1. 46
  1.  

  2. 17

    I agree strongly with the message that the real “problem” with the Mac experience is that app makers no longer feel compelled to get all of the details correct. Back when I started using OS X (10.1), you could even tell which apps used the Cocoa or Carbon APIs based on small details, like how the text input UI reacted when you entered an accented character. The fact that you had to get to those details to tell meant everything else was consistent.

    There were a few early UI inconsistencies (the brushed metal look was supposed to be for apps that modelled real-life hardware, like DVD Player, QuickTime Player and iTunes, but Finder was brushed metal), but even through the Delicious Generation of apps people were making things that were Mac apps, even if they had a distinctive look.

    But now very few developers, including Apple’s internal developers, make things that are Mac apps. Is it a problem? I’m not sure. It feels wrong to me, but I use a laptop every day and many folks don’t. In five years, will “the desktop” be an important platform, or will Macs be the dev kits for iPads?

    1. 20

      The best thing about Electron is that Linux is finally becoming a first class platform for desktop apps. Slack, Git Kraken, Atom, Mailspring and so on likely would’ve never seen the light of day on Linux if not for Electron. Electron drastically lowers the barrier for writing and maintaining cross-platform applications, and I think that far outweighs its disadvantages. I don’t really see any insurmountable problems with Electron that can’t be addressed in the long run as the adoption grows.

      The reality is that maintaining multiple UIs for different platforms is incredibly expensive, and only a few companies have the resources to dedicate separate development teams for that. The value of having a common runtime that works reasonably well on all platforms can’t be overstated in my opinion. This is especially important for niche platforms like Linux that were traditionally overlooked by many companies.

      1. 24

        I think a more accurate description would be that Electron makes every platform second class.

        It is certainly more egalitarian and even an improvement for platforms previously overlooked, but better than before is not necessarily good.

        1. 2

          On the other hand, if the web stack becomes the standard then all the platforms improve together in the long run.

        2. 15

          In a better universe there would be no reason to maintain cross-platform apps. Ideally we would use independently maintained platform tailored apps talking to common protocols. Like, a hypothetical Ubuntu-native VoIP app that could talk to Skype on Windows. Protocols as the point of commonality is far more desirable than a UI toolkit, because a common UI toolkit means that every app works in its own peculiar way on every platform, which sucks.

          Unfortunately, we’re living in this universe…

          1. 4

            Most people prefer applications that are unconstrained by stagnant standards. For example, consider Slack or Discord versus IRC, or web forums versus newsgroups or mailing lists.

            At least when the applications are open-source and API-driven, there’s hope for alternative clients for those who want or need them.

            1. 3

              Most people prefer applications that are unconstrained by stagnant standards.

              That’s an interesting thought, thanks.

              Although I still think that a single entity evolving a standard is better than every app inventing their own UI conventions.

        3. 5

          The arrival of Flutter apps on the desktop will eventually have a significant impact on both Electron and native apps in my opinion. Graphically rich, performant applications on multiple platforms from one codebase is an attractive vision.

          1. 5

            The “Marzipan” apps on MacOS 10.14 Mojave [..] are dreadfully bad apps. [..] I honestly don’t understand how Apple decided it was OK to ship these apps.

            and

            Things like this are canaries in the coal mine regarding the state of the Mac.

            I’m sorry but that canary has been dead for about 10 years now. Right after Apple released iOS, OS X ceased to be a priority for them and Mac turned into a pretty dreadful thing pretty quickly. I’m honestly surprised that despite all evidence, John Gruber still believes that things are only about to get bad every time there’s a new development in this direction.

            1. 13

              but Electron is without question a scourge

              Well, how about a big F* you?

              Sorry for the swear words, but while I agree that in general Electron is not great, calling it a scourge is incredibly offensive towards those who chose to develop with Electron, and have good reasons for it. Sometimes writing native apps is cost-prohibitive, and you’re better off with an Electron app that looks a bit out of place, than have no app at all. It’s cool for you to be smug and elitist and complain about stuff not following the HIG on every single platform the app is developed for, but have you ever thought about the cost of doing so? Yeah, big companies may be able to shell out enough money and pay developers to create native apps and follow the platform’s HIG, but not everyone’s a big business. I dare say the vast majority of app developers aren’t. By hating on Electron and anyone who doesn’t polish their app perfectly, you’re alienating a whole lot of developers.

              Learn to see the other side of the coin already.

