 So thank you very much for joining. I think you can listen to me well yes okay everything is working well so thank you for joining this community call so we are, we know that we are close to an holiday period for some some of us, not for all. So we will, as usual, dedicate some time to some of the updates that we usually give, and we will, we decided to dedicate some time to some studies it was our intention to have in fact the decade theme to two studies the study that our colleagues will present in Andrea and Andrea in Miriam on the availability of repositories and another study that was recently presented that we can also share the link on the, let's say that the readiness of repositories for compliance of the European Commission policies. We can only have one of these studies present today, maybe in one of the coming community calls we can address the other one so today, in fact, we will highlight some of the updates or the work that we are doing related with the open air provide. And then, together with our colleagues from CNR and open air Miriam Baglione and Andrea Manocchi, they will present this study I think it's a quite interesting and relevant study for us to be aware of also a bit of the landscape in the area that we work. As there are lots of things that we need to improve and to modernize and I think this is a relevant contribution for our discussion. So you know all the information is usually we put it in the in this website in the open area and the open air dot u website where we have the information about the community calls where we have the recordings that the presentations you know how it works. We know that there are always new persons new cameras on board in this community calls and the people that join every month. In the first Wednesday of the month so we may repeat some things other know so about the recent news so important to be aware that the last update is on the 4th of April. So, as you usually want to know about this so for sure you will receive some kinds of updates from the broker events soon. If you want to check that information in explore. This is important. Do you remember that usually we share a link where you can find the information about the, the, the news or the changes in the in the graph and in the in the in this update. So usually we presented that we have it here in the in this in this page. So you can access the, the, the last update where we explain the aggregation content workflows. This is a kind of table to report some of them of the updates. So now the technical team is doing this because what is relevant is really on the graph side. Okay, so every time that there are relevant changes in the graph so which means the relevant is included in inclusion of some relevant data source and changing the in the links. The relevant funder included the data from that funder etc. So this information was here but now it will we'll have it in the documentation of the graph website itself and also on the catalog so you will see now the links are in the in the in the slides and then they can also share it in the in the in the in the website so now we have in the in the in the website in the landing page of the service in the open air catalog you can see this change log and if there is something relevant introduce it in the graph that is visible in the date and the same situation. Here in the not here here in the graph website so graph dot open air dot to you, you have this versions and change log part where we everything is properly documented and also in the in the in the landing page of the service in the this the relevant changes will be made available here so be aware of this this is not so critical but as we like to be quite transparent in terms of the dates of the index etc etc and people usually like to know and you can see it in the dashboard also that information in your dashboard, but you can be also informed. The other three news that we want to highlight it's not not so much related with the, let's say features new features of changes in the in the dashboard itself is more about what is going on around that. So, one is the campaign that we are running. So be aware of that, if you are not. If you are, if you are not that the source compliant with with that crease or level three and four of our guidelines, be aware that we are running a campaign for you so for those that are driver and version two compliant we are contacting you asking you to to work a bit on the update of your compatibility level. We are already contacted almost to 400 repositories we are only we are contacting the driver basic compliant repositories by groups of countries let's say so we are sending today the for for Turkish Turkish repositories because in fact is the the the only country that we is missing so we already contacted almost also we will achieve a little bit more than 400 repositories we are already contacted. So, you can contact us via the help desk help desk at opener to you if you have any doubt about this. We provide guides we provide more information there is a guide about that so if you need support please ask us. We will also discuss at the end of the meeting this if you if you have any issue. The other work that we are doing is so we don't need to inform everyone but I think this is relevant. I don't know if you know but the the Nazis that is the national aggregation of the Netherlands is being so we will, we will, they will put it offline soon this the coming months is. And we are aggregating the content to the Dutch repositories content via this national aggregator and now we are working