We have to continue on new fronts. Right now people are making open source silicon, but it is still in its infancy, it is to Intel as Linux 1.0 was to Microsoft: a laughable threat.
We need to find better licenses for AI models, promote the Affero GPL. Forge new tools to protect privacy. Maybe work more closely with governments that have understood that privacy is not just a geek's concern?
Enjoy your freedoms, but don't take them for granted. Fighting for them, forging the tools to preserve them, it is worth a bit of your money and a bit of your time.
Same goes for many other gig economy roles.
so in conclusion even by your test you are not employee
If the work law in US does not protect them, that is another matter.
Sure, you might have more "rights" if you were to tally them up. But human psychology is easily gamed, and things like this end up hurting employees on the whole.
For example by hiring contractors in place of employees.
This reminds me of the Daniel Kahn song 'Freedom is a Verb'. Great listen that argues exactly this.
I think what precludes the final victory of open source is that the world itself changed for the better. The mobile and web revolution brought easy to use software into the hands of billions. These are people who wouldn't have been able to touch a PC twenty years ago, let alone understand or care about what source code is. They have vastly different expectations than the professional software users that created the free source movement in the image of the hackers of the 70s and 80s. Just look at the number of rooted phones sold. It's an appliance model of computing.
In this new consumer environment, the free source development model is less adequate and cannot compete with commercial software that can directly monetize the apps and invest in further development. What is required, I think, is to update the free software philosophy to the 21st century, and relax some of the ideological goals to facilitate development and reach the more substantial goals, like privacy and security.
The walled gardens of Google and Apple can be replicated. What is lacking is a free software business model that can gather the $1 billion or so required.
Maybe we don't live in the same world???
> The mobile and web revolution brought easy to use software into the hands of billions.
Yes, but this software is not empowering the users, it is controlling them, isolating them, radicalizing them and making them feel more depressed.
And:
>In this new consumer environment, the free source development model is less adequate and cannot compete with commercial software that can directly monetize the apps and invest in further development.
What, exactly, is supposed to be new here? I see nothing distinguishing the economics of this "new consumer environment" from 1995, except maybe that Redhat doesn't sell CDs anymore.
What has changed since the 90s is increasingly onerous intellectual property law, anti-reverse-engineering laws, and the continued creep of the belief that ideas should be owned.
And this is precisely backwards:
> What is required, I think, is to update the free software philosophy to the 21st century, and relax some of the ideological goals to facilitate development and reach the more substantial goals, like privacy and security.
If everybody's crazy uncle RMS had been more accommodating, more willing to compromise on ideology, things never would have changed in the 90s. One slightly cynical way of looking at it is that he was could play Bad Cop to ESR and others' more corporate-friendly, accommodationist "Open Source" promotion, somewhat similar (for much lower stakes) to how Huey Newton played bad-cop to MLK's nonviolence.
If you drop the principle of always putting the user first, you'll eventually compromise on everything else.
The free software model failed all the same in 1995 for a certain class of applications. It's great for high performance, challenging pieces of code, for kernels, servers and databases. Things with a large community of devolopers among the users, willing to push patches back. Things that look good on a CV. Software that needs to customized for professional users.
It's less good for mass market software. It's flawed for complete games (as opposed to game engines). It's bad for professional software not used by developers and that does not require customization. For example, it has never been able to displace most Adobe products, despite the numerous attempts and the revolting behavior of that company towards it's clients.
What changed was not the economic realities of free software. It's simply that the type of projects where it shines have grown much slower than the rest of the consumer software ecosystem. The world of computing moved ahead from the needs of programmers to the needs of ordinary people, many of whom are willing to pay .99 to solve them.
Insisting on "free to use" today stops you from realizing the other, more substantial free software freedoms, and deprives the projects from much needed resources in their competition with closed software.
On the computer side of things, for less than $100 ($200 if you need a screen), you can have a full, general purpose computer you can code on, new. And with internet access you can have all the documentation you need to get you started.
And the "web revolution" not empowering users? A simple example: buying an uncommon book used to take weeks, if possible at all. Now a few clicks on Amazon and you get it delivered the next day. Realtime communication to people anywhere in the world is almost a given now, it was somewhat possible 10 years ago but at a much different price. And what about navigation? If you have a working smartphone and data connection, getting lost is more or less impossible.
