In Most of domains, when we create something, we don’t want it only to be eficient at the time of creation, but also in the futur, avoiding obsolescence. In the digital domain, the capacity to find existing datas is also an important preoccupation. It can moreover allow modifications or adaptations to be done. The topic of sustainability of digital collections is as current and complexe as the number of factors affecting it increased significantly during the last decades. The fact that we ignore the needs of the futur strenghten the necessity to focus on sustainability.
The problematic of the annotation is directly related with sustainability and mensionned in one the three selected articles. Indeed, the description of the context is probably as important as the specific description of things within a collection. This is the problematic of the annotation metadata which has not always be structured independently of object metadata. Telling about metadata, it is also to note that their development work miss some long-term considerations. In year 2000, we realised the limitation of the existing tools supporting annotation of digital content. This had had a retarding effect on the evolution of digital scolarship. In response to this gap was created the “Open Annotation Data Model” in 2013, focusing on the creation, capture and preservation of contextual information. Some improvements have also been doing in the redistribution of annotations in existing collections. we also tried to avoid strategies requiring human intervention in large collections because they reached their limits. In a long-term vision, sustainability needs technique of automatisation according to evolution of standards. All of these elements contribute to improve the compliance which is a good point in terms of sustainability.
An other article concerning the sustainability of a digital human infrastructure tells about the HUMlab experience which reveals the need of an elaborate infrastructure of technology, physical space, competences and human attitudes. The collaboration of people from very different fields is seen as an advantage for the elaboration of digital research methodologies and tools.
The aim is to create an attractive meeting place allowing interdisciplinary interactions. This interdiscilinarity requires an open mindset. It is also important to provide approach adapted to special needs of accademia. This leads to the challenge of promoting new ideas, quality, creativity, innovation and individual development. This news ideas will be encouraged if the working process allow mistakes. In this way, scholarly needs will be provided with accessible and trustable technology. Other aspects as understanding and transparency are dealed. The understanding is of course primordial to avoid making mistakes and if some mistakes are still commited, a good transparency will allow it to be noted and corrected. These both things increase the reliability of datas.
Coming back to sustainability, we absolutely need a dynamic, flexible and adapting organization level according to the ever-changing world as the one we live in. Even if it’s impossible to predict the futur, it is important to apprehend it as much as possible and this is why the vision of the futur is very important.
Let’s tell about some people envolving in sustainability: the Luminaries. These people work to get digitals datas accessible and profitable. This allows also to reduce the lack of knowledges in humanities departments. This is seen as an attractive prospect for the universities that want to make a substantial investment in digital humanities. The fact that they have succeded previously gives them a credibility which is an advantage in case of untested practice. Being aware that a sustainable digital humanity community does often not follow from a quick victory, they behave according to this principle so that they guarantee sustainability. However, even the smartest of luminaries is one single person and mustn’t forget to share the leadership. For example, they cannot load an infinite number of students. This is why they are sometimes said to be “evangelists”. An other thing to avoid is to reproduce what was achieved rather than develop new projects. Otherwise, the culture of digital humanities might get more narrow, with a lack of variety and diversity.
 “Improving compliance with evolving standards using computed transformation of digital collections”, Cornwell Peter, University of Westminster, United Kingdom and Dan Granville, Data Futures Ltd., United Kingdom and Alexandra Eveleigh, University of Westminster, United Kingdom and Eric Decker, Heidelberg University, Germany and Christian Henriot, University of Lyon, France
 “Implementing and sustaining a DH infrastructure – The HUMlab experience”, Fredrik Palm, Umeå University, Sweden and Roger Mähler, Umeå University, Sweden
 “The question of the luminary: building a resilient campus DH culture”, Paige Courtney Morgan, McMaster University, Canada and Dale Askey, McMaster University, Canada