Tableau 2023.3 Coming Soon!

0:00Hey it's Tim here, 23.3 is nearly out let's

0:02go through the coming soon page and find

0:04out

0:04what's coming in this release. Let's get

0:06stuck in. Okay so 23.3 is coming soon as

0:09Tableau liked to

0:10say on this page. This page came out a

0:12couple of weeks ago, I'm a little late to

0:14the game but

0:15I'm really keen to go through it. This is

0:18the second release since we had conference

0:21and the

0:21interesting thing about this is that at

0:24conference there was a whole pitch around

0:26new capabilities,

0:27Tableau GPT, Tableau Pulse and this concept

0:30of decoupling. I went through all of this

0:32in my

0:33keynote roundup, check that out if you

0:34haven't seen that. And some of the things

0:38we saw at the

0:39conference are yet to be released and so

0:41what's always interesting a couple of

0:43releases after

0:44conferences to see what's actually shipping

0:47because of course we've only got really one

0:50more release before I think they start

0:52saving features for the next conference. So

0:54this release

0:55and the next release will pretty much have

0:56everything we're going to get this year and

0:59then everything after that will probably be

1:01saved for availability at some point next

1:03year just to

1:04give conference a bit of content that

1:06people can actually talk about and

1:08something that sells

1:09people can actually sell. So a little bit

1:11behind the scenes there. So yeah here we

1:13are with 23.3

1:15let's start going through this. I'd like to

1:16do this before every release is out before

1:18I even

1:18do the videos where I cover them. If you're

1:21new here don't forget that I make videos

1:23covering

1:24all of the features of Tableau as they're

1:26released the most prominent ones not every

1:28single one.

1:29We're trying to address that problem very

1:31very soon but yes I make videos covering

1:32most of the

1:33features that are released and I have been

1:35for three years so go ahead and check those

1:37out I

1:38have playlists for each of them. Let's

1:41start today with Dynamic Axis Ranges. So

1:44this is actually

1:46quite nice because it comes off the back of

1:48Dynamic Axis titles which was released a

1:51couple

1:51of releases ago. This essentially allows

1:54you to map the axis to a specific parameter

1:56so as you

1:57select the data set you can actually get

1:59the axis to change. Now you could kind of

2:02simulate this

2:03with some capabilities around min and max

2:06values and parameters but this is nice

2:08because this feels

2:09more native and so it's really a follow-on

2:12to the previous feature. Now one question I

2:15have for you

2:16is this I've been talking to a few people

2:18and I can't come to an opinion about

2:20whether it's better

2:22that Tableau releases all the features

2:24behind the capability in one go or if they

2:27drip feed these

2:28capabilities. A good example is ask data

2:31another example is the data model when

2:33these were new

2:33features they gave us features over maybe

2:36five or six releases would you prefer that

2:38at the end of

2:39that six release you just got everything in

2:41one go then so you could plan and organize

2:43how you deploy

2:44your features across your organization or

2:47do you prefer this sort of drip feeding.

