Is Multicorn dead? about multicorn

[luss]

Tonight I worked on the code a bit in https://github.com/pgsql-io/multicorn. The code in the master branch now compiles (and probably runs) on PG14 (and still compile and runs on PG10 thru PG13. Later this week I'll do some testing. Who wants get to get involved?

[df7cb]

I will drop the Multicorn packages now.

[luss]

Hi Guys,

There is nothing sneaky or sinister about the pgsql-io fork of Multicorn (https://github.com/pgsql-io/Multicorn). My need, at the time, was for Multicorn to support PG13. Patch #260 seemed to do that so I merged it into my fork and tagged it as v1.4.1 in my fork so I could easily use it. Multicorn is kewl and I don't want it to die.

--Luss

[denishpatel]

@luss Any chance you could able to make Multicorn work with Postgres 14? I tried https://github.com/pgsql-io/Multicorn but getting error during make.

[luss]

I have just added, Issue Tracking, Discussions and Wiki to https://github.com/pgsqlio/multicorn. Now all I need is for other contributors and reviewers and a true long lasting open source project will be born. Run by interested individuals, that is independant of any one company (or individual even). On Mon, Apr 4, 2022 at 11:31 AM Denis Lussier ***@***.***> wrote:

…

Francesco, what is your github user-id? On Mon, Apr 4, 2022 at 11:30 AM Denis Lussier ***@***.***> wrote: > I would definitely love to share the burden with others. It's a great > idea we can start with you and being commiters and then we can develop > process and add more users as appropriate. I just tried to create a new > multicorn user-id in github, but, someone create that a while ago and it > seems to be a dead account. > > > > On Mon, Apr 4, 2022 at 11:23 AM Francesco Frassinelli < > ***@***.***> wrote: > >> Thanks @luss <https://github.com/luss>. Your repository is a bit behind >> the current master, and it is not possible to open issues there at the >> moment. I am comfortable using the current repository, but I wonder if you >> wouldn't prefer to share the burden of maintaining such software with >> others. Upgrading can be trivial, but there are various PR waiting to be >> evaluated/merged, potential improvements and some segfaults. I am no >> Postgres expert, but if there is a small group of people interested in >> supporting multicorn, I would be glad to help, as it can be such a nice >> piece of software. >> >> — >> Reply to this email directly, view it on GitHub >> <#271 (comment)>, >> or unsubscribe >> <https://github.com/notifications/unsubscribe-auth/AAMWOHQFP25FCEQDREZ5PDTVDMCODANCNFSM5IDZRUXQ> >> . >> You are receiving this because you were mentioned.Message ID: >> ***@***.***> >> >

[df7cb]

@frafra: Note that https://github.com/pgsql-io/Multicorn/ just happened to merge #260, it's not "my" fork. They sneakily dropped the "WIP" marker on that commit (without actually fixing the issues with it I guess), which might speak about the state of that fork as well.

[frafra]

@df7cb Sorry for the confusion, and thank you for having clarified that.

[df7cb]

I've spent some hours looking at this. I now believe #260 to support PG 13 is actually correct and should be merged.

But.

• PG14 is a lot more complicated. See #272.
• PG14 moved the internal numbers for error levels around, which unfortunately means the compiled python module isn't compatible with all PostgreSQL versions anymore (i.e. needs to be compiled for each PG major separately). That means at least the Debian packages need major reshuffling (and will likely not be co-installable anymore, which is major trouble when upgrading PG versions), or Multicorn finds a way to fix that somehow.

[df7cb]

Fwiw, I am seriously considering to drop Multicorn from both Debian and apt.postgresql.org, the package has accumulated enough cruft by now.

https://bugs.debian.org/1000589

[frafra]

The company behind multicorn is no more: https://opencorporates.com/companies/us_ca/C4537003
If no one picks it up, it can be declared dead, as well as all the code depending on it.

[luss]

Ahhhh. I didn't realize that. I have hosted a new version of it at https://github.com/pgsql-io. But probably some folks aren't comfortable with that. Perhaps rehosting this project at https://github.com/Multicorn/MUlticorn makes most sense. Most times it's pretty trivial to upgrade multicorn, as is, to a new yearly verasion of postgresql. It so happens, that the latest upgrade is unusually tricky.

