multer
multer copied to clipboard
expressjs
[Discussion] Is this project actively maintained?
Introduction
Hello 👋,
first of all I want to thank everyone who contribute/contributed to this awesome project. I really enjoyed using multer a long time. But now I have some real problems with using this project. The main question of this Issue is as the title suggests "Is this project still actively maintained?". I want to highlight the part actively maintained as it doesn't seem so for users of the project. I also don't want to receive an answer like "Yes, it is." followed by closing the Issue like #592. I want to discuss how exactly it is actively maintained or not and what we maybe can do about it.
My point and opinion
The project has bugs. As we can see on the massive amount of (at time of writing) 179 Issues and 51 PR's the project struggles with multiple Issues. This on it's own isn't really a problem as it's usual for a project with a high user count. The problem comes when this Issues are not fixed and gets even worse the longer this Issues are not fixed.
In example the project has some bugs I myself had to deal with multiple times. Partially needing to directly rewrite project source code so speaking maintaining an own version of the project. Even if there were already open PR for this issues. Examples include: #971 (which is over a year old), #438 (which is over 5 years old) and #259 (which is for over 6 years opened).
For me this doesn't look like an actively maintained project. We're missing important fixes to important problems for in some cases over 6 years.
What is your opinion?
Now that I pointed out what I think about the maintenance status of this project. I really would love to hear what other users of this project think about it. I also would love to hear from some of the maintainers what theire take on this is.
How can we fix this?
I myself don't really have a solution for this as it comes down to how the maintainers deceide to maintain this repo.
This is why I'm most interested to hear from the maintainers about this point. How can we fix this situation? How can we solve Issues that are existent for over 6 years and forced users to maintain a custom version of the project. Which is less then ideal for updates to the main repo?
I think that we do have a problem with a lot of open issues. It's really time consuming to keep them down, as we get more opened all the time. It's my feeling that the majority of the issues opened are actually a problem with their setup, or other libraries. Not an actual bug in Multer. Keep in mind that Multer has ~3.3 million weekly downloads, so from what I can tell, Multer is working for the vast majority of people.
To address your examples:
#971 - I have simply missed this one, but it seems like we would be able to land it. It doesn't include any tests though, so to land it I would like to test it out myself locally which takes time from me. This also means that it can take a while to land since I need to dedicate some time to it...
#438 - As discussed in that issue, I would be open to accepting any fix for that which isn't a breaking change. Since there already is a v2 that has release candidates, now is not a great time to merge breaking changes to the old architecture.
#259 - Seems to be an issue about the previous mentioned PR. As I said, I'm happy to accept a non-breaking PR that fixes this!
If you want to help, something that would be much appreciated would be to go thru all open issues and triage them. I believe that many of them can be closed because there isn't a bug in Multer, or there isn't anything actionable that we can do from our side.
Aside from that I would also really like to hear from people that have tried v2 of Multer! The number one thing to get a release ready would be to hear from people that it's working, or hear what we need to fix before we release it.
It's my feeling that the majority of the issues opened are actually a problem with their setup, or other libraries. Not an actual bug in Multer.
Enabling discussions for this repo would help deal with that. People knowingly create issues in GitHub projects to get help with issues they're having, probably because they're not confident stackoverflow would be of much help for a specific library. GitHub Discussions seems like the logical place to get help.
Discussions are enabled now, but my guess is that people will have a better chance at getting help at StackOverflow since their problem is usually with the entire stack, rather than Multer specifically...
