Bestand wählen
Merken

Handling research software: recommendations to users, developers and research managers.

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
yep so my name from 1 formula shows Institute the neuron and antivirus does Institute I'm the deputy head of TVs Scientific Computing Group
and I am involved for so fair a menu several aspects like I'm using soft and developing I go schemes
and I'm also quite deeply wading through court by myself won't which I find from my lymphatic clients satisfactory of business I like it from OK so for and I will repeat many things that you just talked about in this talk and and I'm still learning also from from what he's saying and the back of this code is still does so just not be of some kind of opportunity to opportunity true to work to want to waste here vendors on all these issues also at DG let's say management level of research organizations and so on and what just came alter just the day before yesterday or really finally went even yesterday it was really fun and published that is at hand I showed some very nice wrote ending on weak and I show just concerning it or if so then on that when we left the fortress of that the recommendations essentially and like from some kind of you give someone the handsome recommendation This research so film and this has been produced by this is just a group of people wasting convened by DDI alliance of the German Research Organization so many leading orders are much as part of all over from from German research funded give and got for and also from this this workshops from the previous month and a a lot of people including myself and also defined Amish the contributed to these things OK and either just but the
just consist of 2 parts 1 is some kind of official where that selfish units were accommodations and of about some some conclusions and thoughts by myself should maybe a little bit more pessimistic than than the year said but you let us OK just to
explain the background of all of 2 so that's that's just kind of German system of let's lots of hierarchies organizations and so on essentially this some kind of organizational structure which some represents together diversity is free and you will show it conference to conference of university rectors too much time form of 11 of some muscle so institutions and the diffusion others I danced abortion portals organizational ends up through the lens of 2 drummers optimizations and this just initiated couple of years ago a parity initiative digital information which is of some obvious thing to do of course because you talk about digitization of in many cases and and they're seeing of course that something really needs to be done also at the institutional level and within this parity initiative digital information bolster convened and other talk working group scientific software which started in 2016 and which essentially with most of these people they're not ornamental fish of Working Group 2 heated pools are from services which will just appliances band director 2018 onto totals intended to and this group just efforts initial meeting next week in Berlin and alive association with this association of in this working group is represented by for and myself and this essentially happened because you have of gelatinous station has seen this it's not very areas of software and you have a master's 2016 and just then he was asked whom can you consent to this bought and the use for us this the mass of a circle and an MMS metric and then justice to scan models OK so that's just concerning the background and essentially made available which produced during this
time was lost his kind of paper and the idea that you really want to we want to continue this work and and both of these and some kind of a more official level and not only that pocket vertical bar The Rabin Group and the this song for what that happened OK OK so now diver
repeat lots of stuff mutual said in this kind of little bit different settings which a essentially in all this kind of you can say OK so universe this is going to give Scott talks and so on but I'll give a TD director of my Institute very artist in OK but he Karaboutis or not until we get some kind of attempts to it just really just make in Venice on this kind of more more efficient institutional level and that's also the reason why our forces agreed true to to addressed and invest my time in in into stings because I see the chance to also true that say it's a make make the rest in Venice of 2 problems of speech to students fodder suffer and so on and so on also put some some kind of high-level so research stuff a central component of scientific work so the simulation models I growth development for for modern for models are growth and development this 1 part generation processing analysis resolution of research that belongs to the use of software and science and also control of devices in experiments so if you just go to lab day all have some laptops around 2 which are some of packed into the systems some graphs going on and some software let's electoral whatever 1 ink and getting older data also part of this kind of stuff the show of so once this is a case
essentially you need to apply the rules of good practices in science to suffer from that means transparency addressability accessibility of usability and applied to soften you have to stand formulation fair principles to some kind but now findable accessible Interoperable reusable that to and in this paper being just have kind of free let's say groups of various that to would toward to to to organize just you'd you'd you'd use to Sir presidents multi-stakeholder management as developers and users of software and infrastructural facilities which are some all have long long on different levels to the use of if different types of software their research called sets the most talking about of course because most of us somehow use them on a menu of holes also developing them they're just frameworks which are used in and research MATLAB should be on what you know just and other courts which which we are just Fuson which are given to us in order to to to to make some because versions X of course as well and services like for instance in order which has been literature access to a total that sales would get up and so on which helps us to to really maintain suffered and just make it kind its and there's different various different relations to suffer Suffolk can be developed it can be used and can also be provided OK and then not just
I just told you stock amended this kind of had transposed to do that and just organize not
arising singing in the talk around stakeholders in mobile to stay out there all induced is to use these this game OK
developments which are essential research stuff engineers said would be dedicated to the little people who were if you have to to do so fell or some all just got into this world but also of course with the academic softly below postwar scientists move riding software by sums themselves liable to identify myself as some academics of development of the context if you still to logics and OK Kalentieva essentially these people being myself as well should should learn to adhere to to best practice standards for coding packaging documentation validation It's important cable interoperability Adobe ability addressed if the idea of a just reduced bell established components because the owes much safer to use something which essentially everybody knows of works than the inviting informed onstage by itself so sometimes can be found never been complication qualification does something which in many cases of some hours selecting that the really can can learn from each other in in this business critical self-assessment myself a really good enough the the minimum would not of course but in the OK and that's that's this kind of thing servitude should freely lie also unusual and of developers and then of course the