2:49The interesting

2:50thing is that with this drip feed approach

2:52my opinion is that most people don't get

2:54the features

2:55until two years down the line because in

2:57essence every organization's sort of

2:59upgrade cycle means

3:00that they get the feature kind of when it's

3:02half baked and when it's half baked you can

3:04't really

3:05do anything with it so you actually get

3:06this lag effect unless you're on Tableau

3:08cloud where

3:09everything comes sort of in real time and

3:12so in real terms I just wonder if people

3:14hold off

3:15developing ways and best practice of using

3:17things until they feel a feature is sort of

3:20fully

3:21realized and as a result Tableau think

3:22people aren't interested in the feature and

3:25they sort

3:25of start barking down the wrong tree. It's

3:27an interesting question I don't have the

3:28data of

3:29course if you're at Tableau and you have

3:30the data in the telemetry let me know we

3:32can chat

3:33but nonetheless let me know what you think

3:35in the comments. Okay let's go to the next

3:37feature custom

3:38data labels now I actually saw this in

3:40Tableau online the other day when I was rum

3:42maging around

3:43I realized that I was in a beta instance

3:44and this was already enabled so I'll do a

3:46video on this

3:47very very soon. It's a really nice

3:48capability essentially it's building off

3:51the previous data

3:52sensitivity feature that was available in

3:54the previous release this now gives you a

3:55place to

3:56manage these so this will be great for

3:58admins who want to start to use this

3:59labeling across the

4:00platform. I think this is going to need the

4:03data management add-on because the data

4:05sensitivity is

4:06part of the data catalog and so this feels

4:08like it's going to be part of that as well

4:10if you have that add-on. Always important

4:13to note that add-ons do affect the

4:15availability of certain

4:16features so whenever you see anything to do

4:18with data lineage or data catalog unless it

4:20's the

4:20metadata API you need an add-on in order to

4:23access those features. Okay let's go to the

4:25native

4:26lightning web component for Tableau. So

4:28lightning web components are from the Sales

4:30force world they're

4:31sort of a bespoke HTML web component that

4:34works really really well with Salesforce.

