 All right. So this has been in discussion for a while that the new user experience is not exactly ideal. We kind of throw people in the deep end and Galaxy. When you first show up, you kind of get bombarded with all of the information, but no real way, no real info on how to actually use it. You have a massive list of types of tools. You don't actually see most of the tools themselves, which can be very overwhelming to somebody who has never used the platform before. The, the masthead. Well, again, once you know what you're looking at can be pretty informative. You don't necessarily know what the, what a workflow is what the different types of visualizations we have is or how to access shared data. You have an empty history panel that like, maybe we want to put a button that not when it's empty that says upload data or something like that. That's a whole other conversation. And the center panel has no real on main at least has no real information on how to get started. This is me. I just created a new account just to see if it was just me or what. Sorry, Jen. But to that ends, there has been discussion on how to revamp the user's first experience to get them into Galaxy faster, let them know how to use it more easily. And also, as a more involved, like how to use Galaxy, if you've been using it for a while, but you don't quite remember something you don't want to jump over to training. So priorities for new user, provide an optional tutorial that you can easily exit on on the start of a new account to very easily differentiate any server for what you can do and what it specializes in very rapidly, make it in depth enough to easily get not enough that the user has to spend an hour and a half getting to know every faceted Galaxy. Anything more in depth needs to go to training. It needs to be really easily modifiable so that we can quickly add in new things on new releases as they come up or on different servers. Make those really easy to make. So this is what we have now. It is, we have built in kind of three levels of welcome. So you have the main one, which is this is what you would see on your first login. Obviously this can like the images can be shifted out. But the most basic version focuses on data tools and workflow. In each of these you have some topics, each of which leads eventually to a slide show that I can show in a moment, but a user can navigate through this very easily and learn on any aspect they they have not yet had the experience with with a workflow that they can do with a slideshow that they can quickly and easily parse three that has text and images that they can skip around and look things up and very easily. So the very the most default version, as I said is data tools and workflows. The data has topics on upload remote retrieval sharing and and the data and metadata modification tools intro to the standard tool form and visualization tools. I haven't made one for it is quite yet. And workflows is how to extract from a from a previously run analysis how to import remotely and how to share. The actual mechanism for this is pretty basic. It is a standard one pulls an npm package that contains a bunch of images and a couple JSON files that it needs to get an imports into as a single based on objects, and serves on the welcome page. I changed this, however, with a new variable in the galaxy email file which is welcome directory by changing that out, we're using web packs dynamic imports to not pulling the npm package and allow a user or an admin rather to create their own chase on files and folders and directors of images and sub directories and help like there's a structure that I can explain more in depth later that they can very easily customize one for any given server and import there. You can also change the existing one by pulling the npm package and altering the files, you can use it as a basis and then go from there. So blocks. We need like pack five is like pack five has a dynamic imports. We can't currently change the imports depending on on the variable. Dan has PR in this, and it is mostly complete he unfortunately most likely cannot join this meeting as he's having a bit of a child emergency. So I'm hoping to get this up and running by GCC. If it were not sure it will be on main by GCC but should be able to be on main by GCC. And I'm also looking for any comments on additional standard contents that we might want to add other than it is still need to write that. Default node package with the default contents is that going to be presumably that's going to be managed via GitHub in a new repository. Yes, I have one that I'm using right now but once once it's up and running I'm going to fork it and just hand it over to Dan and on galaxy. Okay. And so we'll just need to make sure that the version of that package that's pinned with a galaxy version sort of matches the version of galaxy. Yeah, it is. The main reason is because it will massively increase upload times if the user has to download has to pull the images every time. It's, it's, it's not a massive file system like number of files but it is a large group of images, and that can be a bit of a slow down on startup otherwise. Yeah, no. Ideally the whole app would be this way right ideally we could. The whole client would be pulled down from MPN also and yeah, the back end would be decomposed into pieces that we pulled in. Yeah, and that's really cool. I mean I like that I just Once it's in Webpack 5 it becomes much easier. Awesome. Have you know the training people have looked at it you think training should be part of the default content. We already have the link out on the mast head to training that exists a little cap icon. I think that training definitely I will, I think it's probably a good idea to