.
I merged all the changes into the development branch.

[DocSvartz]

@stagep Who will try to publish a new pre-release version, you or me?

[stagep]

@stagep Who will try to publish a new pre-release version, you or me?

You should do it. I am busy finishing up a big project. Need a week and then I will be back and able to contribute.

[DocSvartz]

@stagep Ok

[DocSvartz]

🚀 10-pre01 publish 😀

[andrerav]

Well done @DocSvartz 🙌

[DocSvartz]

@andrerav
Are these tests related to the problem or do they reflect expected behavior?

It actually also have a Race Condition based on the test's construction :)

Parallel.Invoke() -Does not guarantee that they will be launched in parallel or in the specified order.

On my machine, in most cases first run () => { TypeAdapter.Adapt<WeirdPoco>(simplePoco); } and in this case the test will always true (Get Exception), since the mapping settings are not specified.

If need to check for a Race Condition that occurs in the TypeAdapterConfig itself, the test should look like this:

public void Race_Condition_Produces_Error()
{
    TypeAdapterConfig.GlobalSettings.RequireDestinationMemberSource = true;

    var simplePoco = new WhenAddingCustomMappings.SimplePoco {Id = Guid.NewGuid(), Name = "TestName"};

    //first state (i = 0) Must be configured
    TypeAdapterConfig<WhenAddingCustomMappings.SimplePoco, WeirdPoco>.NewConfig()
                       .Map(dest => dest.IHaveADifferentId, src => src.Id)
                       .Map(dest => dest.MyNamePropertyIsDifferent, src => src.Name)
                       .Ignore(dest => dest.Children);

    var exception = Should.Throw<AggregateException>(() =>
    {
        for (int i = 0; i < 100; i++)
        {
            Parallel.Invoke(
                () =>
                {
                    TypeAdapterConfig<WhenAddingCustomMappings.SimplePoco, WeirdPoco>.NewConfig()
                        .Map(dest => dest.IHaveADifferentId, src => src.Id)
                        .Map(dest => dest.MyNamePropertyIsDifferent, src => src.Name)
                        .Ignore(dest => dest.Children);
                },
                () => { TypeAdapter.Adapt<WeirdPoco>(simplePoco); }
                );
        }
    });

    exception.InnerException.ShouldBeOfType(typeof(CompileException));

}

[DevTKSS]

Hi @andrerav,
even though I’m still early in my C# journey — I started in late 2024, after about a year of self-taught VB.NET experience, and before that mainly developed smaller macros to automate CAD applications in mechanical engineering — I would be interested in contributing to Mapster where it’s helpful. I wouldn’t describe myself as an expert, but I’m committed, reliable, and used to working my way into codebases and OSS workflows.

In other projects like the Uno Platform, my contributions have mainly been documentation improvements and smaller technical additions. I also maintain my own samples repository (linked above), where I provide written learning material alongside my German-language YouTube videos whenever I find the time next to my job.

I’m open to supporting Mapster in areas that match my current skill level — whether that’s documentation, smaller issues, testing (I’ve been using xUnit and Shouldly in my PRs and in my work projects), or other tasks that help reduce the workload for the maintainer team. If this kind of contribution profile could be useful, I’d be glad to help and grow with the project.

[andrerav]

Hi @DevTKSS, I've added you to the organization with maintainer level access. Thank you very much for volunteering :)

[ataoliu]

I am a developer passionate about open-source projects and .NET technologies. In the past, I've used Mapster in my work and have been deeply impressed by its powerful ability to simplify object mapping.

I frequently handle complex mapping requirements in my work, so I particularly appreciate Mapster's flexibility and efficiency. I hope to join the Mapster community and contribute my efforts to the project.

[grahambunce]

Hi all,

Firstly, thanks to all for working on this product and stepping up to help @andrerav after what must be a difficult time.

it’s not clear who are now the maintainers on this tool going forward and/or who has merge permissions. It’s something I would like to understand if that’s possible as we are considering this tool but understanding realistic long term support is one of my criteria for open source selection.

can anyone help confirm this?

many thanks

[DocSvartz]

@stagep Have you used the latest pre-release versions of Mapster.Tool? I looked at the current tests for this and apparently the mapster tool is not actually used in these tests🤔😅

I also think we need to define what we consider a stable version to be:
If a stable version is one that works as we intended and does not contain any known bugs that we can fix.
Then all pre-release versions are actually stable.

