As you almost certainly know by now, earlier this week Microsoft announced that it was acquiring Github. There's been plenty of hand-wringing about this among some. Microsoft has a pretty long history of bad behavior and so many of the developers who use Github don't have much love or trust of Microsoft, and thus are perhaps reasonably concerned about what will happen. While I'm disappointed that another interesting independent company is being snapped up by a giant, I'm not completely convinced this will be a bad thing in the long run. Microsoft is a fairly different company than it was in the past, and there are reasons to believe it should know enough not to fuck things up. Alternatively, if it does fuck it up, it's really not that hard for a new and innovative company to step into the void (and certainly, others are already jockeying for position to attract disgruntled Github users).For this post, however, I wanted to point to three different reports in reaction to the news -- because I was fascinated by all three of these takes. More specifically, I found two of them thought-provoking, and one laugh-inducing. And it made me realize just how poorly many non-specialized reporters understand the stuff they're reporting on, while how those who have a really deep and implicit understanding of things provide so much greater insight. Let's start with the laugh-inducing one, before moving on to the thought-provoking. The hilariously bad take is found as an editorial in the Guardian, which has already been corrected once for falsely claiming that Github was open source software, rather than that it hosted open source software (among other things). But the really insane paragraph is this one:
GitHub, by contrast, grew out of the free software movement, which had similar global ambitions to Microsoft. The confused ideology behind it, a mixture of Rousseau with Ayn Rand, held both that humans are naturally good and that selfishness works out for the best. Thus, if only coders would write and give away the code they were interested in, the results would solve everyone else's problems. This was also astonishingly successful. The internet now depends on free software.
Confused ideology? Mixture of Rousseau with Ayn Rand? What the fuck are they talking about? And then after noting how free software has been phenomenally successful, it then says this:
But the belief that everyone coding would solve anyone's problems has been shown up as completely ludicrous. If anything, computer literacy has declined over the generations as computers have got easier to use. In the heyday of Microsoft, almost everyone knew some tricks to make a computer do what it should, because almost everyone had to if they wanted to get anything done. But hardly anyone today has the first idea of programming a mobile phone. They just work. That's progress, but not in the direction some idealists expected. Significant open source software is now produced almost entirely by giant commercial companies. It solves their problems but could be said to multiply ours. Huge cultural and political changes are presented as technological inevitabilities. They are not. The value of GitHub lies not in the open-source software it hosts, which anyone could copy, but in the trust reposed in it by users. It is culture, not code, that's worth those billions of dollars.
The whole piece seems premised entirely on a near total misunderstanding of the reasons why people use Github, the ethos of free software, and well... just about everything. Of course it's culture that's important... but it's so odd that this editorial goes out of the way to insult a strawman culture it believes permeates Github, while then claiming that it's what's valuable.So let's move on to the better takes. I'll start with Paul Ford who is, hands down, the absolute best, most thoughtful, insightful and thought-provoking writer about technology issues around. His piece for Bloomberg Businessweek, entitled GitHub is Microsoft's $7.5 Billion Undue Button is truly excellent. It not only does one of the best jobs I've seen in explaining Github for the layman, but does so in the context of explaining why this deal makes sense for Microsoft. Amusingly, I think that Ford is making the same point that the Guardian's editorial was trying to make, but the difference is that Ford actually understands the details, whereas whoever wrote the byline-less Guardian editorial clearly does not.
GitHub represents a big Undo button for Microsoft, too. For many years, Microsoft officially hated open source software. The company was Steve Ballmer turning bright colors, sweating through his shirt, and screaming like a Visigoth. But after many years of ritual humiliation in the realms of search, mapping, and especially mobile, Microsoft apparently accepted that the 1990s were over. In came Chief Executive Officer Satya Nadella, who not only likes poetry and has a kind of Obama-esque air of imperturbable capability, but who also has the luxury of reclining Smaug-like atop the MSFT cash hoard and buying such things as LinkedIn Corp. Microsoft knows it's burned a lot of villages with its hot, hot breath, which leads to veiled apologies in press releases. I'm not asking for your trust, wrote Nat Friedman, the new CEO of GitHub who's an open source leader and Microsoft developer, on a GitHub-hosted web page when the deal was announced, but I'm committed to earning it.
