While I think downvoting this comment is unfair, it should be noted that if you care about your OS and tools being open-source, you probably shouldn’t be running OSX in the first place.
This all or nothing mentality doesn't often end up being useful in practice. For example, try doing some deep learning on Linux with CUDA without installing closed source NVIDIA drivers. The important detail is who provided the software and in this case most people are going to trust Apple or NVIDIA significantly more than a small third-party software developer they've never heard of.
(FWIW many of us tried running Linux as a desktop OS but found the UI and driver situation to be abysmal so it continues to just power our servers)
Linux is actually good now for desktop use. I install it (= Fedora or Antergos w/ GNOME) on everything now and it just works(tm).
If I got a dollar every time I heard this... I have a Linux machine at home, but GNOME still crashes 50% of the time when I switch the screen off/on. If it switches on without crashing, many applications are LoDPI until I restart them.
The machine doesn't always come back properly from sleep. Sometimes there are display artifacts, sometimes I cannot show windows of running applications.
This is all on well-supported hardware (RHEL-certified Dell Precision, an AMD GPU supported by amdgpu, etc.).
Linux has become pretty great, but it's definitely not a 'just works' experience on a lot of hardware and for a lot of uses.
While Windows keeps crashing and is glitchy, laggy, likes to crash and is generally disgusting and mac OS (Hackintosh) is unreliable and with a not very customizable UI, Elementary OS works very well on the Xiaomi Mi Notekook Air 13 (what a name).
If something breaks I can fix it. I can easily schedule jobs (tried it on Windows, fucking night mare), define my own trackpad gestures, lower the screen brightness to an acceptable level ...
Linux, with all of its flaws, is by far the least frustrating OS, imo.
You’re trying to run MacOS on unsupported hardware and configuring it yourself (instead of letting Apple do it as they always do) which is the only reason it’s more frustrating than Linux. Buying Linux vs. MacOS directly from the manufacturer is a different story and for most users MacOS clearly wins.
It simply doesn’t have the polish, or flexibility that I’d need to effectively stay on it for long. I always end up spending longggg times customizing things only to feel little integration, little support, and a lack of decent third party software. Sketch is Mac only. Spotify barely has a Linux client. Certainly no photoshop, medocire native email clients, zero IM that integrate with any networks I use (sms/iMessage), etc etc. Yes there’s lots of options as a webpage, but that’s a shallow substitute.
If you happen to need only the stuff that’s available, and your workflow is largely CLI-driven, then it’s probably fine. Otherwise it’s macOS for me.
Is it an Electron app now? Last time I poked around on macOS (admittedly, about a year ago), it didn’t look like an Electron app to me, but it most definitely was using a web view of some kind.
Sounds like you're doing graphic design. Unless you're doing animation (in which case https://krita.org/en/ is much better than any Adobe thing I've used) you're definitely better off with a Mac.
Depends on what's required. I installed Linux (Ubuntu and Xubuntu) on an old Mac, and found it extremely hard to get file sharing (with SMB shares) done. I also gave up on installing a VNC server. For both these, I was looking for GUI based options. On the Mac (for longer than a decade), one just goes to Apple menu->System Preferences->Sharing and then turn on screen sharing, file sharing, etc., very easily.
I also had to configure certain other things using the shell. I personally didn't experience Linux being adequate for desktop use.
I’m not going to say I don’t believe you, but I find comments like this one (which I see surprisingly often) funny because I’ve never not had it working perfectly, OOB, with no effort on my own, for as long as I can remember
Instead of buying whatever windows/mac computer looks like a good deal and hoping it runs linux buy with linux in mind and you will have an easier experience.
I don’t disagree with your point but it only addresses part of my problem.
For one thing, the Linux UIs like GNOME and KDE are terrible. They do the job for some people but I would willingly pay more to use OS X without hesitation.
I also have experienced many situations like the following, although typically not quite as extreme...
A coworker decided to purchase a Lenovo with Linux as a developer at a company where everyone else had MacBook Pros. He was very smug the first day or two about the specs of his laptop and it being open source Linux instead of OS X. Within a few days he was having such major issues connecting it to a 30” monitor (along with a few other issues related to Bluetooth/USB as far as I recall) that he called it a lemon and returned it for a MacBook.