[stagep]

I am about to upgrade a .Net 10 project today. Stable requires the acknowledgement that it may break for a very small amount of users, but with the ability to rollback to a previous working version, any lack of stability can be disregarded IMO.

[stagep]

Mapster.Tool works fine. The only change was an empty initializer { } was added after parameterized constructors.

[DocSvartz]

@stagep Great, I don't think this is a critical error.

[stagep]

@DocSvartz it makes no difference, I just wanted to let you know what has changed in codegen since version 7.4

[stagep]

@DocSvartz and myself have repository admin rights. I have helped out with support issues primarily configuration, and managing expectations of what the product is designed to do. I closed a lot of old issues, as they had either been fixed, or were irrelevant. I would classify myself as being "tech support". @DocSvartz was working on a lot of issues and submitted PRs and then @DevTKSS got involved with other improvements. There was a lot going on a few months back and then things got quiet. I have never really dug into the code too much. I have a clone of version 7.4 (the latest release) that I updated to .Net 10 and built and am using as a package source on the applications that I write. I would wait for @DocSvartz and @DevTKSS to respond so that they can give you an update on their status and involvement. I was pulled into the role that I have. I was happy just answering questions, but getting into the code was too high a cognitive load, as I am an architect on a compute intensive, algorithmic and heavily rules based application (i.e. not CRUD). I use the product for the codegen capabilities between entities and gRPC messages, as I want to discover issues at compile time and not runtime. I feel that the biggest issue with the product is that it has feature creep. Too many edge cases were handled. What I feel is lacking is a stable release for the currently supported .Net Core frameworks.

[DocSvartz]

@stagep @andrerav Perhaps the old wiki won't disappear completely if it's deactivated.
If we use the new version, we should add a warning to the main page of the old Wiki indicating that this version is deprecated, with a link to the new Wiki.

[stagep]

@DocSvartz lots (or maybe all) of the links from the readme.md still point to links with /wiki/ in the path. I believe that the links should be to URLs with /docs/ in the path.

[DevTKSS]

@stagep @DocSvartz Sorry for beeing much offline for past Weeks. I am trying to concentrate my capacities to finding a new job engagement, so reduced my whole github and coding activities to bare minimum and none. Hope this is understandable from your POV too, to keep it like this, until I fixed this gap hopefully anytime soon I get a chance.

Beside this, I will have to try to get my Project to a point to get the Rest-API data from Sevdesk transfered to anything I can store (e.g. files) since the export of them is... lets be nice and call it "not usable in production". this month is the last, I have until it will get deactivated and this software they provide is rubish as hell, nothing I will concider paying even more to still get API access. This Project was, if you remember, the reason I found Mapster for mapping and still trying to figure out how to integrate it best.

@DocSvartz not sure, but as you transfered the Docs from the development branch, didn't they have had the links fitted already? Which ones does miss on this? 🤔 Maybe I forgot them without noticing 🤷 For the Reference Docs, the Commits I pushed have had them integrated in docfx.json mapped to /api/ but I think we not pushed them to release, because of unsure Version noting or similar too. So it should be seen as to-get-checked how this should get done here. If we don't care about different TFM / Package Version Docs, this here works in my own Sample Apps Repo quite fine:

https://github.com/DevTKSS/DevTKSS.Uno.SampleApps/blob/db60604420bb2f27251e2992a0d0d86423d14628/docs/docfx.json#L3-L47

but Mapster Repo most likly doesn't need the filterconfig, this is set up there, because of the Source Gen I use in my contained Uno MVUX Apps.

[DocSvartz]

not sure, but as you transfered the Docs from the development branch, didn't they have had the links fitted already?

@DevTKSS The purpose of the transfer was to check whether the site would launch automatically. 😂

Links on the main page new wiki site and repository Readme lead to the old wiki:

I fixed it here:

• Fix(docs) : fix Readme links and local build script #862

in PR links point to pages on new wiki site

[stagep]

@DevTKSS no need to apologize. I completely understand and I should apologize to you. I wish you much success in finding a new position. You must prioritize yourself. As I was giving you an example I see that @DocSvartz already has a PR ready :)