The part about controlling, isolating, radicalizing, and depressing people is debatable. Tech certainly has an effect but is it a net positive or a net negative? It is used to control people but also raises awareness. It makes people stay at home but facilitates communication. It radicalizes people but also expose them to mind opening diversity. It makes some people more depresses but provides outlets.
That's a horribly lopsided analysis. That's not to say that there are no examples or truth to the issues you raise, but it's far from the whole story. People do extract massive value from the communication, education, and entertainment features offered by the software in question.
The world is clearly better in many ways over the last 20 years - more productivity, more literacy, less crime, less poverty, more accessibility, more communication. Those improvements have also brought new problems: more radicalization, more surveillance, more exposure to “keeping up with the Jones’” that might lead to depression.
Are we creating more than we’ve solved? Maybe? There have been radicals and electric eyes online since BBS’ in the 80s, we just haven’t societally kept up with what was obviously going to happen when you drop the cost and barriers to communicating. I’m think we will fix things.
Outside mobile: Linux (the OS not the kernel) won the web.
Way back in the 'Linux on the desktop' days, KDE needed something like gtkhtml so KDE made KHTML and Konqueror. Apple needed a web browser and couldn't make one from scratch, so webkit was forked from KHTML. Chrome needed to fork webkit so made blink. Edge adopted blink too.
KDE's code is now in Chrome, Safari, and Edge. The only major web browser not containing it is Firefox (which is also OSS). That's AMAZING.
On the one hand yes, on the other Apple tried as hard as it could without technically violating the license that KHTML would get as little back from webkit as legally possible.
Reminds me of the 90s when RedHat's source CD wouldn't actually build their binary distro, but since almost noone ever tries, almost noone cares.
Amazing for whom? How much control does KDE have over the code they originally wrote? Last time i checked KHTML and Konqueror were abandoned.
The only major open source browser remains Firefox.
If Chrome etc were actually open source then the vendors would have less control of the browsers and users would have more control. End users could then build Chrome themselves.
Since these browsers are in reality proprietary it would make little difference to end users vs vendors level of control if they used 0% code from open source projects rather than 99% as they do at the moment. No one but Google can build Chrome either way.
So we broadly agree that there is nothing amazing about the current situation.
I however claim that if these browsers were Foss this would be different.
--- Yes I am aware of Chromium. But Chromium is not Chrome. For example Netflix supports Chrome and blocks Chromium.
Edit Removed suprious '?'
I said browser engines initially so I meant the browser engines. I also brought up Electron. Which doesn’t depend on Chrome. But the browser engine.
So I could’ve for a third time repeated WebKit, Edge’s open source Blink, and Blink/Chromium. I’m not sure if that’s the only point of your comment? Based on another comment it seems like being pedantic is your only point.
So what does it matter if the browser itself or as it is now, the engines are open source? Since that is still my question. What difference does it make?
The pedantry doesn’t change anything.
Foss comes with the requirement that you as the user have certain freedoms. With Chrome etc you do not have those freedoms.
Still you have said nothing of substance or use. It’s pedantic in this thread because I had already referred to the browser engines as open source and referred to electron. I didn’t explain what electron was either. You didn’t find a big problem with that.
And I’ll ask again, what differences would there be if the browsers themselves were open source? We already have Chromium vs Chrome. So what would be different if Chrome was open source?
Not to mention responding to anything else.
I wrote about this a few years back:
https://penguindreams.org/blog/the-philosophy-of-open-source...
A) Linus won't stablize kernel ABI
B) Device manufacturers won't release driver code to allow recompilation of an updated kernel.
So this means an upgraded kernel needs upgraded drivers from manufacturers that match the kernel version, but manufacturers don't want to do this.
Not sure where I heard it but I thought that was one of things about Fuschia is the device ABI is a separate, stable layer.
Yes, but this is for very good reasons. https://github.com/torvalds/linux/blob/master/Documentation/...
Where classical Linux drivers are considered legacy (Passthrough HALs) and all new drivers since Android 8 are user space, written in either Java or C++ and talk to the kernel via Android IPC (Binderized HALs).
I wonder for how long we'll be able to use Gmail before it follows the G+ road.