add a link underneath for the primary welcome page which says for more in depth welcome for more in depth tutorials go to training. I think that's probably a valuable thing. Also this is accessible at any time from the health drop down by the way. It would be nice to see how closely we can get that integration so for example you can launch the training materials on galaxy instances that have been set up with that proxy configuration. So that you can click the buttons in the training and fill out the form so it'd be good to make sure that we maintain that nice integration. Yeah, I think this is really good usability enhancements but yeah similar to Dan I just there's all these like images and little, you know, the training material repo just added this new concept of like, like a fact like a people ask questions with little snippets of like just short little quick things and it kind of looked a bit like what was in the how to upload data to galaxy for instance. And I know that there's like a slightly different audience. And so that's fine and slightly different use case but to whatever degree we can sort of, you know, merge with them would be nice, but I don't know if that's possible and I don't I don't think that should this should block anything it's just, you know, a comment. That'd be nice to sort of work on on sort of synchronizing that but this is this looks great I mean obviously this is galtz needs this so that's fantastic. Any, any topics I might have forgotten that might be a good idea for a base welcome. Can you show the topic again. Alex I just had a question about this slide that last one. Yeah. So, um, when you don't include borders to show to show the navigation it's hard to know what this means. You know, it's a hover, sorry, these turn turn gray on hover, but yeah. Okay, so like, but it doesn't show you that you need to click on upload data. And there is a, this is an image. This is an image this isn't the actual model. Right, so I'm wondering if this should include something within it that has like this upload data, and then an arrow and this is what you get. I'm sorry this is part of the slideshow like the next slide focuses on the button. Okay, and then there's a there's a tutorial in the GTN that's specific to it are we going to include links for how to on the slides. So, again, we can we can edit the content as much as we like this is a this is a file that can vary like it's a JSON that we can very easily hyperlink from. That would be really like nice. I know you just added a lot of their slideshow so kind of it just kind of add a little context like okay what is that and then you go in and you can find, but everything looks fabulous this is great. Thanks. Thanks yeah so with that static image would it be possible to some way actually embed the widget, and then use tours. So, the reason originally that we took it this direction was because tours weren't working for a little while. I don't know if they currently are. So, the answer is, possibly, I need to double check if tours currently function. Yeah, because it'd be awesome if you could actually interact with the thing as it's teaching you right. Yeah. So, the way that the way that the way the slides are set up are those in JavaScript like is that something you'd be able to embed a component in. They are through, they're done by a view serving a JSON object. So I don't know that like the way it's written right now that it would be able to just showing it's part of a bootstrap you slide. Hey Alex, I'm, I called him from my phone. Sorry I'm a little late. Yeah, so we could, we could totally do that and embed actual components, the problem is going to be the mounting and context available. So if you wanted to embed like an upload model or something like that have it pop up automatically. So I have to think about how it actually gets strapped in, given the configurability of the component here. In other words, like so if, if, if use gasey.org wanted a tutorial with an upload component and that you did not. How is that specified in the, the extra bundle that gets built for this thing. What's that configuration or what's that a YAML configuration or something. Well, so it's again I mean it's a JSON blob but where do you actually specify that so in in covered, you know use gasey.org for embedding components in markdown there's there's like a library that handles it and plugs it all in and all that. Yeah, I'm not sure how easy it'll be to, to make that embeddable from that end. I think we can do it though. Um, yeah I mean this is awesome so it sounds like a lot of work so make sure you have something for GCC if that's what you're aiming for. Yes, that's exactly it. Yeah, I think we make the first pass today. Yeah. Also just a quick note about the tours. They are currently non functional. I've been working with Helena a little bit just to, there's a PR open on the training materials GitHub repository. We're investigating how difficult it would be to get them functional again there's been some changes in the UIs. The CSS selectors mainly have changed so they don't really work anymore and we're taking a look to see how much effort would be to get them in line and sync with the UI again. And feel free to change whatever plugin or back end libraries being used for that too right I don't think we're. Yeah, so there's there's also there's an open PR on Galaxy for swapping that from bootstrap tour to bootstrap tourist, which is so bootstrap tour got abandoned a couple years ago. And I think incompatibility incompatibility with that and see their jQuery or something is what's causing a lot of the issues right now. So there's a tourist PR open that should fix up most of that. And bootstrap for swap. Awesome. We should coordinate on