use of self like research scientist postdocs PhD students would just say OK of Totoki users piece of frosted would do this kind of analysis or what writer to would assimilation and so on to write a paper and they have a need to develop some functional understanding officer for use in not using usable only as a black box but really try to see what this in this office this of course also has the implication that is often used to be sufficient transparent this in have or provenance that's important of course I can documents after you listen publications including versions configuration of parameters in order to make your results suggesting obtained by this awful reproducible consider free and open source alternatives which this also some point which helps to to to to maintain transparency and visibility and what also just knew just emphasized in this talk user every paltry cost and that's no so that's that's something which which really in directly involved in software development use can can can can do was assessed reviews infrastructure
facilities some may be out a short about distance IT departments Iberis technology transfer units they of course the need to to to to widen the infrastructure also create solutions to quantifiers musicians citations and they need to support developers so needed so surprising and then distance known so and so and they get also some some kind of fluid new rules we need to do as a developed so so so I'm I'm also amorphous stuff for licensing for for certain codes and and and and pricing and all these things and was to have the 1st Austin groomer full force so the development diversions Institute and for it's very much learning by doing and we already started from essentially by also all the prize a Tao which license to be applied and all the ministers and don't really nice to have more on infrastructure level also more skilled people judicial issues concern itself could tool to 0 2 to develop things that you need to develop long-term maintenance strategies consulting and so on and also there are no new possibilities new carrier past sister-something something which which can be a beer also initiated by by the vendor's infrastructure facilities that loaded library instance of God kind of new role also given to libraries which of course some mostrado new digital age to hold the continued in the existence and but there's enough to do management
responsibility did steps lots of things known no and there's also something which which really is interesting to try to to to to put that is just on a table of of ghostly manager scientists leading positions on science ministers signed a deputy group that I'm also scientists in a leading position so it's also concerns myself and science matters like and and the just to try to really was and so on to to to to to make them about these issues also this kind of of refined hand so OK and here's the main point is of course for development just care boat people who developments of hello Nord who qualified developers organize educated scientific and personal organizer clinically Commission addition incentives carrier past for 4 people doing suffer also of course and care about by best practice rules and and and provide organized with provision of state of the art infrastructure but that's something ditch the comes order some some from from year some kind of self CenterPoint at that number needs to do all these things in some sense you also Toros since this is essentially the sum respect to small work so so you had a discussion of the bar yesterday PhD to student just said OK the giver get no you all flagon but we have to do that at just at some the prescriptions what you need to do more than we did before it and 3rd then that steps some really interesting because this perception because history tells us that really it's not only part a to teddy developers all producers what to do it's really also about helping to to them from Buddhists and 2 to provide you do right the 1st of these the culture of using software in a proper way is this this also very important prefer open-source tools commercial services for reasons of economy and it does sovereignty prevent use of ordered it suffered 2 words concerning oversaw it's of
course another point which I'd there you go back to licensing what do we have to see OK
also OK that's not as to not explicitly formulated the urge you on on the slides I'll give a providing is also do do do do your organization a role of course of of also it's necessary to a true to be able to to long-term maintenance of software developing clarify licensing models before publication develop business of financing models sector long-term availability for wider was given up and so on the very much in this what are on the same page concerning open source licensing the formulation of 2 so comment a little bit to let's say in some sense a little bit a careful on this but of course all source code for providing suffer licensing suffers sources also somehow DDD the main the main path to a tool to publishing software but this is of course a problem also in management level because that's another problem concerning the licensing of of offers open-source there is because of the management level somehow often seen as some kind of acid which can be monetarist and and people rooting talking we can give to severe for free so on and that's but now they have of course more and more possibilities when also examples what a view was by not doing all of them will suffer open-source and what can what what what what can happen if you just make so widely available and what kind of recording the recognition people can get from from this OK so management was personal and and that's the summary of OK something which which
OK so this is some kind of fraud will be ofthis is kind of document which might have to go to to also but when used to still raises issues dismissal of the management of the so
far that's not my personal comments now just
going from from from just supporting fixed to
would for me this is some kind of I'm
seeing really some several steps forward on on to suffer issues I'm seeing significantly increased awareness on this topic also many at a low level so so already dis- also made have something to do some kind of generations of change of to understand so so many people are just rising into management positions in science also just as much but out of office suffered and those people moved in with me now before no funding organizations start caring very much about this practices no so there also already funding projects Andromedae was a gift she called suffer sustainability just a spring not high distraction suffer this was body organized by much is countable was also just the 1 of 2 you'd all those of use of these documents and I envisioned that really these feel see also the next step in funding quantization step they also start enforcing certain rules than it comes to a total to funded projects that the software plays a role so this talk and this of course now be applied to a to disseminate this and so we will see how a derivative work and the next step which needs to be taken of course by all of us and also with the knowledge of versions institutes we have of course also already many ideas what to do you need to relive the null organized Detroit's fail to succumb nations and all 4 also but would be an animated plan to to really institutional level rules so if you had an evaluation last year of you have been some of just that this is other things but about what I can say is that just duty software the also fell concept and use it to a teacher also goes along these lines played decried positive roared were do relation for the Institute so this long they
innovate holes from my own experience as so many people tried to do do I think and also of the single so here large grass-root interest in itself issues the it's still of course the I very much Summerland begin yeah that's that's what you've seen of course people want to have to scientists put minimum effort into dealing with software because did you to fight another paper of so I'm I'm repeating what you said the softer addressed getting the total to suffer development stood maybe is the some some character love affair just really also there's been Venice on duty the management level isn't isn't change and this licensing our order said this is learning by doing currently so we don't have many coinciding possibilities concerning concerning the licensing some institutions seem to still have tied licensing polities policies which and dissemination often nested again it's about really a foreground to suffer development and yeah would intentions versus limited resources and it's also something we are starting this also be personally is she let's and bon bon problem is and that's
what I'm saying here concerning of open-source software in fact OK so so so all this also after promises transparency rules ability to work to reach here and you have lots of of toward slice plus plus bias in both systems and so on and so on but systems become more and more complex and you get some some real the level of fragility into all conflicts with systems and software and so the just of writing of courts of 20 years this the most striking buffer the also 1 of the 2 men responsible people and you just working on the bid system of feeding the ball caught fire and in order to make it easy to install with yeah you really using different components from other sources and it's it's always just if you take from from system and to system and plus 1 as always some problem and that's that's something which which which which I see is a really practical problems and it's really toward the right you need lots of skills and suffer design choice of to I tools you would need to think about standardized interfaces down not met many it's a Python becomes not some kind of stand interfacing structure for for for scientific computing and it's 1 thing container things you just also mention this in this talk like Docker container systems data these something of this but the other end justify put myself into contaminants stolid there and nobody cares about it it's s instance transfer and some some some closed-source source code them so maybe you know so I'm seeing so we really need to learn on this level all flats is software engineering a solution of stone long way to to to go on to learn what to do and this is johnny
starting from my point of view because after all but I'm also to buy myself and word for and watched in the framework of different projects also add that all Institute essentially if you really want to see these rules working the need to on love some kind of cultural change and institutions in this this takes 3 long-term some long-term commitment to listings and some some the just this will be easy for my point of view standardization activities between related open source communities that's there some committees which work quite well with others and scientific computing proper of numerical analysis many packages are on which are not really competent to each other you I'm using to read you know I'm using muon could yield a dual role of the from from 1 to the other this was also not easy know so that's something and you have of course now and you have a much more modest on the side so sites exchange forms on concrete experiences and then perhaps you should really go off on this planet level interest still on down to the level of in concrete experiences with PhD students and so on out of a of a softer what you Institute now implemented as recommendations how did you and so on and there's also conferences on this and this also I am also the same idea like mutual to just the a sale can justify nanoscience research after and engineers and you also also would be someone who was really active in organizing community there on these issues which can also have a just learning processes between all involved in developers OK so maybe that's that's enough
this well
Ablöseblase
Topologischer Vektorraum
Ausdruck <Logik>
Einheit <Mathematik>
Mereologie
Gruppenkeim
Mathematik
Modelltheorie
Ordnung <Mathematik>
Übergang
Assoziativgesetz
Kreisfläche
Total <Mathematik>
Linienelement
Minimierung
Fisher-Information
Gruppenkeim
Ruhmasse
Mathematik
Bilinearform
Fisher-Information
Übergang
Algebraische Struktur
Gruppenkeim
Verbandstheorie
Flächeninhalt
Gerade Zahl
Modelltheorie
Modelltheorie
Zentralisator
Subtraktion
Prozess <Physik>
Total <Mathematik>
Relativitätstheorie
Güte der Anpassung
Relation <Mathematik>
t-Test
Schlussregel
Endlich erzeugte Gruppe
Ungerichteter Graph
Auflösung <Mathematik>
Zenonische Paradoxien
Analysis
Übergang
Gerichtete Größe
Forcing
Mereologie
Flächeninhalt
Modelltheorie
Modelltheorie
Ordnung <Mathematik>
Differenzmenge
Analysis
Singularität <Mathematik>
Gewichtete Summe
Gerichtete Größe
Standardabweichung
Extrempunkt
Spieltheorie
Relation <Mathematik>
Validität
Flächeninhalt
Zenonische Paradoxien
Modelltheorie
Standardabweichung
Philosophie der Logik
Resultante
Punkt
Fluid
Quantor
t-Test
Wärmeübergang
Übergang
Einheit <Mathematik>
Standardabweichung
Äußere Algebra eines Moduls
Abstand
Analysis
Parametersystem
Lineares Funktional
Simulated annealing
Freier Ladungsträger
Schlussregel
Lineares Funktional
Topologie
Forcing
Einheit <Mathematik>
Parametersystem
Strategisches Spiel
Ordnung <Mathematik>
Modelltheorie
Term
Addition
Gewichtete Summe
Punkt
Freier Ladungsträger
Ortsoperator
t-Test
Gruppenkeim
Operations Research
Zahlenbereich
Schlussregel
Lineares Funktional
Topologie
Schlussregel
Gerichtete Größe
Standardabweichung
Parametersystem
Mereologie
Endogene Variable
Modelltheorie
Ordnung <Mathematik>
Term
Aggregatzustand
Rechenschieber
Gerichtete Größe
Einheit <Mathematik>
Standardabweichung
Modelltheorie
Modelltheorie
Term
Schlussregel
Übergang
Standardabweichung
Wärmeübergang
Übergang
Modelltheorie
Term
Schlussregel
Quelle <Physik>
Total <Mathematik>
Mathematik
Ortsoperator
Relativitätstheorie
Übergang
Endlich erzeugte Gruppe
Schlussregel
Übergang
Schlussregel
Geometrische Quantisierung
Standardabweichung
Einheit <Mathematik>
Wärmeübergang
Projektive Ebene
Hill-Differentialgleichung
Modelltheorie
Term
Funktionentheorie
Subtraktion
Total <Mathematik>
Mathematik
Krümmung
Kurve
Wärmeübergang
Schlussregel
Übergang
Stabilitätstheorie <Logik>
Puffer <Netzplantechnik>
Auswahlaxiom
Algebraische Struktur
Funktion <Mathematik>
Rechter Winkel
Klon <Mathematik>
Endogene Variable
Freier Ladungsträger
Modelltheorie
Ordnung <Mathematik>
Auswahlaxiom
Offene Menge
Subtraktion
Prozess <Physik>
Punkt
Grothendieck-Topologie
Mathematik
Relativitätstheorie
t-Test
Kraft
Schlussregel
Bilinearform
Schlussregel
Übergang
Wärmeübergang
Näherungsverfahren
Projektive Ebene
Modelltheorie
Eigentliche Abbildung
Ereignishorizont
Standardabweichung

