Thank You, Sal

"The question now is: what are Apple’s plans for automation? Is this another sign of pros being ignored? At best, the company has lost a key advocate for users and link to the Mac community."

At best, indeed. Obviously a very, very bad sign in general for the entire Mac scripting scene, and for pro users.

-----

As an aside, 7 or 8 years ago, I ran into some scripting questions I couldn't find answers to on the web. I somehow managed to find Sal's non-Apple email address, shot him a question, and to my great surprise, got a prompt and detailed answer. He was even willing to maintain the correspondence, and answer more questions. An unbelievable great guy.

I'm hoping this merely marks a change in Apples structure hopefully giving these responsibilities to a team with more resources. While this sounds onimous frankly automation technologies from Apple have been underwhelming for some time. AppleScript Bridge and their JavaScript language were underwhelming to say the least. While I fear this marks a pulling away of very important pro features I'm hoping for the best.

One odd thing I'm hearing though that I've not had time yet to confirm is that Apple deleted all Sal's WWDC session videos. Has anyone else heard this? If so why would they do that?

Apparently the 2016 session 717 "Beyond Dictation Enhanced Voicecontrol for Macos Apps" is no longer available on the website. But Sal's 2015 session 306 "Supporting the Enterprise with OS X Automation" is still there.

if you look at http://dictationcommands.com (official website from apple for dictation command, link was given in the removed wwdc16' 717 session), and look at the video at the bottom comparing Dictation Commands and Siri (http://dictationcommands.com/comparison-1440.mp4) you can imagine there is a *strong* critic against Siri. Maybe this was not appreciated internally?

> I suppose it’s *possible* that there is a lot of internal progress that we just don’t see.

If there's one thing I've learned this month, it's not to trust my own wishful thinking.

"There are smart people all through the ranks at Apple."

It doesn't matter that you have smart people if the process within the company is stupid. I once read an article tying the issues of software development within Apple to the Agility shit they put in place.

"The OS updates bring a steady stream of improvements."

A steady stream of useless features and not enough bug corrections.

"It’s true that we eventually got Swift and APFS"

Swift doesn't solve any issue. And APFS is as available as AirPods.

The eventual abandonment/deprecation of AppleScript and Automator is something I've been worried about ever since App Extensions became a thing on the Mac. They operate in the same problem domain -- allowing unrelated apps to work together -- but AppleScript/AppleEvents is far more thorough and wide-reaching and can be automated, while App Extensions only operate in very specific extension points defined by Apple and activated directly by a user (ie, not automated, barring some wonderful Accessibility hack or something).

The other source of my worry has been how Apple hasn't making its own apps scriptable for some time now. I'd hoped to create a Numbers-to-Motion workflow to turn my iOS conferences spreadsheet into an interstitial video for my livestream, but it turns out Motion isn't scriptable to any useful degree.

Dismissing Sal (and seemingly eliminating his position) kind of ices it, from my POV. At this point, I would expect AppleScript and AppleEvents to be deprecated in macOS 10.13, miserable as that is to contemplate.

@someone

"Swift doesn't solve any issue"

No, it brings a lot of issues to solve.

Or not.

If AppleScript becomes like DDE on Windows--an undeveloped, but still maintained, legacy technology--that would be OK. It's hard to imagine Apple completely removing it, but it's not hard to imagine Apple letting it gather dust.

[…] most of you have heard the news that Sal Saghoian has been let go at Apple. The best roundup of the news is at Michael Tsai’s blog. It’s worth noting that not only was he fired but his entire team was disbanded. At first, like […]

I don't think Swift solves no issues, although I fully understand why people see it as inferior to ObjC for classic programmers or Rust or Haskell for those less enamored of C/Obj-C. The original announcement suggested Swift could do scripting as well but that hasn't panned out for a slew of reasons. Playgrounds in particular are very poor for that sort of thing. (Since the whole program keeps getting rerun -- a big problem for interactive scripting)

To automation, the more I think about it the more I think Hamish is correct. Look at the last 15 years. What was the biggest automation success? Automator. Who do you know that regularly uses Automator? It's basically a forgotten technology that never was successful on its own terms. Everything else like Scripting Bridge is a mess and often inferior to 3rd party alternatives. While by all accounts Sal is a great person with his heart in the right place, by what criteria could we call his team a successful one? They've introduced one failing technology after an other that then Apple gets stuck supporting longer than they want to. At a certain point a good manager would say enough.

@Clark To me, from the outside, it seems like Apple leadership has not been big on automation for a long time. Sal did not get a lot of resources or support, and my impression is that his group was kind of off to the side. So you could look at the results and see a failure to work a miracle. There is obviously so much more than could have been done with these technologies. Or you could say that, given the context, he was amazingly successful at keeping the lights on and even starting a bunch of interesting new projects. It’s not clear that anyone else would have.

Again I'm mixed. Look at ScriptingBridge. That was designed and written when Sanderson's Appscript was available. He even offered it to Apple and was rebuffed. So while I appreciate that Sal by all measures is a fantastic beloved individual, I think the actions of the team are more complex than many are portraying. (Which is likely why Hamish made the tweet you linked to in your post)

I fully agree though that a big part of the problem rests with Apple who simply didn't care about automation. And arguably this was a big part of Apple not caring about pros. It just never was a focus for VPs which means you have people doing stuff on their own with limited long term support. So don't get me wrong, I am fully as angry about that as everyone. And that's not something Sal could have affected.

Yet given the demands of Siri, I'm not sure I'd be willing to say Automation is being neglected. I think we need to wait a little before leaping to conclusions. This may instead be the start of Apple actually supporting the technology and giving it to the Swift or Siri teams. The big worry of course is that the focus of Siri also isn't on pros.

@Clark In my experience, Siri is almost a complete disaster. The API is extremely limited, and the basic features are still unreliable. Even if the plan is to replace AppleScript and related technologies with something better, they still need to be maintained in the interim.

Right, but that's right now. Clearly the Siri team and more particularly key Apple VPs see Siri as going in this direction. Whether what they produce is any better than what's come out of the Applescript team the last 15 years is of course a completely different question. I see Siri as getting better and Apple finally starting to fix management issues.

Certainly I hope they leave Applescript in place until any alternative is mature.

Looking at Apple's recent actions, it seems implausible that they let Sal go because he wasn't successful enough at allowing people to automate the Mac. It seems much more plausible that they simply don't really care about creative professionals anymore.

Ironically, the thing we all wanted - for Apple to be successful - has become the thing that might take Apple away from us. The reason I can't buy a Mac anymore is not because they've gone out of business; it's because they've become so successful that they simply don't have to care about me anymore, and don't have to make hardware and software that fits my needs anymore.

At this point, I'm not even worried about it all that much anymore. I'm more worried about whether I should switch to Linux or Windows as my main operating system, because the only things that still keep me on the Mac are OmniGraffle, Sketch, Pixelmator, and Coda. And I could probably do without them on my non-work Mac.

I created a petition to hopefully convince Apple that there are a body of users who need and want scripting in the OS. Please consider taking a moment to sign it.

https://www.change.org/p/apple-keep-mac-user-automation-such-as-applescript-automator-in-mac-os-x

"Ironically, the thing we all wanted - for Apple to be successful - has become the thing that might take Apple away from us. The reason I can't buy a Mac anymore is not because they've gone out of business; it's because they've become so successful that they simply don't have to care about me anymore, and don't have to make hardware and software that fits my needs anymore."

Es verdad. Welcome to my mindset in late 2010 after Steve-o's Lion "Back to the Mac" keynote.

That was my signal that the policy wasn't really "everybody doesn't need a truck". Instead it was, "we don't want to be in the truck business going forward".

(Similarly, Lukas, you just had a comment on the Touch Bar thread where you mentioned it sucked for the 10% of power users. I almost snarked that it was lucky they weren't putting it on their "Pro" laptop...)

While I, like most here, am angry at how the pros have been treated, my sense is there's a lot of internalized anger looking for any outlet. I just am far from convinced this is indicative of that. Maybe it is. I hope not because when I try to do serious things with Linux or Windows it gets frustrating fast. For all of Apple's flaws for pros it still has a lot of pluses.

However again let's be serious here. Don't you think rolling these functions into more relevant teams makes more sense? Especially if they weren't getting good support in a separate team? I'm hearing for instance that Sal was also in charge of Unix scripting pieces like Bash, Python and so forth. All those builds are notoriously out of date with the default installs. Everyone more or less just goes to MacPorts or Homebrew so it's not that major a problem. However wouldn't it make sense to roll more of these Unix tools to the Xcode team? I'll go out on a limb and say they're more apt to get updated.

Likewise if, like me, you're deeply dissatisfied with Scripting Bridge and Javascript for Automation, wouldn't it make more sense to roll that stuff into the Swift team and perhaps (as most expect) those working on a new app framework based upon Swift?

I'm not saying all that is happening. Just that if things were going to get better it seems like something like this has to have happened. Now don't get me wrong. If Apple is finally designing a new scripting/automation system I have plenty of worries. Some of the design of Swift seemed odd for instance. And if automation is getting wrapped up with the management of Siri's API and application extensions then there's lot to worry about there too. As much as I disliked Applescript the language the underlying Apple Event model was pretty elegant. What I've seen done of late really isn't. That's a bit worrisome.

I think this is very odd. It's certainly true that this could be a mere reorganization, but the fact that Sagohian is out without any good explanation doesn't really bode well.

In Yosemite Apple introduced JXA, JavaScript for Automation which is in the simplest terms AppleScript but using JavaScript. JXA got some pretty big enhancements in ElCap and then nothing in Sierra. It doesn't seem like Apple is ready to abandon Automation, but like many things Mac right now, their commitment is...unclear.

It also seems possible to me that many saw his role as an evangelist and Apple is moving away from some of those roles? I don't really think this is the case, as Safari and WebKit evangelism seem to continue to grow stronger. However, it wouldn't be un-Apple-like to try evangelism out and then not like. Personally, even that sounds like a poor choice given how this particular community feels.

To be clear, I'm not angry. Apple is a company; it's not my friend. They're doing what is best for their bottom line, and I don't begrudge them that. They owe me nothing; I bought their products in the past because they made the best products, not because I had a personal relationship with them. When Android became a better fit for my needs than iOS, I wasn't angry at Apple; I just bought an Android phone instead. I will do the same when it comes to the Mac.

What Apple is doing right now is making a lot of people happy, and works for a lot of people. And that's fine. It's just not for me anymore.

I used to be worried about it, because I like the Mac, and don't really want to switch to Windows or Linux. But Windows or Linux aren't what they used to be in the 90s, either. I'm now using a Surface Book for drawing, and I actually like a lot of the things in Windows. I'm using Linux on a bunch of computers, and, in many ways, it's far superior to anything else.

--

I don't agree that it makes more sense to roll these features into individual teams. You want a coherent way to automate your computer that works consistently across different apps created using different technologies; you don't want every team at Apple to have its own ideas (if you do, you might as well go with Linux :-). But that's a bit besides the point; looking at Apple's actions in the past few years, there's really no indication that they have any interest in scripting. It's just not likely that they let Sal go because they thought that it would help improve automation. Maybe they did; it's possible. It just doesn't seem very likely.

It's much more likely that they let him go because they don't care about automation anymore.

Clark: "Again I'm mixed. Look at ScriptingBridge. That was designed and written when Sanderson's Appscript was available. He even offered it to Apple and was rebuffed."

Nope, you have that backwards. It was Apple who approached me expressing their interest in potentially including appscript in 10.5.

I, on the other hand, was the steaming moron who completely failed to recognize the priceless, once-in-a-lifetime opportunity that Apple were handing me on a plate. So instead of finding out who was backing it inside Apple and establishing direct communication lines so I could evangelize it to their decision-makers and clinch the deal, I just sat around like the big dumb arse I am thinking "Mission accomplished! They're sure to adopt on its technical merits alone!"

Protip: Technical merit doesn't count for shit; least of all in the computing world. What matters is Who You Know, and how well you Sell To Them. So why would Apple adopt the unknown work of some nobody amateur when their very own highly-trained, highly-paid Professional Software Developers can do the whole job, and more?

As the then-maintainer of MacOSForge told me in autopsy, management looked at this sprawl of candidates for 10.5 inclusion—RubyOSA, Python-appscript, Ruby-appscript, ScriptingBridge—and said "Why do we need all of these! One will do!" So SB got upgraded to work with scripting languages as well as ObjC, and the rest got the shove.

Completely right decision, based on all the information available to them.

Shame they didn't have all the information they should've had, but you can thank Mr Stupid here for that. (That's my name: Mr Stupid. Just so you know.)

...

Now, ask me about my last couple years' dealings with Sal specifically regarding JavaScript for Automation and SwiftAutomation, that's a whole different story entirely, and your assessment is pretty well on the nose. Longer versions can no doubt be dredged from the various braindumps still landing on poor AppleScript-Users if you really want to know.

Suffice to say, I only wish they'd sacked him years sooner; when Automation was still in a far healthier state than it is now. When a product manager has more products on his CV than his products have users, do not be surprised at the outcome when Finance politely knocks on the door asking what exactly he's achieved with all that money they've been handing him all these years.

Lovely charming bloke, great user evangelist. Couldn't manage a pissup in a brewery.

Michael Ball: "In Yosemite Apple introduced JXA, JavaScript for Automation which is … AppleScript but using JavaScript."

'Cept it wasn't, never was, and even if it had been the stupid sods on the Automation team never did shit to actually sell it.

JavaScript for Automation wasn't Mac Automation's rebirth, it was the fatal nail in its coffin.

...

So here's the "How To Destroy your Product, your Market, AND your Career in One Easy Step" Apple official handbook, 2014 edition, © Sal Soghoian:

Take the #1 Programming Language in the whole world—with what, Five million? Ten million? users, at least half not yet on the Mac—and turn it into a broken DOA corpse that couldn't even muster TEN.

So how many dollars again did Apple sink into that benighted Automation department over the last 19 years? Five million? Ten million?

I would have given them its first million customers for nothing.

"So how many dollars again did Apple sink into that benighted Automation department over the last 19 years? Five million? Ten million?"

So, let me see if I understand you correctly, has. You have endless complaints about how screwed up Mac Automation has been. And then you mention they might have spent $5-10 million over 19 years on the effort. Do you not see any possible connection between those two items? That dollar number is not just literally spare change in the pocket of Apple, it's spare change in the pocket of the Mac division. It's spare change in their pocket that excludes quarters, dimes, and nickels.

You can't have it both ways. Automation may not have achieved everything it could have achieved, but it did achieve some actually useful things. And hey, if it had been slightly more of a priority in terms of resources, it might have achieved quite a bit more. But now, it's more than likely going to end up EOL'd, even though it remains quite useful for pro users.

@Chucky: That's $10M over 19 years assuming $0.5M p.a. on 2 developer and one middle manager salaries, at maybe 6000 man-hours/year.

That doesn't account, of course, for all the money spent by other departments—QA, AppDev, Security, Documentation, and so on—in keeping that shitshow on the road, so even if my estimate on that one department's budget isn't wildly off (and I won't be surprised if I still managed to lowball it, forgetting things like company health plans and so on), it's still only one part of Apple's total investment in Automation over that period of time.

An initial investment for which Apple has received fuck all back in return.

...

Apple doesn't want to spend money to keep a small, stagnant pool of existing customers happily in place. It wants to spend money and see NEW CUSTOMERS drawn in. If/When that starts happening, Apple can gauge what sort of increased returns to expect if it ups its spending further, and then set its time, money, and manpower budgets accordingly.

Initial investment = some new customers = more investment = more new customers = lots more investment, etc, etc, etc.

Initial investment = no new customers = no more investment. Full stop.

If Sal had wanted a $5M budget to replace his $0.5M one, and a twenty-developer team instead of a two-developer one, he should've proved to Apple that he could pull in the number of new customers it would take to justify that budget increase. The fact that his budget remained perfectly flat doesn't reflect the fact that he failed to do so—it actually reflects the fact that Apple were being extraordiarily tolerant of his failure. Most likely because the cost of Apple of terminating that budget would've exceeded the cost of pulling the plug on the technology already in place and the nuisance of having to deal with the small but loud and angry customer base its termination would affect.

So don't blame Apple, blame Sal.

...

Oh, and if you still don't think $10M is enough money to bootstrap a successful new product that promises to deliver buttloads of new customers in return, just remember that I already proved that I could pull in 1000 new Automator users in a totally new market on a budget of $0 just fine.

A market (Programmers) with another ten million potential Apple customers still to go.

A market that would've synergized the shit out of other Apple products as well, since programmers who buy into Mac Automation as customers are also ideally equipped to provide Apple's AppStore with loads more great new Apps too—not least apps with great scripting support built in!)

You think Apple wouldn't have wrote Sal a blank check if he demonstraed talent that could convert the entire programming industry into Mac users AND devs, simply by hooking them on the greatest, most irresistable 100% Solid GeekBait™ ever devised?

...

Hell, you provide me a $10M budget and I'll deliver you the whole UK/European Food and GM packaging artwork industry markets into the palm of your hand (after I get my big-ass bonus, of course:). Because I've already taught myself how to implement technologically sound, eminently saleable product—and am now also slowly teaching myself the craft of how to sell it too!

Though now I'm probably be required to build it on Windows + Adobe CC, seeing as how Mac-based automation and Mac-based graphics industry might not be hanging around in the Mac-based world for very much longer! COM? Urg! So you know who I'll be blaming for that…;(

"So you know who I'll be blaming for that"

Yes. I most definitely do. But I continue to think your logic is faulty.

Let's stipulate that you're correct that Sal was incompetent at his job. Why didn't Apple terminate him years ago? Why, now, did Apple not just terminate Sal, instead of terminating his job as well? That would be the normal reaction to a manager seen as incompetent.

Obviously the pro Mac market is only a speck in Apple's overall financials, and they obviously simply don't care about it anymore They've been making that excruciatingly clear in a multitude of ways separate of terminating Sal's job and all that implies.

(Personally, I happen to think it's short-sighted behavior on the part of Apple. Continuing to serve the pro market would be marginally profitable, and more importantly, would provide actual benefits to the platform and entire Apple ecosystem going forward. Sure, it'd still be a speck in their overall financials, but Apple's strength in the pro market not only played a part in saving the company last century, but it also played a real part in the company's resurgence this century. Ceding that market completely is just going to open up opportunities for someone to build mind-share like Apple did in the first decade of this century. Influencers are real. Everyone having Mac laptops at tech conferences in '02-'06, for one example, mattered IMHO.)

But given that the core problem is truly about Apple consciously deciding not to serve the pro market, I find it impossible not to think you're clearly aiming your blame at the wrong target, no matter what Sal's merits or lack thereof actually were. And perhaps you might want to consider that your negative personal experience here may be clouding your view of the larger situation...

Oh yeah, One More Thing: Sal never had a two-man team, he had a six-man team!

Other Chris
Other Other Chris
Mark Alldritt
Shame Stanley
Matt Neuburg
me

You know what Sal did with the last four? Nothing. At. All.

--

Rule #1 of selling stuff: Learn what Network Effect is. You don't reach massive success by doing everything yourself, you do it by doing just enough to get your first few customers and supporters sucked in, and then you get them to do all the hard work of sucking in the millions to follow.

As appscript project manager, I made exactly the same mistake as Sal. You go ask what Matt Neuburg what he could have done for appscript, had I not so casually, unthinkingly fucked him about. Difference, I learn from my mistakes, so you're 100% right in one respect: my past personal experiences do inform my current position, because I figured out what I did wrong and how to do it right next time, and I am feeding that knowledge into what I do next.

Which includes (once the current mania safely burns itself out) doing my absolute goddamned best to enlighten Cook and Federighi on just how enormous a blunder they're going to commit next if they don't build their entire Siri Automation strategy on that incredible, once-in-a-lifetime FREE GIFT that 25-years of AppleScript original vision and extant infrastructure actually represents—not merely to Apple, but to every single bugger on this rock!

Because like it or not, we are ALL already (or soon will be) inherently intertwined with the technology that now surrounds and starts to shape our entire world, and our position in it.

And Sal is still absolutely 100% goddamned right about one thing, even if he managed to fuck up execution of absolutely everything else: The ONLY THING that absolutely matters is WHO holds control of the tools that form this ubiquitously connected, technological world.

Automation says: That power belongs in the hands of all users.

Google and Facebook say that power must be held safely in their hands.

Apple is still trying to decide whose side it is on: the users' (e.g. strong crypto even governments can't break), or its own (a locked-down ecosystem where users can only do what Apple wants).

...

AppleScript Automation is, as business propositions go, a failed technology simply on its lack of sales alone. No wonder Apple want to retire it and move onto their Next Big Thing: Siri Automation. But Siri Automation is going to fail a million times harder (and already is—just listen to what users say about it!), if it cannot even learn from the mistakes that AppleScript Automation made, never mind capitalize on the vast extant free resources it has still managed to put in place in doing so.

Application Extensions? Please. AppleScript infrastructure is already a million times more powerful than that tin-plated toyware ever will be.

And Magical AI middleware glue that suddenly knows how to translate real users' Siri requests into meaningful, accurate, reliable actions upon a millions-strong ecosystem of Apps, internet services, and users' information? Even better minds than Apples' have been promising AI "Real Soon Now" for 40 years and we're still only a tiny fraction of the way along.

...

The ONLY way users will retain control is if we retain the philosophy and goals that underpin AppleScript automation, which in turn underpin those of Alan Kay, and Seymour Papert, and Doug Engelbart, and Vannevar Bush before. A vision of computing—personal computing—where it is the user who is in control: not the Facebooks, not the Googles, not the governments; definitely not the bloody computers themselves.

Apple fears ceding control to its users—it has spent the last 15 years erecting meticulously revenue-generating services tollbooths all over its personal computing network, after all; and handing direct control to customers risks those customers choosing to route their way around those paypoints, after all. But it really shouldn't: distributed power and ubiquitous connectivity will generate far more revenue and riches for everyone than even they can ever hope for.

Just look at the internet and the world wide web: it didn't get to where it is by every single vendor fighting to keep the little bit of turf they've got.

This is Game Theory 101, and John Nash was 100 times more crackers than me, and look how right he was.

And now look at the impossible horrors of 2016, UK Brexit and impending Trump government, because that is the product of millions of individuals' response to the terror and uncertainty of not feeling in control of their own lives and destiny.

--

You are thinking on a level that is a billion times too small. Apple are thinking on a level not much better. I'm just seeing the opportunity you've both left wide open here to steal a lead on you all. And not even doing it for myself, but so I can pay something back to the rest of the world for all the incredible opportunities it already has given me in turn.

I've been looking, learning, screwing up, and trying again for 13 years, working this one particular problem in this one particular microcosm: cos I may not know much but I know the way to make a difference is by looking for what everyone is failing to see, and positioning myself to do something about it. I'm not good, but I'm getting better. Who knows, maybe I inherited a few atoms from Steve Jobs after all? I have a chance to pay back now.

Sal failed, because he failed to make Network Effect work for him. I failed, because he failed to make Network Effect work for me. Engelbart, Papert, the same; Kay, not much better. Berners-Lee, doesn't even realize how bad he fucked up. Same problem too, only scale differs. Nor did we fail ourselves: we failed our vision—our hope of making a greater world for all—and thus every human being who would have gained had we not fucked up.

...

Sal is going to spend this month, November, in misery and heartbreak. Been there, done that. Don't mistake my contempt for Sal as product manager for Sal the man. Don't even know him, so how could I feel one way or another about the person? I'm going to give him that time, and then next month I'm going to drop him an email saying: "Have you come to understand the terrible mistakes that you made yet, and can you now admit, accept, and embrace them as the painful but could-yet-be-incredibly-valuable learning experience they are?" And if he can say "yes" to that then once again I am willing to work with the man on that one vision of user empowerment—that vision we both share and the only thing that counts.

--

This is not about "pro users". This is about ALL USERS. Your pro users had no more right to keep all that power to themselves than your Objective-C and Swift Programmers, or your Facebook, Google, and Apple directors who would happily lord over us all.

Breaking your pro users' exclusivity lock on Automation is just the first right step on what will be a very long road. Apple can happily burn its whole pro customer base [1] tomorrow, and I won't care because that is not important. What matters is what comes in its place.

There is no Automation for 10,000 users. There is either Automation for no users, or Automation for all.

And to get the Automation for all, we need Automation to generate its own Network Effect, which will only happen if those few thousands who are already invested in it can bootstrap it by generating Network Effect of their own.

So here's what you do before your next response: you go file those two feature requests I've already said to file, and then you go jump on everyone else you know to do likewise, and so on. [2]

I have already given you, and everyone else who ostensibly cares for this stuff, a real practical, positive direction to push forward, so that you can take this direction to Apple in turn. And if they can be persuaded to adopt it, then all the real work to be done will come after that.

Do it, we'll talk more; don't, don't bother. Hope to hear from you soon.

--

[1] I hope it doesn't, as it'll take me a year to haul all my day job work over to Windows, but its perogative, not mine or yours.)

[2] Some fucking muppet put up a change.org "Save AppleScript" petition for 200(!) signatures, and only got 131! Fuck that shit. They might as well just've banged in a coffin nail as well.

[…] of Mr. Soghoian means the demise of AppleScript altogether, and particularly in iTunes, but many developers, iTunes users, and others are concerned by this […]

> "Though now I'm probably be required to build it on Windows + Adobe CC, seeing as
> how Mac-based automation and Mac-based graphics industry might not be hanging
> around in the Mac-based world for very much longer! COM? Urg! So you know who
> I'll be blaming for that…;("

Even if you're completely right (which you very well might be), Sal's contributions to Apple's recent failure to make pro customers happy is very small. Just the fact that the the professional laptop on Apple's side is the MacBook Pro, while Windows users have access to stuff like the 17" Razer Blade Pro, should be plenty to make pro customers consider switching.

As an aside, I agree with Chucky. This is shortsighted on Apple's part. A lot of people buy Macs because they're aspirational products. Why are they aspirational? Because the most creative people, the kinds of designers and writers and musicians everybody looks up to, use Macs. Why do car brands have luxury or sports models that don't sell all that well? Because it makes the rest of their lineup look more desirable purely by association. Apple is killing their luxury models, driving away the kinds of customers who make them look good by association, and hoping that people won't notice that they've become a pure consumer brand.

But yes, all of that matters only to Apple. What *really* matters is that, in the end, people - not just power users - are in control of their own data, their own devices, their own destiny. Whether this happens with Apple, or against them, is just a sidenote.

Lukas: "As an aside, I agree with Chucky. This is shortsighted on Apple's part. A lot of people buy Macs because they're aspirational products. Why are they aspirational? Because the most creative people, the kinds of designers and writers and musicians everybody looks up to, use Macs."

Just… Oh Dear God, just stop.

My brains are coming out my ears at reading such fanboy toss.

People buy Apple products because no other vendor has anything like them at all.

Apple wasn't merely first to market with iPhone, it made that market in which to sell. From less than nothing at all! Hell, forget six days: Steve Jobs stepped on stage and brought an entire new universe into existence, full-formed, in under six minutes. That is creation for you.

People bought iPhones like mad because Steve Jobs was quite probably the greatest salesman ever to set foot on this Earth. Not only did Jobs know How to sell, he absolutely understood What he was selling and Who he was selling to too. Have you ever had to work with sales? You know how infinitessimally rare it is to have all three such skills inside the same industry, never mind shirt? Not only did Steve sell Product to his Market, he turned right around and sold Market straight back to his Product as well.

And before the world's heads had even stopped spinning, he was prepped and raring to roll with his next "One More Thing…" as well. That's how you do business: you use your new product to build your next market, and then you sell into that with your next product while all your competitors are still figuring out how the hell to copy what you've already sold. And you do it again, and again, and again; cos you don't stop ever, not even to draw breath, or you lose all that unique lead and momentum you've worked so hard to build up, and you're up to your shareholders' ass in butt-cheap Android knock-offs before you know it.

Had two billion years of evolutionary biology not picked that moment to go "Yo, 'm back!" at the unlucky bastard, Apple might still have been on that trajectory as well. Where was Steve going next? "Apple TV." And not that jerkoff dongle on the back—that's selling everyone else's products and experiences, not Steve's—but the whole damn TV set off the wall, content supply chain, and even how we turn it off and on.

Steve only had to rewrite the meaning of "TV" is to "what I know I can—and will!—sell better than anyone else", and he'd have been off again, and all us consumers along with him for the ride, while Steve laughs his liver out as the rest of global industry falls over itself trying to keep up. Just as he did when he served World's #1 Microsoft its ass, not by building a better PC, but by redefining the meaning of "Personal Computing" itself. And what could be moreso than owning your own Personal Supercomputer in your own pants pocket? And he'd synergize the whole show too by stomping your half-dozen remotes and their half-million buttons into the trash, and putting your iPhone or your iPad (or, being Steve, both) into your hands in their place, and selling what he's already sold you once right straight back to you for a second time as well!

...

You think "pro users" mattered even one shit to Steve after 2006? From the moment his new business was released, the only pro-users to matter any more are the ones who put shiny must-buy Apps in Steve's own AppStore! (And nobody else's!)

Steve burned his old useless product (Macs) to sell his amazing new one (Mobile), and unlike muppets like Osborne and Elop he did it exactly the right order too.

None of this means Macs would not endure: you think Steve'd want Windows to supply his App Development platform; just hand over his keys to his basement to the very rival he just stole the whole goddamn castle from? But Macs are just the sandwich guy now, keeping the workers all fed and happy at lunch, and ensuring they stay at their desks and not sneaking out for a liquid one instead. If anyone else still finds Macs useful that's just some extra kitty for the Xmas works party as well.

...

Still, we agree on your final two sentences. 'Cos when you think about what all of Apple's competitors are now selling to all us customersrubes—the Facebooks, the Googles, and so on—no prizes for identifying the next incredible market opportunity being fabricated for free, just waiting for the universe's next Steve Jobs to crack it wide open, not merely by selling the reverse Product to everyone but by selling the 7Bn Market for that Product as well.

>People buy Apple products because no other vendor has anything like them at all.

There's nothing else like Macs?

I'm actually genuinely not sure how to respond to that.

Have you heard of Windows?

>My brains are coming out my ears at reading such fanboy toss. (...) Steve Jobs stepped on stage and brought an entire new universe into existence

Interesting.

I'm actually not entirely sure whom you're arguing with, because nothing you've written seems to contradict anything I've said. Yeah, Apple doesn't care about pro users anymore. We agree. I guess you're saying that people buy Apple products not because of Apple's brand, but because iPhones and Macs are genuinely better than Android phones and Windows nowadays? Even if they genuinely were better (which I tend to disagree with), the idea that people would buy from Apple *because Apple makes better products* is extremely naive. That's just not how most people make most buying decisions.

This comment by BCG http://mjtsai.com/blog/2016/11/21/apple-abandons-development-of-wireless-routers/#comment-2639825 on the AirPort discussion illuminates a curious disconnect in strategy between Apple killing software automation on the Mac while trying to promote home automation on iOS. I'm certainly among the group who has little interest in the latter, and I can't help believe that the availability of general software automation affords much broader opportunities to a wider range of people.

[…] Apple Abandons Development of Wireless Routers, Thank You, Sal, New MacBook Pros and the State of the […]

>The internet has democratized scripting.

I think it goes even further than that. I suck at writing AppleScript, but over the last two decades, there have been hundreds of times when I needed to do something, and somebody had already written exactly the AppleScript I needed, and put it on a website somewhere.

This reminds me of HyperCard. Not everybody who had a Mac in the early days knew how to create HyperCard stacks, but I would wager that almost every Mac had at least one or two HyperCard stacks on its drive, doing something that was niche enough that no "real" software development company would bother to fulfill that need. For example, in the 90s, I wrote a HyperCard stack that taught French vocabulary, tracked which words you didn't know, and automatically generated a cheat sheet based on your weaknesses. I doubt there's a real market for that, but almost every Mac user at my school ended up with that stack on their drive.

">People buy Apple products because no other vendor has anything like them at all.

There's nothing else like Macs?

I'm actually genuinely not sure how to respond to that.

Have you heard of Windows?"

Yes. I hear it's a commodity product that started to sell some years after Macs got there first.

See also iPhone, iPad.

--

"The best way to predict the future is to invent it." (Guess who said that?)

Oooh, my bad: See iPod too.

Bonus point for identifying what linked three of the four together too.

The fact that Ford invented the modern car doesn't mean that there aren't much better alternatives. If you're not even considering using anything except a Mac simply because Apple made the first usable computer, you have very odd priorities. Even so, those are *your* priorities; clearly, there *are* alternatives to the Mac. Other vendors *do* make things like it. Much better things, in many ways.

@Lukas

"I think it goes even further than that. I suck at writing AppleScript, but over the last two decades, there have been hundreds of times when I needed to do something, and somebody had already written exactly the AppleScript I needed, and put it on a website somewhere."

Yup. The web made AppleScript use practical for the pro masses in this exact way. The brilliance of AppleScript is that it's designed to be readable.

I'm not a programmer. I could never get the hang of anything beyond BASIC. But I can find snippets of AppleScript on the web, and either use them as is, or just follow how they work and modify them to suit my needs. As time has gone on, I still can't code, but I can do wondrous things with AppleScript due to its readability and my ability to find snippets on the web. Of course, programmers tend to hate AppleScript due to its human readability focus, but there are a helluva lot more scripters like me out there than there than there are programmers.

(And, of course, that doesn't even get into folks who just download pre-compiled AppleScripts like Doug's for iTunes.)

-----

@has

"There is no Automation for 10,000 users. There is either Automation for no users, or Automation for all."