with all the Dutch repository managers to in order for them to register directly in open air so in fact we have this morning. A webinar with them quite well succeed webinar and we are happy with the results and we will have another one in the coming week on Thursday. So if you want to receive information about that and if you are part of this community contact us if not so we just want to also for you to be aware of this work that we are doing with this specific country with this specific community of repositories. So the other information is that you you you are aware of the open air guidelines so the different templates that we did in the importance of the guidelines they are a standard worldwide use it also there are links with other countries with other regions, not only in Europe it's guidelines that are relevant for the, for example the Latin America and network of repositories la referencia it's also a standard for Canadian repositories for Japan. So other other alignments that we have and the idea is that we create a kind of community driven governance, community driven global governance for the for the guidelines and we are initiating this this year. So the creation of a working group under the MCA the open air MCA organization and then in in June in the open repositories conference we will do, we will have a panel where we discuss this with the different players in the different regions not only we from open air in in in Europe but also with colleagues from the different countries where and in regions where the open air guidelines are being used so if you will be in open repositories be aware of this. So be aware that we are doing this and we can maybe we can share novelties after that panel so where we will discuss how can we ensure this community driven global governance for our guidelines. So and it's, it's, it's, I only have two things so in the last community call we discussed about some of the common dots that we usually receive regarding the dashboard so how to update the IPM interface, the, the, when they are visible set et cetera so if you want to check some of the explanations that we give, give, give to these common dots, you can check the recordings from the past community call, but be aware that there is a tab in the provide dashboard where we can update the IPM interface we can change, or we can have a new one change for the current one that you have the compliance level or have a new one a different one and indicate the level of compliance. Be aware also that if you change the name of your repository or one kind of this change, this is not reproduced automatically in the explorer in the different services you need to wait some time to for this implementation because we do in fact a check in our aggregation manager system that to imply to apply this this these changes. And then there are also that you can also check when when it comes to the visibility of the records the content aggregated from your opposite we do not forget that there is a collection monitor tab for you to check when the content was aggregated when the content is transformed and made available in the in the repository. The last, also, but you can check the recording because we gave detailed demos about this common dots but but but I don't want to spend a lot of time with this. Also the last question that we usually receive about the metadata enrichments. So that is, we don't ensure, let's say, programmatically and automatic updates for all the systems but be aware that DC this space, this space crease, have an implementation of the metadata enrichment so you can have the you can enable the broker events to be visible in your repository and then validate them internally we have already dedicated some time to a previous community call to a case from a this space crease use case from Trieste Trieste University, we and there are there is also a project that the Italian company for science did for the space seven, the committers from the space seven will assess this implementation for the space seven and hopefully this will be also available in one of the coming versions of the space seven, but of course you have the metadata enrichments available. I remind you that you have a set of 100 metadata enrichment for you to check and then if you want to receive. This is a sample of 100 if you want to receive. And to see all the other metadata enrichments for that specific metadata enrichment that you want to receive you need to subscribe. And then you will receive the notifications from a specific. Here's for example links between publications and projects or PID from authors or from publications if you want to receive this kind of things for all the content of your repository you need to subscribe. What you see in the dashboard is only a sample last novelty, as we promised you to share all the developments regarding the link between open air provide and the onboarding process in the Oscar be aware of this. In the onboarding of us of services, resources, that the sources, etc. Open air is already there. So when you want I'm putting gear a screenshot from from a provider that is onboarding a data source, the first step in fact when you want to onboard the data source to us, you can select the data source from open air. Okay, and you have already pre filled the information you just need to fill the fields that are not pre filled by open air. And, and with that, the content is also made available. So but you need to register even if you are an open air you need to register but what is important is that if you are part of open air this process is simplified and the content will be made available and you need to finalize this registration so this is a new novelty that you can access, but you can only see it if you manage a provider of us okay so if he is a provider is already validated you know in the Oscar, you can start onboarding your data source and be aware of that so it's an if it is a relevant information for you and if you want more information we can, we can give it to you. So these are the recent novelties the information we can discuss it. At the end, as there are some constraints in terms of time so let's move to the presentation of Miriam and Andrea. You can also present yourself if you want. This is the study that they will present it's interesting study. Knock knock it's not not knock knock on the Evans door like the song but it's not knock who's there a study on scholarly repositories availability. So I will stop sharing and you can, you can start sharing with our colleagues that are present in this community call. The results of the study and and why it's a relevant study for the community, what were the main findings. So, I'm not not sure who will share the screen if it's okay Andrea. Perfect. Yeah, I am. Can you see that. Yes, yes, yes, yes. Okay. It's like so. Okay. Yeah. Thank you for your friends for having the availability. Yes. So, thanks for the introduction. So I'm Andre Manoj. I'm a researcher at is the CNR is the actually now the, the affiliation has changed the order, apparently, and in this work which is by me Miriam and Paulo, we are testing the availability of scholarly repositories and the main idea is that basically we know that scholarly repositories are a part of some of more than open science practices. As they preserve and research data software and footprints and it's the provide means to site. And these research outputs from, from, from other places, and they can be general purpose and catch all as is in order for example. Or they can be thematic and research community community specific as the one for example for, for, for life science protein databases and so on and so forth. Scholarly registries facilitated discoverability and referencing of scholarly repositories because they provide ideas either persistent or local. They provide each repository with a public profile, which is essentially metadata. And they are. So, I mean, they can affect the work of individual researchers but because in order to discover repositories they, they can, they can turn into registries to discover which are the suitable ones for the disciplines at hand. But they could also use standard search engines like Google for this purpose but definitely registries are an authoritative sources for whomever is running a scholarly infrastructures and scholarly services such as open air. And this is why this work is partially funded by by open air nexus. And I'm saying this because if you need to discover. I mean if you need any posted listing of repositories out there, the best place is to go ask to someone that is doing this, this job of collection and registration. So, the essential research question in natural is that provided the scholarly registries are are authoritative sources of information with the repositories that are registered there in be available be available upon request. We selected for prominent registries. So, which are for sharing rich data opens or enroll and these are the facts about about them. They have been dumped by various means in February 2022. And, you know, I mean, it is not an outstanding list of repositories up there but we are, you know, like, it's over 10,000 close to 15,000 of them. In total. Yeah, okay, 13,356 unique URLs because some of them were registered multiple times in more than one place. So, with all these URLs, we requested them both in HTTP head and HTTP get and having a max allowed a number of redirect sector 30, as well as the number of seconds for time out. Each request we were tracking original URL final final URL after redirection, if any, the final status code ready to read directions chain redirection status codes latency, and we didn't alter in any way. We didn't adjust the regionals contain in the registers and this was the idea behind this is that basically that information is the one that the repository manager put inside the inside the metadata on the repository that they intended to register in the database. So that your that URL supposedly was the one to work. So we didn't, we didn't change that we didn't tamper with that in any, in any way. Data notebooks are available on Zenodo. So you can, you can download everything and reproduce and play with with our analysis and results. From, from our, our requests, 25 of them, 25% of them sorry, activated some kind of problems. 15% were going to into time out and the rest trigger 400 and 500 that 400 is like in the not found category and 500 is basically internal server error family of errors. A slightly different behavior between which means that they are served in a different way on on the repository side, and this is in contrast with the RFC specifications which demand the same the same response. Nevertheless, the, the, the method, and without, you know, considering the method of the request. Also, syntactically, syntactically, okay, requests that we're finishing with ending with a with a 200 class error, they could still be simultaneously wrong for example here in this for this repository. If you follow the URL, it goes on a website and it's supposedly about a magazine or something that's totally unrelated with digital libraries for her, her system education. And