The fact that PBi desktop is free was a great move from Microsoft. You can learn and use all the features at your own pace, as long as you wish, without paying anything, and most importantly you do it locally! That’s how it became popular. I totally agree with you Greg! Thanks for this video!
What attracted me to Power BI was that it reminded me so much of Excel, but without all the VBA that I used to write to mash data sources together. I already had built a data extract process using shell/perl scripts to spin out pipe-delimited files for consumption in Excel. Moving those to Power BI was a no-brainer. Most of the heavy lifting is in Oracle but there are some transformations that I leave to M. Your nocalc method of DAX seems much more straightforward and helps me to "translate" from SQL to DAX. Thanks for the video and thanks for giving your workstation's browser a break.
@DeronHuskey Yep, I originally started using out really just leveraging the data transformation capabilities first in the Power tools in Excel and then Power BI. Still probably the best part of the entire tool.
@@MicrosoftHatesGreg seconding @DeronHuskey on the allure of moving up from Excel BI. This was my exact path as well. You may have alluded to this in "local data model" but I think there are a distinct class of pro developers whose path was enabled by the Data Engine (primary) and Power Query (lighweight ETL)
Great points. I feel like at some point (if it already hasn’t happened), someone somewhere is going to take the cools things out of PBI and put things like DuckDb, Malloy, Polars, and D3/Vega together and make something to tease out the cool stuff. Each of these things do something really cool that Power BI does, but to my knowledge no one has connected these dots.
I really didn't know what I was doing with DAX and specifically Calculate for months, your course DAX for humans helped me a lot thank you I always say that I learned DAX from you I love your opinion, Power BI visualization capability is almost nothing compared to Tableau but I couldn't feel calculated fields in Tableau are as strong as DAX
5. Is already partially possible. A PBIX creates a local SSAS instance which everyone can connect to. The only missing link is, where do the PBIX get hosted?
I kind of feel like Microsoft has already sort of shown the way for this with DirectLake and Fabric. Essentially, they store everything as parquet files and then make those "look" like Analysis Services. With all the different file formats that PBIX can now produce, why couldn't you just write a service that effectively does the same thing but instead of Analysis Services, use some kind of open source analytics engine that your web service then communicates with to display the visuals. I mean, all the information is there in the files for every layer. The first person to figure that out either becomes a billionaire or gets sued out of existence.
For me, #5 is selling point. I work for a medium-sized company and the main value is the ability to serve personalized reporting in a controlled manner with RLS. Could you comment on how this works for other vendors?
Great observation. RLS. That is a distinguishing feature that I did not include. I think that is next tier though, need to meet the base requirements first and then worry about next level functionality. It's a really good point though. For a variety of use cases it is an absolute must.
The fact that PBi desktop is free was a great move from Microsoft. You can learn and use all the features at your own pace, as long as you wish, without paying anything, and most importantly you do it locally! That’s how it became popular. I totally agree with you Greg! Thanks for this video!
What attracted me to Power BI was that it reminded me so much of Excel, but without all the VBA that I used to write to mash data sources together.
I already had built a data extract process using shell/perl scripts to spin out pipe-delimited files for consumption in Excel. Moving those to Power BI was a no-brainer. Most of the heavy lifting is in Oracle but there are some transformations that I leave to M.
Your nocalc method of DAX seems much more straightforward and helps me to "translate" from SQL to DAX.
Thanks for the video and thanks for giving your workstation's browser a break.
@DeronHuskey Yep, I originally started using out really just leveraging the data transformation capabilities first in the Power tools in Excel and then Power BI. Still probably the best part of the entire tool.
@@MicrosoftHatesGreg seconding @DeronHuskey on the allure of moving up from Excel BI. This was my exact path as well. You may have alluded to this in "local data model" but I think there are a distinct class of pro developers whose path was enabled by the Data Engine (primary) and Power Query (lighweight ETL)
I couldn’t agree more.
Great points. I feel like at some point (if it already hasn’t happened), someone somewhere is going to take the cools things out of PBI and put things like DuckDb, Malloy, Polars, and D3/Vega together and make something to tease out the cool stuff. Each of these things do something really cool that Power BI does, but to my knowledge no one has connected these dots.
The person that connects those dots is the next tech billionaire.
I really didn't know what I was doing with DAX and specifically Calculate for months, your course DAX for humans helped me a lot
thank you I always say that I learned DAX from you
I love your opinion, Power BI visualization capability is almost nothing compared to Tableau
but I couldn't feel calculated fields in Tableau are as strong as DAX
@@sajadahmadi595 Thanks! CALCULATE obfuscates a lot of important DAX concepts. I think it's easier to truly learn DAX without it.
Any chance you can review Looker from google in your PBI alternatives series?
100% on the list!
5. Is already partially possible. A PBIX creates a local SSAS instance which everyone can connect to. The only missing link is, where do the PBIX get hosted?
I kind of feel like Microsoft has already sort of shown the way for this with DirectLake and Fabric. Essentially, they store everything as parquet files and then make those "look" like Analysis Services. With all the different file formats that PBIX can now produce, why couldn't you just write a service that effectively does the same thing but instead of Analysis Services, use some kind of open source analytics engine that your web service then communicates with to display the visuals. I mean, all the information is there in the files for every layer. The first person to figure that out either becomes a billionaire or gets sued out of existence.
For me, #5 is selling point. I work for a medium-sized company and the main value is the ability to serve personalized reporting in a controlled manner with RLS. Could you comment on how this works for other vendors?
Great observation. RLS. That is a distinguishing feature that I did not include. I think that is next tier though, need to meet the base requirements first and then worry about next level functionality. It's a really good point though. For a variety of use cases it is an absolute must.
@Greg, where is that Patreon project on No CALCULATE? I'd like to contribute.
Here you go! www.patreon.com/c/MicrosoftHatesGreg. Working on Chapter 7 right now.
your idea is crazy man 🤣🤣🤣🤣🤣🤣🤣
It's crazy now, give it 6 months...