20 years of COVER.INFO – a journey through time

Do you remember what it looked like here 20 years ago? We are proud to show you the originals of cover versions for two decades already. The beginnings are a bit embarrassing today, but we nevertheless look back on them with you.

The origins

In April 1999 two schoolboys, Thomas Wagner and Aaron Praktiknjo, wanted to put a homepage on the Internet, because it was just "in". It should be a site that benefits the general public. And because there was a cover version on the radio – which wasn't unusual in the 90s, because there were many cover versions in the charts at that time – the idea for the topic of the homepage was born: cover versions in the charts.

Brainstorming and an additional screening of the current Top 100 resulted in around 50 cover versions, each of which was compared with the (presumed) original in tabular form with their artist and title. Everything else was done by Thomas Wagner. On April 14, 1999, he finally created a banner with the name of the site on a colorful background and uploaded everything to the web server of his Internet access provider. As the path http://www1.inetmail.de/w9000185/cover/ was difficult to remember, a short and easier to remember subdomain had to be found at a provider where you could redirect to the more complicated path for free. So the subdomain cover.here.de was created. Of course, this address was also registered at two dozen search engines.

The visitors of the website – which in the beginning was in German only  were asked to report additions by e-mail. And it did not take long, until not only the site was found, but after some days also already the first people made submissions, which Thomas Wagner maintained in regular intervals.

Fast growth

From old backups we know today that on August 30, 1999, only four and a half months later, we already had 1,346 entries together. While on the first day everything had been on a single HTML page with olive-green background, this would not have made sense anymore, because the loading times would have been too long. (With the so-called 56k modem one actually reached about 40 kbit/s, similar to when today the mobile Internet was restricted). So the cover version list was divided into 5 lists, which interestingly were sorted by cover artists, not by songs. This was not a real database, but simply a listing in tables on HTML pages. That looked like this – with a newly created navigation frame:

There was no search function yet. One was completely dependent on the corresponding functionality of the web browser. With increasing growth and ongoing subdivision of the cover version list, it became more and more difficult to find the right list to search in.

Finally, on March 5, 2000, the sorting was changed to one by cover titles, and each initial letter (as well as the numbers 0-9 and the rest) got its own list. On this day a search function finally came up,which displayed all found entries from the list.

In the middle of April 2000 the 2000th cover version was already registered. It was the Jetzendorfer Hinterhof-Musikanten with "La Bamba", one of about 40 other versions of this song that were included in the list at that time. (By the way, it is not known why the entry was created later anew and is now in the database with the creation date March 7, 2004.)

Since the end of May 2001 Thomas Wagner was no longer alone with the maintenance of the website. Herbert Zach joined him and has remained a hard-working member of the editorial staff until today. Herbert had gotten to know the website on May 28, 1999 thanks to the presentation in the television show NBC GIGA and had since then regularly contributed to the dataset by e-mail.

The new domain coverinfo.de

On June 9, 2001, paid web space for the operation of the website was rented for the first time. In this context, the website cover.here.de should receive its own domain.

The desired domain would have been cover.de, but it was given to a cover band called COVER. From the news headlines it was known that there would soon be a new top-level domain: .info, which would not start until June 26, 2001. Aware of this news, the wish came up to get the domain cover.info, but you couldn't order it on June 9th. For this reason, the combination coverinfo.de was chosen as domain. Two weeks later coverinfo.de joined the discussion forum of Coverversion.de.

The site hasn't been called "Cover versions in the charts" since then, because we had long ago moved away from the initial list of songs that made it into the charts.

On September 29, 2001 we started to distinguish musical quotations (samples and replayed elements) from the cover versions in the list. Records which are full cover versions now had a "C" for cover and the others a "S" for sample in the right column (from April 28, 2003 instead a "Z" for the generic term "quote" [German: Zitat], which includes samples).

Overloaded technology

Since 2002 coverinfo.de was a little brighter and a little less colorful. In April 2002, 3 years after the launch of the website, we already had over 19,400 entries.

At that time, the whole thing was still not a real database, but a search in HTML pages, which was far from performant, which eventually brought the server to its knees.

Finally a real database

On July 1, 2002, there was finally an end to collecting the data in a large Excel file and exporting it to HTML files for the web server at more or less regular intervals using virtual basic scripts. Our MySQL database was launched. This reduced the response time of the search function and since then, changes to the database go live immediately and not after several days.