in case of redirection, we had 32% of the URLs that were going redirected to somewhere else, which is fine. I mean, it's, it's totally legit behavior, but 300 of them were even after redirection were ending up in errors, which is not great. And, and we also noticed that none of the redirection chains where was longer than five. And we also observed a shrinking phenomenon basically and in the URL space. So, like, after redirection, we had a number of reduce number of final URLs basically like 3000 roughly URLs were colliding to colliding were redirecting into 1500 URLs. So we started checking those manually. Because most of the redirects were super fine like HTTP to HTTPS. They were pre-pending dub dub dub, upending at the end train link slash, but we discovered for example 16 new Norwegian repositories that were written to the same way page. that was telling that the URL was not valid and that and was providing a list of repositories available, which incidentally did not include any of these 16 repositories that we start from. And then again, like six databases at the National Library of Medicine were redirecting to a web page, telling that the information has been integrated into another service, or seven seemingly unrelated URLs pointing to not found page on a domain that was held by a company in the digital libraries market segment. So it was rather unfeasible to check them all manually. So we just, you know, like from the errors we had, we sampled looking for individual examples that could be relevant. And so yeah, this is something that we need to extend in the future. And so like the main source of inspiration for this work is this paper that has been presented at the TPDL in 2020. And we are missing these two bits here, basically including additional headers like cookies, allow cookies in the requests. And we didn't also simulated human computer interaction because basically when there's a way to simulate a human browsing, even though it's everything happens, even though the request happens in totally programmatic way. So this has been done in this work, but it's planned as a future extension of this one. And as well, another shortcoming of this preliminary first study is that one short resolution as the one we did provide just a snapshot, which potentially can be distorted by transient outages and technical issues. As an exercise, we try to support this by checking Ukrainian repositories, which is the UA domain. And like we checked them, and 24.6% of them were unreachable, mostly in Eastern Ukraine, which, you know, highlights that the infrastructure was disrupted. And of course, the repositories availability follows. It cannot be can't hold anymore because of technical issues in the network and probably repeating the same batch of resolution in multiple times over an extended period like one year, that would be what provide a more truthful snapshot of the thing. So repeating, let's say, the same set of experiments like over like one year timeframe would yield more accurate results. So in any case, it's safe to say that one in four repositories registered in scholarly legislatures is not accessible. And this is a lower bound of the actual amount of problems because as I showed before, there are like URLs that redirect to something that's totally legit, but is unrelated. So it works from a technical standpoint, it resolves, but it's not related anymore to the repository in the first place. And also you have soft 404s, which is basically a request that goes okay, in terms of HTTP protocol. So it ends in okay status code. But the web page that's prompted afterwards is basically something that says repositories you're looking for is not available here. You can find it. So it's a soft not found website error. And of course, like there are repositories that have been discontinued, moved and merged. And these increases actually the number of problems that repositories in registering registries can have. There is a chance to extend these analyses and have a more thorough view. And to do these, we should actually probably start using mementos, which is basically go into the internet archive, for example, and check how the website looked back in time to see whether the content went adrift over the years. But of course, it was something that couldn't be done in this first study. And all in all, what the takeaway message is that repositories and IT infrastructure managers have a fundamental role. And they are part of these functions. Because whenever they do something that takes a repository away from the URL declared in the registry, like you can create problems downstream. So it can have an impact on especially on reproducibility and open science practices, less alone the problems that infrastructure silos as as opener have in processing the content of the registries themselves. That would be all. So thank you. Great. Thank you very much. I'm not sure if there are questions. So feel free to turn on your audio and ask questions. If you want to put it in the shot, you can also do it. So it was 25 percent of let's say of errors. So 18 percent of time out and then the rest was what kind of errors? Can you remind me? Because you put it in the slide. It was internal server error, 500 or bad request, 400 families. Okay, okay, okay. Yes, 25 is a lot. Yeah, I mean, considering that you turn into a place that supposedly has to have the truth, the truth about about the repositories, you would expect higher numbers, I would say, not the failure every one out of four. We have already a question here from the 25. So thank you, Marcus, for the question. So it's in our Google notes meeting, meetings of the notes of the meeting, sorry, from the 25 percent of now missing repositories. Do we have information or many of them use it as well persistent identifiers like the OIs? No, we didn't check that. Medium. So we didn't check whether the URL pointed by the thing is that, I mean, the usually the URL that points to the repository is not is not saved in a DOI or the best could be in the case of fair sharing, for example, the repository profile as its self registry as itself a DOI. But yes, so like, yes, so fair sharing provides DOIs for all the registrations they have inside, which is great. Some other registries do not have these, do not provide this benefit. But it is the registry. Is the registry metadata profile the DOI to the registry? Exactly. Then the profile doesn't have, I mean, doesn't have a DOI pointing to the registry itself. Usually it's just a plain URL that goes somewhere to the domain of a university or a research center. But I didn't check, we didn't check for DOIs, let's say in this URL. Those 25%, yes. Okay. This work about DOIs was done instead in the work we were citing. The authors were trying to do this exercise with DOIs that are cited, that are somehow present in literature. So they were resolved and they had problems with those as well. Okay. Thank you. Is there any other question? I have two comments and questions, but if others have, so Marcos is saying thank you. Okay. One of the reasons also that I, so I felt this study interesting and is something relevant for open-air contexts. Are we really, when it comes to the registration process, et cetera, and does open-air use the authoritative sources, the authoritative directories is always critical to have the information proper updated in the directories. And we receive all kinds of issues in the history, in the recent years of open-air. We receive all kinds of issues, double registrations in the registries, wrong links in the last register in the directory, et cetera. So one of the things, for sure, is that we all, part of this community call and our colleagues and repository managers need to ensure that our information, the information of our repository is proper identified in the directories. So from the directories perspective, do you have any suggestion for them? So from the study, so for sure we have this suggest this action. So all repository managers need to ensure that they may make available updates in the directories. But for the directories, is there anything that we can also do if we manage the directory? So this persistent identifier for sure is something relevant. But like tools to do, to run these kinds of checks. Do you have any, from the analysis you did, is there anything relevant that you can share, like suggestions for the directories? Sorry, Pedro, I really need to rush. Maybe Miriam can take the question. I mean, I have some ideas, but I really need to leave. Thanks for inviting. Miriam, do you have any idea, any contribution, what is relevant also for the directories to? They already do something, because when they check, when they verify that there is some problem with the repositories, like for example, repositories are moved or something like that, they, like RetriData does, write an advertise on the page that is relevant for the registry. But if they could do something like we did, just to verify if the URL is responding, if it is, even if they do not check the semantic correctness of the URL, just to have an idea and say to the repository manager, look, your URL is not reachable. It has not been reachable for x days, something like that, this could be really helpful. Yes. Okay, thank you. So one of the things that we need to pay attention is really to keep the records that we have in this kind of services proper up to date. This is one thing that I found critical. There are then some issues in some directories back that they don't have the needed tools to keep track on everything. So even when the repository manager is paying attention, he asks an update and then in open doors sometimes there is the question that they create two entries for the same repositories. Yes, there are also inventory data, more than one entry for the same repository sometimes. And as Andrea said before, we have entries that are the same from several repositories. So the manager register the same repository, both inventory data and Zenode. It is not the case for for sharing because they manage the registry completely by themselves. Yes. But of course, as you said, if they could run a check and see if, for example, the on page of the repository match among the various registrations, it is another work we are doing with Andrea, that is to verify the number of overlaps of registrations within the same registry and among the registry for the repository. And we have found many of overlaps within open door. Thank you. So if it is, if we have a lesson learned from today, and this was also the intention was to highlight that keep your records up to date. So with the right, with the right information, even if it is the sometimes the title, the name, and then important, the URL and even the way IP image interface, for example, in some cases, it also requested. Sometimes for the same repository, also the information about the kind of content is different from one registration to another one, even within the same registry. I do not have the example now. Yes, we know. So the issue is that there are some people that did the registration five years ago, someone that came and do again now. So be aware