Hah, so MariaDB was created because the creator of MySQL got pissed about the acquisition by Sun/Oracle, and now MariaDB might be acquired as well? Making bank while forking their project over and over isn't a bad way to make a living it seems.
As a long term MySQL user I think getting acquired by Oracle pretty good for the product. The old, hard issues that MySQL never had the engineering to tackle themselves finally got resolved.
Atomic DDL (possible because of a complete rewrite of the underlying storage format), proper Unicode by default (did you know that utf8 in MySQL is not really utf8? hit up your favorite search engine for utf8mb3 vs utf8mb4), strict typing by default, plus a horde of new features many MySQL users have been waiting like CTEs and window functions.
I too am a fan of what Oracle did for MySQL. There haven't been any major releases since 2018, though. Since their development process is closed, I am wondering when/if are we going to see MySQL 9 (if it's anything like 8, it's going to be huge).
MariaDB initially went through a process of adding absolutely everything they could into the database (e.g: many additional storage engines like Cassandra and TokuDB), most of which since have been removed. It felt less like it was driven by a strong engineering process, and more like let's throw everything and see what sticks (not much did).
> There haven't been any major releases since 2018, though. Since their development process is closed, I am wondering when/if are we going to see MySQL 9 (if it's anything like 8, it's going to be huge).
MySQL since 8.0 is on a "rolling release" model, where features were shipped as part of 8.0 once they were finished instead of waiting for a 9.0 release. Thus 8.0 got quite a few improvements since 8.0.0.
didn't later versions of mysql 5 also have utf8mb4? Or did they do something else? I also wonder why mysql wouldn't have had the resources for strict typing by default, I'm fairly ignorant of the complexities involved, but since this involves just checking data types on insert, it would seem to my non-expert understanding like not that big of a task to add. I don't say that to disagree with you or disparage oracles work, I'm just trying to understand your point of view better.
The big difference here is that Runa Capital is a venture capital firm, and they already invested in MariaDB 8 years ago. To be honest I don't really understand how this acquisition makes sense in that context.
It's trading at a market cap of around $31 million, which is below even its revenue. Could make sense to buy it, improve operating metrics, and then flip it. (Or buy it, make it profitable, and earn a return that way.)
I'm surprised the employees aren't making a buy out offer at this point. I'd love to own a reasonable fraction of the company I work for at that sort of price and would happily throw in some of my savings.
well how many employees are left? just cause the company has revenues doesnt mean they have the wealth to buy their employer. ofc i'd absolutely love for this to happen. but also fair warning to any mariadb employees seriously considering this, figure out your renewal strategy first
I think that if you're working for a company facing this sort of situation you wouldn't be as bullish. On top of that you'd have ownership through stock options which you didn't need to fork over savings for.
That makes sense, except the numbers may not match what people have/are willing to risk on their employer. $31m divided 500 ways (guessing @ staff size) is $62k each. That's a big bet in one place, and one that is likely to be out of reach for some staff.
Pumping my savings into a struggling company so that when it eventually fails I'm not only unemployed but also out of savings? That's a concentration risk I wouldn't want to have if I were a MariaDB employee.
The key words are “reasonable” and “concentration risk”.
Companies not run by shareholders but employees are often unremarkable but do well in long term. From time to time family owned companies choose that path.
MariaDB is out of cash and losing $12 million per quarter.
Would you also throw in your savings next quarter to pay for the new losses? And the next quarter? And after that?
Paying $30M today isn’t enough when the business is making heavy losses and nearly out of money. You need to have a plan to either raise a lot more money very soon, cut costs, or probably both. Employees aren’t well positioned to do either.
MariaDB already had its exit last December. It went public via a SPAC deal.
Like most SPACs, the stock hasn't performed well. Its market cap is now around $30 million, which is substantially less than it raised in venture funding.
If they want to remain a public company, they'll need to do a reverse-split soon to get the stock price above $1/share, otherwise they'll be de-listed.
Receiving an unsolicited offer compels them to publish this notice, and by their declaration they were not seeking that offer (“unsolicited”).
Technically, I think anyone can make an offer on any company operating in their legal jurisdiction, and that would compel the company receiving the offer to publish an equivalent press release.
I don’t blame them for having nothing to add at this time other than “here is the legally required declaration, please wait”. Even a flat refusal takes time to prepare, in order to meet all the necessary legal forms.
(This is not legal advice, please consult a legal professional regarding the validity of these statements in your jurisdiction(s), etc.)
I am so confused, for the lay person - can someone explain how does something with a copy-left license the GPL end up in this situation of being "acquired"? We can just fork it right?
I have lately been wondering if culture has forgotten how to run any organisation apart from a for profit corporation or a KPI driven NGO. E.g. an association, a co-operative etc. These all require collaboration and entrepreneurship and innovation without it being a competition via KPI or everything justified by the dollar (as is often the case sometimes even in charities)
GPL is what allowed Monty to fork after selling the MySQL trademark and copyright to Oracle. So Monty lost the ability to dual license, yet gained much money. Arguably with GPL2 he could even fork with proprietary plugins created later. Or he may have negotiated a backroom deal with Oracle to keep some relicensing rights.
But at risk because a large company can offer the support as part of a larger offering. For example, if you were a large company on AWS or Azure, would you rather get your MySQL/MariaDB support through MariaDb or through Amazon/Microsoft?
It's a great choice for higher budget applications too. At my job (public tech company, former startup), we use Postgres as our source of truth for just about everything that should go in an OLTP database (e.g. not analytics). I don't think there's another database product that would work as well for us as Postgres.
The way they wrote a full sentence, that seems to be about Postgres, before introducing the word "Postgres" in the second sentence, does make it appear like they were confused and thought this thread was about that. (As the other logical reading would be that the first sentence is complimenting MariaDB before switching tack for sentence two. What they needed to do to not cause the confusion was to start with "Meanwhile, Postgres just gets...")
In Japan, "family" business owners often adopt adult children with the sole intention of them taking over the business without it leaving the "family".
That makes it a bit different than something like Mongo or Elasticsearch suddenly going with a business friendly license.
However, I still would wonder about the health of the committership community independent of this company. The job of a good foundation is to steward the project accordingly, and I wonder if MariaDB has been community or company governed?
I wonder where MariaDB and MySQL will be in like 5-10 years. The whole SPAC thing was weird to hear about and now talk of possible acquisition.
I picked MariaDB for a freelance project recently, because it seems to work well with everything I self-host: Matomo Analytics, Apache Skywalking APM, BookStack Wiki, Keycloak for user management and for any of the .NET (or Java) services that I'm going to be running.
Aside from DDL not being transactional (like in PostgreSQL) it's nice to use, has reasonable resource usage and in some ways is pretty simple (even things like having just databases instead of the database/schema separation, which I don't necessarily need). Also, tooling like MySQL Workbench is great and stuff like Bitnami container images make spinning up new instances easy as can be.
It's good that for most basic use cases switching between MySQL and MariaDB is still doable, yet I still hope that both of the projects will stick around and will remain alive. Admittedly, you mostly hear about PostgreSQL on HN - though a litmus test like Google Trends or the DB-Engines Ranking site both suggest that it might be over represented here a bit (lots of cool features to talk about).
All my best wishes to Michael Widenius (Monty) and David Axmark, I hope they make bank (again!), they deserve it!
After all of these years I trust them to do the right thing with the open source aspect.
I've never met a more competent and energetic technical "salesperson" than David. I got the impression he spent at least a decade travelling full time, evangelizing MySQL (sales) in the 90s/00s. Mad respect. This is what it took. It didn't happen by itself.
The difference is that MariaDB is over ten years old and has pursued revenue growth for years, so hopes and dreams can’t obscure the track record of accumulating losses.
MariaDB has $50M annual recurring revenue but is spending almost $100M in the same time. There’s only a few millions of cash left, and liabilities outweigh assets. It’s on the brink of bankruptcy.
If you bought this company for $30M now, you’d have to pay at least another $30M just to make it through the next few quarters. The investors acquiring the company must be planning big cuts.
Those of you thinking how crazy it is for MariaDB to be valued at just $30M: This company acquired Clustrix[0] in 2018, so it's not even only a MySQL fork.
30M is an absurdly small number for the embodied IP. (I worked briefly @ Clustrix, the tech is legit, ~$75M funding burned)
78 comments
[ 5.1 ms ] story [ 234 ms ] threadEdit: my bad - Sun acquired it, not Oracle. Made me all sad for Suns demise again.
I too am a fan of what Oracle did for MySQL. There haven't been any major releases since 2018, though. Since their development process is closed, I am wondering when/if are we going to see MySQL 9 (if it's anything like 8, it's going to be huge).
https://dev.mysql.com/doc/refman/8.0/en/mysql-nutshell.html
MariaDB initially went through a process of adding absolutely everything they could into the database (e.g: many additional storage engines like Cassandra and TokuDB), most of which since have been removed. It felt less like it was driven by a strong engineering process, and more like let's throw everything and see what sticks (not much did).
MySQL since 8.0 is on a "rolling release" model, where features were shipped as part of 8.0 once they were finished instead of waiting for a 9.0 release. Thus 8.0 got quite a few improvements since 8.0.0.
Recently they announced a new model where they split LTS style versions and increase version number, again, but features still appear as they are done and not for a big release: https://blogs.oracle.com/mysql/post/introducing-mysql-innova...
Oracle is not a tech company but a law firm that happens to have an engineering department.
Founder's investment to start bootstrapped is often opportunity cost of time (years of work instead of earning salary). Maybe some cash invested too.
Or with VC, after years of work a founder's biggest asset is the common shares that are worth $0 if company folds.
Companies not run by shareholders but employees are often unremarkable but do well in long term. From time to time family owned companies choose that path.
Would you also throw in your savings next quarter to pay for the new losses? And the next quarter? And after that?
Paying $30M today isn’t enough when the business is making heavy losses and nearly out of money. You need to have a plan to either raise a lot more money very soon, cut costs, or probably both. Employees aren’t well positioned to do either.
Like most SPACs, the stock hasn't performed well. Its market cap is now around $30 million, which is substantially less than it raised in venture funding.
If they want to remain a public company, they'll need to do a reverse-split soon to get the stock price above $1/share, otherwise they'll be de-listed.
Technically, I think anyone can make an offer on any company operating in their legal jurisdiction, and that would compel the company receiving the offer to publish an equivalent press release.
I don’t blame them for having nothing to add at this time other than “here is the legally required declaration, please wait”. Even a flat refusal takes time to prepare, in order to meet all the necessary legal forms.
(This is not legal advice, please consult a legal professional regarding the validity of these statements in your jurisdiction(s), etc.)
I have lately been wondering if culture has forgotten how to run any organisation apart from a for profit corporation or a KPI driven NGO. E.g. an association, a co-operative etc. These all require collaboration and entrepreneurship and innovation without it being a competition via KPI or everything justified by the dollar (as is often the case sometimes even in charities)
https://news.ycombinator.com/item?id=3546
Source: Used to work there.
But at risk because a large company can offer the support as part of a larger offering. For example, if you were a large company on AWS or Azure, would you rather get your MySQL/MariaDB support through MariaDb or through Amazon/Microsoft?
By all means, use what you’re comfortable with but not having stuff like transactional DDL feels backwards at this point.
can confirm
Or has the acquirer somehow plugged Monty's favorite end-run around the acquisition game this time around?
Recently, the MariaDB board started to pursue additional private equity and a dilution of shares and this is part of Runa's response.
More information is available in this SEC filing : https://d18rn0p25nwr6d.cloudfront.net/CIK-0001929589/2335caa...
https://mariadb.com/kb/en/mariadb-licenses/
That makes it a bit different than something like Mongo or Elasticsearch suddenly going with a business friendly license.
However, I still would wonder about the health of the committership community independent of this company. The job of a good foundation is to steward the project accordingly, and I wonder if MariaDB has been community or company governed?
The core database server will always be GPL though, especially as a fork of MySQL under the GPL.
MariaDB has a community and foundation which (afaik) is independent of the company.
I picked MariaDB for a freelance project recently, because it seems to work well with everything I self-host: Matomo Analytics, Apache Skywalking APM, BookStack Wiki, Keycloak for user management and for any of the .NET (or Java) services that I'm going to be running.
Aside from DDL not being transactional (like in PostgreSQL) it's nice to use, has reasonable resource usage and in some ways is pretty simple (even things like having just databases instead of the database/schema separation, which I don't necessarily need). Also, tooling like MySQL Workbench is great and stuff like Bitnami container images make spinning up new instances easy as can be.
It's good that for most basic use cases switching between MySQL and MariaDB is still doable, yet I still hope that both of the projects will stick around and will remain alive. Admittedly, you mostly hear about PostgreSQL on HN - though a litmus test like Google Trends or the DB-Engines Ranking site both suggest that it might be over represented here a bit (lots of cool features to talk about).
https://finance.yahoo.com/quote/MRDB/
It's at penny-stock-level status with low volume.
https://www.postgresql.org/
After all of these years I trust them to do the right thing with the open source aspect.
I've never met a more competent and energetic technical "salesperson" than David. I got the impression he spent at least a decade travelling full time, evangelizing MySQL (sales) in the 90s/00s. Mad respect. This is what it took. It didn't happen by itself.
MariaDB has $50M annual recurring revenue but is spending almost $100M in the same time. There’s only a few millions of cash left, and liabilities outweigh assets. It’s on the brink of bankruptcy.
If you bought this company for $30M now, you’d have to pay at least another $30M just to make it through the next few quarters. The investors acquiring the company must be planning big cuts.
30M is an absurdly small number for the embodied IP. (I worked briefly @ Clustrix, the tech is legit, ~$75M funding burned)
[0] https://en.wikipedia.org/wiki/Clustrix