Metadaten

Formale Metadaten

Titel Handling research software: recommendations to users, developers and research managers.
Serientitel The Leibniz "Mathematical Modeling and Simulation" (MMS) Days 2018
Autor Fuhrmann, Jürgen
Mitwirkende Leibniz-Institut für Oberflächenmodifizierung e.V. (IOP)
Leibniz-Institut für Troposphärenforschung (TROPOS)
Lizenz CC-Namensnennung 3.0 Deutschland:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen.
DOI 10.5446/35353
Herausgeber Weierstraß-Institut für Angewandte Analysis und Stochastik (WIAS), Technische Informationsbibliothek (TIB)
Erscheinungsjahr 2018
Sprache Englisch
Produktionsort Leipzig

Inhaltliche Metadaten

Fachgebiet Informatik, Mathematik
Abstract A working group initiated by a joint initiative of German research and funding organizations developed a "helping hand" paper providing recommendations on important aspects of research software developmement and maintenance [1]. It is addressed to the different stakeholders in the hierarchy of an research organization. The talk will introduce the main points of this document. Based on the personal experience of the speaker a number of steps necessary to introduce these guidelines into the everyday practice of a research institution will be discussed. [1] Katerbow, Matthias / Feulner, Georg (2018): Handreichung zum Umgang mit Forschungssoftware. Herausgegeben von der AG Forschungssoftware im Rahmen der Schwerpunktinitiative Digitale Information der Allianz der deutschen Wissenschaftsorganisationen. Unter Mitarbeit von Mathias Bornschein, Björn Brembs, Michael Erben-Russ, Konrad Förstner, Michael Franke, Bernadette Fritzsch, Jürgen Fuhrmann, Michael Goedicke, Stephan Janosch, Uwe Konrad, Dennis Zielke.

Zugehöriges Material

Ähnliche Filme

Loading...
Feedback