But perhaps most interesting in Ford's piece is that, while it understands why Microsoft is doing what it's doing, it's also a bit wistful of how he'd always kind of hoped that Github would become something more -- something more normal, something that applied to much more of what everyone did. While it doesn't directly say it, it does imply that that dream probably won't happen with Microsoft in control.
I had idle fantasies about what the world of technology would look like if, instead of files, we were all sharing repositories and managing our lives in git: book projects, code projects, side projects, article drafts, everything. It's just so damnedsafe. I come home, work on something, push the changes back to the master repository, and download it when I get to work. If I needed to collaborate with other people, nothing would need to change. I'd just give them access to my repositories (repos, for short). I imagined myself handing git repos to my kids. These are yours now. Iteratively add features to them, as I taught you.For years, I wondered if GitHub would be able to pull that offtake the weirdness of git and normalize it for the masses, help make a post-file world. Ultimately, though, it was a service made by developers to meet the needs of other developers. Can't fault them for that. They took something very weird and made it more usable.
The final thought provoking piece comes from Ben Thompson at Stratechery, who sees the clear business rationale of Microsoft's decision. Microsoft built its entire business as a platform for developers (who it sometimes treated terribly...). But as we've moved past a desktop world and into a cloud world, Microsoft has much less pull on developers. Github brings it tons and tons of developers.
Go back to Windows: Microsoft had to do very little to convince developers to build on the platform. Indeed, even at the height of Microsoft's antitrust troubles, developers continued to favor the platform by an overwhelming margin, for an obvious reason: that was where all the users were. In other words, for Windows, developers were cheap.That is no longer the case today: Windows remains an important platform in the enterprise and for gaming (although Steam, much to Microsoft's chagrin, takes a good amount of the platform profit there), but the company has no platform presence in mobile, and is in second place in the cloud. Moreover, that second place is largely predicated on shepherding existing corporate customers to cloud computing; it is not clear why any new company or developer would choose Microsoft.This is the context for thinking about the acquisition of GitHub: lacking a platform with sufficient users to attract developers, Microsoft has to acquire developers directly through superior tooling and now, with GitHub, a superior cloud offering with a meaningful amount of network effects. The problem is that acquiring developers in this way, without the leverage of users, is extraordinarily expensive; it is very hard to imagine GitHub ever generating the sort of revenue that justifies this purchase price.
Thompson's piece (among many other good insights) suggests why developers might not need to fear Microsoft's ownership, because of all the potential acquirers, Microsoft probably has the least incentive to ruin Github:
This, by the way, is precisely why Microsoft is the best possible acquirer for GitHub, a company that, having raised $350 million in venture capital, was possibly not going to make it as an independent entity. Any company with a platform with a meaningful amount of users would find it very hard to resist the temptation to use GitHub as leverage; on the other side of the spectrum, purely enterprise-focused companies like IBM or Oracle would be tempted to wring every possible bit of profit out of the company.What Microsoft wants is much fuzzier: it wants to be developers' friend, in large part because it has no other option. In the long run, particularly as Windows continues to fade, the company will be ever more invested in a world with no gatekeepers, where developer tools and clouds win by being better on the merits, not by being able to leverage users.
My own take is somewhere between all of these. As soon as I heard the rumor, I started thinking back to the famed Steve Ballmer chant of "Developers, Developers, Developers!"Microsoft has always needed developers, but in the past it got them by being the center of gravity of the tech universe. A huge percentage of developers were drawn to Microsoft because they had to develop for Microsoft's platform. That allowed Microsoft to get away with a bunch of shady practices that certainly created a bunch of trust issues (Facebook might want to take note of this, by the way). Nowadays, in the cloud world, Microsoft doesn't have that kind of leverage. It's still a massive player, but not one that sucks in everything around it. And, it does have new leadership that seems to understand the different world in which Microsoft operates. So it will be interesting to see where it goes.But, as someone who believes in the value of reinvention and innovation among the tech industry, it's not necessarily great to see successful mid-tier companies just gobbled up by giants. It happens -- and perhaps it clears the field for something fresh and new. Perhaps it even clears the field for that utopic git-driven world that Ford envisions. But, in the present-tense, it's at least a bit deflating to think that a very different, and very powerful, approach to the way people collaborate and code... ends up in Microsoft's universe.And, as a final note on these three pieces: this is why we should seek out and promote people who actually understand technology and business in understanding what is happening in the technology world. The Guardian piece is laughable, because it appears to be written by someone with such a surface-level understanding of open source or free software that it comes off as utter nonsense. But the pieces by Ford and Thompson actually help add to our understanding of the news, while providing insightful takes on it. The Guardian (and others) should learn from that.