Or are you saying you aren’t going to use anything new by them? That is probably the case and I misunderstood. Though it’s not like Google has any major hits this decade anyway. Everything mentioned is quite old.
Basically, yes. I don't even make the effort to log in to use any of their services, except for the Google account on my phone for the purposes of Play Store access. I'm just tired of being Charlie Brown trying to kick Google-Lucy's exciting-new-product-football.
I could be forgetting a few.
ART is being ported to run on top of Fucshia.
That's the real strength of Linux, creating appliances from commonly available hardware. The biggest mistake is therefore to copy Apple and MS and their general purpose approach, it will never work because the community isn't united like a company, every Linux camp seeks to make the best thing for their own purpose, and constantly clashes with other camps that want to do the same for themselves. A general purpose Linux OS is an anti pattern.
There is a need for general purpose computing, and Linux is one of the very few options there. That won't change.
One major beauty of Linux is its flexibility, so it can also provide niche/embedded products.
It's pretty funny that people tried to steal his tool's name. He even has "testimonials" [3] which I guess is some form of dramatic irony.
I think we should look at FOSS more as we regard math or science. Everybody uses it, but nobody pays for it (edit: once it exists).
It is tricky to figure out how to allocate funds, but software underlies our nation's infrastructure, too. Having federal grants for the maintenance of core open-source infrastructure sure sounds like a cheap investment to me.
One of my favourite scientists of all-time is Harrison Brown. He was most known for his role in launching George Tilton and Clair Patterson's dating-the-Earth project. But he had a most amazing ability to convince the cold-war era US government to fund civilian science that had no real military value whatsoever. He was a kind of science-funding Robin Hood, an ethical con artist.
This might be a bit of a tangent, but the US is still at war and has been (in various stages of "hotness") for many decades). The war in Afghanistan will turn 18 in less than 2 months. More on-topic -- quite a bit of research still gets funding from the military, because of their massive budget (not that I think that this is an overall good thing -- I've heard really worrying stories about the sorts of discussions you have when doing military-funded research).
For instance, it underlies basically all modern science and technology, fundamentally.
Since calculus already exists and is in the public domain, nobody needs to buy it. No government needs to invest money in it. (Assuming that calculus is a finished work for the moment).
This is, of course, not to say that calculus is not valuable.
It's always like that. Creation vs Entropy, Hackers vs Systems. One won't work without the other. We can't win software freedom in the current world, we can only fight against losing it and it's going to be a constant battle. But it's the battle enables everyone and raises the tech bar.
For my own hobby code, I don't see any practical benefit to GPL over MIT. I want my code to be reusable in as many cases as possible. I've seen the toxic treatment AGPL code gets (AFAICT, the main value of that license is to companies who offer dual licenses) and the wary treatment GPL v3 (or worse "v3 or later") code gets.
Your comment there is quite powerful - and i believe true! So, ok, what are next steps in discovering similar high quality content like Linux Journal? Because as it happens, my discovery mode for finding good content is hacker news...though half of the things i ignore because they tend to shift the way of startup culture, which while entertaining, isn't as fulfilling as true FOSS/copyleft culture. So, do you - or anyone else here - have any recommendations to find good content like Linux Journal or any other FOSS/copyleft culture content but through an aggregated site like hacker news? (I use HN not because i'm lazy but because of lack of time; again half of stuff discovered is meh, the other half is good.)
Between clouds, binary blobs and SoC devices I think we're losing in hardware freedom on the back of FOSS as well.
At the same time we finally got a phone that is built with some security in mind (Librem).
AMD using an open graphics driver is also a nice development of the last few years.
...
And we can mourn the loss of easily hackable PCs, but a lot of kids these days are just as exposed to tech as 20-30 years ago. (Eg Minecraft redstone programming, mobile app development, high quality YouTube videos about all things electronic, digital.)
So while naturally a higher tech aware populace would be nice, without cultural shifts and accompanying changes in education the status quo is not surprising.
If it ever materializes. It's already more than half a year behind schedule and after preordering it close to when it was announced, I'm very skeptical I'll ever receive it.
Fortune 500 companies, various nation states are willing to use random Android phones. Why shouldn't "we"?
And those license changes mean I’ve lost software freedom. People have decided AWS is the enemy and all changes that counter them are good to the extent that when they released a fully FOSS download of Elasticsearch, they were criticized for contributing to open source.
Software authors ditching open-source licenses to survive doesn't sound like a win for FOSS.
The GPLv3 closed the tivoization loophole, the AGPL closed the SaaS loophole, and now we have Apache-like licenses that close the rent seeking cloudification loophole.
The AGPL is Stallman + OSI + Debian approved, how come these new no-maximo-capitalismo licenses are somehow not?
Corporations are not people, and arguably are not entitled to the same freedoms with software that people have.
There actually used to be a every Sunday workshop where you made one from scratch in 3 hours (including board fab) with a grab bag of components to populate with — within the past couple years. (That shop closed due to lack of patrons.)
You can’t modify simple electronics as easily — but how many people built their own drones in the 80s? I’d guess RC cars and drones have about the same number of DIY whatever, and that we’re not seeing a fall off (because of difficulty — availability and community are different issues).
The simple truth is most people don’t care and don’t want to manage their stuff all the time: I make my own electronics, but my laptop is still Windows with minimal config. My desktop is parts and Linux — but only because I needed it to mirror what I do at work, as a test bed. The company itself found data centers to be too much fuss, and lifted-and-shifted most of it to the cloud.
Why don’t I bother? Because fighting the laptop constantly to keep FOSS working (and secure!) detracts from my time building my own electronics (or other forms of living life).
If you really care, our failure to create broader technical skill is not technical: it’s social.
We allowed RadioShack to be gutted, we allowed ”hacker spaces” to be priced out of business as cities grew, etc.
Radio Shack gutted itself, IMO, albeit probably in response to the reality that they couldn't sell very many single 555 timers for $1.99 anymore once people had obvious alternatives.
As for hacker spaces, if people want them, they have to be willing to pay [enough] to have them. If the local real estate gets progressively more expensive such that the hacker space is no longer viable economically, then it's going to close or move. It was game playing collectives before them. It'll be something else next. Things we like aren't immune from the laws of economics.
I don't have any firm numbers on this, but I doubt that discrete components for building circuits from scratch were ever a significant portion of Radio Shack's business. Seems like most of their sales were in manufactured devices (for example, amps, pocket radios, walkie talkies, computers), accessories (antennas, microphones), cables, replacement tubes for radios and TVs, and, of course, batteries.
I think one of the dominant causes of this is the fact that hackers and tinkerers became viewed by many/most FOSS products not as key users, but as mild annoyances or even outright pests (and the process has been going on for a long time, at least 10 years). As FOSS projects fought with proprietary and limited use options they worked to provide convenience that mass users needed to switch. Then to gain wider acceptance they added polish -- "plug and play", "no configuration needed" that mom and pop users wanted.
There is nothing sinister there, but this step puts projects a stone's throw from corporate IT that offers a lot of support / development money, but wants to lock everything to a few "approved" configurations (because crypto, corporate policy or something else). Project remains open source, everyone can fork it (but good luck making sense of it), but from this point tinkerers become persona non grata. What they want (freedom to change things) is against the payer's desire for a locked setup. And from this point on, the system is free only in name.
If you want your software to stay FOSS in spirit, do not drive away tinkerers as your system matures. Leave enough hooks for them to access low level capabilities, enough knobs for them to do "what no sane user would want" and do not see them as a threat even when they stand between you and that $50k support contract from some megacorp. My 2c.
Just my 2c.
FOSS ate "sell to the end user" software companies. FOSS won not because of open source, but because the price was $0. Companies like Facebook realized that. As long as it doesn't cost anything out of pocket, people will use it, even if they have to give up all of their privacy and personal information.
Now cloud providers give you nothing to run if you stop paying them constantly, and in a lot of cases, say goodbye to your data too.
If you're tracking engaged users [or even growth of revenue and churn from individuals], you're probably not on a path towards rent seeking.
Hopefully no one is using the outlook from then as their mail client at least...
[1] https://www.cvedetails.com/product/320/Microsoft-Office.html...
Personally, in most cases I'd much rather pay directly for a service than get it for free from a company whose business model is "try and monetize it through advertising, mining user data and/or dark patterns and shut it down if all of those fail."
A customer shouldn't care at all for the reason why the service suddenly sucks.
Customer should only care about the quality and value the service provides to him/her.
I don't think I've ever been so upset at a service before... I've been recommending people stay far, far away from Flickr and I advise anyone reading this to do the same.
However, "total radio silence" is not fair. We're very active, and very vocal, about our status updates and progress we're making on the Flickr Help Forum, on our blog, on Twitter, etc. In particular, Flickr now has a CEO (me) who frequently engages with the community and customers to keep them informed, something Flickr hasn't enjoyed since ~2005.
We even have a status page, with frequent realtime updates, which I believe is a first for Flickr: https://status.flickr.net
There is a single feature, Camera Roll, which is still offline undergoing maintenance. No-one is happy it's offline, but it is what it is. It relied on technology internal to Yahoo, which we had to leave behind when we recently exited their datacenters. Given the hard deadline and the vast size, scale, and scope of the Flickr service, we managed to move everything (while keeping it online) except this. It will be returning shortly, but it's not quite ready yet.
I hope you'll stick it out with us, because we're nearly done migrating one of the largest web services on the planet, which means we get to focus on building again instead of just copying. The future is bright.
Disclaimer: I'm the co-founder, CEO & Chief Geek at SmugMug, the company that acquired Flickr from Yahoo last year.
Thanks, WNG Image and Design
I am also an old-time programmer, starting on 6502 machine language, so I understand programming challenges even though I have never dealt with them on this massive level.
Please continue posting updates wherever is convenient for you, I know the users on the Flickr help forums are quite opinionated and fond of their own words so I understand a bit about not posting programming issues there.
Furthermore, calling Camera Roll just a feature undermines its significance. This isn't just removing some insignificant feature like "search by focal length" or whatever. Camera Roll _was_ Flickr for many people. Not most of the total users, sure, but for everyone I know it was. For half a decade Flickr pushed it as _the_ way to use Flickr. It was what you saw first upon opening Flickr, and for any users to Flickr that joined in the last five years it was basically the entire interface.
Then it disappeared, supposedly for a very short time which then became very long... And then the updates stopped. During that entire time, myself and everyone I know that uses Flickr basically stopped using it. This includes people like me who pay for Pro. It became worthless.
The frustrating thing is that I really loved Flickr. I loved it so much I considered applying for a job there as the first job I've applied for in my life (I never apply for jobs, I just get referred to the company usually.) I loved it so much I evangalized it to my friends, family, random strangers.
But then the core Flickr interface, the way everyone I know managed and viewed their photos for half a decade, disappeared. And it's still not back. With the time, and the silence, I grew frustrated, and so did many people. I can't use it without Camera Roll, and everyone else I know can't either.
I really, really want Flickr to succeed. But this whole thing really stings hard.
- The choice to finish moving Flickr or not was literally the decision to keep Flickr on the planet or not. There was no wiggle room in terms of timeline or resources to get the job done. So we had to make some hard decisions, because I, for one, didn't want to see Flickr disappear entirely. For good. This was more likely than most people understand. (See: https://www.vox.com/2019/1/23/18194865/verizon-layoffs-aol-y... )
- Camera Roll wasn't used by most users. When we were faced with the question of "ok, we're going to have to temporarily shut off a major feature to meet this deadline", Camera Roll was the obvious one due to its relatively low usage. I'm sorry you rely on it so much, but it's very clear that's not common relative to other major features.
- Camera Roll being offline hasn't affected usage materially. There appears to be near-zero evidence that many people "basically stopped using" Flickr due to this feature being offline.
- Based on your comment, I checked with long-time Flickr employees who are still employed here, and none can recall Camera Roll being central or core to the Flickr experience, nor it being the landing page. Our usage data, both before and after shut-off, supports this perspective.
- One reason for a smaller # of updates about Camera Roll is simply that it's not done being rebuilt yet. We prioritize time spent building things rather than time spent updating people with "Still no update". I think most customers would rather have us ship it sooner rather than talk about it more and ship it later. The last updates are still true. It's not done, but we're working hard and it is coming soon, which we've said recently, more than once.
If you truly love Flickr, as I do, I hope this perspective is helpful. We (the royal "everyone on Earth" version of "we") nearly lost Flickr forever. SmugMug saved it, but we had to make some hard choices. So we did, and I'd still make them again. Flickr is in good hands, we're excited about the future, and we're excited to get back to building again.
I hope you stick with us.