4:37It makes it

4:38easy for developers to work with the Sales

4:41force platform and so what we have here is a

4:43native web

4:44component for Tableau so you can go ahead

4:46and embed Tableau visualizations inside of

4:49Salesforce using

4:50this lightning web component and the

4:52benefit here is that you can actually start

4:54to integrate

4:55Salesforce and Tableau more closely. It's

4:57kind of felt like all of the embedding API

4:59capabilities

5:00that we've been getting from the last

5:01couple of months has actually been leading

5:03up to this sort

5:04of capability they needed to sort of

5:06implement all the little bits personal

5:08access tokens all of that

5:09stuff needed to come together to enable

5:11this and so you can actually see that they

5:13do call out

5:14connected apps trusted tokens all inside of

5:16this particular announcement so this is

5:18really nice to

5:19see I'm really excited to see what comes of

5:22this. On demand access this is part of the

5:25embedding

5:25story now embedding is just on a march I'm

5:28making a video series about this so be sure

5:31to check that

5:31out very soon over the next few weeks and

5:34in that sort of topic and domain on demand

5:37access has been

5:38a hugely requested feature the context here

5:40is let's say you're building an application

5:42let's say

5:42you're a bank Tableau likes to use a bank

5:45as an example and you want to build a

5:47platform where

5:48people can come in and check their balances

5:50and you want to use Tableau as the

5:51visualization

5:53tool to visualize that data well what you

5:56don't want to do is to go and buy licenses

5:59for all your

5:59customers because what if they never log in

6:02what if they don't actually use the service

6:05the other

6:05thing is and you might actually be on

6:07something called a usage-based licensing

6:10system so there is

6:11an ability to buy tokens and embedding has

6:13a capability to basically pay as you use so

6:16you can

6:16buy a certain number of tokens up front and

6:18then it doesn't matter sort of how many

6:20people log on

6:21you can essentially use up these tokens

6:23over a period of time and so what you need

6:24to be able

6:25to do is say hey I'm going to give all 12

6:27000 people access to this system but I don

6:29't know

6:30when they're going to turn up and I don't

6:31know when they're going to sign in but when

6:33they sign

6:33in I want you to go ahead and create a

6:35username some credentials for them and then

6:37let them log

6:38in and see what they're supposed to see all

6:40in that one step and so what that this is

6:43actually

6:43doing is enabling that when a user then

6:45goes to log in it automatically licenses

6:47them and make

6:48sure they're allowed to see what they're

6:50allowed to see it passes through any sort

6:51of permissions

6:52and general things that your application

6:54needs to control and then you have the

6:56availability to just

6:58go ahead and let them into your platform so

7:00really really a useful feature and the

7:03embedding API NPM

7:04library this is one of these web

7:06development terms but I'll try and simplify

7:09it for you

7:09NPM stands for node package manager I think

7:12I think it's node package manager and what

7:16essentially you can think of it as is a way

7:18of bundling the necessary code you need to

7:20run a

7:21certain set of capabilities in your

7:23application let's say I want to build a

7:25clock I don't want

7:26to have to you know go and do the maths to

7:28figure out the time right maybe I want to

7:30connect to a

7:31computer or a satellite that has the time

7:34what I can do is I can go get a package

7:36that knows how

7:37to call the API for this particular

7:39satellite or clock and I can include that

7:41in my application so

7:42that my application always knows the time

7:44that's a very simple example what they're

7:46doing here is

7:47taking the embedding API and creating a

7:50library of tools and what you can do is

7:52actually use this

7:53library you can call up these libraries

7:55like components and these components allow

7:57you to

7:58quickly and more easily build capabilities

8:01on the embedding platform now the nice

8:04thing about this

8:04is that the package management is all done

8:07for you you don't have to sort of go and

8:09find out I'm

8:10on the right version of this which version

8:12of the API am I using am I using embedding

8:14API version

8:153.5 3.6 the package manager should really

8:18sort of tie all these in and then also

8:20allow you to lock

8:22into specific versions so you really

8:23tightly control the experience and so that

8:26's that's sort

8:26of a really really good capability and the

8:29additional benefit here is that the API can

8:32be

8:32pulled from a CDN so you can go ahead and

8:34call it as you would do with any other

8:36package and it'll

8:37be pulled down and available in your

8:39application so a much easier way to deploy

8:42the embedding API

8:43and the nice thing is I think they do

8:45mention about applications applications on

8:48premise

8:49without the option to connect to CDN so you

8:51can essentially take this package and

8:54include it in

8:54your application without needing access to

8:56the internet which a lot of things kind of

8:58come with

8:59the basic assumption that you do need

9:01especially if you're a Tableau cloud

9:02customer so yeah

9:04really good set of capabilities if you know

9:06more about this and I'm

9:07sort of murdering this description let me

9:09know in the comments below

9:10okay the next feature is stratified