those PR so that's if there's something on the training material site, I haven't seen that. So I have some minor suggestions. Can you show the welcome page again. So I feel like the icons are way too big in contrast to the text. Maybe that's just me, but it's not just you we talked about it when we initially discussed it about two months ago. Data tools. This should be more balanced. Right. And then if you can show. So I would suggest for the right and the next one, where you have the slides. So since these are images. I don't want, I mean, I would suggest that we, that it might lead to confusion because in some cases we just like our two form also has no background. And so it you can't really distinguish visually if you can click on it or not. So I would suggest to put the slides, maybe in a box and with a light gray background or something so that it's that it's kind of differentiated from our regular components. I would help just like the whole thing in a small box and like rounded and then was a slight great background and the return button, but these are minor things one can talk later about it. I would probably put to the upper right left. Right. Yeah. So because we had the issues which with big screens that that the return button would be then at the bottom. It's just like some disbalance but the upper left right would probably be good but here as I said like one round box and then the whole thing in there. Then one would see that it's kind of not clickable. That should be docked at the bottom. Yeah, mostly because otherwise it moved along the size of the images and I spent multiple hours trying to get that to not happen. Okay. Yeah, this great of course this will improve definitely the first arrival for new users on the page. So the font awesome stuff as you can tell but like, that's not strictly necessary if people have preferred images to represent any given thing I'm totally cool to use those two. Yeah, it's first pass right if we have nice sort of professional graphics mocked up for tools workflows data kind of thing that'd be, that'd be amazing. I mean, it seems like a lot of people. I mean like I'm very curious to get in there and like look at the initial content, although I'm not really the person that should do that but we should maybe like have a meeting with I don't know, Jen and Helena and like some developer, presumably Alex or Dan and or somebody so that we could like, you know, have more users hacking on that, the initial content and a meeting to walk through all that might be good. Yeah, so the plan was to have a meeting with the goat group working group to look at it and like the next. Yeah. Sometime this quarter. So that's a good idea and already in the works. Okay, I'm yeah the other thing is that that could be around table to potentially just like go through the content if we have an open weeks slot sometime but the goat group might be more focused and so that conversation might be more productive that way. But yeah, awesome. Okay, anyone have else have any comments. I guess everyone would be happy to. Yeah, I guess I said it when there were only a few people on but this was I told Anton I had a time this is going to be a pretty short one because it's not particularly complex. This is true. But it's very extensible such that if we wanted, you know, I think we should stay focused and get something done, but if we wanted to build on it and add additional vignettes be straightforward to do. And hopes Jeremy would be here because he was the one most excited about this from last few times. Well, he'll see it on main soon it sounds like so this is true. Awesome. I'm going to stop recording then I guess. Any like release related stuff or anything like that we want to talk about do we want to just that to use this time just the testing group still need more people. Yes, I was just about to put out a call for participants. I still need a few people to round out the group so if anybody wants to volunteer and you know potentially spare their mates for my random number generator that I'll use to select the team. If I don't have volunteers that'd be great. Because I'd like to send out invites. You know, ideally tomorrow if I can so we can meet early next week. Maybe actually start the testing next week as well and get done so we can get issues and notes to the dev team to wrap up. You know so it's a small release so you know, if you do it now will be less work the next time around. Yeah, exactly so if you you know volunteer for this one it's an easy release there's only a few issues to test out you'll get off easy whereas next time. You know might be a lot more work so get it out of the way well it's easy feature surge. Or you can volunteer a friend. Don't forget to submit abstracts. Monday. Just to clarify, do we need to submit. We don't need to submit an abstract for the training tutorials, or do we. That's built into the schedule those are happening. Period. I think you've been tapped to do one of the trainings then there's no abstract to submit. It's a different part of the program. Do we have any water stand on, you know how abstracts and registration is going. I guess that's a no. I guess no Dave today so I have a usability question is it possible to download when like you have a tabular file in history and you click download to download it with a custom extension like text or something as opposed to that. You should never download with that. But it would be the predefined one. So the one that corresponds to the data type. It's always one to one mapping. Yeah. I mean if you know the special invocation you can change the header. But it's not super. Wasn't a center working on a tool that would switch the data type of the collection or something like that. I see she's