Behind the scenes we worked with Microsoft Access 2000, which was connected to the MySQL database via an ODBC interface. All the technical solutions that had been in use since 2000 were developed for us by a few hard-working users of the website. We thank Achim Kaiser, Gerd Nachtsheim, Mike Wilhelm and Björn Hutzler. Without them coverinfo.de would never have come this far, which is why we thank them for their work.

It's getting more professional

On September 1, 2003, the website presented itself for the first time in a professional looking design, which was donated to us by a user of the site, Marcel C.

In the meantime, our team had been enriched by further editors, so that after 5 years, in April 2004, almost 57,000 cover versions and quotations could be collected.

Holger Kung's work

But on October 4, 2007, the website was relaunched for the first time completely by the editorial staff itself, more precisely by its member Holger Kung. The essential design elements of the site designed by Marcel C. were preserved.

Technically, the site was completely redeveloped. This was accompanied by an improvement in the database search function, which was now more intuitive to use than before. Another new feature was that, to make it clearer at first glance, the database now displayed covers in bold type and musical quotations in lean type.

For the first time, there was a web interface for the editorial staff with which the database could be edited, so that platform-dependent additional software was no longer necessary, but a browser was sufficient. Nowadays such a thing is almost taken for granted.

After 10 years of continuous work, we had collected more than 181,000 entries in April 2009. We suffered a bitter blow on April 23, 2010, when his girlfriend told us about Holger Kung's death. His work – the redevelopment of the website and database in 2007 and a list of artists that could be mixed up – lived on for many years. His entries, which he contributed to the database, will hopefully last forever.

Unfortunately, Holger did not leave any detailed documentation of his development work. It was therefore not practicable to continue his work. However, as the size of the database increased, the solution he developed came up against its limits. On September 14, 2010 we listed 200,000 entries, on November 5, 2012 already 250,000, on June 16, 2015 even 300,000, on July 15, 2017 impressive 350,000 entries. With workarounds, the system's limitations could still be pushed back for some time, so that it continued to run with hardly noticeable limitations for the user. Towards the end of 2017, however, it was no longer possible to maintain a reliably functioning search function. Increasingly, users were only able to see part of the data records matching their search.

Today's COVER.INFO

Fortunately, at that time our editor Falko Rickmeyer, who had meanwhile been trained as a software developer, was already in the process of completely redeveloping the website – together with Adrian Semmler, who helped to design the website, and Thomas Wagner, supported by the rest of the editorial staff, who contributed ideas for redesigning the data output. For the first time in the website's history, there was no longer a table view with cover versions on the left and originals on the right, but a chronological presentation that made it possible to display chains such as cover versions that have an original that samples another song which for his part quotes another song. For the first time, this new design also features a mobile view for usability on smartphones.

On May 6, 2018, the redesigned website went online, and finally under the domain name we had dreamed of since 2001, but which had already been taken at an early stage: COVER.INFO. A domain other than coverinfo.de was necessary for the new concept anyway, because from now on the site was to be internationally oriented and therefore also made available completely in English (besides German), so that a regional domain extension would no longer fit.

The new presentation has a disadvantage, which is why there were complaints from the users: It was no longer possible to see at a single glance which songs were covered by which artists or who covered their songs. You had to click each song individually to find out. We fixed this with the extended artist view that went live on July 19, 2018. It now offers on artist pages under the link "go to table view..." again a tabular representation of all songs of an artist – but now according to the new chronological concept the other way round than before with originals on the left and cover versions on the right.

COVER.INFO is still undergoing technical development in order to increase the range of functions and user-friendliness both for the public and for the editorial staff who maintains the data. Since June 1, 2018, the task of coordinating this has fallen to the non-profit association COVER.INFO n. e. V., who took over the responsibility for the maintenance of the website from Thomas Wagner. In this way, the dependence of the website on an individual is to be minimized to make it future-proof. In addition, tax-privileged donations for the maintenance of the website become possible.

The anniversary year at COVER.INFO

We have planned to present you more background information about us here in the blog in the next months.

But now we're asking you! What from the history of 20 years of COVER.INFO do you still remember? Since when do you know us? Write it down in the comments under this article. The comment function is the replacement for the discussion forum which was closed on May 24, 2018.

EU copyright reform approved: less diversity on the Internet

The European Parliament today adopted the Copyright Directive, rejecting any amendment to the latest draft. This provides the basis for the destruction of diversity on the Internet.

