All the important parts of Qt are available under LGPLv3. This only makes a difference to you if you want to do Tivoization of your software (Gtk uses LGPLv2, this being the only major difference AFAIK). In which case you still have the option to buy a commercial license.
Yeeeeahhh.... at the moment. They have gotten really aggressive about commercial licensing recently, and even a conservative extrapolation should give one pause.
Today they might limit themselves to forced registration, SEO, and spamming business contacts with carefully crafted statements designed to stir fear, uncertainty, and doubt around free licenses by strongly suggesting (without actually claiming) that commercial use without a commercial license is illegal. But tomorrow? Also, keep in mind that a business partner who isn't already familiar with Qt and LGPL is going to be about 10x more susceptible to the FUD. That's the whole idea.
My guess: 30% chance of an ugly fork and lots of drama in the next few years. Then, absent a change in direction, another 30% chance in the few years after that, and so on.
Yes, there is a lot of FUD about the licensing for Qt when you search online, definitely not helped by how unclear the Qt site's explanation of it is (to me at least)
My boss looks at this page, he sees that open source programs can use Qt for free and that commercial programs need to pay. That's not the case, but the page is carefully worded to prevent him from confidently coming to the correct conclusion.
If this were the extent of the shenanigans, I wouldn't be mad. I like having a "help me sell this to my boss" page. But it isn't the extent of the shenanigans. They went around me to shake someone down on my behalf (as I perceive it). Last time it was my boss. Next time I choose a GUI framework for an open source side project, I'll primarily worry about it being my users.
>he sees that open source programs can use Qt for free and that commercial programs need to pay.
Where does it say that? Can you mention what he is having trouble with? I just took a quick glance at that link in GP and it seems to spell out the obligations of the LGPL pretty clearly on the right side, which are somewhat specific and notably don't include a requirement for your program to be open source. That requirement is only for the GPL components, which is included in the small print on the left.
So, by your own admission, the critical piece of information my boss cares about must be inferred from the fact that it is absent from a sizeable pile of relatively technical details and from the fact that no detail (especially the GPL sub-component callout) implies it in turn.
Making this inference requires you to not only have outside knowledge of open source licenses and the Qt licensing situation, but to be rather confident in said outside knowledge. That's what he had trouble with.
No, my point is that it doesn't have to be inferred, and that the relevant information does seem to be all there. IANAL and this is only based on my own understanding of those licenses, but it seems to do as good a job as a short marketing page probably could. It's not a place where they can reasonably address every common misconception about the LGPL that someone might have. Your boss can't be helped if he is going to assume the worst just from reading what amounts to a single powerpoint slide.
I think that detail about GPL could be easily missed because it's written in small print, but that's a different problem. There are a lot of Qt components so it doesn't make sense to list them all on that page. A full list that can be filtered by LGPL/GPL status is here: https://www.qt.io/features
I'm assuming by outside knowledge you mean the fine print of the license: how is any company supposed to prevent your boss from having to go over that with a lawyer? This isn't even an open source thing, it applies to anyone in the software business. And this is ignoring that the GPL is probably one of the better understood open source licenses at the moment.
> the relevant information does seem to be all there
The key piece of information is absent.
> it seems to do as good a job as a short marketing page probably could
It could have done better in a single short sentence by stating the key piece of information, rather than leaving it to be inferred laboriously from a pile of details.
> It's not a place where they can reasonably address every common misconception
No, but it could have addressed the single largest misconception. It chose not to.
> So Macs are out of question... Can't be signed can they?
... you know that even Apple ships some GPL software on every Mac right ? GPL is 100% fine mac hardware. There are even GPL apps on the appstore. Signing does not prevent you to upload a new version of the app to your own device.
> Am very surprised, after reports a while back of GPL (etc) apps not being approved.
it was an issue until Xcode 7 : prior to that you had to pay 99$ to Apple for the right of uploading something to your own iDevice. Since Xcode 7 this is not necessary anymore.
> Apple ships some GPL software on every Mac right [..] re are even GPL apps on the appstore.
To my knowlage and IANAL, this is only possible with older pre-v3 licenses. v3 licenses specifically prohibit tevoisation, something that Apple's App store TOS effectively mandates by placing restrictions on what App users are allowed to do. I believe this is why apple doesn't ship recent versions of Bash.
While users are still allowed to run software of their choosing on the Mac I would think it would be perfectly acceptable to provide software that could be built and run outside of the mac app store.
Qt being C++ also makes creating language bindings often harder so not all languages will let you build a Qt interface. GTK is C so it's more straightforward.
Both reasons are why I'm trying to develop an OSS alternative with easiness & platform-independence of the web but much lower system requirements than what electron has.
This is ultimately why I went with gtk in a recent Rust project. Qt bindings are a pain, and the license stuff was coming out. But most of the native Rust GUI libraries don't even try. Maybe its a maturity thing and they'll get there but I picked GTK for now
Note though that GTK has no accessibility support on Windows and Mac. If you're only targeting GNOME, then that's fine. For something that's more or less accessible cross-platform, I'd go with wxWidgets or something Web-based.
Borland made a lot of mistakes with Delphi, but the best part of delphi was VCL was shipped with the source, it was a great introduction for the Win32 API world.
Oscar Toledo is very talented to write minimal code, but it does mean either his entire family is a society of wizards or they can build a computer or develop browsers, these "achievements" are urban legends without any proof.
> Unfortunately, they live on the outskirts of Mexico City, not Sunnyvale or Boston, so the public accounts of their achievements have been mostly written by vulgar journalists without even rudimentary knowledge of programming or electronics.
It would be a valid reason in '80s
One of the first web browsers in the 90s was written by one dude in four days, with the whole GUI. A browser doesn't necessarily mean CSS and JS.
Sometime in the past couple months, there was a link on HN to a project also by one dude, who's implementing a CPU on breadboards, with wires. (Can probably be found by the score filter from the ‘undocumented features’ of HN: the post got >2000 points.)
Also, implementing browsers (or operating systems, or cpus) isn't about needing to write amounts of code that would take a long time write - it's about whether you know the subject area and do you already have a plan/architecture in your head that you are going to create.
All of that sounds very interesting, but as edgarvm points out, it's not exactly evidence. Reads more like a product brochure than a research paper detailing how their claims are possible.
Well, if they didn't build their own computers, they certainly went to a lot of trouble to fake up photos of the family members doing electronic prototypes. And I think some of those photos may be older than PhotoShop, so faking them would have been more work than just building a home computer, though validating the photos’ apparent age might require digging up 1980s Mexican computer hobbyist magazines. (There's a convenient list of computer-magazine articles stretching back to 1981 on the Biyubi website; if you're in Mexico, maybe you could try to find them in a library or a used bookstore.)
Also they offer a course in building your own electronics that you can sign up for if you go to their workshop. I can't go to Mexico so I haven't signed up for it; it's at least theoretically possible that the course doesn't really exist and will be canceled if you do sign up for it. But if that doesn't happen you can go there and take the course and see the G11V3 and presumably Fénix and Biyubi in person. I've seen comments in news-website comment threads from people who claim to have done so and claim that it's real, and no comments from people who claim to have done so and then been rejected, so it seems most likely that the course is for real.
Together with Óscar Toledo G.’s amply demonstrated skills at programming under extreme resource constraints, skills which require a great deal of practice to acquire, I think the evidence strongly favors the hypothesis that Fénix, Biyubi, the G11V3, the electronics courses, and the rest of it is real, if maybe a bit oversold.
Many people go to a lot of trouble to perpetrate fraud, look at that guy claiming to have invented bitcoin for example. That is not to say that I believe this site to be fraudulent.
I could not find any information on a course available, could you point me in the right direction? It is a bit odd to me that none of their products are available for purchase from their website (I saw the $99 computer with minimum order of 1000 units but when clicking through I was redirected back to the home page).
Yeah, ......._ (what a strange nickname!), they might have a good-faith belief that the software is bug-free and without security holes, but such a belief is likely mistaken. It's not absolutely guaranteed to be mistaken — earlier in the thread, someone compared BootOS to some Woz code, and as far as I know nobody has ever found a bug in Woz's Apple Integer BASIC — but it's very likely to be mistaken.