9:13sampling in prep now stratified sampling is

9:16one of these

9:17things where I'm always like I bet you

9:19people don't even know this is a problem

9:21and this is

9:22going to fix it for them so when you open

9:24up Tableau prep by default it samples over

9:27a certain

9:28number of records it basically doesn't load

9:30all the data in and when you're working on

9:32certain

9:32data sets this can actually affect the

9:34spread of data that you get inside of the

9:37way you build your

9:38flight and so you get the scenario where

9:39you build a flow you think you've solved

9:41all the problems

9:42you hit run the full data set comes through

9:44and actually you discover more problems

9:46because

9:46something you designed isn't quite working

9:48for the specific subset that you were using

9:50when you

9:51built the flow because of sampling what

9:53stratified sampling allows you to do is it

9:55essentially allows

9:56you to pick a field and it will use that

9:59field to I guess better proportion the way

10:01it samples

10:02so it kind of uses that field to say hey I

10:04'm designing my workflow around this column

10:07this

10:07is really the absolute sort of you know

10:10most important level of grouping in my data

10:13set

10:13I want you to pick a representative sample

10:16of data using this grouping and that's

10:18essentially

10:18what this is doing so it's a really nice

10:21feature I'm glad this is coming in most

10:23people probably

10:23won't know this is going to help them but

10:25it's a great option to have in there

10:26especially if

10:27you're working on data sets I kind of wish

10:29it was default I kind of wish Tableau just

10:32figured out

10:32the best field to use and just went for it

10:35you know figure out the level of detail and

10:37just use

10:38this as a way of sorting things out I'm

10:39sure something like that already happens

10:41but nonetheless

10:43really really nice okay fill down and

10:45running calculations so multi-row

10:48calculations were added

10:49in the previous release of Tableau in this

10:52release they're adding a few more options

10:54to fill down and

10:55essentially do running calculations what I

10:57loved about this feature when it was

10:59launched is that

10:59it's got this nice beautiful interface and

11:01I like that they're adding more

11:03capabilities to it

11:04another example of something that I think

11:06this should have all come in one go in this

11:08release

11:09rather than being the last release because

11:11in real terms when you want this feature

11:14you kind of want

11:14the full gamut of what it can do but hey in

11:16this particular instance I think this is an

11:19exception

11:19because even the previous set of features

11:22delivers value out to the gate next up set

11:24a header and

11:26data start row for CSV and text files again

11:28another enhancement for prep it feels like

11:31prep is

11:31starting to sort of edge a little bit

11:33closer to altrix I don't think it's

11:35anywhere near a sort of

11:36taking altrix market share or anything like

11:39that I just do think it's going to make

11:42those altrix

11:42use cases a little bit more difficult to

11:45justify because all the sort of ground

11:47level capabilities

11:48can just be done in prep anyway set header

11:51and data start row for CSV slash text files

11:54and this

11:55is a nice feature it's kind of an

11:56expectation really for CSV and text files

11:58just being able

11:59to choose the row you start from but it's

12:01finally here and so we're going to be able

12:03to use this

12:04okay the next feature is write data sets to

12:07data cloud from tableau prep now this is an

12:11interesting

12:11one because of course salesforce are

12:14pushing salesforce data cloud tableau genie

12:17cdp all these

12:18things I never understand what they mean

12:20but at least for now you can use tableau

12:22prep to push

12:23data to the data cloud in real times I don

12:25't know what that means it does suggest here

12:28there's an

12:29ingestion api connector I think this

12:32specifically allows you to pull to certain

12:34objects inside of

12:36salesforce but honestly I don't know enough

12:38about salesforce yet to be able to talk

12:40about this but

12:41that's definitely something we're going to

12:43try and address long term still working my

12:45way through all

12:45those knowledge bits and bobs that I need

12:48to kind of get through but salesforce is

12:50dense and salesforce

12:51is a very complicated platform especially

12:54if you approach it as like an analytics

12:56person coming

12:57from the world of tableau and you're trying

12:58to understand salesforce there's just a ton

13:00of things

13:01where you're basically sifting through

13:03marketing terminology rather than actual

13:05sort of core

13:06features so and yeah that's that's just

13:08something to be aware of okay virtual

13:10connections support

13:11for all connectors in tableau this is nice

13:14so they've been slowly adding capabilities

13:16to

13:16virtual connections what this means is that

13:19they are pretty much done adding all of the

13:21capabilities that I think you can access

13:23through the web edit features so the

13:25connectors for

13:26one drive apache spark cloud adara cloud

13:29era hadoop hive the data rama dreamy all of

13:33these

13:33have been added along with a list of other

13:35ones I'll do a separate video on this

13:36because there's

13:37been a few virtual connections additions in

13:39the last release and in this release so I

13:41'll kind of

13:41round up everything that we've got so far

13:44maybe do a 2023 update for virtual