Don’t get me wrong, I’m waiting for the amazing developer Linux laptop that replaces my MacBook Pro. But I just don’t see it coming for a while.
If you define not working like OSX I guess its terrible seems pretty functional to me.
Personally I prefer i3wm and keyboard driven workflow and find the way mac handles multi monitor loathsome. This doesn't mean that it is objectively bad it means I find it unsuitable as a matter of taste.
The fact that he had hardware issues hardly means linux is terrible in fact its a fairly nonsensical way to qualify an entire OS.
You could I'm sure find someone who bought a singular mac and had issues with it in a workplace full of linux users and try the equally nonsensical if opposite conclusion.
As I said, the point wasn't that one person had hardware issues. Its that I constantly see people buy Linux laptops and have hardware/software issues as the two are quite related on Linux due to bad driver support and inconsistent hardware configurations. I relayed an anecdote that summarizes many of the problems I've seen. This one involved an extremely senior engineer with successful exits who was beside himself trying to make a brand new Linux laptop work.
Here is another anecdote, no hardware involved. Ubuntu 10.04 to 12.04 upgrade was busted and several friends couldn't boot at all after upgrading via the GUI option. Sure that could happen on a Mac but its way, way less common.
Since day one, every time I boot up there has been some sort of display error. The networking has never worked correctly.
Desktop linux needs to do what Apple did. Pick a hand full of laptops and desktops (or just ditch desktops) with some combination of hardware and just support those.
Until then, it's always going to be an "it works for me" situation.
Your methodology appears sound but if I can offer you some improvements.
Ubuntu's non LTS releases seem to suffer far more issues comparatively. I would suggest you stick with LTS releases and for software which you need more recent versions you may look to individual ppa's for that project. It's a common misconception that such releases are years out of date. By default they are for say the kernel meanwhile user oriented software like say your web browser is easily quite up to date and can even be bleeding edge if you don't mind adding a few repos for the software most important to you.
Unlike Apple there is no singular entity that called Linux that can opt to do anything like say abandon the desktop. This is rendered doubly insane when you realize that desktop workstations actually work fantastic right now and are an area where Linux makes a ton of sense. The hardware is standard and interchangeable. Unlike a laptop if a part does not have good Linux support its trivial to swap out just that part for something that does. Business users care about stability and a small finite number of apps. While users are turning to laptops or even mobiles some workstation users will continue to need the significant horsepower that a desktop provides. Triply nutty when you consider that there is nothing whatsoever to be gained by abandoning the desktop. You are thinking in terms of a company narrowing its focus to enable it to devote increased resources to a smaller group of products but desktops but there is nothing about this analogy that actually works.
While improving support for particular laptops is a laudable goal I'm not sure it makes any contextual sense. People working at improving Linux are presumably worried about broad projects and subsystems not device specific hacking. Presumably part of improving things is taking bugs from users but how do you propose they focus on a small subset of laptops? Privilege the tiny subset of bugs that come from those users?
Developers are already paid to focus on particular laptops. Those paid by the oem to support those sold with linux reinstalled. There are dell machines sold with Linux preinstalled as well as a number of smaller vendors. If you really want this to be a happening thing the logical thing to do is to support vendors that sell Linux machines with your money.
In short we can continue improve Linux support on a particular subset of laptops without abandoning the desktop and you can help.
I'm using a HP Elitebook 840 running Ubuntu 16.04 LTS. Didn't work with 14.04 LTS either.
I hope you appreciate that you wrote 4 paragraphs on how I could improve on picking a laptop from the Ubuntu Certified list to have fewer problems.
>If you really want this to be a happening thing the logical thing to do is to support vendors that sell Linux machines with your money.
I've been using Linux since 99. It's never worked without problems. It's been 20 years and I can't pick a laptop off a list of Certified hardware and be confident. I abandoned linux for home use a decade ago which makes me sad because I wish my side project laptop could be a linux machine. Development is so much easier on linux.
Limiting the number of entities you need to trust with closed-source software is perfectly sound wisdom, but that number doesn’t need to be zero to be useful.
