Steve Ballmer was an underrated CEO

Azure's success was specifically set in motion under Ballmer. Owed to the fact that it was developed to Microsoft's strengths (enterprise support) that it didn't piss off too many of their partners and sales channels. Same with Office 365 and all of their other successful services. None are glamourous - but all are impressive with how not awful they are given their design constraints.

Even things like Surface, while considered a failure, did its intended job of getting hardware partners to get their act together and make better consumer products.

We should resist that temptation and judge him on the results he delivered. MS was the essential tech company, king of the world, and under his leadership their innovation stalled, they lost in markets where they were leading, the stock stagnated, and huge piles of money were vaporized on acquisitions that were poorly planned or executed.

He tried to buy Yahoo for $44 billion! Only Yahoo’s greater idiocy saved him from that gargantuan mistake. And that was just one of many.

We later transitioned into volume licensing, that also was simple and straightforward. The business side of Microsoft was a streamlined unstoppable train at that point.

The technical side, not so much. Microsoft was still trying to be the only software company in the world, and it was pushing all kinds of WPF, WCF, and other WTFs. So they completely missed hyperscalers and the growing market of Linux-based servers.

Dated 2013, a year before Nadella became CEO:

https://www.change.org/p/the-microsoft-board-of-directors-as...

Buying Yahoo would have been a bet that didn't work out, probably, but I don't think it goes against the point in the article.

Basically the cloud for everyone but the tech companies

Developing OSes and software was clearly an unsustainable business. It's obvious in hindsight that cloud infrastructure was the way to go. But at the time placing a lot of different bets to find a few successful product-market fits was the best you could ask for.

In this era of Python development, Microsoft Windows still feels a step or two behind as far as using a Windows laptop for coding in the cloud. Python is the language of AI - not Asp.net, not C#. Ballmer would never have seen the writing on the wall. He would've pushed something wierd, like VBA .

You could literally get certifications in Microsoft licensing. There were experts whose only job was Microsoft Licensing consultants.

MS’es licensing was so bad you would get different quotes from the same person within a week of asking because almost no one understood it.

And most companies are not tech companies. This is something that tends to be lost in HN discussions (not saying that applies to you, specifically).

I've spent a lot of time in the Microsoft world. I worked for AWS as well. In general, Microsoft executes on platform and ecosystem in a way that works very well for a lot of customers. In general, AWS executes on products better, which tends to appeal more to those who are focused on technology, specifically.

I think that Ballmer's management can take a lot of blame for that. I think a different CEO could have executed and possibly have kept Microsoft in that market with success.

The Apple App Store by itself is a trillion dollar ecosystem. Microsoft being able to gain even a sliver of that size would be worth quite a lot.

We might give Apple similar criticism on the other side of this coin by saying that it's somewhat insane that Apple hasn't tried entering the public cloud market, especially given the fact that they now design their own ARM processors that are essentially the market leaders in that architecture.

I'm even using some Azure Tables for backend services just because it was easier to deal with.

Cloud infrastructure has become a commodity though and you can replace your cloud provider easily (theoretically, lol). What moat can MS or anyone else build around cloud infrastructure? Compare to OS' where MS may never have had a competitor catch up if they'd kept up speed on their OS teams.

Same with video games these days. Adding in digital casinos may seem nice but now you're just the same as every other digital casino offering.

Yahoo.com remains the 8th most visited website on earth[1] (I had no idea until I read that on HN some months back). It sits between Wikipedia and Reddit.

[1] https://www.similarweb.com/top-websites/

But OTOH AWS generally works and usually does what you think, whereas I'm never surprised when Azure breaks or some random Azure API works nothing like we expect.

A lot of this sounds like Google under Sundar's leadership, although I'm not sure if there is a parallel to the failed acquisitions, and some of the rot had set in well before.

I mean, Microsoft was too early and too late on smartphones. I never cared to look into the pre WP7 history.

But the more recent Windows Phone died with WM10, which I don't think is fair to blame on Balmer. WM10 came out in public beta in Feburary 2015, and Balmer was replaced in February 2014. Microsoft eliminated their legendary testing program in August 2014, and the WM10 betas and release in November 2015 had very poor quality. On my phones, I had to choose between annoying bugs in notifications in WP8 or WM10 with a subpar, laggy experience with mobile Edge that managed to be worse than mobile IE. They did manage to get a decent final release together in 2020, although mobile Edge was still crap. You can blame Balmer for not letting Firefox on their app store, I think; a browser that didn't suck would have helped me stay on WP longer anyway.

Still, I think Continuum with an x86 phone could have gotten market share, but Intel cancelled atom for phones in April 2016.

MS did not just have bad luck, they lost to competitors.

Yahoo management was looking to reduce the cost of running web search and advertising platforms, but ended up still having a large expense to crawl the web and basically do web search in order to enhance Bing results. And then the Microsoft ad market managed to be worse in all sorts of ways (for advertisers and publishers) compared to the existing yahoo one, plus Microsoft took a cut of the revenue. Some of that should have been better if it was one company; plus, I bet Microsoft would have sent Yahoo employees an Xbox360 or something. (I worked for Yahoo Travel from 2004-2011)

We needed only the basic stuff: Windows Server, Exchange, MSSQL, a bunch of XP licenses. And this all was straightforward. We also got MSDN subscription basically for free.

There are only 3 significant providers and the needed investments are a gigantic barrier to entry but sure it’s a commodity.

Until you have to call Google. Google business services are awful.

The biggest difference IMO is in how they're handled by large organizations and how prod permissions are provisioned by them. In Azure you have one user account and one org, with subscriptions for your user account to activate to get permissions. You can have multiple subscriptions but they're under the same login/user account and you can have multiple active at the same time. In AWS, you get access to an account or accounts that have different logins, so you get to juggle those with login/logout, even if there's SSO. In GCP, there are multiple projects, under a single login, but you may have to juggle projects.

The other aspect is how regions are dealt with. AWS global resource index/search thing is useful, but it totally feels like I spend more time juggling regions with AWS. Azure's regions themselves are, let's just say, interesting. GCP is better at it than AWS, and less interesting than Azure (which is a good thing).

> Basically the cloud for everyone but the tech companies

is referring to the idea that tech companies are competent enough to run their own computers, not that Azure is missing something specific.

the old ‘embrace-extend-extinguish’ model is what it _truly_ is, f.e. , you cannot take extensions from m$ store and use it.

there have been large number of discussions around this topic, and folks have highlighted these concerns more articulately than i could ever hope to do.

take your pick.

It’s commonly promoted as „open source“ and this seems to be commonly believed. Pretty much everyone I tell that the official builds of VSCode are proprietary (and how proprietary they are) is pretty surprised.

He feared it as a threath to Microsoft's business model and revenue streams.

With this mindset, what could MS possibly create that wouldn't make you say this?

I must have very different needs. In my perception, AWS is the best of the three, Azure is second, and Google would be #3. Depending on your unique news, you might choose different CSP's, from Digital Ocean to Oracle or IBM (the only place you can get AIX, IBMi, and z/OS)

Surprising. As a startup I just couldn’t understand how to subscribe to MS Office, seems like it required a hotmail account or something, it always bored me before completing the steps.

I was a Windows Phone user during 8 and 8.1. There was a short period where I felt like some traction was taking place. My bank even had a Windows Phone app, until they didn’t.

Windows 8.1 was the most competitive version against contemporaries, but then the delays and issues surrounding 10 really took the wind out of those sails.

My perspective as a consumer is that Microsoft buying Nokia seemed to have made Nokia worse and delayed their phone development process. I found myself without any upgrade path, while Apple and Samsung users could get a pretty significant upgrade in capability every year at that time.

Nokia was also better at making low-end phones and had very few flagship products that were iPhone and Galaxy competitors.

On the business side Microsoft didn’t focus on having their entire lineup available on all four US carriers. They had all these weird carrier exclusives where getting a new Windows Phone would mean switching carriers.

I have to think that the break in compatibility between Windows 7 and 8 really screwed over developer relations as well. On the Apple side they were delivering an experience very familiar to Mac developers, and on the Android side the experience was an open source free-for-all playground.

Microsoft does something shitty? See, they're a terrible company.

Microsoft does something awesome? Well, we're currently in the "embrace" or "extend", so they're a terrible company.

I'm as (or more) pessimistic than the next guy about the state of tech and capitalism, but at least give credit when and where credit's due.