13:46connections containerize

13:48bridge for linux so what does this mean um

13:49you can basically make your tableau

13:51architecture more

13:52streamlined with support for linux and

13:55deploy tableau bridge within a container in

13:57linux now

13:58containers are an interesting concept in

14:00computing if you take your computer as an

14:02example and it's

14:04basically right in front of you you can

14:06always access it let's call that a bare

14:07metal it's

14:08physically in front of you the next concept

14:10is you take your computer and then you chop

14:12it up into bit

14:13but then you say each of those bits are a

14:15small bit of your computer in that case you

14:17could take

14:18your laptop and say it's in six bits you

14:20could say you have six virtual machines

14:23they all run on

14:24one machine one physical machine so bare

14:27metal one machine virtual machine six

14:31computers okay

14:33a container is when you take an application

14:36and you put it inside of a box typically

14:39inside of

14:39a virtual machine and it has no awareness

14:42of anything else okay so it knows that it's

14:45running

14:46on windows or it knows that it's running on

14:47linux but it doesn't have any concept of

14:49any other

14:50application that could be running on what's

14:52the value of this well when it breaks you

14:55can just

14:55replace it you can just set up more and you

14:57can in fact take the exact same settings

15:00and put it

15:00in another virtual machine or do do

15:03something else with them the nice thing as

15:05well though is because

15:06it's not aware of anything else you can

15:08actually run multiple containers on a

15:10specific virtual

15:11machine which makes it much more portable

15:13and it means that you can manage your

15:15virtual machines

15:16and your physical hardware separately from

15:18how your applications work you could update

15:20let's say

15:20your operating system without really

15:22worrying too much about the containers and

15:24the applications that

15:25are running on that system so it just sort

15:27of modularizes things and it makes it

15:29easier for

15:29it professionals to manage infrastructure

15:32in a more sort of a flexible way and so

15:34being able to

15:35run tableau bridge in a container on linux

15:37is actually a pretty good capability

15:39because it means

15:40companies can do a couple of things firstly

15:41it's linux which means you don't have to

15:43pay for a

15:43windows license just to keep tableau bridge

15:45running and then secondly the container is

15:47the preferred method because again

15:49containers are just a little bit more

15:50flexible and portable

15:52okay the next update is actually quite

15:54useful bridge support for embedded data

15:57sources now make

15:58connecting your data easier with bridge

16:00enhancements um tableau bridge now supports

16:02embedded data sources and published data

16:04sources you might think what on earth does

16:06this mean well

16:07when you go from tableau server to tableau

16:09cloud tableau bridge is what you use to

16:11update data

16:12sources that sit within your firewall that

16:14need to be pushed to tableau cloud for

16:17availability

16:17and workbooks and other data refreshes

16:20until now you couldn't do that with

16:22embedded data sources

16:23and embedded data sources simply a data

16:25source that only exists inside of a work

16:28book it is not

16:28separately published and so what that means

16:30is that when companies were going from

16:32tableau server

16:33to tableau cloud they'd get this sort of

16:35painful awakening that their embedded data

16:37sources would

16:38no longer be supported and so you'd have to

16:40work out a whole new data flow to get data

16:42sources

16:42stood up on side of tableau cloud so inside

16:46of tableau cloud so this update me means

16:49that

16:49essentially the effort required to migrate

16:52from tableau server to tableau cloud goes

16:53down

16:54unfortunately though you'd have to go to

16:56this release as soon as it's launched to

16:57realize that

16:58benefit so in real terms in two years time

17:00this benefit will actually come to life and

17:03an existing

17:04content requiring bridge can now be

17:06migrated as is rather than converting

17:08embedded data sources

17:09to publish once it takes a whole load of

17:11effort it seems like a feature that's been

17:13designed to

17:14make migrations easier rather than a

17:16feature that's been designed to bring

17:17parity that's

17:18sort of how it's being framed here um but

17:20yeah a nice feature i think a lot of people

17:23doing

17:23server to cloud migrations will appreciate

17:26this activity log enhancements the activity

17:29log has

17:29been the place where tableau has been

17:31adding capabilities to allow you to see

17:33what's happening

17:34from your tableau service perspective or

17:36tableau cloud perspective think of it as a

17:38really detailed

17:39auditing capability they've added a hundred

17:42types of usage events the activity log

17:44including access

17:45create update and delete events across all

17:47content types so that's basically if you

17:49take each of those

17:50and you split them up there's probably 20

17:52things in each and you've got them all

17:53covered through

17:54the activity log i need to do a video on a

17:56whole range of activity log this is sort of

17:58the release

17:59where i realize i've got so many small

18:00technical bits that have been accumulating

18:02over a while

18:03that i've not covered and i should probably

18:05cover them in one go admin insights tokens