              (I ranted about this on my blog recently, also explaining why Electron was chosen over developing native apps.)

              1. 37

                complain about stuff not following the HIG on every single platform the app is developed for

                Do you know why human interface guidelines exist? They exist because humanity is imperfect and accessibility is really important.

                Electron apps are not accessible. They’re often completely opaque to a screen reader, they’re often completely opaque to assistive speech recognition software, they often don’t respond properly to keyboard navigation and if they do, they don’t behave as expected. They don’t often respond properly to text scaling, they don’t understand system-wide options like increased contrast or reduce motion.

                To whole classes of users, your Electron app is worthless and unusable. What are we supposed to do? Congratulate you on your accomplishment? You need to shrink your ego and take some time to understand real world users, rather than throwing words around like “smug” and “elitist”.

                Also your blog post doesn’t even mention the word “accessibility” once. How disappointing.

                By hating on Electron and anyone who doesn’t polish their app perfectly, you’re alienating a whole lot of developers.

                At the risk of sounding controversial, what does it matter if we alienate some developers? Developers should do better. They need to do better. I’m fine with alienating developers for good reasons.

                Electron is not the answer. It’s a shortcut at best and a bandaid at worst. This isn’t a secret, so there’s really no point in acting surprised that people don’t agree with your choices.

                1. 10

                  I think that we who care about accessibility need to avoid taking a condemning, vitriolic tone, and meet developers where they are. That way, we’ll be more likely to get results, rather than just alienating a large and growing group of people. It’s true that Electron apps often have accessibility problems. But I believe we can improve that situation without calling for people to throw out Electron entirely. Frankly, we need access to these apps more than these developers need us. So we have to work with what we’ve got.

                  No, I haven’t always lived up to this ideal when communicating with mainstream developers, and I’m sorry about that.

                  1. 14

                    I think that we who care about accessibility need to avoid taking a condemning, vitriolic tone, and meet developers where they are.

                    The problem is that these developers don’t want to meet anywhere else - that’s how we arrived at Electron apps in the first place. It’s the easy way out.

                    1. 4

                      Most trends in IT are driven by herd mentality, familiarity, leveraging ecosystems, and marketing by companies. All of these seem to contribute to Electron use just like they contributed to Java and .NET getting big. They sucked, too, compared to some prior languages. So, it’s best to identify what they’re using within what ecosystems to find an alternative that’s better while still being familiar.

                      Then, see how many move to it or don’t for whatever reasons. Iterate from there.

                      1. 1

                        Developers don’t want to listen because by the time they start publishing something (based on Electron, for a number of reasons), what they meet with is pure, unconditional hate towards the technology (and not just the tech! towards developers using said tech, too!) that enabled them. It is not surprising they don’t want to listen to the haters anymore.

                        You’re alienating new developers with this kind of mentality too, who would be willing to listen to your concerns. You’re alienating them, because of the sins of their fathers, so to say. I’m not surprised noone cares about accessibility, to be honest. When we’re told the world would be better off without developers like us, we’re not going to be interested in working towards better accessibility.

                    2. 5

                      Do you know why human interface guidelines exist? They exist because humanity is imperfect and accessibility is really important.

                      I’m aware, thank you. I’m very much aware that Electron is not… great, for many reasons. That’s still not a reason to unconditionally call it a scourge.

                      To whole classes of users, your Electron app is worthless and unusable. What are we supposed to do? Congratulate you on your accomplishment? You need to shrink your ego and take some time to understand real world users, rather than throwing words around like “smug” and “elitist”.

                      You, like the article author, ignore circumstances, and generalize. Yes, my electron app is going to be useless for anyone using a screen reader. It will be useless for a whole lot of people. However, it will exist, and hundreds of people will be able to use it. Without Electron, it wouldn’t exist. So ask yourself this, which is better: an application that is not usable by some people, but makes the life of the vast majority of its intended audience easier; or an application that does not exist?

                      Here’s the situation: there’s a keyboard with open source firmware. Right now, to change the layout, you need to edit the firmware source, compile a new one, and upload it to your keyboard. While we tried to make the process easy, it’s… not very friendly, and never going to be. So I’m building an application that lets you do this from a GUI, with no need for a compiler or anything else but the app itself. I develop on Linux, because that’s what I have most experience with. Our customers are usually on Windows or Mac, though. With Electron, I was able to create a useful application, that helps users. Without it, if I had to go native, I wouldn’t even start, because I lack the time and resources to go that route. For people that can’t use the Electron app, there are other ways to tweak their keyboard. The protocol the GUI talks can be implemented by any other app too (I have an Emacs package that talks to it, too). So people who can’t use the Electron app, have other choices.

                      So, thank you, I do understand real world users. That is why I chose Electron. Because I made my due diligence, and concluded that despite all its shortcomings, Electron is still my best bet. Stop smugly throwing around Electron hate when you haven’t considered the circumstances.

                      At the risk of sounding controversial, what does it matter if we alienate some developers? Developers should do better. They need to do better. I’m fine with alienating developers for good reasons.

                      Well, for one, a lot of our customers would be deeply disappointed if they weren’t able to use the GUI configurator I built on Electron. “Developers should do better”. Well, come here and do my job then. Get the same functionality into the hands of customers without using Electron. I’ll wait (they won’t).

                      Electron is not the answer. It’s a shortcut at best and a bandaid at worst. This isn’t a secret, so there’s really no point in acting surprised that people don’t agree with your choices.

                      I agree it is not the best, and I’m not surprised people disagree with my use of it. I can even respect that, and have no problems with it. What I have problems with, is people calling Electron a scourge, and asserting that native is always best, and that anyone who doesn’t follow the HIG of a given platform “should do better”. I have a problem with people ignoring any and all circumstances, the reason why Electron was chosen for a particular product, and unconditionally proclaiming that the developers “should do better”. I have a problem with people who assert that alienating developers because they don’t (or can’t) write apps that match their idealistic desires is acceptable.

                      Before writing off something completely, consider the circumstances, the whys. You may be surprised. You see, life is full of compromises, and so is software development. Sometimes you have to sacrifice accessibility, or native feel, or what have you, in order to ship something to the majority of your customers. I’d love to be able to support everyone, and write lighter, better apps, but I do not have the resources. People calling the technology that enables what I do a scourge, hurts. People asserting that I should do better, hurts.

                      Until people stop ignoring these, I will call them elitist and smug. Because they assume others that chose Electron have the privilege of being able to chose something else, the resources to “do better”. Most often, they do not.

                      1. 1

                        Electron apps are not accessible. They’re often completely opaque to a screen reader, they’re often completely opaque to assistive speech recognition software, they often don’t respond properly to keyboard navigation and if they do, they don’t behave as expected.

                        I haven’t written an electron app, but I’ve done a fair share of web UI programming with accessibility in mind. You communicate with platform accessibility APIs through web APIs. It’s not technically complicated, but it does require some domain expertise.

                        Does it work the same way on electron?

                        1. 8

                          Electron embeds chromium, which doesn’t connect to the native a11y APIs (MS are working to fix this on windows).

                          As a result electron apps are as inaccessible as chrome (screenreader users tend to use IE, safari or firefox).

                          1. 3

                            Huh, this is a real surprise if true.. i tested our web UI with screen readers across macOS and windows in chrome, FF, and IE, and the only problems that occurred were due to bad markup or the occasional platform bug. Reaching the accessibility API was not a problem i ran into with chrome

                            1. 2

                              Chrome is definitely accessible to screen readers. I’ve spent more time that I’d like getting JAWS to read consistently across IE, FF and Chrome. From memory, Chrome was generally the most well behaved.

                        2. 7

                          When Apple was first developing the Mac, they did a lot of research into human/computer interaction and one of the results was to apply a consistent interface on the system as a whole. This lead to every app having the same menu structure (for the most part, the system menu (the Apple logo) was first, “File” next, then “Edit”) and under these standard menus, the actions were largly the same and in the same order. This would lower training costs and if a user found themselves in a new app, they could at least expect some consistency in actions.

                          I’ve been using Linux as a desktop since the mid-90s (and Unix in general since 1989) and to say there’s a lack of consistency in UI is an understatement. Some apps have a menu bar at the top of the window (Macs menu bars are always at the top of the screen per Fitt’s Law), some you have to hold Ctrl down and press the mouse button, some the right mouse button will cause a pop-up menu. I’ve been able to navigate these inconsistencies, but I’m still annoyed by them.

                          Further more, I’m used to the CLI, and yet even there, the general UI (the command set, the options to each command) still surprises me. I wrote about the consistency of GUIs and the inconsistencies I found on the Unix CLI over the years and while I still prefer the CLI over the GUI [1], I can see where the consistency of the Mac GUI makes for a much better experience for many.

                          As I’ve stated, I think it’s wonderful that PHP has enabled people to create the dynamic websites they envision, but I wouldn’t want to use the resulting code personally.

                          [1] One can program the CLI to do repetitive tasks much easier than one can do the same for any of today’s GUIs. There have been some attempts over the years to script the GUI (Rexx, AppleTalk) but it still takes more deliberate action than just writing a for loop at the command prompt for a one-off type of job.

                          1. 5

                            I think the case where it makes sense to go Electron is not the point of Gruber’s rant. The point was that many, many developers today are easier with writing Electron app and using it on all platforms instead of putting time and effort into polished Cocoa apps.

                            The core of this blog post is how Mac really was different in terms of UI/UX. During the time I started using Mac (10.5) it was really differentiating itself by having “different”, “better looking and feeling” apps. Electron definitely made Mac feel less unique. Critics were pointed towards macOS app developers. Don’t get so offended by simple blogpost. Your reasons are fine, but that simply isn’t the case most of the time. Most people decide to go with electron because of plethora of mediocre JS devs, that can chunk out a lot of code that does something, and then you get slow junk like UX. In minds of 2000s Apple fans that is a big no.

                            Have a nice day, and move on.

                            1. 7

                              I think the case where it makes sense to go Electron is not the point of Gruber’s rant.

                              Correct.

                              The point was that many, many developers today are easier with writing Electron app

                              Incorrect.

                              Please just read the article.

                              His point is what he says: it is bad news for the Mac platform that un-Mac-like apps far worse than those that were roundly rejected 15 years ago are now tolerated by today’s Mac users.

                              It happens to be the case that Electron is the technology of choice for multiple prominent sub-par apps; that’s a simple statement of fact. (It also isn’t purely coincidental, which is why I agree with his characterisation of Electron as a scourge. If someone like @algernon who builds apps for Electron is bent on interpreting those statements as a judgement of their own personal merit, well… be my guest?) But Electron is not singled out: Marzipan gets a mention in the same vein. On top of that, Gruber also points out the new Mac App Store app, which uses neither. The particular technologies or their individual merits are not his point.

                              His point is, again, that a Mac userbase which doesn’t care about consistency spells trouble for the Mac platform.

                              1. 2

                                Marzipan gets a mention in the same vein.

                                Electron is the only one that gets called a scourge, and is singled out in the very beginning. It’s even in the title. It’s even the first sentence, which then continues: “because the Mac is the platform that attracts people who care”

                                If that’s not an elitist smug, I haven’t seen any.

                                1. 6

                                  Once upon a time, Mac users were a ridiculed minority. In those days, Microsoft-powered PCs were better in just about every way. They were much faster, they had a more technically advanced OS (even crappy Win95 was far ahead of MacOS Classic), they had more applications, they had boatloads of games… just about every reason to pick one computer over another pointed in the direction of a Microsoft PC. You had to be special kind of kook to want a Mac regardless. It was inferior to a PC in basically every dimension. The one reason to pick a Mac over the PC was the depth of consistency and care in the UI design of its software. Only users who cared about that enough to accept the mile-long list of tradeoffs went for the Mac.

                                  1. 1

                                    Elitism is often a good thing. It’s how we get from the mundane to the truly excellent.

                                2. 3

                                  The point was that many, many developers today are easier with writing Electron app and using it on all platforms instead of putting time and effort into polished Cocoa apps.

                                  My beef is not with the author wishing for apps that would look more native - I share the same wish. My beef is with him calling Electron “without a question a scourge”. How about I said MacOS is without a question a scourge, for it jails you in its walled garden? You’d be rightly upset.

                                  There’s a big difference between wishing apps would be more polished on a particular platform, and between calling a technology (and by extension, developers who chose to use it) a scourge. It reeks from privileged elitism, and failure to understand why people go with Electron.

                                  Most people decide to go with electron because of plethora of mediocre JS devs, that can chunk out a lot of code that does something

                                  No. Most people decide to go with Electron because it provides a much better cross-platform environment than anything else. Please don’t call a whole bunch of people “mediocre JS devs”, unless you have solid data to back that up. Just because it is JS and “web stuff” doesn’t mean the people who develop it are any less smarter than native app developers. Can we stop this “only mediocre people write JS/PHP/whatever” bullshit?

                                  1. 13

                                    There are more bad developers writing webshit because there are more devs writing webshit period.