…

On Mon, Apr 4, 2022 at 10:59 AM Francesco Frassinelli < ***@***.***> wrote: The company behind multicorn is no more: https://opencorporates.com/companies/us_ca/C4537003 If no one picks it up, it can be declared dead, as well as all the code depending on it. — Reply to this email directly, view it on GitHub <#271 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMWOHVNNWUL4HUBOSVW6NDVDL7TFANCNFSM5IDZRUXQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[frafra]

Thanks @luss. Your repository is a bit behind the current master, and it is not possible to open issues there at the moment. I am comfortable using the current repository, but I wonder if you wouldn't prefer to share the burden of maintaining such software with others. Upgrading can be trivial, but there are various PR waiting to be evaluated/merged, potential improvements and some segfaults. I am no Postgres expert, but if there is a small group of people interested in supporting multicorn, I would be glad to help, as it can be such a nice piece of software.

[luss]

I would definitely love to share the burden with others. It's a great idea we can start with you and being commiters and then we can develop process and add more users as appropriate. I just tried to create a new multicorn user-id in github, but, someone create that a while ago and it seems to be a dead account.

…

On Mon, Apr 4, 2022 at 11:23 AM Francesco Frassinelli < ***@***.***> wrote: Thanks @luss <https://github.com/luss>. Your repository is a bit behind the current master, and it is not possible to open issues there at the moment. I am comfortable using the current repository, but I wonder if you wouldn't prefer to share the burden of maintaining such software with others. Upgrading can be trivial, but there are various PR waiting to be evaluated/merged, potential improvements and some segfaults. I am no Postgres expert, but if there is a small group of people interested in supporting multicorn, I would be glad to help, as it can be such a nice piece of software. — Reply to this email directly, view it on GitHub <#271 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMWOHQFP25FCEQDREZ5PDTVDMCODANCNFSM5IDZRUXQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[denishpatel]

Thanks, @luss ! You had wrong link. Here is the correct link Link

[luss]

the correct link https://github.com/pgsql-io/multicorn I am @luss on github, my work email is ***@***.*** and my personal email is ***@***.*** I read and respond to them both in the same place from my Gmail account.

…

On Mon, Apr 4, 2022 at 4:13 PM Denish Patel ***@***.***> wrote: Thanks, @luss <https://github.com/luss> ! You had wrong link. Here is the correct link <https://github.com/pgsql-io/Multicorn> :) — Reply to this email directly, view it on GitHub <#271 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMWOHQVDEP2Y675K2QRTYLVDNENJANCNFSM5IDZRUXQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[frafra]

I might help with CI/tests and/or with Python code, but I do not think I can help much with Postgres specifics in the short term.

[luss]

Kewl, I think I've done the PG code mods such that they will work. I use multicorn on the matthewfranglen <https://github.com/matthewfranglen> / * postgres-elasticsearch-fdw <https://github.com/matthewfranglen/postgres-elasticsearch-fdw> *project, but, I am also interested in writing a DataWrangler FDW for AWS that would also use Multicorn. How do you use Multicorn. Lets not try and do too much in our first release such that we don't have a first release. :-)

…

On Tue, Apr 5, 2022 at 3:54 AM Francesco Frassinelli < ***@***.***> wrote: I might help with CI/tests and/or with Python code, but I do not think I can help much with Postgres specifics in the short term. — Reply to this email directly, view it on GitHub <#271 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMWOHURBCFM5OTBLVJFLIDVDPWT7ANCNFSM5IDZRUXQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[denishpatel]

Tonight I worked on the code a bit in https://github.com/pgsql-io/multicorn. The code in the master branch now compiles (and probably runs) on PG14 (and still compile and runs on PG10 thru PG13. Later this week I'll do some testing. Who wants get to get involved?

See pgsql-io/multicorn2#4

[luss]

Francesco, what is your github user-id? On Mon, Apr 4, 2022 at 11:30 AM Denis Lussier ***@***.***> wrote:

…

I would definitely love to share the burden with others. It's a great idea we can start with you and being commiters and then we can develop process and add more users as appropriate. I just tried to create a new multicorn user-id in github, but, someone create that a while ago and it seems to be a dead account. On Mon, Apr 4, 2022 at 11:23 AM Francesco Frassinelli < ***@***.***> wrote: > Thanks @luss <https://github.com/luss>. Your repository is a bit behind > the current master, and it is not possible to open issues there at the > moment. I am comfortable using the current repository, but I wonder if you > wouldn't prefer to share the burden of maintaining such software with > others. Upgrading can be trivial, but there are various PR waiting to be > evaluated/merged, potential improvements and some segfaults. I am no > Postgres expert, but if there is a small group of people interested in > supporting multicorn, I would be glad to help, as it can be such a nice > piece of software. > > — > Reply to this email directly, view it on GitHub > <#271 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AAMWOHQFP25FCEQDREZ5PDTVDMCODANCNFSM5IDZRUXQ> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> >