here. Yes, I'm here. Um, it's not ready yet, but you can change the data type of the collection using the converter tools right now. That's just for like a whole collection of known with a uniform homogenous type right this. And this is asking about a direct file conversion from a to B or something like that. So just a single data set for a collection. Again, it's once a single data set right. It was a single data set. Yeah, but like just on download not. Not within galaxy not actually change the data type just that download time downloaded as a custom data type. You should be able to get the same file. Yeah, extension. Yeah, so you can. So now when you hover over it it shows you. But if you hover over the download button it says to X something something so I think since last release, you know you could type in the extension you want and then you get the extension back that that's. Yeah. So the time it should be that is when it's a actual data formatted data type. Right. Yeah, it's that it sounds like something's going wrong. Yeah, download a band file you get a dot band file. Yeah, if you don't something's broken. Thanks just double checking here. But in that regard is always annoying is when you download a tabular file and you know your setup knows what to do with TV with CSV where it doesn't know what to do with the tabular right. Wait, would it make sense to download a tabular just to default to download on a tabular to a CSV or something. Yes, we better. Yeah, TSP is what I would say but because we actually have a CSV right. We have CSV CSV and tabular and the tab the TSP is more strict. So, I'm not sure it's 100% win if we change that the TSP is more strict. I would expect TSP and CSV would be identical except for tab instead of a comma right. No that that is identical and I think they use. Okay, just the CSV parsing in Python but the our tabular data type is not a strict I think. And that's why we have them separately which is, yeah, we should have unified that at one point, but at this point I don't think they're exactly the same so we'll be tricky to do it. Is it white space handling or something. Okay, I think I think it is. Yeah, something like that. Okay. And like also, you know that TSP parsing in Python is a bit. So you have to choose a dialect and it's. You can do that stuff in the client now though I mean we could, we could very easily stream the text and do a file object and manipulate it and then, you know, as part of the I mean if it's something you really want and it is something that we could build specific converters for we can look into doing it in the client. I think most of what we're talking about here that wouldn't be reasonable for the client to. So we're talking about gigabytes only be good for like, simple previews. Yeah, yeah, yeah for like previews and stuff absolutely. Sorry. No, no, no, go ahead. Okay, data types with like a default download extension and so we could like, yeah that that's the file that's a galaxy extension by default but for just for tabular we could download them as TSP and then if we had something else like, I mean there's other places we have dots in there like raw.gz or something where we might, you know, we might want that to be slightly different. I would have a download extension on the data type or in the data types XML file or a list though. Is it, is it maybe this is the like the display applications 2.0 kind of thing we were talking about right where you can have each data type has multiple ways to display it's all linked in and different visualizations, whether it's charts or tracks or whatever. And then download is just a different version of that that has multiple output types. Yeah, I just, I just wanted to point out that the modern client you can in fact stream all that data into a file object and even if it's large you could pass through dump it to a link that you click on it's it's the kind of thing you really can do in the client nowadays, even for big files, maybe not gigabytes but pretty big files. It doesn't have to be all or nothing. It doesn't have to be a simple trick text transformation modern the modern native JavaScript objects are pretty, pretty beefy. Something to think about if you really want it. I mean somebody wants to give it a shot I'm not sure it's a really thankful task to do it. And it's tricky to get it right. No suggestion is not bad but like what happens if we use the downloads tabular file tabular data type file gets a T is the absolute uploads it back again, and then selects T is the things that you can just go on because he may not be able to read. So what is the difference. Like why, why do we use that I mean is if we, if a tool is to output a TSP what it displays equally in Galaxy. Display yes but because they're not exactly the same. They don't inherit from each other so they're not, you know, unless the two said I accept both. It's not necessarily working. Looking at the code the tabular data type definition does a lot more with metadata than the TSP type definition so there's a difference there. I'm just saying it's tricky. Right. I mean it's not something you can do in an afternoon. I think I'm fixing tabular to download as a TSP. I mean you're right that you could, you would upload it and it might not strictly be a TSP. I don't. That's a weird who is downloading files from Galaxy and re uploading them and not sniffing like it's. I don't know that that's a problem, I guess is what I'm saying. Your TSP web browser extension or it not opening in Excel or whatever seems like the bigger problem than like upload and download sort of matching. And then that extension could be I mean I could do that and you know it's it's like four lines of code right just like when you, when you like look at the data set to generate that extension to get that download link, you know you could change it. It's just a monotype right. I mean the content attached. This is completely configurable right I mean it's not even four lines as a single line, but. Yeah, so the mind type and the extension are different. But yeah you just set the extension that sets for the download. But I'm not so sure we wouldn't get into user issues that are subtle. I mean, just being able to click on it and I have it open in whatever thing your computer knows how to display TSP is also worth it. So I don't know. I mean a dot tabular extensions kind of worthless right like that doesn't exist outside of the galaxy only concept right like how is it. How is it ever useful to have that that tabular type I don't. Yeah you would think it would be either TSP or CSV. And then tabular is an abstract parent that can have a concrete. Yeah. I think it's easy to have like a fixed number of columns right also is that my making that up whereas I don't think the galaxy tabular type needs to have a fixed number of column. Yeah, I mean the tabular data type can definitely do something like, you know, one column first to come second 15 the third, not complain about it. Tabular can be whatever we want it to tabular is whatever Dan decided tabular was 15 years ago. The other problem with CSV files is there is no real CSV standard everybody that writes a spreadsheet comes up with their own, how they're going to handle embedded columns or embedded quotes or what not so it's CSV files are very much a moving target depending on what you want to program you're trying to load it into. Yeah, just to make it easy we display CSV files to click on the icon icon as tabular. So you have to add is on the peak. Yeah, it's in the peak view it's correct but if you display it it turns into tabular. I mean that's a good thing right because who would want to look at commerce in your table. It's a good thing for someone who knows what they're doing. But if you're not sure, like, I guess I just answered so many questions about people who don't understand, but like, I'm looking at my data, they're like it looks like as columns are trying to append it into a tool it doesn't accept. Or maybe we should start parsing comma separated data to see as speed data as tabular when it gets input to tools. If it's expecting a tab file. I mean we do I mean if it's possible we do that right but you know galaxy does a lot of these things like we display bam files, as you know the same representation. We display compressed files without you know showing them garbage. I think that's a good thing we should keep doing that. Maybe the UI should reflect like hey this is a display you know this is a view. This is raw data, because the problem is is people are going to select all copy put into a work document and then try to run Sam tools on it. I need to kind of get back to the thing I mentioned before the next version of display applications of visualization forever you want to think about it. Maybe that that makes that delineation more clear where look you're viewing data but this isn't the raw data that you download. Yeah, I mean this will be amazing because I also make it much easier to have like custom display. Yeah, you know, you know, the eye can can be one of the visualizations right that's what you're saying. Yeah, yeah exactly right. I'm just playing notebooks directly, but will be my first thing I do. Also, we should, you know, maybe see if we want to do something about the peak, because the peak is not stories HTML and it's in most cases we don't actually store HTML in the peak in the database so we actually either have like, I don't know, we have, I don't know, CSV or we have Jason. I don't know, maybe that's something we can think about that, you know, that's not super hard to like to declare what what that content is, and then client render it appropriately. If we want to do what I just described for the peak. I don't like passing HTML back and forth over the API either. It would be good if, if we could do exactly what you just said, if I could feed into a file object and then render it in some kind of preview. You know, on the client word, that's where the client rendering should happen. Right. Yeah, that's exactly what was pushed for and a recent PR. We should totally do it. Do we sanitize at HTML. It's on the client side right. No, I'm sure the HTML we're putting in the database and we want to get away from that anyway so that's yeah. We want to get away from it immediately or we probably do to some extent right so only I mean that's only generated by the data set class. So hopefully we haven't screwed those up. I mean, even if the peak is gone. I mean, let's put it that way most of the peak, you know show peak implementations actually generate the peak if it's not there. Of course that's terrible because it happens in the front end. In the webpanger. Right we don't want file access during that operation if we can avoid it. But you know we do that actually. When we, well, I am not sure we don't actually touch every file object we display when we click on the details view currently. So if we do we should fix that. Yeah, we have that was part of the reason for having a peak accessible from the database in the first place. It could also be just when you're administrator I don't know, I didn't check. I don't know, maybe a priority for 21. What, I don't know, the last to your title. Yeah. Well that'll be 2201. Well we will have one more starting this year right. Yeah after 2109. So it'll be 2201 is the next. Yeah he has the release yeah. I think that's great. I'll plan anything you want for next year. All right then, I guess all. Now I'll stop recording. Oh, I thought we did.