The dangerous folly of “Software as a Service”

Comes the word that Saleforce.com has announced a ban on its customers selling “military-style rifles”.

The reason this ban has teeth is that the company provides “software as a service”; that is, the software you run is a client for servers that the provider owns and operates. If the provider decides it doesn’t want your business, you probably have no real recourse. OK, you could sue for tortious interference in business relationships, but that’s chancy and anyway you didn’t want to be in a lawsuit, you wanted to conduct your business.

This is why “software as a service” is dangerous folly, even worse than old-fashioned proprietary software at saddling you with a strategic business risk. You don’t own the software, the software owns you.

It’s 2019 and I feel like I shouldn’t have to restate the obvious, but if you want to keep control of your business the software you rely on needs to be open-source. All of it. All of it. And you can’t afford it to be tethered to a service provider even if the software itself is nominally open source.

Otherwise, how do you know some political fanatic isn’t going to decide your product is unclean and chop you off at the knees? It’s rifles today, it’ll be anything that can be tagged “hateful” tomorrow – and you won’t be at the table when the victim-studies majors are defining “hate”. Even if you think you’re their ally, you can’t count on escaping the next turn of the purity spiral.

And that’s disregarding all the more mundane risks that come from the fact that your vendor’s business objectives aren’t the same as yours. This is ground I covered twenty years ago, do I really have to put on the Mr. Famous Guy cape and do the rubber-chicken circuit again? Sigh…

Business leaders should learn to fear every piece of proprietary software and “service” as the dangerous addictions they are. If Salesforce.com’s arrogant diktat teaches that lesson, it will have been a service indeed.

220 comments

  1. Worse when it is subtle or sneaky like Microsoft.

    Gab was kicked off Azure, and they now specifically say they will cancel your account if they find hatespeech in your Office 360 docs.

    https://blogs.microsoft.com/on-the-issues/2016/08/26/new-resources-report-hate-speech-request-content-reinstatement/

    First they just had bad products. Then they added the online connectivity for things like Cortana. Office is now cloud “sharing”. Then they shove Windows 10 down everyone’s throats – “free” because they do the surviellance capitalism and you are now the product.

    Zoho Docs, Libre Office, Linux.

    1. The worst of it is that the surveillance will be automated, with no human review, in which case “I hate Nazis because they believe that Jews should be sent to extermination camps” becomes the same as “Jews should be sent to extermination camps.”

      Add a little white privilege, plus the occasional racist moderator, and you have the current situation, where F*cebook is more likely to ban an anti-Nazi account than a Nazi account.

        1. Down that path, down this general path the totalitarian tech Left is taking us, which is the ostensible topic we’re discussing, one of the best outcomes is “And then one day, for no reason at all, people voted Hitler into power.” (Note that the Left claims that’s already happened….) If you think that’s a worse case, see Cambodia if the Left wins, Rwanda as one way the Right can win, in the sense of avoiding a Cambodian fate.

          1. Because any discussion of the reasons the people voted Hitler into power would necessarily be interpreted as support for electing Hitler, the topic itself ist verboten.

            That this itself can lead to support for electing, say, Trump (who of course is Literally Hitler™) from people tired of being told they’re Deplorable by their Betters, is an irony lost on those Betters.

        2. Are you the original Tron Guy? I think it’s so cool I found you awesome as a 16 year old, and we ended up having similar politics.

    2. Since I haven’t needed anything from Microsoft since 1984, this just confirms my conviction that they’re a vendor I can do without.

  2. I’d emphasize and amend that you should avoid like the plague anything you possibly can that’s based in the California Bay area like Salesforce. The area’s hive mind makes it socially punishing for people to avoid piling on when a Two Minute Hate starts, which has resulted in crime thinkers losing all their services in a matter of hours. Not that Seattle for example is tremendously better, Azure and Amazon retail are suffering from social justice convergence, the latter now outright banning books in all forms, not just in the Kindle silo, but doing business with Bay area firms is by far the most risky option.

  3. You only have to care if your business remotely caters to the right-leaning side of the political spectrum.
    Otherwise you’ll totally be fine.

    1. Give us half a dozen examples of companies of medium or larger size that don’t “remotely cater to the right-leaning side of the political spectrum”, including selling to them. And are invulnerable to purity spirals like the current one, which history tells us will not stop until lives are lost.

      Maybe also some thoughts on what it’s like to live your life on your knees?

      1. >Maybe also some thoughts on what it’s like to live your life on your knees?

        *snrk*

        I know Garret. Your sarcasm detector is impaired.

        1. Your sarcasm detector is impaired.

          Poe’s Law tells us there’s no such thing today, especially if you don’t know the author. And for some people it doesn’t hurt to spell it out in detail.

      2. > Give us half a dozen examples of companies of medium or larger size that don’t “remotely cater to the right-leaning side of the political spectrum”, including selling to them.

        Apple
        Google
        REI
        Patagonia
        Nike
        Gillette
        Facebook
        Twitter

        Now, some of those will *sell* to you, mostly because they’re hypocrits and want your money. But they don’t go out of their way to *cater* to you, and make it clear they despise you and are ONLY doing you for the money.

        1. No, they won’t *sell* to you. They *LICENSE* to you, then revoke your license when you irritate them.

    2. “When they came for $X, I said nothing because I wasn’t $X …”

      Good luck with that.

      The purity screws will continually tighten. They have to, else the Outrage Industry has nothing to do. Even people like Feinstein and Pelosi and other once-darlings are beginning to look insufficiently extreme.

    3. They also tend to ban nudity, despite not all nudity being necessarily about sex, thinking skyclad Wiccans or simply art. And “cultural appropriation” is a very handy way to fuck with even very liberal people. Precisely those people who do not have a strong ethnic identity or ethnic ingroup preference, are the people who are likely to “appropriate” things from other cultures. A conservative tends to stick to his own culture. The real reason for campaigns against “cultural appropriation” is just standard guild-monopolism, established producers in a fairly small market field like to try to shut the entry of new competitors out. https://quillette.com/2019/04/08/how-the-campaign-against-cultural-appropriation-came-back-to-haunt-canadas-indigenous-peoples/

  4. I think there already was an example of an Open Source project with a similarly restrictive license, but I forgot, which project. But I guess if someone changes their license to something of this kind, the project can be forked.

    1. >I think there already was an example of an Open Source project with a similarly restrictive license

      I know of one attempt to add a politically restrictive codicil and blogged about it. It was quickly quashed.

  5. There’s only one problem: there doesn’t seem to be an open source competitor to Salesforce, and I’m not sure that a project of that scope would be viable.

    1. Yes.

      An example that has frustrated me for 2 decades is finding an alternative to Intuit’s QuickBooks. It would not be easy to write & maintain something of that scale and with the necessity of frequent time-sensitive updates (e.g. payroll tables).

      1. If you could separate the code part from the information part, it would be fairly simple to open source the software and then sell the tax data.

        1. But how valuable would that be? The software is mostly useless without the data, and the data is totally useless without the software. Why would they ever be kept in two separate places, when interoperability is a key requirement for both? This basically just opens everyone up to twice the risk, because they need both the open-source team *and* the data retailers to play nice in order to keep their books. I see no actual advantage here.

          1. This is a problem that has been solved for decades. A rarely mentioned discovery from the depths of the CS departments, but the solution does exist.

            I think the technical term is a “file format”.

            1. This is not a technical question. As you say, the question of converting the data from one provider to another is trivial. This is a business question. Or rather, two business questions.

              1) How does it benefit the sellers to generate a file in your format, instead of just selling the software instead?

              2) How does it benefit the users to be just as reliant as ever on a vendor, but also face all the problems of using open-source software where the support contract is “hire your own coders to figure it out”?

              If you bootstrap a whole ecosystem with multiple vendors and a robust support community, sure, it can work. But that’s not a thing anyone can do(aside from a firm like Google trying to screw a competitor, but nobody with sufficiently deep pockets hates Intuit that much). Respect path dependence.

        2. The tax data ought to be provided free by the IRS and state departments of revenue. That government should provide information necessary to comply with the law for free should be seen the same way as the government’s obligation to provide legal counsel in the event of criminal proceedings, as a fundamental protection against arbitrary state power.

          It might require a literal Act of Congress to get all of them using the same data format, of course.

          1. It might require a literal Act of Congress to get all of them using the same data format, of course.

            Or possibly a literal Act of God. I don’t think even Congress has enough horsepower to get every town government’s tax laws to be posted at all, never mind them all being in the same file format.

            1. All they have to do is invoke the Interstate Commerce Clause, which SCotUS defined in Wickard v. Filburn to basically mean anything Congress says it means. (They ruled that a farmer growing crops to feed his own animals, which wasn’t even intrastate commerce, because he didn’t buy the feed from anyone else affected the interstate price of the feed crops, and therefore was covered under the ISC.)

              1. I’ve been mocking Wickard for years. But lower levels of government aren’t as easy to push around as family farms, unless you have grant money that can be cut off (cf. federally mandated drinking ages and speed limits, enforced by threats to highway funding.)

          2. You know better than that. They would get to specify the format, and it would be a terrible format, and then they would half-ass delivering the data and it would not match the published format standards and so on.

            Rather it should just be some kind of cooperative association of professionals. Over here there are good experiences with that: https://en.wikipedia.org/wiki/DATEV only issue is they would not use Linux. But neither would the government.

          3. It might require a literal Act of Congress to get all of them using the same data format, of course.

            Congress is lobbied extensively by Intuit and H&R Block to explicitly not do this.

        1. I’ve looked at many over the years – a lot of which are no longer around.

          As best I can tell none of them have a full-featured payroll module (with auto updated tax tables) – which is why I singled out that feature. They all say “payroll management” but that is not the same.

          Feel free to correct me if I’m wrong.

        2. And do any of them have the integrated scope of function that Salesforce does? That system is as widely used as it is for a very, very good reason: it works well and manages a very large set of problems for a company. I’d be really surprised if there was an open-source CRM tool with its set of capabilities.

      2. Methinks everyone misunderstood my point. Payroll tax tables change sometimes as often as monthly and *must* be shipped timely. And there are many, many jurisdictions that influence these. The point is not that the code & data must be separated – we’ve been doing that forever. The point is that it is a huge and messy human burden to keep these tax tables current and ship the updates on a very inflexible schedule.

