Hi all,<p>geee ... a friend pointed me here and I have to admit that I'm quite overwhelmed by the wave my post generated. But very happy it seeded some good discussions.<p>However I should point out that only I said that I would stop providing free community support. I am only one of the contributors. Even if I am probalby one of the most active ones, there are still others probably going to step up and tackle the one or the other issue someone might come up with.<p>If the project was just me, Apache would probably have already moved the project into the atic, which is far from happening :-)<p>Now I'll continue reading all of this ... but I had to set this straight first.<p>Chris
I truely feel for the guy.<p>Reading the blog it seems like the age old story of having a superior product (drastically cut cost, better performance, etc.). However getting managers at customers to stick their neck out for it is hard. Because they know (and to a degree I can't say I blame them) changing such a critical piece of core infrastructure is going to be stressful. And what's in it for them if the company saves some money it is currently happy to spend? If it fails they sure do know what's in it for them... So big chance they'll just stick with the properietary solution they already have.<p>EDIT: I should add I think there is also another dimension to this. Things that currently work and "only" cost money are a problem solved for a company. If they have to change it they have to spend their "intelectual budget" on it. Meaning resources like technical people which are probably already scarce and working on other project need to be assigned to it.
Developments like this one and the faker.js and colors.js fiasco from last week have made me rethink my position about licensing.<p>Now it seems to me that releasing code you wrote for free under a <i>permissive</i> open-source license is somewhere between ill-advised and unethical. On one hand it will not benefit you in any way, on the other, it will be incorporated in a company's proprietary project thus lowering the expenses necessary to develop it. In the end, you've contributed to closed source and on top of that you got nothing in return... And that is the best case, as we saw with log4j, you can be held responsible in the court of public opinion in case the project you developed and maintain for free (with no warranty or liability as per license) causes an issue...
There's quite a bit of OPC UA bashing across this project. So let me chime in to keep the "balance in the universe".<p>OPC UA is a protocol to interact with an object-oriented information model. Basically <i>CORBA done right</i> to use object-oriented principles and reuse software components in industrial automation.<p>Since OPC UA is a protocol, its performance depends mostly on the implementation. Some PLCs may be crappy. But that doesn't translate into bad performance overall. My experience goes to the exact contrary.<p>Full disclosure, I lead-develop and maintain an open source OPC UA implementation that sees quite a bit of use by the big guys in the automation domain. We use C for performance. And we do have funding from the industry.<p><a href="https://github.com/open62541/open62541" rel="nofollow">https://github.com/open62541/open62541</a><p>But yes, it is hard to break into this world. Especially since solutions have to be maintained for 20+ years. A solo developer usually cannot ensure that this will still be usable some years down the line.
I have observed two common reactions in companies not willing to sponsor open-source:<p>1. "Do you mean we will pay money to develop software for our competitors? Murder!"<p>2. The OSS donations are not seen as business expense but as charity. "We don't do charity."<p>What we found effective so far is an OSS project, a consulting company and a consulting arrangement where any modifications to the core project will be upstreamed. This way, the company pays money to solve their problem (and not competitors') and it's not seen as charity (it's consulting), while an OSS project goes on. Moreover, the OSS project is itself used as a selling point for the consulting services, as a proof that we will not lock the customer in with some proprietary framework.
In case you're wondering what PLC4X is, it's a way of gluing different industrial IoC things together. <a href="https://plc4x.apache.org/" rel="nofollow">https://plc4x.apache.org/</a>
Just quoting this snippet from the landing page:<p>"No license costs, no restrictive licenses.
Being open source, PLC4X is free of charge. Start experimenting without worrying about license costs or shady license agreements. The Apache 2.0 is one of the worlds most trusted and adopted open source licenses out there."<p>Big companies need to grow up and start contributing with serious amounts. I wonder how this could get regulated with a law, for example that any company with more than x in revenue must pay y percent to an open software foundation. Maybe one needs to be created, one which monitors the space and ranks the projects on some metric like pull requests or issues closed. Then the rank is a ratio of the amount they get.
"Like with typical Machine-Learning and AI projects, the job of the data-engineer is the least fun part and everyone wants to be a data-scientist, that works with a magically curated data-lake. This is the same with industrial data-integration. Most focus goes onto the projects that allow you to do great things with industrial data, however almost nobody is willing to work on the part of actually getting the data."<p>To me, this seems like the inconvenient truth of this latest ML wave. There's lots of activity around using various ML libraies and tools, but they all assume you have input in the right format, that someone is collecting the data you need in the first place, etc.
This is dangerously close to being off-topic but it would be good to hear from people interested in PLC4X:<p>I would really like to see home appliance manufacturers extend their products with a simple 2.n IO pin interface at the back. Absolutely minimal. For example: a kettle could have an input pin to turn it on/off and an output pin that just writes whether it is on or not. A simple 4-pin interface on the back in this case. An ice-machine that has three size options and an on/off switch. The classic case of the bean-to-cup coffee machine.<p>Does anyone know if there is a standard (for manufactures) for this sort of thing? If this was a standard that added a dollar or 3 to appliances one could plug-in some IOT board that allows for Home-assistant or similar integration.<p>Of course I've completely ignored the question of whether or not a market exists.
The sad consequence of OSS is that the cost of software is fast approaching $0. Only SaaS barely manages to turn in a profit these days and I think that's simply because server-side is the only software still seen as costly. And even then, a common response from customers will be "why do I need to register". People are happy to marry a giant and sign up with Apple/Google/etc. Everyone else is expected to give their work for free and even support it. The world would be a better place if devs collectively stopped this
Glad he gave up. This is the kind of project industries should be funding. It seems like a bunch of factories got his labor for free. I love open source but its better if its something that scratches you own itch, or someone funds it.
A good compromise for this kinds of things is usually dual licensing:<p><pre><code> - If you're big enough to make real money and you need my attention then pay me this much please
- If you're some rando with a toy project, then we have a github issue tracker
</code></pre>
A project this size and age should always have a lawyer at hand just in case, and they should be able to walk the maintainers through the process of dual-licensing everything.
I think this developer fails to address a major concern - companies looking for industrial control software want support. From the page, it looks like your support options are two guys, one in Germany and one in Poland. That would be an immediate non-starter for our company.<p>The alternative would be to hire one or two developers on our own for support, and then the cost savings would be out the door. Maybe this library is more performant, but it's a (relatively) small piece of the industrial control scheme. A massive shift like this is just not something companies will take a risk on.
There’s a reason why the majority of costs in running a tech business are often in the non-technical parts.<p>Don’t open source unless you can afford to. Invest in non-technical bits of your organisation. Avoid making choices and then hating the consequences.
The title is misleading -- one developer (a major one, mind you) is stepping back from offering free community support. This is not "Apache PLC4X" announcing an end of community support, it is one of the core developers announcing <i>he</i> will be doing so.
An important Open Source lesson: it isn't really a business model in and of itself. It's worth it if you're scratching your own itch: running it on hardware that you have anyway, to do things that you want done. Doing it purely for others is not sustainable.
> On the one side there’s no point in convincing the IT folks that what you’ve got to offer is awesome, if the OT folks don’t know about it. The OT folks are the ones doing the decisions. Everything in the OT world is pay-to-play. You’ll never get an article posted in an OT magazine, if you don’t pay for it. In the OT world there are also almost no conferences, like in the IT world. Everything is about industrial fairs with extremely expensive booths. The budgets the big players have at their disposal are simply unbelievable. As an open-source project you have no chance of being noticed at all.<p>This is a key lesson. Open Source has not won in this part of the technology world, and it's a HUGE walled garden.
I'm assuming [1] didn't work very well then:<p>[1]: <a href="https://plc4x.apache.org/users/commercial-support.html" rel="nofollow">https://plc4x.apache.org/users/commercial-support.html</a>
Maybe I missed something, but I had to search for the repo : <a href="https://github.com/apache/plc4x" rel="nofollow">https://github.com/apache/plc4x</a>
I thought this blog post ironic since it said stuff like "free trial of open source" and "Industry 4.0."<p>I also think it's crazy that the government forces businesses to be profitable.
I have mixed feeling - almost all vendor-specific stuff in this space is a mess.
Definitely "yesterday's technology, tomorrow", done in a way that is often very hard to use.
This is a space that mainly wants stuff that has worked and will work for 20-30 years.
Monitoring/etc is often still an optional feature.
If I am ever rich enough to engage in philanthropy for open source projects, I’ll let an admittedly petty pet peeve drive some of my decisions about which projects to shower cash onto: If you define your acronyms and tell me up front what tf your project does without making me google it, I’ll support you.
I think building the stuff YOU (or your company) need for free, and offering paid support to build stuff other need is exactly how open source should work. If no one has interest in paying the paid support, I guess there is no business opportunity, and it's better to focus on something else.
This is not really as much about open source as it is a larger issue: people are sheep. That includes developers and especially managers and executives. Few people are capable or willing to make decisions based on technical merit. They generally follow trends mindlessly or make decisions about tools based on political expediency or following whatever their local hers leader is doing.<p>This has caused me quite a lot of frustration over the years because I like to make tools and solve problems. But often building a tool or coming up with a framework for solving a problem, even if it's technically really challenging, is only a fraction of the battle. It's getting any of the sheep to pay attention, notice, or decide your solution can benefit them in their sheep power struggles that is actually the biggest challenge.
> In Germany a company needs to be profitable. If you run your business at a deficit for too long, you get your business closed, and then you usually even have to pay back a lot of taxes.<p>What is this about? The government forcibly closes companies that lose money?
This strikes me weird. This software seems to be under the "Apache" umbrella. But this guys last plea for support is on his own blog. Beyond the "how do we fund open source" thing in the piece, I'm left with too many "what's going on" questions around this. Does Apache support it? Why or why not?<p>Getting back to the blog, he says the comms piece is his unique selling point. OK, so people don't want to support the open part but yet it's missing a rather important piece. Is that correct? The whole thing just left me a bit confused.
It seems like something that should be backed by an industry consortium; member companies pay money to the foundation and that employs the author and other folks to work on the code etc.
<i>Believe it, or not. We still failed at getting customers to go beyond our incredibly successful POCs. I’ll write a separate article on this. What we were told off the record was, in every case things didn’t proceed due to politics and not our performance.</i><p>Sorry to hear it didn't work out, Chris - looking forward to your article. PLC4X was definitely one of the more interesting projects at codecentric.
It's not surprising that people and especially companies are not going out of their way to pay for a free product.<p>Instead of being hurt by how ungrateful the world is, maybe it would be better to find a way to make the rich pay and let the poor (like me) still get the software for free.<p>It's not like it's unheard of in the open source world.
open source is a great code license scheme, but a horrible business model.<p>Shame it's not worked out for him, but when it's voluntary most businesses hear "free tool"
If you don't want people to rip you off without giving back, then use GPL and not Apache license.<p>People in favor of the Apache license are people who like to "pay you in exposure".<p>Turns out you can't eat exposure.<p>Well you wrote that code, you want people to use it, right? Then attach no strings!<p>Turns out I also need to pay rent and and buy food.
Guys this is just beggining, I am working in cyber security since 2013, After the discovery of log4shell, almost whole internet is at risk of supply-chain attacks, Hackers are continuously targeting open-source packages because it's really easy way to hack big organization.
We literally track 100's of attack on daily basis. Services like vulert.com can play a vital role, let me know what do you think.<p>Best