Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Consider renaming project. DataFuse is too similar to DataFusion. #654

Closed
andygrove opened this issue May 28, 2021 · 35 comments
Closed

Consider renaming project. DataFuse is too similar to DataFusion. #654

andygrove opened this issue May 28, 2021 · 35 comments

Comments

@andygrove
Copy link

This project appears to have similar goals to Apache Arrow DataFusion, contains code from DataFusion, and has a very similar name.

The names "DataFuse" and "DataFusion" only differ by a few characters and this could cause confusion about the relationship between these projects.

On behalf of the Apache Arrow DataFusion community, who have put a lot of work into building the DataFusion software and brand over the past three years, I respectfully ask that you consider renaming this project.

@github-actions
Copy link
Contributor

🎉 Thank you for opening an issue! 🎉 One of the maintainers of the project will respond as soon as possible. Seeing that you are new here, please familiarize yourself with our code of conduct and contributing to datafuse

@l1t1

This comment has been minimized.

@BohuTANG

This comment has been minimized.

@andygrove
Copy link
Author

Hi @BohuTANG and thanks for the response.

Yes, the use of the DataFusion code is absolutely fine as far as I have seen. The issue is just the confusion with a very similar project name. The projects are similar in that they both implement query engines in Rust based on Apache Arrow format.

DataFuse looks like a really interesting project but in my opinion, I think it would be better to use a name that differentiates it from DataFusion.

@onzo-bryanfeeney
Copy link

Name is cheap, what matters is the code and what problem it solves!

If the name is cheap it should cost you nothing to change it, so why not make the change?

@BohuTANG
Copy link
Member

BohuTANG commented Aug 5, 2021

Name is cheap, what matters is the code and what problem it solves!

If the name is cheap it should cost you nothing to change it, so why not make the change?

Hi!
In terms of projects name, the datafuse project has an earlier name than datafusion.
datafuselabs/datafuse(27 Mar 2021): b335630
arrow-datafusion(04 May 2021): https://arrow.apache.org/blog/2021/05/04/rust-dev-workflow/

So, we just keep using the name.

@Dandandan
Copy link

@BohuTANG

That's incorrect.

DataFusion is a project that was started end 2016 / beginning 2017 by @andygrove with that name.

https://github.com/andygrove/datafusion/commits/master?before=219ca3f61f0d751779b4f30c02aa5fdcf8a40d12+385&branch=master

Early 2019 it was donated to Apache Arrow already:

https://arrow.apache.org/blog/2019/02/04/datafusion-donation/

@BohuTANG
Copy link
Member

BohuTANG commented Aug 5, 2021

@BohuTANG

That's incorrect.

DataFusion is a project that was started end 2016 / beginning 2017 by @andygrove with that name.

https://github.com/andygrove/datafusion/commits/master?before=219ca3f61f0d751779b4f30c02aa5fdcf8a40d12+385&branch=master

Early 2019 it was donated to Apache Arrow already:

https://arrow.apache.org/blog/2019/02/04/datafusion-donation/

Thanks for the correction.
Who have a better name ? 🤔

@gauss1314
Copy link
Contributor

Hi @BohuTANG ,how about Basefuse ?

@andygrove
Copy link
Author

I think that there are plenty of smart and creative people here that can help with coming up with a name. It would be helpful to know the motivation/meaning behind the name DataFuse.

@BatmanAoD
Copy link

Why was the previous name (as seen in that commit from March), FuseQuery, changed?

@l1t1
Copy link

l1t1 commented Aug 5, 2021

try datamix?

@BohuTANG
Copy link
Member

BohuTANG commented Aug 6, 2021

@andygrove @BatmanAoD @l1t1
Hi!
There are some challenges here, the project datafuse creator company also called datafuse, it's hard for us to change it.

@houqp
Copy link

houqp commented Aug 6, 2021

if you are going to change the project name eventually, it's better to do it sooner than later so you can build the brand. Changing the name later after will impact you negatively on SEO and marketing etc.

@BohuTANG
Copy link
Member

BohuTANG commented Aug 6, 2021

Why was the previous name (as seen in that commit from March), FuseQuery, changed?

FuseQuery is one part of Datafuse, it's a query engine:
https://github.com/datafuselabs/datafuse/tree/master/fusequery

Datafuse is a DBMS, it consists of several components:

  1. FuseQuery -- distributed query engine
  2. FuseStore -- distributed storage engine
  3. FuseCLI -- a deploy and management all in one tool for Datafuse

@andygrove
Copy link
Author

Given that all the parts start with Fuse, I would probably start the product name with Fuse as well. The most obvious choices to me would be:

  • FuseDB
  • FuseData
  • FuseBase
  • FuseIO
  • Fused

I think the company name is less of a concern to the community than the product name, but that's just my opinion.

@junli1026
Copy link
Contributor

How about DataForge, which has the same init with DataFuse (DF). Thus we can keep use the df prefix in the code :)

@BohuTANG
Copy link
Member

BohuTANG commented Sep 3, 2021

@junli1026 Thank you.
As it stands now, datafuse has some (minor) popularity, we're not considering changing the name and we don't want to waste any more time on this topic, this issue will be closed this weekend.
If you have any ideas, feel free to open a new one.
Thanks all.

@BohuTANG BohuTANG closed this as completed Sep 4, 2021
@BatmanAoD
Copy link

@BohuTANG For what it's worth, this seems pretty underhanded to me. It seems clear that you never intended to change the name, and that your request for suggestions was not sincere. The fact that the name is popular is, if anything, an indication that you probably have benefited from choosing a name so close to DataFusion, which I am still not convinced you didn't know you were doing.

@BohuTANG
Copy link
Member

BohuTANG commented Sep 5, 2021

@BatmanAoD

That's not true, datafuse team is under pressure because of this ISSUE, we have discussed it many times and we don't want to continue.
The name datafuse does not harm any community and our team will not do so, and not intended to leverage anyone's halo, if so please let us know.

So why are we currently struggling with the name?
Anyway, The name we believe is not wrong and does not harm others, if so please let us know!

@Dandandan
Copy link

@BohuTANG

In my opinion, keeping the name as datafuse is harmful, both for the future of DataFusion and DataFuse communities and product, as the names are too similar for very similar projects (sharing a large portion of inspiration, dependencies and even some code reuse from DataFusion) and leads to confusion.

The current "resolution" (not to change the name) is disappointing to me, given the discussion and facts presented earlier and discussed alternatives.

@PsiACE
Copy link
Member

PsiACE commented Sep 5, 2021

Hello everyone. It's great to see your suggestions in terms of name changes.

But I'd like to say a few extra things.

  1. we are not trying to get into the light of datafusion. Despite the similarity in name, we are not trying to divide and harm the community. We are trying to actively contribute to upstream projects as well, and we are contributing to the prosperity of the entire ecosystem. It would be unfair to stereotype us as doing something harmful just because we have a similar name.

  2. we recognise the similarities with datafusion, but we are doing something different and trying to do it well. we don't want to get caught up in controversy, which would make everyone feel bad. We had some serious discussions about changing the name, but it was difficult. On the one hand, we have put just as much effort into building the brand; on the other hand, if we give up datafuse today because of arrow-datafusion, will we need to give up even more tomorrow?

I believe that the Rust community is inclusive and full of goodwill. I also hope you will give us some time to work through these things. datafuse and arrow-datafusion are like twins, we trust and appreciate arrow-datafusion's efforts and try to work side by side with them. Let's put aside the name debate for a moment and work together to build a great era of data.

Good luck, Rustaceans, the friendly community blessed us.

@sundy-li
Copy link
Member

sundy-li commented Sep 5, 2021

It would be helpful to know the motivation/meaning behind the name DataFuse.

When we started this project at the beginning, its name was fuse-query. But when we noticed that its goal is not just a query engine but a data warehouse, then we started another repo named fuse-store. And then we noticed it's better to have them into one repo for better issues/prs management, so we changed it into Datafuse, the motivation is really not related to datafusion. We did not notice it's very similar to datafusion until this issue is opened.

It seems clear that you never intended to change the name

We were also searching for an easy way to rename it again. But due to many irresistible reasons, currently, it's really very hard for us to change into another name.

@l1t1
Copy link

l1t1 commented Sep 5, 2021

if rename the product, the company name and domain name need to be renamed too.
but great product doesn't relay on name only.
eg. after tencent renamed oicq to qq, it is much more successful than icq now.

@houqp
Copy link

houqp commented Sep 5, 2021

I have seen many contributions from the datafuselabs team, especially @sundy-li and @BohuTANG, to upstream projects like arrow2, sqlparser-rs and datafusion. Based on my observation so far, I do believe the team has good intention and are good citizens to the open-source community. The naming issue is perhaps due to language barriers. In English, datafusion is really similar to datafuse, so it will definitely cause confusions to your potential customers outside of China.

Anyway, The name we believe is not wrong and does not harm others, if so please let us know!

@BohuTANG I think all the other comments in this thread is the community letting the team know that the name does harm both sides :) It not only confuses datafuselabs customers, but also paints a bad look to the company as demonstrated by some of the comments in this thread.

if we give up datafuse today because of arrow-datafusion, will we need to give up even more tomorrow?

@PsiACE I think you are over thinking this :) The community is only asking for a rename that will benefit both sides. That's it. I would personally stand up and help defend other unreasonable requests if any comes up. But I really doubt you would get any in the future.