The main criticism of the reform is Article 13 (after the final numbering Article 17). It is addressed to online content-sharing service providers – in particular YouTube and other social networks, but also forum operators. In the future, these providers will have to ensure that authors can put their works, for which they do not want to grant a licence to the provider, on a blacklist and achieve that these contents cannot be published on the platform. Some call it an upload filter, while others call it recognition software.

Article 17 makes service providers liable for copyright infringements insofar as they do not reasonably prevent such infringements. Those who do not have their own recognition software or cannot have it developed, will have to buy such filter services in order to avoid the liability risk. The platforms for which this is not profitable are very likely to close.

Another possible scenario is that instead of developing filtering software, providers will only allow selected, trusted users on the platform who they believe will not commit copyright infringements. YouTube could also do this. The company already has one of the best detection softwares in the world, but only for movies and music. The new Copyright Directive requires all works to be recognized, including texts, photos, sculptures, performances, and so on. Whether YouTube will further develop its software for the European market or instead delete small YouTube channels is not known.

For COVER.INFO, deleting small YouTube channels would mean losing many of the over 200,000 audio clips of songs hosted on YouTube linked to our database. Not all songs were uploaded by the copyright holders themselves, but there are also many record collectors who make sold-out records available to the interested public via YouTube. This archive of cultural assets threatens to be deleted.

In order to exclude this danger, another regulation would have been preferable, for example one that excludes providers from liability if they have concluded licence agreements with the major collecting societies, without, however, excluding the possibility that authors conclude licence agreements directly with the providers.

The detection software (or upload filters), however, also bring with them another great danger: the problem of false detections, which could also block legitimate content (so-called overblocking). For example, videos with rights-free piano music from composers who have been dead for a long time can be blocked because a record company has put a similar-sounding recording of the same piece on the blacklist. Or a song contains the purring of a cat and users can therefore no longer upload cat videos.

No matter how the providers implement the new specifications: The copyright reform will probably result in a noticeable decline in the variety of offers on the Internet.

/TWA

Save your Internet from the new Copyright Directive! [UPDATE 03/22/2019]

The EU wants to endanger freedom of expression on the Internet with an inadequate copyright directive. The German government gave the green light for this in the Council of the European Union even in violation of the coalition agreement. The implementation of the directive could lead to censorship on Internet platforms such as YouTube, Facebook, Twitter, Instagram, but also in Internet forums and comment areas under newspaper and blog articles in order to rule out copyright infringements. It's not too late to protest. On March 26, the European Parliament will vote on the directive.

The final draft of the Copyright Directive has triggered a huge wave of protests on the Internet, especially among young people, which has largely gone unheard by traditional media such as television, radio and newspapers. This only changed this week, after several thousand people in different mainly German cities had been bringing their protests from the Internet onto the streets and demonstrating mainly against Article 13 of the directive for a good two weeks. The demonstrations will continue this weekend.

In its German section, the website www.savetheinternet.info announces the current dates for demos in Germany.

Article 13 is the main point of criticism of the new DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on Copyright in the Digital Single Market. COVER.INFO first reported on November 8, 2018. Meanwhile, the final version of the draft has leaked and will be submitted to the European Parliament for a vote on March 26.

Unfortunately, this is the worst possible version of the draft, as it threatens the fundamental right to freedom of expression. Currently, platforms such as YouTube, Facebook, Instagram and Twitter work in such a way that users are responsible for protecting copyright. If the user infringes third-party rights when uploading, he is liable, but not the platform, if it deletes the content immediately as soon as it becomes aware of it.

Article 13 now requires online content sharing service providers to ensure that content can only be published if the respective platform (!) has permission from the copyright holder to do so.

Each platform should now try to acquire the rights from the authors. On this point, the directive requires something that is practically impossible. Every platform would have to try to conclude contracts with every author in the world, regardless of whether they were pictures, music, texts or other works.

Where the rights cannot be obtained, the platform should make best efforts to prevent the publication of works for which the rights could not be acquired. Otherwise, the platform shall be liable for copyright infringements.

But here, too, the proposal for the directive demands something impossible. How is a platform supposed to know whether a user upload infringes copyright? The best efforts it can make is to use so-called upload filters. This means that an algorithm compares uploaded content with known works and blocks the content if it matches. Such a computer program cannot detect whether the use of a third-party work was permitted and would therefore also block permitted use such as in satirical works or as a quotation. Especially expressions of opinion live from dealing with the opinions of others and quoting them for this purpose. Upload filters thus attack freedom of expression.

However, these filters can only block content they know. In order to store and synchronize the entire works, immense storage and computing capacities are required that only large companies can afford. Nevertheless, a liability risk remains if the filter fails because it does not know works and therefore does not recognize them or for other reasons.

YouTube already has such a filter for comparison with copyrighted films and music. It is regarded as the best and most expensive upload filter in the world, but even this filter would not meet the requirements of the directive and would have to be developed further. For example, it would have to have images and texts in its repertoire and would also have to be able to detect them in videos. Nevertheless, anyone using YouTube's filter would certainly meet the requirements of the directive to make best efforts. Therefore, we have to fear that YouTube will make the use of its filter available to other providers and thus become even more powerful, because now the contents of the other platforms are transferred to YouTube who can collect even more data.

But not every platform will be able to afford upload filters. According to the directive, however, even a newly founded platform will have to meet the requirements after three years at the latest – no matter how profitable it is. The Internet will probably be unrecognizable, especially if Article 13 of the directive is adopted, because many platforms in Europe would have to close their doors. The big ones that remain may only allow selected people to publish content; those they can trust they will not infringe copyrights.

The young people who are taking to the streets in Germany these days are protesting mainly against the parties CDU/CSU and SPD. The reason for this is that the representatives of these parties intend to vote for the copyright reform with their controversial Article 13 in the European Parliament on March 26, even though they had stipulated in the current coalition agreement that they would oppose upload filters because they considered them to be disproportionate. The German government, on the other hand, has already voted in favor of the reform in the Council of the European Union because it is predominantly reasonable. It accepted that Article 13 was not sufficiently clearly formulated.

Another criticism of the reform, by the way, is Article 11, which could also lead to the Web soon looking different from what it does today. In many social media it is now common for a hyperlink to be posted with a thumbnail and a short excerpt of the link's destination. Result pages of news search engines are also based on this principle. Article 11 requires that press publishers be given the right to demand payment if the link target is their site.

The press therefore has an interest in presenting the copyright reform as something useful. It would probably have preferred not to report on it at all, as long as heated discussions were conducted on the Internet only. Now that people are taking them to the streets, the press feels obliged to report, but in some cases, the press presents it as if the demonstrators were just children and adolescents who had been instrumentalized by YouTube. CDU politicians have even denied that the many complaining e-mails they received about the copyright reform came from people. They accused Google, as the operator of YouTube, of having generated and sent these e-mails through bots, i.e. computer programs. This can be seen from the fact that the mails were predominantly sent from Gmail addresses. The fact that this is simply because YouTube users as Google customers also have Google Mail accounts is ignored consciously.

While there will be demonstrations in some German cities this weekend and next, Europe-wide demonstrations against the copyright reform are planned for March 23. As Julia Reda, member of the European Parliament of the Pirate Party, told us, Manfred Weber (CDU), the leading candidate of the European People's Party, is said to have tried to accelerate the vote on the directive, so that the European Parliament can anticipate the citizens' protests and create accomplished facts. In any case, the version of the directive to be submitted to the Parliament has not yet been translated into all the official languages of the EU. However, the English version of the working paper was not officially made available to the public either, so that a substantive debate on the directive could only take place after it had been unofficially leaked to the outside world.

The idea of reforming copyright law, adapting it to today's realities and providing authors with appropriate remuneration is to be welcomed. Most opponents of Article 13 also see it that way. However, the provision in Article 13 is quite simply unsuitable for achieving this objective. On the contrary, it threatens to destroy the Internet in its present form, with which many creative authors earn their living. Not only would a good entertainment network be lost, but if every platform were to filter on a large scale for fear of liability for copyright infringements, freedom of expression being one of the fundamental pillars of our democracy is endangered.

That is why it is important to raise our voices and protest against Article 13 of the Copyright Directive. The locations and dates of the demos can be found at https://savetheinternet.info/demos.

Update March 22, 2019:
The draft of the directive has been finalized now with new numbering. Article 11 is Article 15 now (page 116 of the linked PDF file), and Article 13 becomes Artikel 17 (page 120). The term "online content-sharing service providers" in this last article is defined in Article 2 no. 6 (page 90). The European Parliament will vote about it on March 26. The translations into all official languages of the European Union are finished now. Julia Reda explains how to find them.

/TWA