18:08uh data

18:09source so this is a strange one when i saw

18:11this the first time i sort of thought

18:12thought hey why

18:13do we not get the workbook as well in

18:16essence admin insights is a data source and

18:18what they're

18:19adding is a tokens data source a specific

18:21data source that tells you about your

18:23access tokens

18:24and how they're being used it also has uh o

18:26auth client tokens and oauth database tokens

18:29so all of

18:30your tokens will sit in one place tell you

18:31when they're being used when they're exp

18:33iring i hope

18:34you get this dashboard with it i hope you

18:36don't have to build that out of the gate

18:37because this

18:37dashboard looks useful but i'm also certain

18:39people will be able to build better ones so

18:41if you're

18:41sitting out there thinking of a good work

18:43book that you could build this would be a

18:45great one to kind

18:46of get stuck into suspend extract refres

18:48hers for inactive data sources oh god this

18:51is like one of

18:52those low-hanging freaks that i'm glad they

18:54've solved um if an extract refresher is

18:56running and

18:56no one's using the data source why is it

18:58running it's one of those things where if

19:00you do an order

19:01of tablo server every tablo server has a

19:03certain number of workbooks that sort of

19:05meets this

19:05criteria so thank goodness this is being

19:09enabled i think you might be able to

19:11control um when a data

19:13source is labeled as inactive i think there

19:15's a there's either a variable that you can

19:17sort of

19:17apply or there might be a way of um sort of

19:20toggling that a little bit so i'm not sure

19:22if

19:22you have control of that through something

19:25like tsm or if you're able to set a

19:26threshold but that

19:27would be really good to see once the

19:29features launch i'll have a look around on

19:31that feature

19:32once we get it through and see what we can

19:35do autosave enhancement so autosave is the

19:39capability that allows you to save it is

19:41only web edit specific don't get tricked

19:43into thinking this

19:44is coming to desktop but you now get more

19:46support from tablo autosave to improve

19:48productivity the

19:50latest enhancement autosave automatically

19:52modifies wait this is weird automatically

19:55saves

19:55modifications made while editing a new

19:57never published workbook to a draft so

20:00basically as

20:00soon as you open a workbook it's

20:02automatically saving and they're saving it

20:04to a draft which

20:05means you no longer need to publish your

20:06work immediately to be able to get the

20:08ability to

20:09save draft essentially so this is really

20:11nice and the update allows you to pick up

20:13where you left

20:14off conveniently and this is great if you

20:15're working in the browser because

20:17ultimately it

20:17means if your browser dies or your

20:19connection drops off you've already

20:20automatically got like

20:21a backup that you can fall on the login

20:24flow has been enhanced essentially this has

20:26been redesigned

20:27it's a little bit nicer and a learn more

20:29link for the site uri has been created i

20:32had to create

20:32a video on what the site uri is because

20:34when tablo cloud kicked into action lots of

20:37people

20:37didn't know what their uri was because it

20:39got removed from this login page that was

20:41typically

20:42there back in the tablo days so um this is

20:45now uh been updated to sort of make it a

20:48look nicer

20:49look smoother whoever this is uses last

20:51pass you can see the last pass thing

20:53logging in

20:54that should really be first photoshop that

20:56's free advertising for last pass but anyway

20:58um uh really nice to see this what i think

21:01would be nicer is that if tablo just

21:04centralized all

21:04its login um there's there's currently i

21:07currently have four tablo logins and they

21:09did sort of merge

21:10them all but the problem is is that because

21:12i've had years and years of people using

21:14separate

21:14accounts for separate things it was an

21:16absolute mess so i still just keep

21:18everything in separate

21:19accounts because it's easier than trying to

21:21merge everything into one especially for

21:23people who have

21:23a work profile and a personal profile and

21:26are trying to figure out what to do and

21:28also if you

21:28have an email tied to tablo cloud

21:30unfortunately it's so hard to move things

21:32off that account

21:34that it just becomes impossible so um yeah

21:36never never really sign up to anything that

21:38you want to

21:38keep personally using your work account

21:41sign up using your personal account and

21:43those can move

21:43with you including tablo.com as well so

21:47just bear that in mind okay editable viz

21:50all text is now in

21:51desktop this is not really new this is a

21:53feature coming from web edit to desktop so

21:56this is nice

21:56the ability to edit this in desktop is now

21:59available um command cancellation this is

22:02one

22:02of those things where if you clicked cancel

22:04when a data source was updating it wouldn't

22:06actually

22:07cancel it would do some other weird thing

22:08and so what they've actually done is made

22:10this do what it

22:11says and um it also has an ability to pause

22:14and uh enable auto updates as well baked in

22:17so they've

22:18kind of done everything you'd want to do

22:19when you hit this feature so you don't get

22:21it spinning again

22:22automatically as soon as you go to load the

22:24workbook which is kind of nice.

