Martin Aschoff

Im EOS Directory geht es nicht um Open Source Software, sondern um Enterprise Open Source Software (Enterprise OSS). Das ist zwar im Namen nur ein kleiner, in Wirklichkeit aber ein sehr großer Unterschied! Der Namenszusatz "Enterprise" bedeutet, dass die Software für den professionellen Einsatz in geschäftskritischen Bereichen eines Unternehmens geeignet sein muss. Und die aus diesem Einsatzbereich resultierenden zusätzlichen Anforderungen sind signifikant, wie die Praxis zeigt.

So beinhaltet der Einsatz einer Open Source Software in Unternehmen auf der einen Seite funktionale Anforderungen wie

• Schnittstellen für die Integration in eine existierende IT-Infrastruktur
• Berücksichtigung von Zeitzonen
• Mandantenfähigkeit (sofern erforderlich)
• eine mehrsprachige Benutzeroberfläche (sofern eine GUI vorhanden ist) mit Unterstützung verschiedener Zeichensätze
• die Unterstützung älterer Browser-Versionen wie Internet Explorer 6 (für Web-Applikationen)

Auf der anderen Seite kommen nicht-funktionale Anforderungen hinzu, wie zum Beispiel

• ein stabiler und nicht zu exotischer Software-Stack, der zum Betrieb der Enterprise OSS erforderlich ist (Application Server, DBMS, ESB, o.ä.)
• die einfache Skalierbarkeit der Software (z.B. durch Caching, Verteilung auf mehrere Server, etc.)
• die Sicherheit der Daten vor Hack-Angriffen (nicht nur von außen, sondern auch von innen
• eine gute Usability, die die Bedienung der Enterprise OSS auch Gelegenheitsnutzern erlaubt (oft ein Schwachpunkt bei OSS)

Über die bloßen funktionalen und nicht-funktionalen Anforderungen hinaus haben Unternehmen oft Bedarf an zusätzlichen Dienstleistungen rund um eine Enterprise OSS, und zwar umso dringender, je geschäftskritischer die eingesetzte Software ist. So erwarten Unternehmen auf folgenden Gebieten Service-Leistungen durch externe Anbieter:

• Schulung: Dokumentation, Trainingskurse, Experten-Hotline, etc.
• Consulting: Workshops von der Spezifikation bis zur Implementierung, Unterstützung bei der Installation und Konfiguration der Software
• Service&Support: Sicherstellung des Betriebs der Software und Beseitigung von Problemen im laufenden Betrieb
• Wartung: Anpassung, Pflege und Aktualisierung der Software
• Entwicklung: Umsetzung von Änderungs- und Erweiterungswünschen (und zwar so, dass diese ein Software-Update überleben und nicht neu programmiert werden müssen)
• Infrastruktur-Tools: bedienungsfreundliche und zeitsparende Monitoring- und Administrations-Tools (die im Gegensatz zum Kernprodukt oft nicht Open Source sind)

Sicher ist diese Aufzählung noch nicht vollständig. Doch es ist bereits jetzt eine umfangreiche Aufstellung, die demonstriert, wie hoch die Anforderungen an eine Enterprise OSS im Vergleich zu normaler Open Source Software sind. Und meine persönliche Erfahrung zeigt, dass eine Software, die die oben genannten Punkte nicht oder nur unvollständig erfüllt, für den Einsatz in Unternehmen wenig geeignet ist.

Doch wie ist Ihre Meinung dazu? Welche Anforderungen stellen Sie an Enterprise OSS? Ich freue mich auf Ihren Kommentar!

Der Autor ist Gründer und Vorstand des E-Mail-Marketing-Dienstleisters AGNITAS, Maintainer der Enterprise OSS OpenEMM und Vorstandsmitglied der Open Source Business Foundation, die das EOS Directory betreibt.

In all "silence" both Knowledgetree and Pentaho have gone beyond 20'000 project page views on EOS Directory, Alfresco will soon be following too. It's probably not by accident that the interest in Open Source Document Management and Business Intelligence is that high. It will be interesting to see how quickly they can reach the 30'000.