        1. Payroll tax tables seems like something that needs a project similar to the IANA’s tzdata/zoneinfo project. Of course, I’m not sure IANA’s the appropriate body to run it – but then again, just why is the IANA keeping up with time zones?

          These types of databases – payroll tax information, sales tax information, ISBNs, UPCs, etc – need to be kept separate from the applications that use them, to prevent vendor lock-in. Ideally, in such a manner that government agencies could submit the changes for their jurisdictions, and then simply have their ordinances point to the database for the specifics.

        2. Does Intuit accept liability if they accidentally have a typo in the tax tables they ship with their product? I doubt it. Every time the government updates the tax code you probably have to spend some time verifying that the tax tables you got from Intuit are correct. You could just as easily spend the time transcribing the information yourself, if your accounting software could read the info from a file in a well-known format. Or you could subscribe to a service that did it for you. You could even ask the government to save everyone the time and effort by supplying it in that format to begin with. Since it costs Intuit time and money to create these tables, they would have already gotten the ball rolling on this if it weren’t for the advantages of vendor lock-in. They only realize those advantages if their customers allow them to.

        3. Yes the regulatory compliance burden translates to rent, to whoever is able to help you meet it.

          This is not an accident.

  6. Aside from being cut off for no business reason — which can be an existential threat to your business, no question about that — there’s the cost. Microsoft, Adobe, and a couple others have seen their profit curve inflect upward since they switched to the SaaS and subscription models. At least some of them showed a per-customer profit increase. It’s not a great leap of logic to conclude that the customers are paying more for the software. This isn’t conclusive — SaaS may save overhead costs to the vendor, or each customer might be buying more licenses. However, based on the pricing offered to me as a small business owner a few years ago, subscriptions became more expensive than outright purchases in less than a year and SaaS running on vendors’ systems would have flipped even quicker because of the various fees which weren’t part of the advertised price.

    (A couple caveats and clarifications: I’m very technically savvy and able to administer workstations and servers. My business was small and focused on creating software and performing a variety of business services relating to software creation. We didn’t need 100% compatibility with an ever-upgraded variety of office productivity apps and the like.)

  7. It’s 2019 and I feel like I shouldn’t have to restate the obvious, but if you want to keep control of your business, you should avoid doing any business with the USA. The example of Huawei is just a reminder.

    1. Huawei is, like most companies of any size in China, effective an arm of the state.

      It was engaging in large scale industrial, corporate and *probably* government espionage.

      Bad example.

      And they always had the option of *not* selling into the US. There’s 6 billion people on the planet, the US is 320 million of them.

      Then again, coming from Europe and complaining about how the USG prevents you from controlling your business, pot kettle, black.

      1. USA companies are forbidden to SELL to Huawei. Huawei loses access to important products it needs to produce products.

        As for espionage, it is the NSA that spied on Huawei and planted backdoor in their products.
        https://ithealthcare.computerworld.com/article/2488962/nsa-hacked-into-servers-at-huawei-headquarters–reports-say.html

        For the rest, the first proof of espionage by Huawei has yet to be published. Meanwhile, there are libraries full of reports and evidence of the USA spying on their friends, including breaking into Nato networks (with thanks to the UK).

        So, all this espionage hubris is propaganda to kill off foreign competitors.

          1. Winter kind of has a point. While the right fulminates about the fate of their beloved Marvel superheroes (who aren’t really theirs; Stan Lee was a cosmopolitan, New York Jewish lefty), the U.S. government is Trump-ing up charges against people it doesn’t like (a.k.a. people who exposed the skeletons in the U.S.’s closet) and putting them in jail.

            It’s kind of like how right-wingers are worried about Antifa committing “violence” against them with Molotov milkshakes — when 100% of the extremist killings in 2018 were committed by members of the far right. Didn’t Jebus, the ultimate right-wing superhero, say something about not pointing out the mote in your brother’s eye before casting out the beam in your own?

            1. “While the right fulminates against [left wing interest]”

              Meanwhile, women are stripped of the right to control their own body in large parts of the USA.

              The Freedom so much talked about by supporters of politicians on the “Right” seems to be only worth defending when these are interesting to men.

              1. >Meanwhile, women are stripped of the right to control their own body in large parts of the USA.

                That’s a tendentious way to put it.

                I think anyone who is not a fanatic on one side of the issue or another recognizes that there is a real question about when the developing fetus becomes something other than part of the mother’s body. Right-wing fanatics (ugh) say it’s at the moment of conception. Left-wing fanatics (ugh) say it’s not until the moment of birth, or (recently) after birth.

                Claiming, either explicitly or implicitly, that the issue is only about women’s control of their own bodies is an attempt to sweep that issue under the rug. I am pro-choice, but I value honesty and integrity more than taking any political side; I think it’s dishonest and disingenuous of the pro-choice side (my nominal allies) to try to limit the framing of the issue in this way, and I must speak out when I see that dishonesty being committed.

                Honesty demands that we face up to the real issue here: when is aborting a fetus taking a human life?

                I have an opinion about this, but my opinion is not relevant to the point I am now making. That is: we should not spread lies or be evasive about the central moral point here. There is a conflict between the right of the mother to autonomy and the right of the unborn child – when it becomes a human being and not merely an biological subsystem that will someday become a human being – to be treated as a fit subject of words like “murder” and “crime”. We should face this conflict squarely.

                1. I’ve been saying for years that the entire debate about abortion is people talking past each other: one side considers only the mother, the other side considers only the baby.

                  So long as the two sides keep talking past each other, nothing will be settled.

                  And this demonstrates the real problem with jamming “solutions” down the people’s throats by court order: they’ll fundamentally have a much harder time accepting it.

                  1. I’ve been saying for years that the entire debate about abortion is people talking past each other: one side considers only the mother, the other side considers only the baby.

                    You’ve been telling yourself a lie. See for example the most unprincipled “pro-life” position, making exceptions for rape or incest. As I recall, even the Catholic Church changed its position to favoring the life of the mother when her condition would likely bring death if the baby was brought to term about the time medicine got to the point where a mechanical abortion wouldn’t also likely be a death sentence to her, late 19th Century wasn’t it?

                    1. The doctrine of double effect has been in Catholic theology since Aquinas. All the ability of women to survive abortion did was make it relevant.

                2. Another piece of disingenuosity is freaking out over the idea of a woman being arrested for having an abortion.

                  It is cast as being some uniquely horrible tyranny, when it is the blindingly obvious conclusion of Abortion == Murder. Under that definition getting an abortion is no different than hiring a hitman.

                  1. I wouldn’t say it’s disingenuous, they just think that “abortion == murder” is a bit of disingenuity cooked up to justify tyranny that’s already been decided upon. In other words, when pro-lifers say “abortion == murder” they hear “abortion = murder”. And given some of the shit men have put women through in history, it’s hard to blame them.

                    More disingenuous are the nominal pro-lifers that say abortion is murder and wouldn’t personally get an abortion, but oppose every single practical step towards abolition, but even in the cases, they have concrete reasons for their position, related to concerns of prior restraint, presumption of innocence, liability avoidance culture in the medical industry, etc.

                    For my part, it’s a non-negotiable that abortion eventually be defined legally as murder, and that fetal personhood be affirmed in the law. But I am also prepared to grant, and even insist upon, fairly broad concessions, such as full and automatic pardon for the mother in any abortion case (punish abortionists only), affirmation of the mother’s guardianship over the fetus and final say in all matters related to the pregnancy short of abortion, explicit affirmation that no person causing the death of a fetus is to be convicted of murder unless intent to kill is established, etc.

                    1. “For my part, it’s a non-negotiable that abortion eventually be defined legally as murder, and that fetal personhood be affirmed in the law.”

                      Your reasoning does illustrate my point neatly:
                      Forcing your personal religious opinion on other people by law is the very opposite of Freedom.

                3. The limit in Alabama is put at 6-8 weeks. That means, 2 weeks after a missed period. By the time a woman suspects she is pregnant, it probably is too late. Especially if she needs to arranged the medical side.

                  In practice, abortion has become banned if the law is uphold.

                  It also holds in ALL cases where the life of the woman is not immediately in danger. Including 11 year old victims of incest or rape.

                  But the freedom of women to decide themselves whether they want to bear a child are not a priority of the right.

                  1. Let’s put it this way: I do not believe that forbidding abortion deprives women of any right that society could not reasonably expect *me* to surrender to save *your* life.

                    1. “that society could not reasonably expect *me* to surrender to save *your* life.”

                      As far as I know, the “Right” in the USA does fight any law that puts any obligation onto you to act to save a stranger’s life.

                      However, the Right wants to strip women of all rights to control what happens to their body if she might become pregnant.

                    2. You keep conspicuously leaving out the part where the Right wants to protect another human life. This point has been brought up enough times, even in this very thread, that I’m convinced you’ve rendered it invisible to yourself, as if with some sort of Beeblebroxian lens. You could address this argument with something at least plausibly valid, but you never do – you just leave it out as if it never occurred to you, and consequently leave the discussion hopelessly stuck in a rut while the rest of us have moved on. I’ll even coin a term for it: you’re simply life-blind.

                      I’m this close to just replying with this every time you bring up another life-blind abortion argument.

                    3. “You keep conspicuously leaving out the part where the Right wants to protect another human life.”

                      Where “human life” is defined by religious dogma to include what is an inseparable part of a woman’s body. The fact that something might develop into a human being in the future, does not make it a human being NOW.

                      So, either come up with rational and objective criteria for defining a human individual, or admit this is religion forced upon those who do not believe.

                      Meanwhile, no law in the US forces a man to spend a year of his life, and risk his health, and life, for another human.

                4. The central moral question here is — do you have a uterus? If not, then sit down and shut up because this is not a moral dilemma you will ever have to face and you have absolutely no dog in this hunt. Your opinion doesn’t matter.

                  The only person in a position to decide whether a mother-to-be has the wherewithal — in terms of health, energy and resources — to go through with the nine months of pregnancy and childbirth, followed by the 18 years of raising the child (because let’s face it, that will ultimately be the mother’s responsibility; fathers can be coerced into making monetary, but not time or energy commitments) is the mother-to-be herself. Anyone arguing against leaving that decision solely up to that woman, throughout the term when her life and health are at risk (i.e., nine months of pregnancy) is arguing for coercion.

                  Those who favor coercing women to go through with childbirth would have more of a leg to stand on (albeit admittedly still not much of one) if they also favored the government paying 100% of that mother’s healthcare and hospital bills throughout the pregnancy and also favored laws that mandated long terms of maternity leave. Unfortunately, that side also tends to call such measures “Eurosocialism”. Conservative hypocrisy strikes again.

                  1. followed by the 18 years of raising the child (because let’s face it, that will ultimately be the mother’s responsibility; fathers can be coerced into making monetary, but not time or energy commitments) is the mother-to-be herself

                    I must say, it’s refreshing in a way for the Left to drop its mask and make it crystal clear they worship Moloch with their full throated advocacy of infanticide (and beyond). Then again, plenty of us remember our 20th Century history.

                    1. >Your opinion doesn’t matter.

                      I hate that tired attempt to shut down a conversation.

                      The “your opinion doesn’t matter” bit is what we used to do to black people because they weren’t white or to women because they weren’t men.

                      Imagine if gun-owners told everyone else that “Your opinion doesn’t matter, you don’t even *own* a gun!”

                      Would that work for you?

                  2. >The central moral question here is — do you have a uterus?

                    No, that’s the dishonesty again. I’m pro-choice, but I have no patience for bullshit arguments, and this is a bullshit argument. If a gun-rights advocate advanced a fallacy, I’d call that out too.

                    If (or when) killing a fetus becomes murder, all the standing I need to object to it is being a human being. A human being who objects to murder being legitimized. Women do not get to give permission for murder just because they have uteruses.

                    I judge the right-wing “moment of conception” position to be wrong. But at least it’s more honest than this sort of special pleading.

                  3. The central moral question here is — do you have a uterus? If not, then sit down and shut up […] you have absolutely no dog in this hunt.

                    Wrong: I was a fœtus once, therefore under a non-broken decision theory, I have a legitimate interest in “what happens to fœtuses”, even though I will never be one again.

                    Like Eric, I’m generally pro choice but find your arguments dishonest.

                    Those who favor coercing women to go through with childbirth would have more of a leg to stand on if they also favored the government paying 100% of that mother’s healthcare

                    Here’s a thought, how about the person responsible for the pregnancy paying the costs of it. Which is either the father (in genuine rape cases; as a tort, the damages clearly include the healthcare, lost wages etc.) or the mother. J. Random Taxpayer didn’t inseminate her, why is he on the hook for the costs? But of course Jeff can’t conceive of costs falling on those responsible, everything has to be paid for “by the government”.

                  4. This from a “man” who supported the rape and murder of Sarah Palin, and the burning down of her church with old women inside. because she didn’t abort her son with Down’s Syndrome.

                5. The abortion issue is fraught because, given the scientific facts, there just isn’t a stable middle ground between the positions you call fanatical. Either the potential of a human embryo to develop into a rational being is sufficient reason to protect it, or it isn’t. If it is, the Catholic position is the most lenient one possible – the potential is there from the start. If it’s not, well, human infants don’t reach rationality until months or years after birth, so it would be just fine to abandon inconvenient babies in a wilderness as the Greeks once did, if they weren’t old enough to speak.

                  Pointing to a specific stage of fetal development and saying “there, that’s when we have to treat this organism as human, but not before then,” if that stage isn’t the appearance of reason itself, prompts the objection that other organisms also reach that stage and can be killed for our convenience, so what makes a human fetus special? I can’t see any answer to that, except that a human fetus eventually becomes rational – which is true at every stage, even the earliest.

                  1. There are a couple of candidates for a middle ground. About 14 days is where twinning can occur, and there’s a go/no-go check where a lot of natural abortions occur. This doesn’t allow C+D, but does allow potentially abortificant methods of birth control (IUD, and hormones).

                    The second is viability, at 20-24 weeks where ex utero survival is possible. Where the fetus has become actually individual in function rather than merely potentially so.

                    1. There’s also first detectable brain activity, which is at about 6 weeks.

                    2. >There’s also first detectable brain activity, which is at about 6 weeks.

                      Chimpanzees have detectable brain activity in a neurology very similar to ours. Thus, I don’t think that’s sufficient. But I do think it’s the right kind of criterion.

                    3. >>The second is viability, at 20-24 weeks where ex utero survival is possible.

                      The problem with viability is that there are some cases where you have two rational minds that are individual in cognitive function, but not in independent bodily viability: namely, significantly conjoined twins.

                      At this point, my general feeling is that a) There is some transition from non-person to person at some point during pregnancy, but exactly where is uncertain.
                      b) Humans have a proven tendency to deny the personhood of others when they feel threatened, or sometimes merely inconvenienced, by the implications of the person in question actually being a person, so it pays to be cautious.
                      3) It is unlikely that the point of personhood precedes the first differentiation of neurons, so this is probably a safe point to draw the line, however, this does, to my understanding, precede the first missed period, so early abortificant birth control is probably OK, but abortion after the pregnancy is manifest is not.
                      4) I am religious, however, and am thus uneasy about the possibility of personhood beginning spiritually before physical structures capable of supporting rational thought exist, but nothing in the Bible jumps to mind that plainly weighs in on this one way or the other, so I’m reluctant to put too much weight on this consideration, though it does incline me to further caution in where to draw the line.

                    4. These ‘what constitutes a full human, with the rights accorded a person’ are dangerous.

                      Because they can and will be applied to more than just fetuses. Euthanasia is becoming very popular in certain parts of the world……

                    5. >Euthanasia is becoming very popular in certain parts of the world……

                      And for people in a persistent vegetative state with sufficient cortical damage, I’m OK with that. What made them human is gone past retrieval. I don’t think the slope to euthanizing people who might recover ia actually that slippery. Except in one case…

                      …which is that euthanasia creep is one reason I’m opposed to socialized medicine. I’ve noticed where the horror stories about this sort of thing emanate from.

                  2. >Either the potential of a human embryo to develop into a rational being is sufficient reason to protect it, or it isn’t.

                    I don’t think it is. I prefer to try to understand when the fetus can actually support human-level mentation, then add some safety margin. I think Roe v. Wade accidentally gets this about right, though it’s a terribly-crafted piece of law in other respects.

                    1. That is a direct path to euthanasia for the any biological human not, for whatever reason, able to support human-level mentation.

                    2. >That is a direct path to euthanasia for the any biological human not, for whatever reason, able to support human-level mentation.

                      And if not just the actuality but the biological potential is gone, I don’t see that as a problem.

                    3. The obvious next question is, where do you draw the line on ‘human level mentation’.

                      Seriously slippery slope. (I have close personal experience with the long-term after effects of a major stroke…..)

                    4. >The obvious next question is, where do you draw the line on ‘human level mentation’.

                      Somewhere pretty conservative, in the sense of putting in healthy bumpers against killing a human.

                      I would say absence of organized neocortical activity is right at the present state of our knowledge. But reserve the right to change my position as we learn more about the neurology of human consciousness.

                  3. The only way I can square the woman’s right to bodily autonomy with the right of the fetus to live is to say that she has the right to have the fetus evicted from her uterus, but that the abortion process be a non-destructive deliberate premature delivery of the fetus. That means no tearing it limb from limb, nor piercing its skull and sucking its brain out to make it easier to deliver its corpse (in cases of anencephaly there is no brain, so the procedure would be allowed). Once removed from the womb, one doctor would tend to the fetus while another would suture any wounds suffered by the woman.

                    The fetus can then demonstrate its viability ex utero. If it can survive without the uterine support system, it has earned the status of “separate human life”, by being separated from its mother and still functioning without her.

                    Hard-core RTL folks will tell you that taking the fetus off uterine support is killing it, but I can’t justify sending Men With Badges And Guns to punish that as “murder”.

                    Hard-core RTC don’t like the idea that, should a woman wait choose to terminate a pregnancy in the 2nd or 3rd trimester, it may be difficult to do so non-destructively without adding some pain and discomfort (episiotomy, caesarean) to her. But because the larger the fetus is, the greater its chances of ex-utero survival, this seems like a reasonable tradeoff.

                    1. So how does this demonstrating viability ex utero intersect with a premature birth which will not survive ex utero that the medical decision maker (the mother) authorizes medical intervention to save it?

                    2. @JohnOC I would expect that a procedure attempting to save a fetus thought to be non-viable is going to refrain from tearing it into pieces or otherwise destroying it in the process. If the clear intent is to save it, it’s pretty much the opposite of the procedures that would be restricted if my thinking were adopted.

                  4. My personal outlook has always been to try to find a reasonable cutoff as to when a potential human (hard to view a fertilized egg as a person) becomes a *person* with rights that need to be protected.

                    It’s becoming more difficult over time. Particularly in light of the ‘pro-choice’ side increasingly openly supporting things like post-birth ‘abortion’ (seriously, look at what the VA Gov said).

                    A fetus, at whatever stage, can plausibly be looked at as a biologically human person who is profoundly disabled. (That is essentially the pro-life position, so help me.)

                    I have a terrible, terrible fear that any line where we determine that ‘hey beyond this point this profoundly disabled biological human can be freely killed without consequence’ will inevitably be applied to more than just the fetus. (Seriously, LOOK at what the VA Gov said. Deliver, then decide. Fuck.)

                    I have been trying to figure out for some time why on earth the progressive chorus is so obsessive about their ‘right’ to kill. The only thing that is making any sense to me is that this is camel’s nose in the tent, leading to the death camps.

                    Because ANY criteria you specify that makes it OK to kill a fetus can, and I fear will, (actually in some contexts like the NHS likely has) be applied… more broadly.

                    1. >I have been trying to figure out for some time why on earth the progressive chorus is so obsessive about their ‘right’ to kill.

                      Yeah, you know, that’s starting to seriously creep me out too. I have principled reasons to remain moderately pro-choice, but I do not want to be associated with people who hold up signs saying “MY ABORTION WAS FABULOUS” or advocate it after birth for any reason.

                    2. >A fetus, at whatever stage, can plausibly be looked at as a biologically human person who is profoundly disabled. (That is essentially the pro-life position, so help me.)

                      That is *exactly* the pro-life position. It’s a person, and its deficiencies in cognitive power or viability with respect to an adult human do not change that.

                      Depending on metaphysical assumptions, pro-lifers may or may not hold that a fertilized egg is a person, but will generally agree that personhood begins early enough that there are no circumstances under which abortion can be guaranteed not to kill a person.

                6. How do you reconcile the debate on abortion law here and the content of your text?
                  In your text, you explain that a private company could not use moral decisions to stop your activities (I guess they can decide to not work with you as long as you can take your business to another partner).
                  But on this discussion, it looks like you agree that there should be some moral rules that apply on society members despite what they personally want to do.

                  I agree on the discussion about saas being able to capture, but my comment targets in particular this quote:
                  “Otherwise, how do you know some political fanatic isn’t going to decide your product is unclean and chop you off at the knees? It’s rifles today, it’ll be anything that can be tagged “hateful” tomorrow – and you won’t be at the table when the victim-studies majors are defining “hate”. Even if you think you’re their ally, you can’t count on escaping the next turn of the purity spiral.”
                  Why this segment does not apply when it comes to abortion (and it’s not about being pro-life or pro-choice, it’s about agreeing that it’s legitimate that a law on the subject could even exist)? Why deciding that, for example, imposing gun distribution restriction opinions are the result of extremists and that deciding what a pregnant woman can do is not (again: it does not matter if this decision is pro-life or pro-choice, the question is about the legitimacy of a law applying to all citizen)? I guess the answer is “it’s not the same, because I think the second is reasonable but the first is not”, but this is just a question of ideology.

                  1. >But on this discussion, it looks like you agree that there should be some moral rules that apply on society members despite what they personally want to do.

                    Well, yes. But all those rules need to derive from prevention of initiation of force or fraud, and no other ground of “everybody” rules is proper. What Salesforce is doing is not preventing force or fraud. it’s interfering in mutually voluntary interactions. That’s where I draw the line.

                    Now, you can call this “ideology” if you want. But it’s not possible to avoid “ideology” in the sense unless you’re willing to lives in a society that has no rules against murder, rape, and slavery. And my version is tuned to result in the minimum of avoidable harm.

                    The question about abortion is whether it’s like murder (thus, a fit subject for universal rules) or not. There aren’t many behaviors for which this is a difficult decision, abortion is one of the thorniest.

                    1. Thanks for the answer.
                      I agree that ideology will always somehow be there. It’s not a problem for me.
                      But it implies (and this is not an attack, just a constatation) that people who set the line slightly differently are no more or less legitimate.
                      It also implies that the following paragraph is as legitimate as yours that I have quoted:
                      “Otherwise, how do you know some political fanatic isn’t going to decide your product is unclean and chop you off at the knees? It’s women bodies today, it’ll be anything that can be tagged “part of the sjw propaganda” tomorrow – and you won’t be at the table when the victim-studies redditers are defining “sjw”. Even if you think you’re their ally, you can’t count on escaping the next turn of the purity spiral.”

                      But again, it does not change the main message of the text: saas can capture.

                7. It is called a fetus from week 10 (8-9 weeks after conception). Brain activity starts around week12-13.

                  Movements become “complex” around the end of the first trimester. Which is not coincidentally the limit for abortion in most European countries.

                  In practice, using the first trimester as a limit gives women enough time to decide on whether or not to carry on and also precludes any chance of aborting a thinking being.

                  I also wonder why people who object to a law that obligates men to act to save a fellow human being (good samaritan laws), even if its picking up a baby drowning in a puddle at their feet, are willing to force women to risk live, health, and future to bear a child.

                8. I believe that life beings at conception, and that killing that life is murder. I am a Catholic. I also believe that one can be an atheist Libertarian and hold that same view.

                  But… that does not mean I think a democratic society should make early abortion murder. The laws and the morality need not necessarily be perfectly congruent. The law should be a result of a democratic process, while morality is not.

                  But,if we are to err in our determination of when the unborn becomes a human life, how about erring on the side of the right to life?

                  Also, what happened to the rights of the father? That is almost never mentioned.

                  1. @John Moore
                    “I am a Catholic.”

                    Religion is a private matter. When religious rules become law, that is the opposite of freedom.

                    Btw, life does most certainly not equate human. Algae are alive too. So, the question is not whether life starts at conception (life really started long before, an egg is alive, and so is a sperm), but when a living thing becomes a human.

                    “Also, what happened to the rights of the father?”

                    If it was his body and his risks of giving birth, we would welcome him to take them on. Until then, it is the woman’s body, her health and her risks.

              2. Meanwhile, women are stripped of the right to control their own body in large parts of the USA.

                You seem to have misspelled “EU” at the end there, champ.

                Sure, a bunch of US states have recently passed no-abortion-after-six-week laws that were immediately blocked and are not in force.

                On the other hand, of the 28 EU states, 1 bans abortion outright, 1 severely restricts it, and 23 only allow it in the first trimester (except under stringent conditions). That leaves only 3 EU states where access to abortion approximates the actual legality in the US.

                Oh, and of the mere three EU states with American-style legality of abortion, the only large one is currently in the process of leaving the EU. If and when that process completes, 94% of the 27-member EU’s population will live in states where abortion is far more restricted by law than it is in the US.

                1. “Sure, a bunch of US states have recently passed no-abortion-after-six-week laws that were immediately blocked and are not in force.”

                  The point is that the “Right” is fighting to get the ban in force, with a good chance of winning, thanks to the current Toddler in Chief.

                  What counts is that the “Right” has installed laws to strip women of the right to get an abortion. And a limit at the 6th week is before many women even find out they are, unexpectedly, pregnant.

                  And even if a woman finds out in time, trying to find a willing doctor not yet sued into bancruptcy for performing abortions in time will be “difficult”.

                  Fyi, Ireland is busy overturning the ban on abortion. For the rest, this is not about the precise limit, but on being able to get an abortion after you find out you are pregnant.

                  1. Fyi, Ireland is busy overturning the ban on abortion.

                    Um, no. Ireland has already legalized abortion, and was one of the 23 first-trimester-only states I mentioned.

                    Your patent ignorance of the actual conditions in the US (like claiming there’s a good chance the “Right” will win on six-week abortion ban bills because of Trump) on its own is understandable. As a foreigner it’s entirely understandable that you don’t know enough about American conditions to reliably distinguish truth from propaganda.

                    But combined with your just-demonstrated ignorance of the actual conditions in the European Union, a polity of which you are a citizen, it makes it clear that you are not actually interested in abortion rights; you simply are an anti-American bigot who seeks out propaganda to regurgitate.

                    But, please, go ahead and embarrass yourself further by continuing to display your ignorance.

                    1. I was indeed wrong. As of 1 Jan 2019, abortion is legal in Ireland. I did not update my info since the outcome of the referendum in May 2018.

                      The Irish were faster than I expected.

                    2. PS

                      The point was not whether the Right actually will succeed in stripping women from having control over their bodies. The point was that the Right is fighting a decades long war to strip women of control over their bodies. AND that the Right claims to fight for freedom. Obviously, the fight is not for the freedom of women.

                    3. “Obviously, the fight is not for the freedom of women.”

                      Is the American Right really out to destroy the freedom of women, though?

                      First off, I have seen a good case be made (by a woman that used to be pro-choice, at that) that legalized abortion actually hurts a woman’s right to keep the baby: an inconveniently pregnant woman is surrounded by people pressuring her to have an abortion.

                      Second, you completely gloss over the issue at hand, where we are talking about the fate of a separate human — the fetus — and it’s an issues fraught with thorns. In this one little sentence, you make it seem like this isn’t a concern of anyone at all, when it’s the driving force of people who are pro-life.

                      Third, where else is the Right trying to limit the freedom of a woman to do as she pleases, but doesn’t restrict men in the same way? I think you would be hard-pressed to find an issue.

                      It is precisely this kind of disingenuous debate that Eric and others were complaining about — both sides do it, but both sides hate it as well.

                    4. “Is the American Right really out to destroy the freedom of women, though?”

                      If you do not have the final say over your body, but are forced to endure what others tell you to do with it, what worth does the word “freedom” have?

                      And denying women rights to protect them against coercion is an extremely devious argument.

                      “the fate of a separate human — the fetus — and it’s an issues fraught with thorns. ”

                      The embryo nor the fetus are separate. They are an integral, inseparable part of the woman’s body. And it is on her that all the risks and costs fall.

                      “Third, where else is the Right trying to limit the freedom of a woman to do as she pleases, but doesn’t restrict men in the same way? ”

                      What more fundamental right is there than the right to decide over your own body. It affects everything you do, from your health to your professional career.

                      In some jurisdictions, pregnant women are also legally restricted in their behavior if it might affect the health of the fetus, or child after birth.

      2. They would still be in business if they had an adequate alternative to google playstore, google maps, and google mail.

        Google mail is a treasure trove for spies. We know they go through all your emails looking for sale receipts, and use that data for targeted advertising.

        We know they ratted out general Petraeus, giving Obama the the metadata from his emails calls. But that particular metadata probably would not have warranted attention unless they also looked at the content of his emails.

        What else are they doing?

        Viber now permanags you to upload all your viber messages in the clear to google, where they will presumably get the same treatment as General Petraeus’ emails and every business receipt that gets mailed to your google email account.

    2. Don’t even land in the USA. If you have to fly across North America, route your layovers through Canada. Unless you like being harassed and having your stuff seized and all your data rifled through by U.S. customs…

    1. >WikiLeaks were the real heroes of free speech, and the USA does everything in their power to shut them down.

      Are you supposing that anyone here thinks BoA’s action isn’t wrong?

      1. The silence about that issue, and the moaning when a company takes action against gun sellers or racists, tells a different story.

        1. There are enough people already squawking about WikiLeaks that our voices aren’t needed in that cacophony. Whereas if we don’t defend the gun-sellers, no-one will.

          Price is the intersection of supply and demand.

      2. It was definitely wrong, and I’ve seen other wrong actions by credit-card companies of the same type/scale.

        The issue here is that all decent people want to maximize freedom, while all decent companies want to maximize profits…

        1. while all decent companies want to maximize profits

          I wish we could go back to that. Nowadays companies are falling over themselves to see who can kill their profits faster in order to achieve some alleged moral good.

    2. Much as I appreciate Wikileaks’ release of information about the venality of the Clinton campaign, I can hardly consider them to be heroic. Wikileaks took information that was properly classified, that protected the lives of American soldiers, and of our allies in dangerous places, and released it with no regard for those lives. Assange admitted that it had blood on its hands, and was okay with that. In that, Wikileaks knowingly abetted not only enemies of the US, but terrorist enemies of freedom and western civilization.

      Wikileaks solicited that information, and it aided a traitor in getting it.

      Those are facts we know. There are also credible reports that Wikileaks was operating in conjunction with Russian intelligence to discredit the US.

      Wikileaks is a vile organization, operating against freedom.

      1. “Wikileaks took information that was properly classified, that protected the lives of American soldiers, and of our allies in dangerous places, and released it with no regard for those lives.”

        Wikileaks took evidence of war crimes committed by US soldiers and released them. Whether information about war crimes is considered “classified” is immaterial.

    1. This is a very intelligent bit of thinking. I’m going to have to consider the implications a little.

    2. I think that something is, pushing to make the standard of enforcement disciplinary action upon suspicion of unsavory behavior, not evidence of it. Evidence being, technically, only necessary when the government intends to punish you.

      This reasoning ties somewhat loosely into how I have been describing to people why I think we are living in the cyberpunk distopia of ’80s s sci-fi: “My war with the nation of Facebook proceeds well, and has recently established critical successes on numerous fronts.”

      They laugh, of course—part of my goal here is to “gently” introduce a very, very wonkish topic by being funny, after all—but I then point out: what exactly defines a “nation”? If it is legislative power, Facebook has that (we call it “terms of service” rather than law, though). Perhaps it is judicial authority, but Facebook has some of that too–including an entire internal appeals process. Maybe it is some form of border control–but again, Facebook can exert various forms of that (both in terms of denying service to undesirable users, and if their algorithms might prioritize on-site vs. externally linked content). Really, the only thing that most people could point to that “real” national governments have that Facebook lacks might be an army … but then again, how do most people describe how many moderators work there?

      All possible joking aside, I do honestly think some corporations, (such as Google, Facebook, Salesforce, ADP, Visa, etc.) whose potential to influence or control our lives mirrors that of “real” governments, should be viewed as governments when evaluating the ethical, social, or legal ramifications of their actions.

  8. If the software is a communication program, its utility scales as the square of the number of users. So running your own private instance of e.g. the Signal server is not nearly as valuable as accessing the SaaS one the default build uses, even though it’s open source. The corresponding dictum in this domain is to prefer federated protocols! If Moxie Marlinspike decides you aren’t welcome on his server, you’re out of luck; if your email provider tries to boot you, it’s easy to take your business to a competitor. A non-federated protocol benefits the developer as he doesn’t have to worry about maintaining compatibility with instances running outdated versions, but a federated protocol benefits the user.

    1. Yeah, yeah. Federated protocols work nice until network effects kick in, after which “Find us at AOL keyword X” — or its modern equivalent, “like us on Facebook” — becomes much easier for the masses to deal with. This process becomes more rapid when you account for spam, which tends to consolidate people on the few major nodes/sites/servers equipped to deal with high-volume spam, hastening their transformation into an oligarchy. You cannot set up an email server capable of routing email to randoms on the internet, not without investing the time, effort, and money into filling out the right forms and paying the right people to whitelist you. So most people just get a Gmail — or Facebook Messenger — and be done with it.

      “Prefer federated protocols” belongs in the same bin as Postel’s law and the Unix philosophy: a rule of thumb shaped more by hacker wishful thinking than reality.

      1. >“Prefer federated protocols” belongs in the same bin as Postel’s law and the Unix philosophy: a rule of thumb shaped more by hacker wishful thinking than reality.

        It’s that or be owned. Controlled. Again, Salesforce.com’s move will be doing us a favor if it teaches that brutal reality to people who aren’t hackers.

        1. Major businesses are being hurt – it is not just woke capital cracking down on Chick-a-fil

          Every business that uses Google Analytics gets burned, Huawei got burned.

          If someone else silos your business data, they own your business.

          1. It’s like letting someone own your brain. Why would any intelligent person allow that?

        2. There does seem to be an ebb and flow between people federating their protocols to avoid being co-opted, and then sinking into a comfortable proprietary silo to get Good Stuff without having to think too much.

          It may be coupled with people not thinking of a proprietary protocol as valuable until it is. Everything probably saw Facebook as a way to share fun stuff until it occurred to enough people to use it to share stuff for money. (Salesforce, not so much – FAICT, people should’ve seen that one coming.)

          It’d be nice to have this cycle self-moderate, but that would appear to require discipline in far too many different domains.

      2. Federated protocols work nice until network effects kick in, after which [submitting to some 800-lb gorilla] becomes much easier for the masses to deal with.

        Yeahbut –

        The nice part about federation is – you don’t have to see anything you don’t want to. Just put EeebilUnPerson#N on your block list, and they vanish into the void. If I want to interact with them, I can. And I can be seeing them from a no-holds-barred server, and you can see (or not see) them from a completely-holier-than-thou, fully SJW-converged Totally Woke server. As long as the servers are federated, only local rules/norms must apply, but we can agree to disagree and still pass around pointers to each others’ traffic. (You don’t even have to “risk” hosting anything you might think is “toxic”, as long as you are willing to at least hold your nose and point over there at “The Cesspool”.)

        “Prefer federated protocols” will become the norm when a few things happen – [1] it becomes trivial to set a server up, so that there will be a plethora of them, and it won’t matter which one you’re on, [2] some set of companies realize they can make (a rather small, but non-zero) amount of money providing such a service (including “spam” filtering and kicking off “undesirables”, however they want to define either), and [3] “most people” (the non-techie ones who would just “get a Gmail – or Facebook Messenger”) are taught that if they don’t like their current provider, they can migrate trivially to one more to their liking.
        Out of the abundance of the ones that will exist.
        Without losing any of their information.

        Win-win!

        (HINT: look at the history of email server deployment….)

        1. The problem here is that the federating authority can and will enforce Totally Woke rules. See Mastodon, which says up front that they will not federate with anyone who does not actively moderate against racism, sexism, and homophobia.

          1. > See Mastodon, which says up front that they will not federate with anyone who does not actively moderate against racism, sexism, and homophobia.

            You mean the specific server mastodon.social? *Shrugs* They’re not the only player in town.

            If they extend that lack of federation transitively, they’ll end up with their own (small?) Woke Ghetto, and the rest of us will be interacting elsewhere just fine.

            Yeah, this ghettoization just reinforces the “echo chamber” effect. That’s a problem with people’s psycho-sociology that I don’t know how to solve (and that probably can’t be solved by technical means).

            1. That’s it exactly: their terms of federation include explicitly making it transitive. They’re making a Woke Ghetto, all right, and we need to stop pushing them as an example to emulate.

              1. I don’t think this is correct – I have an account on an instance that specifically refuses to block any other, and the flagship instance doesn’t block the one I use. There do exist instances that block the one I’m on for not adhering to their criteria for acceptable speech, but they don’t include mastodon.social. The fairly minimal anti-harassment policy in place (if you’re a pest on one account, anybody who doesn’t like it is expected to mute/block you; if you create multiple accounts to evade that, the admins will ban you properly) seems to be a happy middle ground between getting a good ideological cross-section and avoiding the kinds of trolls that make other instances want to ban you.

                There’s just not much point in making those blocks transitive; posts don’t get mirrored through third-party instances, which means that the neutral instances don’t really pose a threat to the “safety” (in the sense of “safe space”) of their federated peers. Given this, the blocklist used by the flagship instance seems to be pretty good at distinguishing between actual troll habitats and mere ideological divergence. The former get blocked and the latter are tolerated, if not exactly welcomed.

        2. [3] “most people” are taught that if they don’t like their current provider, they can migrate trivially to one more to their liking.

          I think the key word here is “trivial”. Gmail and facebook have won so far because they don’t make the user do much of anything to get set up. Software as a service in general wins because it doesn’t require the user (or sometimes the admin; we’re not immune) to install or configure anything. Web forums and blogs won out over Usenet in part for the same reason.

          When AIM died, I set up my own XMPP server for the folks I still talked to on that service. It works, but even so I lost a couple contacts in the transition — to the inconvenience of creating an account, or installing a client, or in one case to the inconvenience of starting a client that had already been installed.

          Network effects and product differentiation, as with Facebook or Twitter, compound the problem. Facebook could open-source their software tomorrow for anyone to serve and it wouldn’t improve matters much. Switching would be mildly inconvenient. Figuring out how to use not-Facebook would be more so, and finding alternate ways to keep in contact with people on the old service would be worst of all. As for convincing others to follow you….good luck. The most insidious form of vendor lock-in is simple user habit.

          Never bet against convenience. When the total inconvenience of dealing with one’s comms provider outweighs the total inconvenience of switching services, federated protocols will have a chance. When the total inconvenience of dealing with one’s SaaS provider outweighs the total inconvenience of running something local for the same purpose, local software will likewise have a chance.

          1. This leads us into the Standard Monopoly Problem: as soon as a monopoly starts being overtly nasty it is creating opportunity for competitors to undermine it with every action it takes.

            In the grimdark of $CURRENT_YEAR that means banning as many unwoke people from your platform as possible, which is far more inconvenient for them than the learning curve of a new system.

            1. I for one prefer the option that involves making non-SaaS more convenient.

              Local software is always going to be at a disadvantage here, though. It’s easy to be better than web applications, because they pretty much all suck, but the main value of SaaS isn’t that the software is *good*, it’s that you can use it from anywhere without installation or admin rights.

              I don’t know how to fix this.

          2. Tangentially, an observation: Email still works perfectly well, it’s probably even more convenient than Facebook et al (at least with providers like gmail), and everybody has an address so the network effects should be stronger. Yet social media is still the dominant form of communication. That should be surprising, because the network effects of email should be stronger.

            Why, then, does so much seem to revolve around the state of SM?