That was Bill Gates. Bill Gates founded the company on BASIC and seemed to remain a fan of the language even as the rest of the world moved on to other languages.

Ballmer wasn't technical so appeared to have no skin in the game of which language "won", so long it was Microsoft Developer Tools like Visual Studio developers used to work on it (and what would become VS Code, which as many point out did start under Ballmer's tenure). That "Developers! Developers! Developers!" meme was directly an "I want to support developers wherever they are and however they want to work". Sure he was a huge Windows cheerleader and would want those Developers working on Windows machines, but he really did seem to want to see Windows be the best platform for developers to code for anything (including/especially the cloud).

In terms of Python specifically, IronPython was active and interesting during Ballmer's tenure and Ballmer helped form a team that was actively contributing to open source projects like Python (and Node and Redis and others) to make them all run better (sometimes much better) on Windows. Ballmer may have been afraid of open source as a business model, but he also seemed to realize the usefulness of open source for bringing developers (back) to Windows and he did start efforts in that direction.

I very briefly tried WP10, and it actually seemed a step backwards; in their desperate attempt to somehow unify the desktop and phone thing, be it in code base or user experience or whatever, they tried too hard.

The project was in my team (CSA labs) but was cross-funded behind the scenes by Kevin Johnson, the president of Server & Tools. KJ & I did this because there was passive-aggressive resistance to a 'cloud first' design/architecture philosophy from within his org, where there was a deeply-rooted belief that enterprise servers and ops management tools would adequately scale-up.

KJ bought in and was all-in, as was the 'tools' part of his org (Soma & ScottGu). SteveB initially didn't quite know what to make of my desire and myriad efforts to fundamentally transform the company from packaged products toward services, and he had to cope with some of the wake I was leaving. It wasn't all smooth. But he believed in me and helped me to recruit internally, which was essential.

My explicit cross-funding agreement with KJ, my peer, was that when I decided it was the 'right time', I'd hand off my Azure org and it would be re-merged into S&T in more-or-less a 'reverse merger', with cloud leadership taking over server.

I launched Azure at PDC 2008 with what today we'd call lambda's (functions-as-a-service based on .net) & blobs & cloud database as the core services. Why no linux or windows VMs? They were absolutely part of day 1 plans, but a major political ploy from within KJ's team ('this will kill the server business') resulted in an active decision (mine) to defer until post-launch. It wasn't a technology issue, nor was it an OSS issue; the team believed in OSS & Linux. But shipping was top priority, and we shipped.

When I ultimately left the company in 2011, it was time to do the reverse merger that KJ and I had planned. A proven, super-talented manager from Bing that everyone loved, Satya, was chosen to lead the org as it was moved into S&T upon my departure. James Hamilton, the architect of Azure's relational DB, left for AWS. Ultimately, under Satya, ScottGU & co ended up re-plumbing much of the original code with a by-then-ready Windows hypervisor, VMs & Linux, and all that you see today. By then the org finally was aligned and 'believed', and SteveB was genuinely 'all in'.

Getting products from 0 to 1 is sometimes a challenging process involving incredible people and stamina from believers at every level. In this case I'd say it was worth the effort.

Do you do cross-region work in AWS? For me that is just bad. I usually want to work at a service level and see across all regions, where as AWS wants me to work at a region level and see everything I have running there.

BTW, I hate Terraform with a passion, but I don't have the energy to write something that can describe infrastructure in Lisp.

In any case, all three big ones are pretty good for almost anything, and I don't think there is anything that's doable on one that's impossible to do in the others (IBM is the exception here - if you need AIX, IBMi or Z, they are the only ones who can offer that).

Microsoft's documentation wouldn't even acknowledge the present of Linux, I kid you not, till maybe 2012 or so. For example, pathnames like "My Folder" (spaces in folder names) - which are a no-no on any kind of server code (leave alone the block letters). This was as someone pointed out, Gates, since he was the tech architect (and hated linux or feared it ). In a sense, Linux rescued Azure, and Microsoft. Quite ironic, today we see Gates smiling (cluelessly imo), but Window's is still not a good environment for development - be it C# (a fine language), or Web. My colleagues on Windows struggle to run any Python code - all you need to do is git clone, followed by Pip install - that's still a challenge comparatively on Windows.