See, this totally misses the point. Automation, which is mainly scripting, but also is utility type things like text macro utilities, is only for "pro" users. It's always only been for pro users, and pretty much by definition will always be only for pro users. Civilians aren't going to spend time and effort in order to save more time and effort in the future - it's just not how they think.

Now, how many Mac users are "pro" is some meaningful way? 5%? 3%? No matter how conservative you get here, it's still more than 10,000 by orders of magnitude.

And I'm casting "pro" users with a wide net. It's not just folks who write AppleScript. It's also folks who use 3rd party utilities dependent on AppleScript, like Michael's superb EagleFiler, and many, many more. It's folks who download iTunes AppleScripts. It's folks who operate in some collaboration where someone has written AppleScript for all.

The "civilians" on the platform are never going to go in for Automation. They don't know about it, don't want to mess with it, and/or don't think they need it. It's the same as civilians who aren't ever going to buy a Mac Mini unless they want to use it as a cheaper desktop machine than an iMac. The supplementary Mac Mini is only for pro users, just like automation. But they still sell. There's still a massive amount more than 10,000 pro users out there.

I normally think you have quite interesting things to say, has, but I continue to think your personal closeness to the situation here has pretty much completely blinded you to the actual larger issues.

[…] to pile on to what’s already been said about the state of the Mac — @mjtsai is doing a bang-up job of that — but when even long-time Mac fans like @flargh say that the message is […]

It's always only been for pro users, and pretty much by definition will always be only for pro users. Civilians aren't going to spend time and effort in order to save more time and effort in the future - it's just not how they think.

This is a contemptible attitude. Kindly go read Papert's Mindstorms before spouting futher, else you are no different to all those Real Programmers who declare all users should not be allowed to code.

"This is a contemptible attitude. Kindly go read Papert's Mindstorms before spouting futher, else you are no different to all those Real Programmers who declare all users should not be allowed to code."

Everyone should be allowed to code, of course. But how many do?

Everyone should be allowed to use automation, of course. But, again, how many do?

I'm not trying to restrict anyone by merely stating what should be a pretty obvious fact that the percentage of users who are going to go to the trouble of investing a small amount of time and effort in automation in one way or another to save much more time and effort in the future is a small percentage of total users. That's just human nature. Make it as simple as you want, and it's still human nature. What percentage do you think ever do even the simplest "automation" like customizing a toolbar or a keyboard shortcut? Most users won't even change simple defaults on their computing devices, as I'm sure you're aware.

I don't understand why that should be controversial, let alone contemptible. Nor do I understand why it should have any relevance on Cupertino's decision to stop serving those users who do make use of automation. Apple's automation efforts may or may not have had great faults, but it's bizarre to think that they weren't genuinely serving a not-insignificant number of pro users in a very helpful manner. And while that number may be small in percentages, it's still pretty large in terms of actual people. You create an absurd strawman when you use 10,000 for that number instead of millions.

After re-reading the thread, I really don't understand where you are coming from on this topic, has, other than just venting your own personal frustrations with your involvement in Cupertino's efforts. So given that I don't think you are engaging on the topic, I think I'll just let this failed effort at dialog lapse on my end.

> the percentage of users who are going to go to the trouble of investing
> a small amount of time and effort in automation in one way or another
> to save much more time and effort in the future is a small percentage
> of total users.

I do think that Apple has the power to shift that percentage, though.

HyperCard was extremely good at increasing the number of "citizen developers." AppleScript failed to do the same, and I think its poor language design is to blame for that.

AppleScript is readable, true, but it's very difficult to write, both for professional coders and for amateurs. Part of the problem stems from the fact that you can't easily deduce the underlying language rules just from reading AppleScript code, something you *can* do with many other languages aimed at normal people (e.g. HyperTalk, BASIC). Also, unlike HyperCard, there is no easy entry into the language. In HyperCard, you could gradually learn more, but do useful things right from the start; you could click together your database, then have HyperCard auto-write some code for you, then start extending that, and, without even really noticing it, go from a casual user to a pretty sophisticated developer.

With AppleScript, there's the recording feature, but that really doesn't work. And there's nothing else.

So: poor language design, and extremely steep initial learning curve. These are not good design choices for a system that is intended to allow casual users to automate their computers. Hence, most people don't bother.

"You create an absurd strawman when you use 10,000 for that number instead of millions."

For benefit of everyone else: It's called doing your basic homework. Ballpark figures for various languages are easy to find. Total developers are estimated at 20M. Java users are estimated at 10M. Python users get estimated at 2-3M (Windows + Mac + Linux); personally I think that's high so call it 1-2M. My estimate: AppleScript has maybe 1% as many users as Python has, which gives us 10-20K. 10K, 20K, doesn't make much difference; we're just concerned with magnitudes here.

AppleScript doesn't even get name dropped in Tiobe's top 100 any more (pretty sure it used to), which means it has virtually no presence online. Look at activity on AppleScript Users, MacScripter.net, discussions.apple.com; that's pretty much your whole AppleScript community right there. Compare to activity in Python community. Look at how AppleScript books are selling: last one came out in 2010; I asked Apress a couple years back about chances of doing a 4th edition of Learn AppleScript: bugger all. (Honestly, I'll be surprised if it broke even: by the time it went to print the AS market was already starting to slump.) 3rd edition is at #900,000 on Amazon; Sal's book #600,000. Compare to 3-year old Learning Python at #14,000; Python Crash Course (2015)at #3,883. (And they're in a much more crowded market too.)

Oh, and compare the Automator community, books, etc; cos it makes even AppleScript look good.

So, there's two possibilities: 1. Mac Automation is so easy and intuitive and problem free that there's a vast invisible silent army of AppleScripters and Automatorers out there who are happily using it for and by themselves and never having questions or problems and never telling anyone else how great this stuff is, or 2. Anyone who thinks it's my numbers that are two magnitudes out is really, really desperate to believe their own pet technology is not already completely fucked (and has been for quite some time). No customers ⇒ no team ⇒ no product.