The former aide, James A. Wolfe, 57, was charged with lying repeatedly to investigators about his contacts with three reporters. According to the authorities, Mr. Wolfe made false statements to the F.B.I. about providing two of them with sensitive information related to the committee’s work. He denied to investigators that he ever gave classified material to journalists, the indictment said.[...]Mr. Wolfe’s case led to the first known instance of the Justice Department going after a reporter’s data under President Trump. The seizure was disclosed in a letter to the Times reporter, Ali Watkins, who had been in a three-year relationship with Mr. Wolfe. The seizure suggested that prosecutors under the Trump administration will continue the aggressive tactics employed under President Barack Obama.
The war on unofficial transparency continues -- this time ensnaring a reporter. The indictment [PDF] shows Wolfe was in regular contact with four unnamed reporters and the classified info leaked apparently related to the investigation of Carter Page. (The indictment refers only to MALE-1.).Despite all the dots connected by the Justice Dept. after hoovering up email and phone records of four reporters, none of the charges brought against Wolfe involved mishandling classified info. All three charges listed are for lying to the FBI, not exposing secret info. While the info obtained may have been necessary to prove Wolfe lied to investigators, it does seem like a serious breach first amendment boundaries for nothing but vanilla "lied to the feds" charges. Those charges are mostly there for the government to punish people when it thinks it can't nail down more serious charges.And it is a breach of expected norms, if not a reliable indicator of how many civil liberties the government is willing to doormat to hunt down leakers and whistleblowers.
Under Justice Department regulations, investigators must clear additional hurdles before they can seek business records that could reveal a reporter’s confidential sources, such as phone and email records. In particular, the rules require the government to have “made all reasonable attempts to obtain the information from alternative, non-media sources” before investigators may target a reporter’s information.In addition, the rules generally require the Justice Department to notify reporters first to allow them to negotiate over the scope of their demand for information and potentially challenge it in court. The rules permit the attorney general to make an exception to that practice if he “determines that, for compelling reasons, such negotiations would pose a clear and substantial threat to the integrity of the investigation, risk grave harm to national security, or present an imminent risk of death or serious bodily harm.”
It's not clear all these steps were followed. But there are a whole lot of exceptions available to the FBI to bypass these steps meant to protect the First Amendment. No one seems to have been notified beforehand, and it was far more than call/email metadata that was obtained. The indictment cites the content of encrypted messages -- suggesting yet another area where the FBI's "going dark" rhetoric is overblown.
After the story was published, WOLFE congratulated REPORTER #3, using Signal, stating "Good job!" and "I'm glad you got the scoop." REPORTER #3 wrote back, using Signal,"Thank you. [MALE-l] isn't pleased, but wouldn't deny that the subpoena was served."
Going after reporters' records may become standard operating procedure. The Obama Administration prosecuted more leakers and whistleblowers than all previous presidents combined. This administration appears ready to dwarf Obama's numbers.
Attorney General Jeff Sessions said last year that the Justice Department was pursuing about three times as many leak investigations as were open at the end of the Obama administration.
If the DOJ isn't going to give the First Amendment a wide berth, it's not going to be much friendlier to the rest of them -- like the Fourth. Aggressive pursuit of leakers -- and the attendant collection of reporters' communications/data -- will continue. The DOJ may have guidelines meant to limit investigators from obtaining journalists' records, but they're not much practical use when they can be waived to preserve the "integrity of the investigation."