That’s not to say Marta’s author has any ill intent, because I doubt he or she does.
I've been using Nimble Commander for a while now. Snappy, similar to Total Commander on Windows and I love it. The developer recently made it open source and published the code on GitHub here: https://github.com/mikekazakov/nimble-commander
I'm not sure if that's the kind of open source you were looking for, but may be worth checking out.
Total Commander is the best piece of software I ever used! I even bought a license for it, they sent me a cd with an installer and the license and I promptly lost it. At some point i even had tc as a replacement for the Windows shell. These days i dont care that much about where my files are, I have a "work" folder and thats that. My music is on spotify, my videos on netflix/amazon prime/now tv, i dont take that many meaningful pictures anymore and I dont see a reason to care about files anymore. But Total Commander was my life for a very long time, along with piracy :-P
If that's the case (and I take it the absence of a source code repository is how you're reaching your conclusion), that's ironic. The Marta webpage linked to this story includes "Because I want to own what I pay money for." yet one never truly owns proprietary software regardless of price; one can never really tell what it is doing, and users are prohibited from inspecting the program to find out, as well as sharing copies (modified for sure, often verbatim copies too).
I agree with you that the open source software has higher level of trust than the proprietary one, but I afraid I just can't maintain Marta steadily for a long time if it was open-sourced (and then forever free). And I doubt the community will do that instead of me. In case if you're interested, I expressed my thoughts about this here [1] earlier.
I do promise, though, that Marta will never collect personal data, display ads or do some other bad things. I just don't see any sense in that, and I'd rather have a good reputation.
Personally, I'm happy to pay for high quality software as I understand someone needs to be paid in order to maintain it, but I like having access to the source in order to reduce risk and maintain more control over my environment.
I think it's possible to release the source and still sell the app, Textual [0] seems like a good example of that. The make my favorite IRC client. Although I'm not sure if the author is that well off or not.
Sidestepping the actual open-source issue, from a morning's use I think you've done a wonderful job with Marta. I'll be more than happy to pay for it when the time comes.
Actually one doesn't need to trust free software because one has the freedom to inspect and modify the software to make it do what the user wants. Whether users choose to do these things is their choice and besides the point; the freedoms remain.
Trust comes into play when users are denied these freedoms; users have nothing else on which to make an informed decision. Therefore users are left to evaluate promises such as the promises you made. Belief in those promises essentially boils down to uninformed trust. It's no secret that many other proprietors do spy on their users (Microsoft purposefully changed Skype protocol to better allow spying, according to http://www.guardian.co.uk/world/2013/jul/11/microsoft-nsa-co...). I'm pretty sure Microsoft wants to have a good reputation despite their behavior.
As you said in the post you linked to, "Although this is my current mindset, it can change in the future.". This could apply to the promises you made (or strong convictions you posted about on github.com) as well.
The question of maintaining Marta is another issue I think is misstated in your response here and on github.com: with free software users can maintain software too. They could even choose to hire you to do the job. And they can certainly host their own discussions regardless of Marta's license. The underlying issue has to do with whether users will be able to make the software they run on their computers do what they want the software to do (which is the issue with which I raised my first post on this thread) and thus retain control over their computer.
Is there a reason to not license it under a dual license?
Have the source code open with a non commercial license and also distribute the binaries under a commercial license.
I’m not sure that it will have a significant impact on the revenue since most people would likely opt out and purchase the binary rather than building it on their own.
According to what I know, in case of desktop applications, the revenue drops almost to zero, unless you have a Free and Pro versions. "Free edition" then lacks almost all features that make the product competitive, and it's not really good.
It's reasonable, though, if your potential user base is totally different. For example, IntelliJ IDEA has a Community edition that lacks enterprise frameworks support, but that's not a problem for "single users" because they basically don't do any enterprise. But it looks like it won't work for Marta.
So I'd rather choose the Sublime Text strategy (though I think that Sublime costs too much).
Don’t listen to them. The HN crowd wants everything to be open-source, but likely wouldn’t want to pay for anything even if it was dual licensed. This seems to be a great product, and you deserve to charge for it and keep it closed source.
I might even suggest running Linux. Oh my :)