                                    Native apps tend to outperform Electron apps and use less memory, because to do the same things that don’t bring in a browser and language runtime.

                                    Elitism is, in this case, warranted. The only really performant app (usually) in Electron I’ve seen is VSCode, because MS really does have sharp people working in a domain they’ve been leaders in for decades.

                                    1. 7

                                      There seems to be a shift towards less attention paid, and value given, to the experience of the user. This makes me very sad.

                                      When people talk about why they use Electron, they always phrase it in terms of “developer productivity”, and that’s where I find the most elitist bullshit to be. Developers talk about using Electron so they didn’t have to learn a new platform, or so they only had to test it in one place, or it was faster. They talk about lower development costs (which they wildly overstate, in my experience).

                                      But the questions I’d like use to start asking: what are the costs of the shit user experience? What are the costs of people having to learn new tools that don’t behave quite like the others? When we save money and time on development that money and time is saved once, but when we save time for our users, it’s saved repeatedly.

                                      Maybe calling Electron shit is elitist bullshit. But I’ll take that over having contempt for one’s users.

                                      1. 3

                                        Contempt is a strong word. Would all these people be users in the first place if the app doesn’t exist for their platform? Go ahead and write a native Cocoa app for OSX, but that sure feels like contempt for Windows or Linux users. “Buy a new machine to use my stuff” vs. “deal with menus in the wrong order”?

                                        1. 1

                                          I never said “buy a new machine to use my stuff.”

                                          From extensive experience: for most small applications, I can develop them natively on Mac, Windows, and Linux* faster than someone can develop the same thing with similar quality using a cross platform thing.

                                          (*) “native” on Linux is less of a sticky thing that on Mac and Windows.

                                          1. 3

                                            Here’s a challenge for you: https://github.com/keyboardio/chrysalis-bundle-keyboardio (demo here).

                                            Go do something like that natively for Mac and Windows. It’s a small app, some 3700 lines of JS code with comments. You do that, and I promise I’ll never write an Electron app ever again. You can make the world a better place!

                                            1. 1

                                              Thank you for your interest in my consulting services.

                                              1. 4

                                                Thought so.

                                                FWIW, the app, like many Electron apps, were originally built in my unpaid free time. Complain about Electron apps once you built the same stuff under the same conditions.

                                    2. 9

                                      How about I said MacOS is without a question a scourge, for it jails you in its walled garden? You’d be rightly upset.

                                      I wouldn’t. I might point out that you absolutely can bypass their walled garden, on MacOS, but those objections for iOS are not only valid, but are honestly concerning.

                                      Electron is a scourge. iOS is a scourge. Facebook is a scourge. The feature creep within the browser is absolutely a scourge. There are loads of scourges. This is not an exhaustive list. I pray daily for a solar flare which delivers enough of an EMP that it utterly destroys the entire technology landscape, and gives us an opportunity to rebuild it from the invention of fire onwards, because we’ve fucked up, and our technology is bad.

                                      And I say this because I want technology to be better. The Web is an awfully complicated way to render a UI. Our systems are overly dependent on a few corporations who don’t have our best interests at heart. Our computers are slower to do less work than they did two decades ago. Pretty much the only way anybody makes money in tech anymore is by abusing their users (see: Google, Facebook, etc). Or their employees (see: Uber). Or both (see: Amazon).

                                      1. 1

                                        In the EMP scenario we’d be too busy trying to get essentials back up to care about doing it right

                                      2. 7

                                        You’d be rightly upset.

                                        No, I wouldn’t be rightly upset. I am not the technologies I use, and neither is that true for you.

                                        Electron is the technology used in multiple highly prominent applications that are written with little or no regard to platform conventions. Their success is bad for the native platforms. Those are statements of fact. If you have good reasons to use Electron, then there is no need for you to relate those facts to yourself and take them as a statement of your merits as an individual.

                                        1. 7

                                          Agree. The notion that your identity is somehow linked with the tools you use is toxic. It is fine to like your tools, but once you get comfy with them you should be pushing beyond them to expand your taste.

                                          1. 5

                                            your identity is somehow linked with the tools you use

                                            I used QBasic, Visual Basic 6, and later FreeBASIC. If my tools define me, I feel like such a shallow person with no depth or skill. Such a sinking feeling. I think I’m going to re-install SPARK Ada and buy Matlab to feel like a mathematician. Yeah, that will be better… ;)

                                    3. 4

                                      Thanks for sharing the blog post.

                                      I think your case is quite different from, say, Slack. I have worked on major cross-platform apps at a company, and it’s not a huge deal, when you have a few people working on it, whose collective knowledge covers those platforms. All apps used the same core libraries for the non-UI parts, and each app added native UI on top. A company with tens, or even hundreds of well-paid developers should be able to do that, if they care at all about accessibility, performance (which is a different kind of accessibility issue,) resource usage, and all those things.

                                      1. 2

                                        It is still a big deal, even for a larger company, because there’s a huge difference between employing N developers to develop a single cross-platform application (with some of them specializing in one platform or the other), and between employing a set of developers to create native applications, and a set for the core libraries. There may be overlap between them, but chances are that someone who’s good at developing for Windows or OSX would not be happy with developing for Linux. So you end up employing more people, paying more, diverging UIs, for what? Paying customers will use the Electron app just as well, so what’s the point of going native and increasing costs?

                                        Yeah, they should be able to do that, yes, it would improve the user experience, yes, it would be better in almost every possible way. Yet, the benefits for the company are miniscule, most of the time. In the case of Slack, for example, or twitter, a uniform experience across devices is much more important than native feel. It’s easier to document, easier to troubleshoot, and easier for people who hop between devices: it’s the same everywhere. That’s quite a big benefit, but goes very much against making the apps feel native. And if you forego native feel, but still develop a native application that looks and behaves like on any other platform (good luck with that, by the way), then the benefits of native boil down to being less resource hungry. In the vast majority of cases, that is simply not worth the cost of the development and maintenance burden.

                                        In the age of mobile devices, I do not feel that apps looking “native” is any benefit at all, but that’s a different topic.

                                        1. 5

                                          I built them in the past myself. I’ve read write-ups about what it takes for others. If designing program right, most of the code is shared between the different platforms. The things that are different are mostly in the front-end that calls the common code. A lot of that can be automated to a degree, too, after design is laid out. For main three platforms, it basically took a max of three people two of whom only worked on UI stuff here and there mostly focused on shared code. That isn’t the minimum either: it can be lower if you have 1-2 developers that are experts in more than one platform. In mobile, many people probably know both iOS and Android.

                                          The cross-platform part will be a small part of the app’s overall cost in most cases. It will mostly be in design/UI, too. That’s worth investing in anyway, though. :)

                                          1. 5

                                            Our experience clearly differ then. I worked for companies that made native apps for the major platforms (mobile included), and each team was 10+ people at a minimum, with little code shared (the common code was behind an API, so there’s that, but the apps itself had virtually no code in common). Not to mention that the UIs differed a lot, because they were made to feel native. Different UIs, different designs, different bugs, different things to document and support. A whole lot of time was wasted on bridging the gaps.

                                            If they made the apps feel less native, and have a common look across platforms, then indeed, it could have been done with fewer people. You’d have to fight the native widgets then, though. Or use a cross-platform widget library. Or write your own. And the writer of the article would then complain loudly, and proclaim that multi-platform apps are the worst that could happen to the Mac (paraphrasing), because they don’t follow the HIG, and developers nowadays just don’t care.

                                            1. 3

                                              each team was 10+ people at a minimum, with little code shared (the common code was behind an API, so there’s that, but the apps itself had virtually no code in common).

                                              “If they made the apps feel less native, and have a common look across platforms, then indeed, it could have been done with fewer people. “

                                              I said native look on multiple platforms minimizing cost. You example sounds like something about the company rather than an inherent property of cross-platform. You usually need at least one person per platform, esp UI and style experts, but most of the code can be reused. The UI’s just call into it. They’ll have some of their own code, too, for functions specific to that platform. Mostly portable. Just sounds like the company didn’t want to do it that way, didn’t know how, or maybe couldn’t due to constraints from legacy decisions.

                                              1. 1

                                                My experience mirrors yours almost exactly.

                                        2. 3

                                          What about sciter? Companies doing stuff like anti-virus have been using it for a long time with way, way, less, resource use. It’s licensing scheme looks like something other vendors should copy. Here’s a comparison claiming a simple editor is 2MB in sciter vs 100+ in Electron just because it brings in less baggage.

                                          How many people using Electron could use the free, binary version of sciter? And how many companies using Electron could afford $310 per year? I mean, that’s within reach of startups and micro-businesses, yeah? I’m asking because you said it’s use Electron or not cross-platform at all cuz too costly/difficult. I’m making it easy by using a comparable offering rather than, say, Lazarus w/ Free Pascal or otherwise non-mainstream languages. :)

                                          Note: I also have a feeling lots of people just don’t know about this product, too.

                                          1. 4

                                            I actually evaluated sciter when I got frustrated with developing in JS, and wanted to go native. For my use, it wasn’t an option, because the app I write is free software, and therefore so much be its dependencies. For closed-source use, it’s probably fine. Though, you’d still have to write plenty of platform-specific code. AV companies already do that, but companies that start off with a web-based service, and develop an application later do not have that platform-specific code already built. What they have, is plenty of JavaScript and web stuff. Putting that into Electron is considerably easier, and you end up with very similar UX, with little effort, because you’re targeting a browser still.

                                            Oh, and:

                                            With Sciter, changing the front end of your application involves just altering the styles (CSS), and probably a couple of scripts that do animations.

                                            Yeaah… no. That’s not how changing the UX works. It’s a tad more involved than that. Not sure I’d be willing to trust my UI on an offering that basically tells me that changing between, say, Metro and Material UI is a matter of CSS and animations. (Hint: it’s much more involved than that.)

                                            Since we’ve already decided we’re not going to follow any platform HIGs (and thus, have the article author proclaim we’re a scurge), what’s the point of going native? Less resource use? Why is that worth it for the company? People use the app as it is, otherwise they wouldn’t be writing an app. Writing native, and making it look and behave the same has non-negligible cost, but little to no benefits. Faster and lighter is in many cases not a goal worth pursuing, because customers buy the thing anyway. (I’d love if that wouldn’t be the case, but in many cases, it is.)

                                            1. 1

                                              I appreciate the balanced review of sciter. That makes sense.

                                        3. 2

                                          I don’t see anything that prevents anyone from fixing these issues on Electron. What is all flame war about?

                                          I am not happy with Electron apps right now, but it does not mean that they cannot get better. The language, the VM, and the SDK, they are all open source and widely adopted, actively developed.

                                          Isn’t it exciting that in the near future whether we use Windows, MacOS, Linux, or BSD won’t matter anymore?

                                          1. 4

                                            IMO, sure, but only if that doesn’t mean that they’re all dragged down to the lowest common denominator.

                                            1. 3

                                              I think VSCode proves that with enough effort, Electron apps can be very, very good. I’m not sure how much of the MS tweaking can be applied generically to the platform itself, though.

                                              1. 1

                                                Isn’t it exciting that in the near future whether we use Windows, MacOS, Linux, or BSD won’t matter anymore?

                                                There is no official support for any BSD in Electron…

                                              2. 2

                                                The worst thing to happen to the Mac is Steve Jobs dying (okay, it’s also the worst thing to happen to Apple). He was the only one with enough clout and confidence to say “we aren’t shipping this until it’s fixed.”

                                                1. 18

                                                  Unfortunately if you count unfinished things that they shipped on his watch, you find a different story: the cracking-case G4 cube, Mac OS X 10.0, the iWork rewrite that removed features from existing documents, Apple Maps, iAd (we’re not an advertising company), the iPhone 4, and MobileMe are the ones I can think of now.

                                                  I’m not arguing that quality is better now (I think it isn’t), but I will argue that Steve Jobs is not the patron saint of problem-free technology products. Apple has, like all other technology companies, shipped things that they thought (or maybe hoped) were good enough, under all of their CEOs.

                                                  1. 1

                                                    Didn’t they charge for the updates to fix their broken software on top of it while Windows did them for free? I could be misremembering but I think Mac folks told me that.

                                                    1. 4

                                                      All I can recall is charging for a patch to enable 802.11n wireless (as opposed to draft) but the explanation was that sarbanes oxley prohibited delivering a “new” product because they already booked the revenue for it, but then the law was clarified and software updates are ok now.

                                                  2. 2

                                                    It’s not so much about Steve Jobs, but about people that were behind products and development process, which of course were brought and held together by Steve Jobs. I would say that departure of Bob Mansfield is one of the mayor impacts on Apple’s final products.

                                                  3. 1

                                                    I just hope that progressive web apps end electron at least so I don’t have worry about 5 different–and the unmergable memory pages of, chromium versions.

                                                    In terms of “native feel” I don’t think such a thing will ever be attainable as long as every platform has a special snowflake toolkit or a plethora of alternatives.