After listening to episode 321 of The Azure Podcast[0] my understanding was that Azure DevOps would eventually be phased out; question begins at ~10:30 and at about ~12:00 a rough timeline of 5 years was given with guidance to select GitHub if just starting out.

0: http://azpodcast.azurewebsites.net/post/Episode-321-GitHub

The writing is on the wall... TFS is going the way of Silverlight.

Very valuable comment / insight via the podcast – thanks!

As somebody who uses Pipeline (well, VSTS Releases, we're not on Azure Devops yet) professionally, I've got to pick up GH actions now. Hadn't gotten around to it.

That said, like 90% of my Pipeline actions are "screw it, I'll do it all in PowersHell"

Please deprecate Azure devops repos, which my company uses, to let me go back to GitHub. I absolutely hate the UI and miss GitHub greatly

I get that you guys want to say that publicly, but let's be real. No company would invest a massive amount of money in a duplicate product. One product will eventually starve.

I guess it is up to us to guess. Anyone?

I see GitHub being the unmovable giant here. Microsoft is publicly developing on it, as opposed to Azure Dev Ops. It has a very large mind-share. More developers are willing to use it without having the Microsoft stigma that some nix people feel.

> No company would invest a massive amount of money in a duplicate product.

I don't mean to be rude, but have you worked at a very large company like Microsoft or Amazon or Google? Redundant products are par for the course because of the byzantine internal politics and funding structures of big companies.

Big companies like Microsoft and Google like to burn products with little notice too.

Google sure, but Microsoft? The company that kept the Zune service alive for 4 years after the product was EOL and with a userbase likely measured in the hundreds of thousands?

https://www.wired.com/2015/09/what-to-do-with-your-zune-rip-...

The company who STILL supports 16-bit apps?

https://www.groovypost.com/howto/enable-16-bit-application-s...

Ya... I would hardly say MS is known for killing stuff early - more like they've spent years being ridiculed for carrying baggage forward for decades longer than anyone else.

MS might be bad at a lot of things, but I'd hardly say they're known for "burning products with little notice".

Have you done any development work on .Net in the last 10 years or so. I've been buggered at least 5 times by massive discontinued chunks of stuff and the several reorganisations that got rid of my entire selection of enterprise customer and MS connect cases conveniently.

Then again there is this list of 346 discontinued Microsoft products, some of which had very short lifespans: https://www.versionmuseum.com/history-of/discontinued-micros...

Yes, I would definitely hate to trust Microsoft with my enterprise software build pipeline because of how they refused to support Microsoft Bob.

Well, probably not because of Bob, but their cloud based offerings have make me wonder about trust.

- Business Contact Manager for Outlook, Outlook Customer Manager

- Microsoft Invoicing, Listings etc.

And these are critical applications for a company.

Have a look at Sharepoint which is widely used and has an uncertain future. Or the strategy behind Lync, Skype and now teams.

But we'll see. Microsoft has shifted in a good way in the last couple of years but their track record in keeping legacy operating system APIs for decades is not necessarily a good indicator of the stability of their other product lines.

Business Contact manager is still fully supported - it's just not supported on the latest version of outlook. On Outlook 2010 you've got support through the end of 2020. For Outlook 2013 they haven't announced an end-of-support date yet.

Microsoft Invoice has transitioned to a cloud-based product, so again, they didn't end support. You might not like the new purchasing model, but that's very much different than them burning the product to the ground.

https://einvoice.microsoft.com/Default.aspx?MSIStateKey=f513...

Sharepoint is the backend for onedrive for business, and fully integrated in to Teams. What on earth would make you think it's going away?

Sharepoint has an uncertain future? I had never heard of it a year ago, but as I got to know the "enterprise" space, it seems every large company is heavily invested in it. What might replace the need to share documents across a company in the MS world?

well a lot of things in the business section had a different production which could directly import the data from the old one or different migrate the data. like business server essetnial or dynamics marketing most often the new stuff was more expensive. Even skype for business online is upgradable. some stuff has less features, like hotmail which could use all custom domain names and not only godaddy ones like outlook.

What about Atom/VSCode? Atom development looks dead to me:

https://news.ycombinator.com/item?id=21142934

https://github.com/atom/atom/graphs/contributors

...and small companies go under or radically morph their products.

There's this irrational demand vocal on social media that large corporations keep their products forever.

Sears is doing it!

That is true for Google, but certainly not for Microsoft. Microsoft's support for legacy software is pretty amazing actually.

It's terrible. AppFabric, WCF, WWF, windows phone. I could go on for hours...

WCF is still supported and a lot of stuff works on .net core 3.x and more is coming in 5.x. webforms on the other hand... (which should die a more faster death)

See this: https://news.ycombinator.com/item?id=22872556

> No company would invest a massive amount of money in a duplicate product.

Google's text messaging and video chat apps didn't get that memo.

ADO is widely used inside Microsoft, with a variety of internal extensions to integrate with our internal build & deployment solutions.

AFAIK, there aren't any plans in Azure to give up ADO in favor of GitHub. If anything, with the push to standardize builds internally, it wouldn't make sense to move to GitHub for at least another 2-5 years.

Obviously, I don't speak for my employer and leadership may have other directions in mind.

Even then... I don't expect Github actions to go away any time soon. I would expect a lot of the underlying systems, build agents and workers to be the same over time though.

Azure DevOps and Github largely cover different, though overlapping market segments.

I would be slightly more concerned about Github Enterprise and Devops co-mingling over time, as I think that may be inevitable, which makes me concerned over the public/free resources that Github offers in the long run... even then, migrating to Gitlab is an option should that time come. My only hope would be better discoverability and social coding with Gitlab to better match Github over the interim time.

Even then, it's just a possibility and somewhat unlikely that MS would burn this much karma.

They clearly capture different markets and are both doing well. Why is is it inevitable that one will starve? I feel like that's only likely to happen if a new CEO comes or something and decides to shake things up.