            (my hypothesis, without much evidence because I’m still an SM refusenik: There is no communication on social media. What looks like communication is actually a facade for tribe-bonding and image-crafting, and it’s important to the speakers that it be public. To neurotypicals of any stripe, what you say is less important than being seen to say it. Emailing someone is communication, facebooking them is performance. You can’t do the performance by email. Hence the real open-protocol competitor for social media was usenet, and it lost the convenience war. People have their actual meaningful conversations out of sight.)

            1. The whole point of virtue-signaling is to signal your virtue to as many people as possible. Thus, it must be done where the largest audience is available.

              This is why SJWs MVST control social media. They need their public platitude professions, like a crack addict needs that rock.

            2. Federated protocols are preferable, but that preference is subject to being outweighed by better features. Facebook has a lot of advantages that more than make up for its lack of federation in the evaluation of its users.

              Email is sort of a modernisation of writing actual letters to one another, but letter-writing isn’t the only form of communication throughout history, and it makes sense that people prefer platforms that don’t constrain them to that particular model. Facebook, for instance, gives users a default monolithic chat with each of their contacts, like how a cell phone handles text messages. This simple form of communication without subject lines, complicated in-reply-to trees, signature disclaimers, etc., fits with their positioning as a less formal medium than email. Further, the ability to publicly post content without sending it to anyone in particular, and making the history of those posts accessible to anyone who cares to look, is not really achievable by email and more akin to a blog.

              Sending media is another area where Facebook beats the pants off many other means of communication. Instead of the letter metaphor where you can include some non-letter object in the envelope with your letter as an attachment, Facebook treats common content types as fully equal to text, allowing the user to view them without opening a separate program, generating thumbnails and website previews, etc.

              Group communication is also improved relative to email. Whereas email “group” membership is mostly achieved by adding or removing people from a thread on a per-message basis, Facebook allows newly added users to review post history before they joined, and a user can unilaterally leave a group (ever been on the receiving end of a several-thousand-person email where people start replying-all?). Further, the privileges to post, moderate others, add/remove users, etc., can be finely controlled in a way that would require setting up a listserv to accomplish via email – a prospect not really practical for casual users.

              So Facebook combines many of the advantages of email, SMS/MMS messages, web forums, and blog posts. Now, you could run many different services that each fill one of those niches and fill it well, but integrating them lets your users link their pictures in public posts, share posts in groups, refer to groups from private messages, and so on – and they don’t need to find each other on each service on which they want to communicate. The ability to see one’s contacts’ contacts and the use of real names rather than addresses are both advantages to discoverability that Facebook enjoys over other communication media. If a user sets his profile picture to a recognisable image of himself, then his full name is sufficient information to contact him – and if, as is often the case, the person attempting to find him knows who their mutual acquaintances are, it becomes that much eaiser.

              Finally, I’ll remark that a limited form of federation does exist in Facebook, in that messages can include recipients not registered on the platform. A gateway service translates the message to email or MMS and masquerades in the native representation of the group, posting any messages received on the foreign interface. From a user point of view, Facebook beats email on features, and beats federated social media things like Friendica, diaspora*, etc., on n² scaling largely by being first to the market. All that’s left is for Facebook to federate with those others ;)

            3. > Email still works perfectly well

              Email is thoroughly broken, and I quit using it for anything important some years ago.

              Every single hop now has to negotiate someone’s spam filters, admins who block entire class A networks, and blocklists, blacklists, whitelists, and who-knows-what now.

              Mail I sent seldom made it to its destination; mailing lists I had been on for a dozen years, I was getting maybe one message in three. I changed mail hosts several times, battling the problem as it got continually worse, and then gave up.

              I have a couple of throwaway freemail accounts I use for places that demand an email address; I don’t care if they work or not. I opened a Gmail account and a Yahoo account to be able to exchange messages reliably with a few people on their systems. And… after a quarter of a century as a thousand-message-a-day email user, I quit.

              Spam, I could deal with. But the anti-spam admins have broken the underlying transport mechanism so thoroughly, the ultimate effect is to make mail too unreliable for anything important. Near the end I didn’t get much spam, because I didn’t get much of *anything*…

              1. >I changed mail hosts several times, battling the problem as it got continually worse, and then gave up.

                I self-host and do not see these problems.

                1. How do you handle spam control? I’ve considered self-hosting for my own mail, but keep concluding that I don’t want to burn limited life-hours keeping up with the spam war.

                  1. >How do you handle spam control?

                    Graylist and fail2ban. I used to use Bayesian filters, but training them was too much of a hassle.

                    My spam receipt rate isn’t zero, but it’s acceptably low.

      3. I’m sorry but having to fill out forms and get whitelisted to run your own email server is just not reflective of reality. If you setup your email server up to snuff with regards to SPF and other standards, your email is perfectly fine.

    2. I’m missing something here. Where is the network effect in Salesforce.com’s service? Isn’t their business e-commerce Web sites for small retailers? Retailers don’t need to talk to each other; their need is to find customers, and that involves dealing with search engines and social networks. Google and Facebook have powerful network effects working in their favor; but Salesforce is at best piggybacking on them, with their API layered on top of those companies’ sites.

      1. Salesforce’s business is customer relations management. It has the ability to literally manage the entire sales process from lead generation through quotes all the way to closing the sale and supporting the customer afterwards. My employer used to use it, and it does no commerce over the Web.

          1. It’s not matter of network effect. Salesforce is simply a huge software suite, well integrated, and something a business will latch on to and make heavy use of. Once you’ve committed to it, extricating yourself is difficult – especially since there are effectively no competitors, certainly not in the open source world.

            1. This illustrates another problem I often see in OS critiques. People complain that OS keeps playing catch-up with CS. One motivator is clear: companies make money by selling feature-rich software. OS makes nada from coding such features (unless they have a rich altruistic patron). Another apparent motivator is that the problems that attract CS companies to sell solutions are often problems that OS stalwarts simply aren’t as interested in solving. They like writing emailers and browsers and chat apps and website templates and text adventure games. They’re nowhere near as interested in coding up customer relations solutions or restaurant management suites.

              1. I would actually like to create a point-of-sale device, both because I want something open-source in this space, and because I want to experiment with accepting multiple currencies (in particular, American dollars, gold and silver).

                To do this, though, I’d have to get myself my own register or two, (which I’d imagine is *somewhat* expensive, but not terribly so, especially if I found something used), and I’d have to have the time to work on it.

                Aye, there’s the rub! There’s only so much time in the day, and this isn’t a high enough priority to pursue it, even as a hobby.

                For better and for worse, a *very* good substitution for time is getting paid. Hence, SalesForce.

                Come to think of it, it isn’t just SalesForce that’s playing catch-up. If I wanted to create a SalesForce-like system from the ground up, it will, like any open-source equivalent, necessarily have to start out with something that simply doesn’t have the features that SalesForce has. The only way out of this problem is to find a feature that SalesForce doesn’t has, and isn’t likely to add in the future, and then build from there to something similar to what SalesForce has….

                It isn’t necessarily hard, but it’s definitely non-trivial.

            2. It would help somewhat if SalesForce made it easy to export their data. I have no experience with SalesForce myself, though, so I have no idea whether this is even an option.

              SalesForce certainly doesn’t have any incentive to provide this, *particularly* if they don’t have any competitors. Even then, they’ll likely want to make it easy to import, but difficult to export….but then, counter-intuitively, making it difficult to export could prevent people from wanting to adopt in the first place, because sometimes you have forward-thinking customers who don’t want their data to be trapped in one location….so, who knows?

              Of course, being able to export your data is futile if you don’t have an easy way to use that data, once it’s been exported….

    3. If the software is a communication program, its utility scales as the square of the number of users.

      Note that this is disputed; some claim the true scaling is n log n (as n grows, the probability that the marginal user has anything interesting (and new) to say falls).

    1. Oh, fuck those people.

      I mean, the lawsuit, not Salesforce.

      (Sure, fuck them too, but not for this.)

      “You let Backpage use your software, and people used Backpage to do sex work [TBF, that was its purpose], and SEX WORK IS HUMAN TRAFFICKING.”

      Bullshit on that, and the “woke” should know better, even by their OWN standards.

      1. I’m willing to give the lawsuit a bit of leeway, though, considering that allegedly it’s Jane Does who are initiating the lawsuit. If they genuinely were being forced into work that they didn’t want to do, then I think they have a good case that they were being trafficked.

        Whether or not SalesForce should be held into account for that is a different issue. One I’m not willing to try to resolve at the moment.

  9. Unfortunately, most companies aren’t going this way. And though open source libraries and frameworks are thriving in corporate America, we’re seeing the death of federated protocols. Centralized proprietary services have easier discovery. Using SaaS means having to hire smaller IT staff to manage everything. And frankly, there aren’t enough skilled UNIX sysadmins to go around, and they’re expensive. Or developers for that matter.

    I agree with you, though, that owning the infrastructure and running entirely open-source software is the best long term solution overall, but corporations aren’t thinking about the long term. The big talk now is IaaS — infrastructure as a service. At least when done well, that’s more of a commodity play; though AWS and Azure are sexy right now, there’s many, many providers of cloud servers on which you can use Kubernetes or Terraform or whatever to build our your virtual infrastructure. But then, the chief value is delivered by software, not hardware. And that’s what makes renting your software so dangerous.

  10. And I forgot to mention the evil of google analytics. Seems that every business runs google analytics on their website and fails to wonder how google is so accurate at directing their major competitor’s ads to their customers.

  11. “And you can’t afford it to be tethered to a service provider even if the software itself is nominally open source.”

    Yes. This is why I won’t touch Go/Golang with a ten-foot pole. It’s a Google language, by Googlers, for Google purposes. Its technical merits are irrelevant, since basing something on Go is living in the shade of a giant who might step on you at any moment – a risky proposition at best.

    1. What part of “software as a service” did you not understand?

      Go is not that, so what you are quoting does not apply.

      1. Isn’t Team Golang at Google wiring a bunch of Google web services into the basic use of the language?

        And Teddy’s point about Go being analogous to SaaS still stands, right now Go is still tightly under the control of Team Goland (generics, anyone??) and thus Google, without the potential for a practical fork at the moment. They also registered various Go marks.

        After walking away from Clojure for the above reason, Oracle’s bloody-mindedness about Jave/the JVM, and worse governance, I also am avoiding Go after seriously considering over the last few weeks. I don’t disagree with our host’s conclusion that it’s a great intermediate systems language, but I don’t find it suitable for my personal use.

        1. >Isn’t Team Golang at Google wiring a bunch of Google web services into the basic use of the language?

          Not that Ian or I have noticed. Can you be more specific?

          1. It was asserted in one or more of the recent Hacker News discussions of Golang governance, but not in the context of continuous integration and in general the resources Google is devoting to the development of Go itself, vs. programmers using Go. I didn’t verify it because it wasn’t necessary to make my decision.

            1. Hacker News isn’t really a reliable source.

              Golang is pretty neutral, even though the standard library does somewhat reflect a bias towards Google’s own needs.

              1. Go is opinionated AF. It’s of the opinion that all of the great work that’s been done in type theory and static lifetime management over the past 3 decades or so is unnecessary cruft and we should just pretend it’s 1975 and none of that stuff ever existed.

                This makes it great for writing very small, simple programs in. It lets bootcamp kiddies write microservices with a bit more zip than JavaScript can provide. However, should you wish to build a large, complex system in Go, you will probably drown in all the boilerplate you have to write, thanks to the Go team’s total punting on parametric polymorphism because “muh compilation speeds”.

                I don’t know what that has to do with being SaaS, but Go is clearly aimed at a specific use case of Google’s (allowing junior-tier programmers to rapidly develop small network applications) and undergoes serious metaphor shear when you try to put it to heavier duty.

                1. Go is opinionated for sure, this cannot be doubted. Perl (both the 5 and 6 variants) might be one of the few languages that isn’t opinionated.

                  It is also irrelevant to the discussion. The assertion was made that Go specifically promotes Google’s own services and there isn’t evidence that it does; we haven’t found it, at least. Please provide some if you can find it.

                2. There’s equivocation going on here. Golang is very opinionated in the software design sense, yes. However, it has no callouts to any Google services built into it, or given any particularly special privileges, so in that sense it is as free of opinons as Python or Ruby or something.

      2. Go is controlled by evil people who hate you.

        You write software in Go, your software is controlled by evil people who hate you.

        Example: Go’s architecture makes it very easy to support gui, since gui is event oriented and Go is event oriented. It would be very easy to integrate wxWidgets with Go, so that Go supported easy to write, highly portable, gui programs. But it has not happened. Go has no gui support because they want you to use the browser and the server as your gui.

        Go is a great language. I will not touch it with a ten foot pole.

        1. >Go is controlled by evil people who hate you.

          Well, first, there’s a Go compiler maintained by people who aren’t Google, so I’m not worried.

          Second…I’m pretty sure Ken Thompson and Rob Pike don’t hate me. Ken sends friendly answers to my mail. I doubt he’d be hostile to my apprentice.

          1. > Well, first, here’s a Go compiler maintained by people who aren’t Google, so I’m not worried.

            I think you’re missing a hyperlink.

              1. Except that is maintained by Googlers–more specifically, by Ian Lance Taylor, who happens to be none too fond of our host.

                1. >Except that is maintained by Googlers–more specifically, by Ian Lance Taylor, who happens to be none too fond of our host.

                  Hm. I didn’t know that. It changes my risk assessment some.

                  BTW, if Ian dislikes me he’s does a pretty good job of hiding it during our go-nuts interactions.

            1. I think you dropped the “t” in “there” and turned it into “here”. It’s really bad form to preface such with “>”, pretending you’re quoting someone verbatim.

          1. There are lots of folks on car forums who swear by it, so I’m going to assume it does. I have a couple of jobs for it on my 1987 560SL that I just haven’t gotten around to yet.

            But I was using that in a metaphorical way, in the sense of metaphorically getting rid of the language and its zealots (apparently there’s something about Rust that turns its advocates into Stallmanite-class zealots).

      1. Rust is great.

        But it faces a design problem, in that figuring out what code is in fact safe is equivalent to solving the halting problem, so with great regularity one runs into situations where the compiler forbids you to do the perfectly safe action that you intended to do.

        Making a safe language is hard. And it needs to be done in way such that the compiler’s reasoning about safety is transparent to the programmer, so that he can know what the compiler will accept as safe. The rust compiler’s reasoning about safety is sometimes opaque.

        1. >The rust compiler’s reasoning about safety is sometimes opaque.

          TIL I learned that Jim is in fact capable of humor by massive understatement.

        2. Making a safe language is hard.

          And don’t I ruddy well know it! Working on the eBPF verifier has been a very interesting experience (particularly the variety of approaches to the question of loops).

          OTOH, we’re now getting to the point where people are musing seriously about writing device drivers in eBPF (if IR protocol decoders count, then more than just musing), which has the potential to be a Really Big Deal.

        3. People only tend to get flummoxed by the Rust compiler’s reasoning if they see things from a C perspective — memory as a long string of bytes, and object references as raw pointers into that byte array. In order to grok Rust, the C perspective must be abandoned entirely. Learn to view objects as things with definite lifetimes, references to which are owned and must be borrowed. Make the borrow checker your friend, and Rust becomes an absolute joy to use. Complex programs can be written that have the performance and memory-usage characteristics of C, with the confidence that comes from not having to worry about the problems that plague all but the most trivial of C programs. I’ve seen it happen over and over.

          The people most able to do this, it seems, are the Ruby and JavaScript kiddies who’ve had no prior exposure to C, and thus no real experience with C’s understanding of memory, the CPU, and concurrency (which old-time hackers will tell you is the way computers “really work”, even though this too is a lie). These are the people who’ll be writing the next major OS kernel and other hardware-platform-level software. (Yes, Jay, even on microcontrollers — Rust can target anything LLVM can.)

  12. Eric S. Raymond: Is this opensource game worth a damn ( http://www.moddb.com/games/chaosesqueanthology )? I want your thoughts since you are a NetHack programmer. It’s… not a software as a service… but didn’t know where to ask you. Lots of people say this game sucks and they hate how it looks etc. Author just responds : 200+ weapons, pnd snd commercial wagies: As if making a living from programming was “less noble” than doing it for free for no reason as a hobby.

      1. Looks like MikeeUSA recognizes you as an alpha, and is looking to you for validation, so he can prove to himself that he is something besides a miserable little troll.

        Hey Mikee, don’t you have legal clients to attend to? Maybe a buddy of yours from bible study who got charged with statutory rape, something like that?

        1. >Hey Mikee, don’t you have legal clients to attend to?

          Not at the moment. I like to program opensource projects more. Maybe when the weather gets worse again: that seems to be how it goes. Pro-bono of-course. I kinda dread the email that a client needs help: then it’s a scramble and I’m occupied with that for weeks. Always the same few issues.

          >Maybe a buddy of yours from bible study who got charged with statutory rape, something like that?

          The 1st amendment is basically just treated as another freedom of association clause. It does not protect against state-enforced equally applicable prohibitions. I wish it did. There is no real “freedom of religion” in the USA: only freedom to obey the state religion.

      2. I believe this was the lame game someone was shilling hard on /g/ recently. They’re making the rounds to all programming forums it seems…

        1. nobody: What’s so lame about an expansive opensource game? Because you disagree with one of the contributors that makes it lame? This is why your going to get license rescissions eventually: you cannot (will not) separate the person from the work: one of the foundational principals of opensource. Have you tried the game? No. Yet you have an opinion on a contributor and thus impute that upon the work as a whole. The entirety of the opensource world will collapse because of this: Men will NOT work for free just to have the fruits of their work stolen (CoC’d) from them and their expression curtailed.

          But perhaps that is your aim. Free Software was fine when GCC destroyed the compiler cottage industry. Opensource was great when Linux destroyed SUN. But now that opensource threatens all commercial software spheres aswell as governmental control and censorship… we have CoC’s and gate-keeping committees (often under the name of [Project]-Women) introduced: to crack the consensus (achieved) and drive away the honest engineers and hackers (mostly achieved).

          1. I find it lame only in that it’s being shilled everywhere, which suggests that it can’t merit attention on its own, and requires external help. (This is assuming ‘nobody’ is being truthful, of course. Otherwise the shoe’s on the other foot.)

  13. Riddle me this. If your SaaS vendor goes belly up and doesn’t pay his last set of bills to Amazon/Azure/Google, how do you get your data back, and how much is your contract with the SaaS vendor worth in bankruptcy court?

    1. Interesting. Unfortunately, this also creates an unfortunate network effect in SaaS. If you must you SaaS, you want your SaaS vendor to be the one least likely to go belly up.

  14. This post made Slashdot…and the Stallmanites are saying “it’s about time ESR agreed with RMS!”. Gah.

    1. Hey, you have to take whatever made up victories you can get! Otherwise you might have to think about HURD, and no one wants to do that.

  15. Time for somebody to startup a business providing Salesforce services to other businesses.

    *They* aren’t a business that deals in guns, but their customers are.

    Let’s see how recursive SF wants to go with this bigotry. Let’s see what happens to their stock price.

    1. Won’t work. Software companies don’t like you reselling their stuff — in the 80s they manipulated the law in such a way to prohibit resale of packaged software (something that would otherwise be permitted under the first sale doctrine). In the SaaS era, reselling the service will be met with a swift banhammer.

      1. OK…well…if SF is a publicly traded company, then how is turning away lawful business *not* considered a violation of its fiduciary responsibility to its shareholders? Ripe for a class action lawsuit?

        1. They’d probably argue that 1) serving those nasty eeeeevil gun sellers carries reputational risk that outweighs the benefits of their business, and 2) they need to be a socially responsible company.

          1. Nowadays, you can’t discount the possibility that shares in a company like SF are splattered all over millions of peoples’ retirement plan portfolios.

            If so, ~75% of those people might object to SF willfully diminishing their profitability for fanciful reasons that have no basis in law…which is ultimately what ‘fiduciary responsibility’ is meant to protect.

  16. … who’s gonna make an open-source “Salesforce”?

    Or an open source point-of-sale system, say? (This is my personal area of domain expertise.)

    I mean one any business would ever want to actually use?

    A search of “top N open source POS products” reveals a dumpster fire I would never trust with my business, with useless UI, or “hey, we only support this card processor we get kickbacks from”, or “free version supports one user, one store” … which, hey, might suit some people, sometimes, but ain’t gonna touch most market segments.

    One key factor is “we can only make money to develop our OSS product – and make our own salaries – by selling support and setup”.

    Problem: Awful incentives, to make it need support and expert setup…)

    UX is one thing FOSS can’t seem to ever manage to not make awful.

    UX is also the heart and soul of all commercial software outside of the systems/system admin arena.

    … which is where FOSS rules the roost.

    But Salesforce is all UX.

    1. Presumably companies that use POS systems have some development budgets, and it seems like the sort of project they could, in principle, cooperate on. And they’d have incentive to make it not suck.

      (but I know nothing about POS anything, so…)

      1. Yeah. Hence my earlier comment. OSPOS is unlikely to happen unless we luck out and there is someone who’s big into both OS and POS, and has enough skill to write the next POS VisiCalc.

        Might just be a matter of time, but in that case, OS is racing closed source, and, well, SalesForce already won that race… In general, it looks like OS will always be the enemy of “good enough”.

  17. There’s a related section of TAOUP that has never sat right with me: “The Web Browser as Universal Frontend,” particularly this line:

    The way that browsers decouple front and back ends has larger implications. On the Web, locking in consumers to closed, proprietary protocols and APIs has become more difficult and less attractive as this trend has advanced.

    It’s not…wrong, exactly, in that the use of html/js/etc is “opener” than some alternatives. But that doesn’t seem to have bought us much more than tcp/ip being “open.” The important stuff just got hidden on provider servers instead of opaque binaries. I still can’t dump my data unless they let me, nor can I swap a crappy web client for a local one more to my liking.

    1. And worse, unlike data locally jailed by proprietary software, you don’t even have a copy you can back up, attempt to reverse-engineer, or just archive until a competitor reverse-engineers the jailing file format for you.

  18. Yeah, but there aren’t any good open source CRM out there. Nor even any good close-source on-premise installed ones. I looked and looked when Microsoft CRM when cloud-only i.e. SaaS.

    You see there is software people want to use. There is software people don’t want to use but it is their job to use. And there is software people should use at work, but it is not their job, so they won’t. CRM is in this category. Most CRM developers think in the framework of salesmen sitting down in front of a web form and entering I met so and so and we discussed this and that. They won’t. And no sales manager ever will punish a salesman with good numbers for not doing the “paperwork”. And we cannot afford hiring assistants for this.

    So Microsoft CRM did the sensible thing and integrated it into MS Office so it at least captures the emails and meetings automatically, which is something. I could not find any that would do this.

    Things are getting weird in ERP as well. Microsoft, SAP moving to the cloud. Unacceptable to me. But I am not at all convinced open source ERP like Compiere or Odoo has really proper localizations, like, filling out all the important tax and other forms in major European countries, which the accountants expect their ERP to do. I was in Italy in 2012 when a dottore commercialista told me their local accounting software vendor sends them an update CD every single month, capturing the tiniest changes in tax rules and the forms that report them. This is something neither Microsoft, nor SAP, nor any open source ones do. And of course I need all subsidiaries to use one integrated ERP in all countries so that I can automate corporate analysis properly, I cannot let them use local software.

    I don’t know what is going to happen. Maybe some large multinational corporation gets pissed and invests $100M into localizing Compiere everywhere or something. I cannot imagine everybody will just accept their core financial data being stored in the cloud. It is not even really legal. If any personal data we have, like a customer’s address, gets stored on a server outside the EU, that is a GDPR violation. I suppose S-Ox and suchlike has similar rules in the US.

  19. I sense something happening quite soon now. Let me use a parallel. It is more efficient if all the factories in a country are right next to each other. But WW2 happened and many realized that having dispersed factories makes it harder to bomb them from the air.

    The point is, the economically most effective solution, towards which markets naturally evolve, is not necessarily the best solution from a security viewpoint. The most efficient way to use the computer for most purposes would be to merge Windows, Google products and Facebook into one Saas blob coming pre-installed on every tablet with an attachable keyboard. The most secure way is the Gentoo way, every app you use you compile from source you or other people you trust have eyeballed.

    I don’t know if markets are capable of overcoming their inherent orientation towards efficiency and implement something less efficient but more secure. I don’t know if those factories got bombed by non-state actors, would their owners on their own, without a state ukaze, have dispersed them. This is the part that worries me as my parallel is a bit off: it is the factory-conglomerate itself that is bombing its customers, at the request of the state. So you cannot rely on the state to organize it.

    The difference between efficiency and security is in time-scale. Every customer and every business notices lower costs are better rather immediately when there is a chance to go lower. But they tend to care about security only after they were burned once.

  20. This is the world that we helped create, by pushing so hard for Linux and open source everywhere that all the money to be made simply writing software has dried up, and now companies have to tie it to services and cloud just to stay afloat. There is one exception: the Apple ecosystem. There, people have money to burn and will gladly pay generously for a quality software package or even an iOS app.

    This essay you wrote is exactly why I was skeptical of cloud computing. The cloud hypetrain left the station while I was working at a robotics company, and unfortunately today it’s the case that a journeyman programmer usually has two choices: work in the cloud or starve.

    I don’t see SaaS going away anytime soon for this reason, and also because business owners are already well aware of the risks — and willing to take them. Remember Dahl’s law: the only thing that matters in software is the experience of the user. It has to be easy to use and it has to Just Work. Otherwise business will not adopt it. It doesn’t matter how much more secure you are, how much more protective you are of the user’s interests over your own corporate interests. If the shit’s hard to use, it’s unacceptable. Period.

    The lack of a reasonable alternative to SaaS solutions also goes unaccounted for. Forget Salesforce — there are still no open source alternatives to Excel and Photoshop, both of which are subscription-based and partly cloud-hosted. GIMP and LibreOffice Calc don’t count except for the most casual of users. For professional workflows, Excel and Photoshop are the gold standard, of which the open-source offerings fall far short in terms of usability and features.

    1. >This is the world that we helped create, by pushing so hard for Linux and open source everywhere that all the money to be made simply writing software has dried up, and now companies have to tie it to services and cloud just to stay afloat.

      The reward for solving problems is that you get to deal with the problems that follow from your solutions. This is not news.

    2. >Remember Dahl’s law: the only thing that matters in software is the experience of the user. It has to be easy to use and it has to Just Work. Otherwise business will not adopt it.

      No. In my circles there is a term “demoware”. Demoware is that kind of software that looks very good on a projector. Full of eye candy like 3D graphs and green blinking balance scorecard lights. Demoware does not care about things like the ease of data entry or ideally automatically collecting data.

      While the real purpose of software in business is cut human hours needed to do a job. The best “user experience” is not needing a user at all. But many managers don’t get it.

      Story. Some dudes developed a financial collection add-on to an accounting software. The idea was that consultants set up everything in configuration tables and the actual user interface is just pushing two buttons. It could have been fully automated and running in the background, but this compromise was necessary so that users see and report bugs, error messages, if it is fully automated they would be hidden in logs nobody reads. Do you know what was the managers of a potential customer said at the first demo? “Why do we have to pay so much for just two buttons?!” It was excellent software, but not demoware.

      >I don’t see SaaS going away anytime soon for this reason, and also because business owners are already well aware of the risks — and willing to take them.

      Not around here. In the greater Germanosphere business owners are paranoid about potential US spying and are generally conservative with data. It should be stored on-premises, and while the app may be closed source, it should be stored in an Oracle or MS SQL database where any IT guy can read, export or if really needed modify it with SQL queries.

      It is not decided yet. SAP, Microsoft pushing the cloud, customers pushing back by sticking to old versions.

      1. Concern about demoware reminds me of the curse of the demo: it’s easy to make a UI mockup that demonstrates everything you want to demonstrate, but if 90% of the work is the back-end stuff that makes the UI meaningful, the managers and investors suddenly wonder why things don’t just work.

        After all, there’s an interface for it!

        I have sometimes wondered if interface proposals should be made to look like they were created in crayon, to give the interface an unfinished look, to illustrate the parts that can’t work just yet….

  21. customers pushing back by sticking to old versions.

    And vendors pushing back against the pushback by saying “Oops! The license to your old versions just expired! Contact a sales rep now for upgrade options.” Adobe is currently doing this with its Creative Suite. I have friends who still use CS5, but if they weren’t pirating it before, they technically are now.

    1. If they’ve done that with CS6, I haven’t heard about it.

      For me, though, the death of CS6 isn’t likely to be anything Adobe does, but what Apple does: when 32-bit programs stop working on OS X, then my CS6 will stop as well. While Photoshop itself is 64-bit, the license manager isn’t.

  22. CERN gets the message.

    Money quote:

    A prime example is that CERN has enjoyed special conditions for the use of Microsoft products for the last 20 years, by virtue of its status as an “academic institution”. However, recently, the company has decided to revoke CERN’s academic status, a measure that took effect at the end of the previous contract in March 2019, replaced by a new contract based on user numbers, increasing the license costs by more than a factor of ten. Although CERN has negotiated a ramp-up profile over ten years to give the necessary time to adapt, such costs are not sustainable.

    1. Another take on that:
      https://www.theregister.co.uk/2019/06/13/cern_microsoft/

      As such, CERN has been quietly working on a project called Microsoft Alternatives (MAlt) to develop migration paths away from the commercial software offered by Microsoft and like-minded vendors.

      CERN is the place where Berners Lee developed the WWW (HTTP/HTML) and they also have their own Linux distributions (CentOS and Scientific Linux CERN 6). As for “cloud”, “networking”, and Big Data, they have a reputation to lose.

      In short, a huge organization, with a huge budget, and very, very smart people.

      What could go wrong for MS?

      https://www.theregister.co.uk/2019/06/13/cern_microsoft/

      “While the Microsoft Alternatives project is ambitious, it’s also a unique opportunity for CERN to demonstrate that building core services can be done without vendor and data lock-in, that the next generation of services can be tailored to the community’s needs and finally that CERN can inspire its partners by collaborating around a new range of products,” said Ormancey.

  23. Economics in everything. Technically you do retain control even if you use something like Salesforce. However, there’s a risk that you might suddenly have to replace the service, and that risk times the cost of doing so can be estimated with sufficient accuracy.

    Using technically untrue absolutes like the above will sometimes trigger intuitive markers that cause unnecessary rejection of the message.

Leave a comment

Your email address will not be published. Required fields are marked *