22:25Accelerators are getting more enhancements

22:28they keep tweaking this sort of design

22:31and I think they're trying to make them

22:32more and more prominent you can kind of see

22:33it's

22:34sort of slowly creeping up you've got quick

22:37start and it's taken up half the page now

22:39I'm sure very soon you'll let you load up a

22:42start page and ask you what do you want to

22:45do and it'll

22:45say here are three accelerators ready to go

22:48and off you go I think that's pretty much

22:51what's

22:51well where this is heading it's um it's

22:53pretty obvious to me but hey let's let's

22:56wait and see

22:56where a tablet end up with this um if we go

22:59to the next one table fonts for web author

23:02ing

23:02uh this is interesting because obviously

23:04web edit is still catching up with desktop

23:06fonts were one of those areas where it's

23:08behind and now you can edit the fonts for a

23:10table

23:11inside of web edit which is kind of handy

23:14red hat enterprise linux that's what rel

23:17stands for it says

23:18so right here red hat enterprise linux

23:20version 9 is now supported for tablo server

23:23so you can now

23:24update to that version of linux or put

23:26server on that version of enterprise linux

23:29dynamic parameters in table extensions so

23:32table extensions are to me this is this is

23:36easily

23:37the um the most uh capable part of tableau

23:41that I think people can use chat gpt for

23:44because

23:44previously you really needed to understand

23:47python and art in order to interact with

23:49table extensions

23:50I now believe it's possible in chat gpt to

23:52basically code your way through this and

23:55feel

23:55like you know what you're doing dangerously

23:58obviously um and build some really

24:00incredible

24:00capabilities um what you need to do is

24:02understand how the table extensions

24:05capability works

24:06but I'll definitely do a video on this

24:07because I've had great success trying a few

24:09things out

24:10that I just never really had the balls to

24:12try before so I'm really interested to see

24:14how this

24:15works and so we can kind of push on with

24:17this and and see where it ends up but when

24:20you have dynamic

24:22parameters inside of table extensions what

24:24you can essentially do is dynamically

24:26change what the

24:27table extension is doing and that means you

24:30can change the query you can change the

24:32call being

24:32made in python and r and you can also

24:34change the way in which you're interacting

24:36with the dashboard

24:37so because these can be part of a data

24:39model by changing the parameter you can

24:41also do things like

24:42change the relationships and how they work

24:44so that's a really really nice set of

24:46features but

24:48that's it that's 23.3 I'm not gonna lie it

24:51does feel a little lighter than usual I

24:53think tableau

24:54gpt and tableau pulse are stealing a lot of

24:57exogen for upcoming releases all the effort

24:59is going into

25:00making that work making that ship I'm sure

25:03tableau just pivoted tons of dev work

25:05towards

25:06ai there's probably a company-wide memo if

25:08it's not ai related I don't want to hear

25:10about it or

25:10something like that and but also we can see

25:13through these little little features here

25:15and there that

25:16there is something there what I like to do

25:18is go to desktop like this and I just like

25:20to look at

25:21this and say hey what's new in desktop

25:23there's a little less that's appreciated if

25:25I go to tableau

25:26cloud there's a little bit more and so what

25:29you start to realize is that yes tableau

25:32really is

25:32heading more towards the cloud I just don't

25:36understand how anyone can not agree that

25:39tableau

25:40cloud will be the best version of tableau

25:42to use because that is the version that

25:44gets just a

25:46boatload of releases every year and tableau

25:48server is getting a bunch of features in

25:50this release

25:51that were also available in the previous

25:53release remember this is the server release

25:55so it's

25:55actually available in this release it'll

25:57get all the previous features not all of

26:00them because some

26:00of them are cloud specific as well as these

26:02releases so this will actually sort of make

26:04its way to tableau server now as well as

26:06everything in the previous release anyway

26:08thanks so much for

26:10watching it's been really good to kind of

26:12go through these features dream forces next

26:14week

26:14look out for that I'm pretty excited to see

26:16what tableau has in store for us at the

26:17dreamforce

26:18keynote the tableau dreamforce keynote as

26:20well as salesforce of course and what they

26:22have to say

26:22about everything and maybe do a reaction on

26:25the salesforce keynote and the tableau

26:27keynote just

26:28to see what they have to say about ai and

26:29chat gpt but I don't really know much about

26:32it so we'll try

26:32and figure out what's going on there and I

26:34'll catch you in the next one thanks for

26:36watching

26:36and I'll see you soon

26:38you

26:48[BLANK_AUDIO]