...

Number of developers worldwide is estimated at 20M. Xcode gets maybe 10M d/ls, so call it ~2M Mac developers. JavaScript is the most popular language in the world, so quite a few million there. Swift, exciting new geeky language currently seeing unprecedented growth. Mac Automation should be absolutely irresistable geek bait to a million geeks easy. Every other shiny toy is, and this one can do real work too. Secure that market, you ensure the self-supporting growth of the free App infrastructure that makes Automation worthwhile.

Then you can turn attention to figuring how to make Automation accessible, relevant, and useful to all those lowly commoners that Real Programmers and Real AppleScripters alike both like to look down on. People like my mum, my dad, my sisters, their kids, and so on. Which, as I've said elsewhere, is just begging for a simple command-response interface that can be driven by text, by GUI, by speech. And integrated into every single app that does any kind of notifications, and floated up to the top, not buried under 3 layers of Preferences dialogs, with simple prebuilt commands bundled as standard to get them started, and the ability to customize and add their own part of that interface too; not some opaque Script Editor app buried in Utilities completely disconnected from everything they do.

Handling system notifications and user-triggered actions are exactly kind of repetitive reactive secretarial scutwork that will make Mac Automation relevant, accessible, and useful to the tens of millions of Mac users who aren't task-driven automation workflow builders like me or you. And prove it can be sold to a decent fraction of those 80 million, there's 800 million more on iOS to port it to as well.

Lukas: Agreed. AppleScript's real value now is in understanding what it got right and what it did wrong. Ditto Logo. Apple now deal in markets of hundreds of millions, they don't even get out of bed for a few tens of thousands; that's just an pain in the ass for them, and distracts from far more important work.

Language design is part, marketing is part, placement is a huge part. If you want people to use it, is has to be integrated fully and ubiquitously into what they already know—Apps—and make itself relevant and useful to their lives there. No more script editors or crap like that. It needs to be zero interface, like Spotlight: just type, it works. Or speak, it works. Or click a button to drill down a level, and assemble sequences of commands using auto-complete text or draggable GUI boxes. Never leave the user's application though.

...

Entoli, if/when I get to fully work, should have a good chance of addressing the first of those problems, and since it's the blocker to the other two it needs to be cracked first. Entoli (Greek phonetic for "command" or "instruct") takes the good ideas from AppleScript and Logo and fuses them into something novel and new, so you can say stuff like:.

On morning alarm call: say "Yay, another fine day!", get up, make breakfast [coffee {sugar: 2lumps, milk: none}, conflakes]; eat {while: reading the paper}, walk the dog.

Real words, real phrases, real punctuation. But it's all just sequences of unary commands, with the option to import syntactic sugar (operators) on top where the added convenience justifies it (e.g. on alert: action). Very easy to add autosuggest, autocorrect, autocomplete, and all the other modern tools that allow instructions to be assembled even without knowing how to code.

Logo-style composition of commands (behaviors). AppleScript-like lists and records for composing complex data structures. Scales from super-simple, switch off, all the way up to as complex as you like: like Logo, it's a stealth-Lisp with homoiconic syntax, code as data and data as code, and able to manipulate data symbolically as well as algorithmically.

The language has legs, in other words. Unfortunately it also has me (slowest, dumbest, laziest user-programmer on Earth), but I'll get there eventually. Just hope there's still an end-user automation ecosystem for it to drop onto and into when that day comes.

@Lukas

"AppleScript is readable, true, but it's very difficult to write, both for professional coders and for amateurs. Part of the problem stems from the fact that you can't easily deduce the underlying language rules just from reading AppleScript code, something you *can* do with many other languages aimed at normal people"

Personally, I'd rather have had BASIC. I was able to script the newly "cross-platform" Microsoft Word in '94 rather easily using WordBASIC, back when I was avoiding AppleScript. Same deal once VBA was introduced. (I might have been the only Mac user on the planet who was happy with that un-Mac-like version of Word.) Part of that has to do with my having prior experience with BASIC from a young age, but I suspect that it also had to do with BASIC being an easier language for non-programmers. (Though frankly I'm not fully sure about that. For all I know, folks who grew up without exposure to BASIC might not have had an easier time. I can't tell from my own experience; I get the sense BASIC exposure for kids has pretty much been phased out; and I'm not sure how much easier the learning curve would be for newbies. I also have the lone anecdote of a creative I knew well in the '90's who couldn't get his head around BASIC, but did many very cool things with AppleScript in very short order via the 'search web and read' method.)

So, yup, I would have personally preferred a BASIC-like scripting system over AppleScript. But even if we stipulate that BASIC would've been more accessible to more newbies, I still don't think that would've moved the needle all that much in how many users took part in active scripting.

And while I would've made that different initial decision on what kind of scripting language to use, I think you somewhat undersell the usefulness of AppleScript's readability. Programmers all hate the language of course, because it's so unlike a programming language. But as you noted upthread, I did manage to get the hang of AppleScript without any real trouble precisely because of the ability to find snippets on the web, and use the readability aspect to customize for my purposes, which made the learning curve painless and quick for me.

In short: I think I agree with you in principle, but I just don't think it mattered all that much in terms of the size of the active audience.

(And while I'm trying to avoid more non-productive dialog with has, I will note that my personal experience, along with those of peers, was that Logo was pure Greek for non-programmers.)

[…] Thank You, Sal […]

[…] Previously: Thank You, Sal. […]

[…] Previously: Thank You, Sal. […]

[…] Previously: Tell Us Your Mac Automation Stories, Thank You, Sal. […]

[…] Previously: Thank You, Sal. […]

[…] Thank You, Sal, Omni’s 2017 […]

[…] Thank You, Sal […]