I also hope you will give us some time to work through these things
We were also searching for an easy way to rename it again. But due to many irresistible reasons, currently, it's really very hard for us to change into another name.

I am really approaching this from datafuse the companpy's point of view. If the company plans to rename the product in the future anyway, then doing it sooner than later will cause a lot less chaos. As @andygrove mentioned earlier, the community is not being unreasonable and ask the company to do a full rename, but just renaming the datafuse project/product to something else that would benefit both sides. I also agree with @andygrove that datafuselabs is a perfectly fine name for the company.

If the company really cannot change the project name due to irresistible reasons, I think it would be beneficial to share that with the community to end the discussion. If those reasons are really irresistible, I believe the community will understand and move on. To be fair, extra work of changing the domain name is not a very convincing argument.

but great product doesn't relay on name only.

@l1t1 I fully agree with you, that's why I think renaming datafuse to a more distinct name now will not harm the product and will in fact benefit it more in the long run. It would even benefit the product in the short run because rebranding a product is a good opportunity for an other round of marketing push to gain more publicities.

eg. after tencent renamed oicq to qq, it is much more successful than icq now.

In my opinion, qq is very different compared to oicq because they don't share majority of the prefix prefix. Like I mentioned earlier, perhaps this is due to language barrier. A better comparison would be qq v.s. qqs or another company launched a product called datafuser that does similar things compared to datafuse.

@sundy-li
Copy link
Member

sundy-li commented Sep 6, 2021

Hello, thank you for all the feedbacks so far, it does seem like there are some misunderstandings due to language barriers. We never had the intention to profit off the datafusion brand from the beginning and we have been collaborating closely with the community since the inception of the project.

We decided to have another round of discussion with our legal to search for potential solutions. We hope to give an update to the community in a few days.

@sundy-li sundy-li reopened this Sep 6, 2021
@sundy-li
Copy link
Member

sundy-li commented Sep 10, 2021

Hello everyone, here comes a good news:

We are going to change to a new name!

Our team has discussed this issue many times internally for months. Although the process was very tortuous, we finally reached an agreement.

The main reason is that, as @andygrove said, keeping the name datafuse will cause misunderstandings in the community.This is detrimental to both repos and the entire community. Sorry that this long-thread issue has bothered my friends from the community a lot.

Now, the process of renaming the project has started. Rebranding requires a lot of process work (domains, trademark registration, documents, codes ..), but this is all worth it, I expect it can be done ASAP.

After the new name is determined, we will make it public ASAP. No matter what the new name is, we promise the product will keep 100% free, open-source and community friendly. Thank @andygrove for pointing out the problem kindly, so that many potential risks can be prevented.

We also thank many community friends for their suggestions, especially @Dandandan, @houpq, @BatmanAoD, @junli1026, @gauss1314, @l1t1 etc.

Lastly, the open-source ecosystem needs everyone's efforts to be better. As datafuse noted in README, Datafuse is inspired by ClickHouse and its computing model is based on apache-arrow.

It's already standing on the shoulders of giants. We are also proud of that, datafuselabs will continue to embrace the whole open source community and contribute back to upstreams, especially the big data projects under the rust ecosystem.

Best wishes, may you have a good weekend.

@kingice

This comment has been minimized.

@kingice
Copy link

kingice commented Sep 10, 2021

@BohuTANG
That's incorrect.
DataFusion is a project that was started end 2016 / beginning 2017 by @andygrove with that name.
https://github.com/andygrove/datafusion/commits/master?before=219ca3f61f0d751779b4f30c02aa5fdcf8a40d12+385&branch=master
Early 2019 it was donated to Apache Arrow already:
https://arrow.apache.org/blog/2019/02/04/datafusion-donation/

Thanks for the correction.
Who have a better name ? 🤔

Why not fusedata?

@BohuTANG
Copy link
Member

@kingice Hi!
We have considered all the names have provided, this is a difficult choice as it involves trademarks as well as domain names etc.
Thank you

@andygrove
Copy link
Author

Thank you @sundy-li @BohuTANG and everyone else involved! This is really great news for the open-source communities involved. I know this caused additional work for your team and I am very grateful that you took this on and resolved the issue.

@datafuselabs datafuselabs locked as resolved and limited conversation to collaborators Sep 13, 2021
@datafuselabs datafuselabs unlocked this conversation Sep 13, 2021
@andygrove
Copy link
Author

Resolved by #1816

@BatmanAoD
Copy link

Resolved by #1816

I believe you mean #1814!

Thanks again to @sundy-li and @BohuTANG.

@houqp
Copy link

houqp commented Sep 13, 2021

Thank you @BohuTANG @sundy-li and the rest of the datalabs team for all the efforts you put in :) Go Databend!

@Dandandan
Copy link

Thank you! Databend sounds great 😁

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests