Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I’m an arm length removed from all this drama having not used Wordpress in a while, but to be honest this opinion feels overblown. To an outsider, it just looks like some legal issue between two entities irrelevant to my concerns on whether I’d use or contribute to Wordpress in the future. Something that happens between corporations all the time

My guess is there will be some settlement, one party will walk away with a better position than before, and that will be that



One of the two entities completely controls the plugin ecosystem and wielded that control against the other entity (the largest WordPress host except, possibly, Automattic itself) to block them and all of their customers out of the ecosystem over this dispute.

That's why this matters to average developers. WordPress is the plugin ecosystem, and messing around with it does as much damage to the WordPress ecosystem as left pad did to npm—it's not unrecoverable, but it's a major setback that could quickly become unrecoverable.


Note that one of the most popular plugins, to the point many WordPress people suggest it's the only plugin that should be part of core, is acf. Which is a plugin contributed by wp engine.


I believe WPEngine is actually larger than Automattic. This is an attempt to get the bigger (and likely better) company to pay the bills of a company that took a bit too much VC money and now needs the revenue to support it.


Does Tumblr make Automattic much money?


I doubt that a website they bought for $3 million from Yahoo is a significant contributor to their net income.


Is WPEngine actually the biggest Wordpress host? I'd have bet that GoDaddy had more, or maybe even BlueHost/HostGator/EIG.


The more you press the nuclear button, the more you damage your reputation. I’m of the opinion that the circumstances here are exceptional and rationality will prevail


The damage to your reputation comes because pressing the nuclear button even once is an unhinged thing to do. The reputational damage is because once you've pressed it once we consider it more likely that you'll press it again because you've shown yourself to be the kind of person who presses that button.


Yep. We no all know not to trust anything that Matt or any of his businesses or non-profits are involved with.

Unless WordPress, both the open source codebase and the plugin/theme publishing/distribution channels, are completely free of Matt's influence - they are now a serious risk any business using Wordpress needs to address and mitigate.


Matt seems to have jumped on WP Engine essentially because they were making a lot of money. So now, any other company that is making a lot of money (or hoping to make a lot money) with Wordpress may wonder whether Matt will target them too.

What's the criteria? Is there some exact revenue or profit number a company needs to stay under to avoid this sort of attack? Does Matt only get mad at hosting companies, or do other companies making a lot of money with WP (e.g. big creative agencies) need to be concerned?

Without clarity, it's hard to quantify the risk. And companies might decide to shift their CMS work elsewhere rather than deal with it. The drama undercuts one of the big advantages of WP: it was free and permissively licensed.

I agree it might not have much effect on random people using or contributing to WP. But open source projects actually need a lot of investment to grow and survive. And anything that depresses that investment can depress the overall project trajectory.


Making a lot of money and contributing effectively nothing back to Wordpress (the open source project).

In WP Engine's own complaint they say that they've "invested hundreds of millions of dollars and 14 years of hard work building a successful business to serve that community", and nothing about contributing to the project itself. Which to me indicates that they indeed have contributed very little to the project they've made a huge amount of money from.

You can argue WPE doesn't need to and that this is a feature (not a bug) of open source. But Matt Mullenweg seems to disagree and wants to use the fact that they are also capitalizing on a trademark they have no license to in order to press the issue.

From my perspective, however, it seems unlikely Automattic will win this. Automattic seems to only be enforcing their trademark as a response to WPE not contributing back to Wordpress, which is not something WPE is legally obligated to do. Trademarks need to be defended universally and without reservation.


You’re welcome to your opinion, but my company is in the market for a large enterprise CMS management contract, and this situation has taken every WordPress option off the table due to uncertainty about the business and technical stability of the ecosystem. I highly doubt we’re the only ones feeling that way.


What alternatives is your company considering now, if you don’t mind sharing? Genuine question


Probably Drupal, maybe Joomla, or maybe something more custom if in PHP-land. Otherwise one of the other CMS / DXP platforms (some require eye-watering enterprise budgets): https://cmscritic.com/key-takeaways-from-gartners-2021-magic...


DXP: When you have nine products, one logo, and a hundred salesbros.


Heh didn't know you were on HN too, this court case is like a bat signal for anyone involved in PHP historically.


Isn't it just! Karma comes around.


If you were considering Wordpress for a "large enterprise CMS", you shouldn't be making those kinds of decisions.


Can you explain your reasoning? Wordpress is pretty scalable, if done right. Even Microsoft uses Wordpress in some of its microsites. I know some of the new federal government sites even use it (for instance whitehouse.gov as part of the USDS project). It's highly supported and has been tested vigorously for decades now. It can at least be considered for various projects, even large ones.

Sometimes you don't need a complex solution for various projects, this thought process reminds me of people building highly scalable configurations for projects that just don't need it then overrunning in costs and overengineering it all.


> in some of its microsites

Exactly. That's not a large scale enterprise CMS. WordPress is fine for that.

But if you wanted to do something that requires fine-grained access control, publishing control, audit logs etc, you're not going to use WP, or you're building a CMS on top of WP. I'm doing that all day because I work in an industry that loves WP and everybody knows WP, so it's easy to collaborate. I've built dozens of plugins to accommodate for our various needs because you will hit some hard walls if you scale to large amounts of URLs, and you will run into problems with the code quality in popular plugins when you're going beyond "I just want it to look nice and work okay".

I don't hate WP. I'd consider myself a WP veteran, I've worked a lot with WP over the years, I've contributed code to core, I've found various bugs. I wouldn't ever consider it for anything that I'd call "enterprise" or "large". It's like when somebody talks about building an enterprise data management system and then says "the fact that the inventor of the CSV format liked pineapple of pizza will make sure that I won't consider CSV as the data store for my system". If they considered CSV before finding that out, they really shouldn't be making decisions in that type of project.


I guess I don't really understand what you consider "large" or "enterprise" there are very large sites that use Wordpress. I just gave you enterprise level consumers using it. You obviously aren't going to use it to build the next eBay or Amazon, but if you're a publisher or need a CMS that does publishing well? I don't really see your issue with it, if it's the right fit. Too many people try to reinvent the wheel. In my experience that more often than not leads to massive overhead and long run problems when it's in venues where it's unnecessary.

I adopted someone's node.js project once, overengineered and cost the company tons of money to run, I rebuilt it in Wordpress for a fairly large regional grocery chain and not only was it vastly cheaper, but the end product was also better and more reliable. So, I'm leery of anyone that writes something off, right off the bat. Pick the best tool for the job, not the ones that is suddenly in vogue to avoid judgement from random devs on HN or Reddit. You can't pigeonhole solutions, be it Wordpress or whatever you decide on.


Large enterprises using something doesn't make them suitable for large or enterprise-class use-cases.

When I hear "large enterprise CMS", I don't hear "a CMS that can be used by large enterprises" (because everything could, even if it's based on manually editing HTML files), I hear "a CMS suited to be used in large environments with complex requirements and no room for error".

"Enterprise" as an adjective is something that is tailored to the needs of very large entities that, due to the nature of their size, activities and legal environment, have very complex needs, and who also need to deal with things that normal website owners rarely do: legal compliance, different threat levels, audit-logs, fine-grained access privileges, publishing workflows etc.

WordPress isn't the right fit for that. WordPress isn't, and doesn't claim to be, an Enterprise CMS or targeting very large installations. Similarly: while you can manage data in WordPress, if someone suggested building a banking system on top of WordPress, I'd shake my head just the same. But that doesn't mean you can't set up a few post types and add some fields with ACF and have something you can use to organize and document your gardening efforts. They are just _very_ different requirements, and WP doesn't meet (and doesn't aim to meet) those that you associate with "enterprise" (adjective).

Like I said, I don't hate WP, I use it all the time and I know it very well. And for the vast majority of the internet, it's perfectly fine and usually the right choice because it guarantees that you'll always find somebody who can take over maintenance for your project, you'll find plenty of editors that are already familiar with your system, and there's a bajillion themes you can use and be done with it.

But if you need much more than that, you really shouldn't be using WP. Yes, you can (and I do), but you will build so much custom logic on top of it and wrestle it into behaving appropriately, that you'd be better off just not using WP. But 99% of WP sites never hit that ceiling, so for them WP is a fine choice.


Why?


Hopefully you’re right.

I expect long-term effects for both entrepreneurs and enterprise.

Companies that would’ve started in the ecosystem just got a clear signal that success comes with a tax.

And the risk-management department of every enterprise will use this* as their logical basis for choosing something with a better license.

* Matt unilaterally turning off WP updates for millions of WordPress sites is a major risk signal.


> irrelevant to my concerns

This is very relevant to anyone that cares about open source.

Companies being able to host any OSS without the threat of a trademark dispute is vital to the software industry.


Kind of a similar situation as https://en.wikipedia.org/wiki/Debian%E2%80%93Mozilla_tradema...

This didn’t kill Mozilla in the end or really affect the OSS community for Firefox or Debian


I don't think there was ever any expectation that the Firefox/Iceweasel trademark dispute would "kill" either organization. About the most serious it got, IIRC, was some heated discussions on the Debian bug tracker; there were never any legal threats involved, and the situation was ultimately resolved to everyone's satisfaction by the name change.


Mozilla didn't cut Debian users off from the plugin ecosystem or any other Mozilla services.


Matt is usually the lead developer at WP.org you can find his references in almost all changelogs. I still think, attention and resources will be diverted.


I mean sure if you want to rely on a tool where this guy will disable chunks of it because he's had a hissy fit, I guess it "won't affect you".




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: