Okay, excellent.

So this presentation is more or less thought as a starting point for a discussion about

metadata and taking it a little bit beyond just referencing, citing and giving credit

to the authors of who wrote the software to see what we need to do in the software catalogue

to actually make the software in the catalogue discoverable and also usable.

So we've seen yesterday and today already a couple of software metadata standards and

some of them are built on the others.

But these all concentrate on referencing and citing software.

Some of them have a little bit of additional metadata about, for example, what language

software is written in, where you can obtain the software.

But it's mostly about linking the authors and specific versions of software to papers

that use that software.

So to think about what we could do in the software repository, I phrased a couple of

questions.

So one of the questions is if I have a bit of data and I need to process it and I'm not

too familiar with that particular data, how do I find out what bits of software in the

catalogue I can potentially use on that data?

And actually I think you can take that a step further and also say, okay, say I want to

make an image of an astronomical object.

I have some input data, but now I want also to produce a specific output.

So I think it makes sense to base this decision on some description of the data formats that

the software takes as input and the data formats that the software takes as output.

The science platform that's developed in another work package in SK could, for example, use

this metadata to offer the user choices about what virtual environments to offer.

A problem here is that I don't think that all communities within escape have well-defined

or standardised data formats.

But at least some of them have, and also, of course, the virtual observatory has their

data formats.

So while thinking about this, it seems that the media types as defined by the RFC number

I've written here that are used in the wider community for software, this is, for example,

what operating systems use to decide which application to use with a specific file.

If you click on a file in your browser, it asks you to open a specific application.

And these used to be better known as MIME types, I think.

Some of the older people might recognise that from that.

So there's a registry of these.

So you can formally register media types with the internet science number.

I don't know what the A for stands for, but it's an organisation.

But there's also the namespace for unregistered formats.

So that gives you the flexibility to, if you don't want to, go through the process of formal

standardisation to at least adopt this standard.

And at least within the astronomy, we have the FIT standard for files.

There's actually two media types registered for that, and there is an official RFC for

it.

And these media types are already used in many VEO protocols.

So these media types would offer a good link between data formats and what our software

accepts and produces.

So the other question that we all have is, OK, now I find my software, how can I actually

run it?

So how do we create an environment where this software runs?

And I said without spending a day to install it, because I think we've all tried building