What is a bad result of greed

Donation greed? Why Wikipedia is more successful


“That is a really ingenious code, but unfortunately the risk we buy with it is too high for us. We have therefore decided on a different investment. ”So that you don't have to hear this or a similar sentence when using your product or an exit, you should avoid the following five coding errors from a legal point of view.

1. Be careful when using open source.

The source code of open source software is provided free of charge. This is especially interesting for young programmers and is often integrated in their own code, but caution is advised here.

Licenses for open source software often contain so-called "Copyleft" clauses. Such clauses oblige licensees to also make edits of the source code available free of charge. The best known copyleft license is GNU General Public License (GPL)

So if you build your code on open source, which is, for example, under GPL, the copyleft effect is applicable to your code. You would therefore have to make your source code available free of charge and you could have difficulties exploiting the code (e.g. when demanding license fees).

Key takeaway: So check who integrates (open source) code into their own software.

Don’t forget: Of course, you also have to check the license conditions for paid software to find out whether exploitation is possible.

2. Code documentation - more than just "comments"

Yes, the main thing is that your code works, but accurate documentation of the code is often just as relevant for investors.

In order to impress potential investors, it is not only advisable comments to the code, but to document the individual coding steps in more detail, for example in a README file. Depending on what the application area of ​​your code is, API documentation may also be appropriate.

Key takeaway: Documentation is key - also from a legal point of view.

3. Coding together - a smart move?

Programming together, whether as an employee or as a business partner - which sounds so nice, is not always a smart move from a legal point of view without "safety precautions".

The copyright law basically provides for programmers by (i) providing copyright protection for computer code and (ii) stipulating that if employees code for the employer, the rights of use to the programmed pass to the employer ... So everything is easy? Unfortunately not!

The UrhG - and thus also this favorable position for the employer - only applies if the employee programs a computer program within the meaning of the UrhG. If only part of it or only individual algorithms are programmed, this provision may not be applicable. In this case, it is necessary to make a supplementary agreement with the employee (this can happen in the employment contract, for example).

If you code together with your business partner, it may be possible that you both have what are known as "Co-author“Of the code. You could only use the code together - you should keep that in mind. Your business partner will often want it that way anyway. It could get legally tricky if a friend helps you significantly with coding and you haven't thought about cooperation / rights / exploitation until then.

Key takeaway: Think early on who should have which rights to the code and secure this legal position contractually.

4. Attention in connection with input data

The development and training of algorithms requires data - only through examples can an algorithm learn to recognize patterns in data.

Depending on which data you feed your code, you have to observe further regulations. As soon as you get to workout personal data must process, the provisions of the GDPR apply.

Personal data is all information that relates to an identified or identifiable natural person, such as name, age, personal preferences, email address or photo. It is often not entirely clear what personal data can represent (for example, the European Court of Justice has held that in certain cases IP addresses are also personal data). Before processing your data, it is therefore important to clarify whether these are personal.

Data processing is only lawful under certain conditions, in particular with the consent of the person concerned or for the necessary fulfillment of a contract. In addition, the processor has extensive obligations (data deletion, correction, organizational precautions, etc.). Failure to comply with the GDPR is threatened with heavy fines.

In the run-up to development, it should therefore be taken into account which data will be processed by an algorithm and how the requirements of the GDPR can be complied with without additional effort (contractual basis, obtaining consent, etc.). Otherwise there is a risk of additional costs later due to the subsequent adjustment or even penalties.

Key takeaway: Check which data you are feeding your algorithm during development in order to avoid nasty surprises and time-consuming reprogramming later.

5. Contractual confidentiality

As already said before, it is possible that your code or parts of it (e.g. mere algorithms) do not enjoy copyright protection; But there is also a solution for such cases.

Program codes can namely be a "Trade secret“And thus be protected from competitors under competition law - but watch out, it is important here too to the protection early on to think. In order for a code to be a trade secret within the meaning of the UWG, it must meet the following requirements: lack of disclosure, will to maintain secrecy and interest in secrecy.

Lack of disclosure means that information is not generally known and is actually kept secret. They may only be made known to a limited group of people (e.g. employees, testers). Therefore should Nondisclosure agreements met with people who have access to the code to prevent further spread. Often a mere confidentiality agreement in the service contract is not enough; but this has to be clarified in each individual case.

The Secrecy must be expressed from external circumstances. Nondisclosure agreements or technical protective measures (Access restrictions, etc.) indicate the will to maintain secrecy.

Contractual or technical protective measures allow that too Interest in secrecy detect.

What is the point of protection as a trade secret? The unfair exploitation of trade secrets for competition purposes is a criminal offense. In addition, claims for damages and injunctive relief can be asserted against competitors. Especially when codes do not meet the requirements for copyright protection or a patent, competition law is of particular importance. In addition to technical precautions, contractual confidentiality obligations should also be provided for.

Key takeaway: If you want to protect your code (also) as a trade secret, think about it early, Which contractual and not contractual (e.g. technical) activities you are going to put to achieve such protection.

In order to protect your code as best as possible and then use it, there are a few things to consider, whereby the most important of course remains the functionality of your code and the fun of programming.

About the author

Martin Hanzl is a Senior Associate at EY Law Pelzmann Gall Groß Rechtsanwälte and advises clients there on issues relating to new technologies, among other things. He is also in the project management of the Blockchain and Smart Contracts project of the European Law Institute and publishes regularly on legal topics relating to new technologies, blockchain, smart contracts and digitization.