Many people have asked the question whether an Enterprise operated only using Open Source software is possible. Well, today, certainly this is possible and many of these companies do exist. It's not really the question whether the right Open Source software exists, it's more the question whether the decision makers and users are ready to adapt to it. Of course, changing from MS Excel to OpenOffice Calc takes time and doesn't happen effortless. And this is enough for many people to decide to not do it. Same with ERP or CRM. As we know Open Source ERP or CRP is not on the same functional and scalability level as SAP or Oracle/Siebel. But do companies really need that level of perfection? Couldn't complexity be reduced by focusing on smaller units and creating intelligent interfaces for consolidation and reporting instead of trying to model all the complexity in one tool? Most applications areas offer a lot of software choice, including Open Source alternatives. Focusing on Open Source asks often for braver decisions and more know how allocated internally. This is often difficult to justify in front of the top management, the board or even the shareholders. But, interestingly, even government often hasn't the guts to move ahead with Open Source and it's mostly because of the unwillingness to adapt!

Talking to Enterprises and Organizations across the world it becomes quite clear that Open Source is understood by most of them. They see the advantages, but also the shortcomings. And they are willing to to take decisions and to use Open Source in their enterprise architecture! Even large and conservative organizations are using Linux and JBoss application servers, MySQL databases or Joomla! or Drupal content management systems. Often new technologies are taken in when new business use cases asked for new solutions. In these cases it usually is possible/easy to calculate the financial benefits and to judge the risks. Very different is the situation when a migration is needed to use a new Open Source technology. Migrations are often difficult, expensive and even risky. No wonder, that rarely a technology switch can be financially justified when it implies a significant migration, regardless how much cheaper the new technology may be. Migrating to Open Source is pretty similar to what you have experienced in the past with proprietary technologies. The only difference may be that you can analyze the new (Open Source) technology with more ease and you may actually be able to use part of the new code for easing the migration yourself. And the secret on how to cheaply and easily migrate from A to B is well kept.

Looking back on the Open Source year 2009 there are some clear trends and patterns that can be highlighted:Governmental push and push backWe have seen US administration pushing Open Source in health care and we observed the Swiss government preferring Microsoft without even looking at (Open Source) alternatives.Rise of new Open Source services playersLucid Imagination has been formed around Lucene, Acquia continues to grow on the basis of Drupal. Thousands of new services players have been created around the world, some with, many without VC backing. Most of them are happy with their healthy pipelines despite the financial crisis and the recession.Consolidation within Open Source has continuedTerracotta has acquired EHcache, SpringSource taken over multiple companies, many smaller niche technologies got eaten up. A lot of what is happening is not even transparent.Commercial vendors have continued to acquire Open Source companiesThere is a big appetite for Open Source in commercial companies, i.e. VMWare bought SpringSource, Oracle is taking over Sun (partially an Open Source vendor), to just name two well visible acquisitions.The jury is still out on whether Open Source companies actually can make big moneyRecent management changes, for example at SugarCRM, indicate that the climate is getting hotter even within the companies widely seen as successful. But outside of RedHat and a few companies that got cquired for good money, there's still no proof that all that VC money will pay off. Which doesn't mean that many of these companies are not making money. And, while the business case may not add up for the VCs, it certainly does for the Open Source user/buyer. According to most surveys they are happy with the cost savings and the won flexibility thanks to Open Source.Open Source Business Applications continue to increase in importancePentaho just took the lead on EOS Directory, Open Source Business Intelligence is seen as a hot niche by Gartner, Open Source ECM and CRM are able to beat established commercial technologies. The number of reference cases is growing rapidly. But not just large enterprises are selecting Open Source, also the SME world is more and more convinced that Open Source is an alternative to established but often inflexible and expensive commercial offerings.Linux has continued to increase its presenceLinux wins the consumer hearts through mobile phones and netbooks and continues to run on victory lane on the server side.The battle on the client is won piece by piece According to Statcounter Firefox 3.5 has a bigger market share than Internet Explorer 7. OpenOffice.org continues to grow, Chrome OS and Android are becoming real contenders together with Ubuntu to make endusers switch.Open Source events are suffering from decreasing interestBe it ApacheCon or many other Open Source focused conference, the interest of both the community as well as the enterprise buyer has decreased. This may have caused partially by lack of budget but mostly because of the growing experience and adoption of Open Source technologies.Open Source projects continue to bring innovation to the IT landscapeGoogle presented Wave, Chrome OS, new ideas for better browsing rather come from Mozilla than Microsoft, Alfresco has been quicker in adopting CMIS than most commercial players, cross category integration is implemented faster within Open Source projects than outside.Did we miss something? Don't hesitate to add to the list!