of that. So check the way that your repository and your services present in these directories because this is relevant for other kinds of services, like OpenAir, that we are relying on alternative registry. So for repositories, OpenDoor, Re3Data, and fair sharing, also, there is the for Cree systems, the Drees directory that is important also for you to be aware. I think it's also, I think if we have this action, the call for action to pay attention for these kind of things that are critical. This is not a formal study, but in one of the meetings, it's just something that I want to share in one of the recent meetings that with another colleagues under the Confederation of Open Access Repositories, one colleague presented a study. Maybe I will invite them and then I can make this publicly available. But they did some tests with the OIPMH interfaces. What else from the OIPMH interfaces available in OpenDoor? And the number of non-resolved links were much, much higher than you presented. So it's around 46 of the OIPMH URLs were not resolved. 46% of the links available in OpenDoor. So this doesn't mean that 46 of the real repositories are available in terms of OIPMH interface because there are old records. So double records from the same repository. Maybe one is correct and the other is wrong. But of course 46 is already give already a number for us to think about. And then when we want to, when we are working on OpenRepositories to make available the scholarly content to the world and we have this kind of problems with the doors that open the content for the world. So we need to think about that. So pay attention to the records. Do you have a URL that you can share? No, I suppose this will be presented soon. I will invite the colleague to present in one of the community calls. I want to challenge. It was just really just a quick testing downloading the dump from OpenDoor and run the test. But yes, this is critical. I found it quite critical and I want that to be presented for us for sure in OpenAir but also in this community call. So I think this study was useful because of this call for action of all of us to ensure that we have the content proper updated for those that don't know about the fair sharing directory. So you can also register there your repository, OpenDoor, then Re3Data for research data repositories. It's important to just to make clear that Re3Data was the first one to give DOIs to repositories but fair sharing is also providing that DOIs. I suppose that there is some kind of gap between the DOI from Re3Data and the DOI then in fair sharing but I'm not sure. Maybe you can give or receive some clarifications from them and then for pre-system DOIs. So ensure that your rep or your system is properly registering these directories. Miriam, any final remark? No, just to when you register your repository check before if you have already registration update the metadata and if something changed please go back and change the information just no more than this. Okay, perfect. Thank you very much. So we are coming to an end so be aware of that those novelties that we shared. So the last update for first of April in terms of content from OpenAir and also this work around the onboarding of YOSC. So we will for sure dedicate in one of the upcoming calls more time to this onboarding in YOSC for you to be like with a demo of the process etc but be aware of that. This is also a critical issue and it's important to highlight that the provide of OpenAir is linking to the onboarding of YOSC. We have some problems during the morning if you access the provide with AII the login. They are already solved. I was just checking so we had some time one hour or a bit more that the system for when we try to log in was not available but everything is working well now so if you try that be aware of that. Is any other question any other topic that you want to discuss any feedback any issue with your repository that you want to report that we can solve it now? So about the dots the common dots that I also highlighted if you want to have more details I did some demos in the last community call so please check the recordings in the OpenAir website they are available there. So medium please share the slides with us in order for us to share it with the colleagues. I will send them to Andre. Perfect thank you very much and then Andre that is organizing this session with me will send them the slides for all the participants will be also made available in the website. So do not forget that we have the we have the newsletter that you usually receive. We highlighted some interesting novelties in the last newsletter that we sent out yesterday and before so if you don't receive subscribe in openair.eu slash newsletters. We highlighted some important novelties it's not this one but be aware of some of the highlights that we did about the campaign about the campaign with the interoperability guidelines also about this study that we have highlighted and about some guidance that we are providing also for the positionary manager so please subscribe the newsletter and follow our community calls. Our coming community call will not be on the 5th of May but on the 3rd of May okay so we will have the community call on the first Wednesday of May that will be on the 3rd of May okay. I'm just confirming in the calendar. Thank you very much for joining. Do not hesitate to contact us via the helpdesk at openair.eu if you have any question. So have a nice day and if it is the case to have some holidays in this Easter also enjoy in this period okay. Bye bye all. Thank you.