Asking old time Flickr employees is good, but that kind of misses the point. Old time Flickr users generally stuck with the interface they knew, from what I can tell, and did not use Camera Roll as much. "New" Flickr users, aka those using it for less than a half decade (at least those that I knew) used Camera Roll nearly exclusively and never used Organizr. There was never any other way to view large numbers of photos or organize them on any non desktop device other than Camera Roll, for example, and for anyone who used Flickr on mobile it was not merely central to the user experience - it _was_ the user interface.
I'm extremely shocked that Camera Roll isn't commonly used relative to other features. I wouldn't believe it from anyone else :) Can you say roughly how often it was used compared to Organizr before the switch? (as those are the two only ways to organize your photos on Flickr.)
Regardless, I know first hand that a not insignificant number of people used Camera Roll for everything. Personally, Camera Roll's absence has prevented me from providing photos of important things to several people. I simply don't have the time to dig through Organizr to find and organize a selection of photos that was easy to locate before but is now nearly impossible.
I appreciate the work you all did to save Flickr, and I look forward to using Flickr again once it's usable for me.
The "Cloud" is certainly a clever means to extract value from GPLv2 software stacks that many of the services are built upon, frequently without having to give much of anything back.
How many "cloudy" services run on Linux and other GPLv2 software?
I sometimes wonder what the world would look like had more free licenses (such as MIT/BSD) been just a bit more popular during critical points in history, or had less free options like GPLv3 or AGPL come a bit sooner as viable options.
Would we have to pay for more software, and if so, would the privacy equation have turned out differently? Would software simply not have advanced as quickly as it did? Would the direction tech is moving still be a web-centric one, or would it be a bit more of a desktop/local centric world with more shrink-wrapped software?
And really paying for software is no guarantee that you won't also be the product - I'm not sure where that naive notion came from. Paying for the software isn't what keeps your privacy safe - keeping your privacy safe does. That it is an inane tautology is the point - they are separate aspects.
It isn't a guarantee, but not paying for software means that the developers have a strong incentive to figure out other means for extracting value out of you. So it isn't so much that paying for software wont make you the product as it is not paying for software will make you the product (note that this is when businesses are concerned - personally as an individual i give a bunch of stuff for free and i do not care about getting any value out of them, but at the same time i wouldn't build a business around giving out stuff for free).
It's just less than trivial to get it running yourself.
Reminds me I need to check in on https://Sandstorm.io and see how they're doing. I've tried to install it a couple of times and bounced off of it a couple times, not always due to things that are their fault (i.e., they do sensible things with DNS but my managed DNS provider won't let me set up the requisite records in their "easy" interface, and I'm not quite motivated to go all the way back to running my own bind).
It would have stayed as it was and how we are now back to, thanks to adoption of MIT/BSD and daemonising of GPL.
Public Domain, Sharewhare, Demoware, whateverware.
They too used to occasionally provide source code for the free/demo version.
I use my computer to do work. I don't have time for my computer to be some social battleground first-world-problem and quite frankly I don't care. RMS is right but he has the answer to an irrelevant question that nobody asked.
I mean, getting everyone's grandma (and her java server, I guess?) on Linux, is the goal - right? Surely the well behaved user will fight for openness?!
If any project accepts that its goal is to gain users by displacing other products (commercial or not) then that project is committing to a fight it can't win simply because of the nature of open source projects - in the present day world, whoever has the most money tends to win.
Open source needs to find a way to exist as partially commercial software, or else projects won't have longevity, which inevitably means they'll be displaced by commercial products because while the motivation to develop open source isn't constant, the motivation to find a way to sell software for money IS constant.
The ideal that Stallman and others have always aimed for - that all software should be free - is impractical in the real world until the real world changes significantly, like most utopian ideals.
It is like playing on the street, yes you can make it for the supermarket, but don't expect to pay for rent and pension out of it.
It doesn't seem likely to in the near term either.
"It became clearer than ever to me that while Linux and FOSS had won the battle over the tech giants a decade before, new ones had taken their place in the meantime, and we were letting them win."
It's true. We won it all, but we somehow still lost, and that's a difficult and sad thing to realize. Even though FOSS ate the world, we didn't win software freedom, we just enabled a bunch of new tech giants to put us into consumer roles with little to no freedom once again.