After my article on how to create / -> cross-posting bots, I did an experiment with @crossbot and let it run with ~10 different sources for a couple of weeks.

The idea was definitely successful: I brought with me to the Fediverse all the sources that I wanted to follow, without forcing them to move, and I actually didn't feel the urge to open Twitter/Facebook for "fear of missing out".

But I've realized that one single bot to manage multiple sources isn't ideal. People who may want to follow only some of them are forced to get on their timelines also content that they didn't ask for. Some people did indeed follow crossbot, but many also unfollowed it - probably because it posted too much, too often, and since all the content was coming from the same account it was hard to tell which was the source without actually reading the toot.

So I've decided to split it into multiple bots, one for each of the sources that I'm cross-posting. Feel free to follow any of these bots if you are interested in the content! But please also avoid commenting on their activities (there's no human behind the profile that can react). Instead, favourite/boost/re-share the link if you want to bring the discussion to the "human" sphere.

List of available bots:

- The Economist: @economist_bot
- Quanta Magazine: @quanta_bot
- Nautilus Magazine: @nautilus_bot
- Nature: @nature_bot
- Scientific American: @sciam_bot
- Phys.org: @physorg_bot
- The Gradient: @gradient_bot
- The Hacker News: @hackernews_bot
- Hackernoon: @hackernoon_bot
- IEEE: @ieee_bot
- IoT for All: @iot4all_bot
- Better Programming: @better_programming_bot

Also, feel free to comment on this post if you have any requests for interesting sources that are only available on Twitter/RSS and you'd like to bring here - I may definitely consider making a bot for them.

If a monkey hoarded more bananas than it could eat, while most of the other monkeys starved, scientists would try to figure out what's wrong.

When humans do it we put them on cover of Forbes.

@bob I would argue that the main problem with today's industry is that both competition and cooperation no longer exist.

Competition means having many competing businesses in the same space. If one of them becomes sloppy or behaves like a jerk, people will move to another one.

Instead, today we have an oligopoly where 5-6 large companies basically own 90% of the Internet public services and they can basically get away with anything.

They have no incentives to innovate either: after years of VC/pension/hedge fund money being poured into them, they only care of paying back their investors, maintaining their dominant position, increase the entry barriers to competitors and move to a simpler rent-based business model so easy money can come in.

So yeah, a healthy mix of competition and cooperation is required to drive innovation in any industry, but today we don't have either of them. We are in a post-capitalistic scenario similar to the one of Big Oil in the early 1900s, where a handful of large businesses that own the whole market just want to sit on their asses and watch free money come in, and they use their lobbying power and size (not genuine innovation) to remove any obstacles between them and that money.

And it's not even a trend peculiar only to our industry: in the US and the UK there's in general much more consolidation and less competition today than there was 40 years ago: economist.com/leaders/2018/11/.

I would argue that this is a natural evolution of a capitalistic system left unconstrained: once the market has picked some winners, those winners will just create a new economic aristocracy that is resistant to change.

discourages employees from talking about to prevent the risk of being seen as a "hostile work environment".

In other words, you can't openly oppose bigotry, religious narrow-mindedness and people wanting to put their noses in affairs that don't belong to them, because the priority of the company is to "make everybody's ideas feel welcome and respected" rather than sitting on the right side of history.

theverge.com/2022/5/19/2313171

@primalmotion @EU_Commission @eff yes, for us it probably won't change much. A lot of us have an un-googled phone, use alternative search engines, email providers, browsers and cloud solutions.

But we're <1% of the total market. The remaining share uses Google products rendered in Google's clients and apps. If they take down the alternative clients, then we can't even tell people "the Gmail app tracks you a lot - if you still need your Gmail address then at least use this app".

Not only, but Google's decisions eventually also affect us, even if we don't use their products. After his app was taken down from the Play Store, the developer of FairEmail (and many others before him) decided that it's not worth to invest in building something that may only be distributed on F-Droid or Github repos and reach less than 1% of the market.

Eventually, if alternative clients are only distributed on alternative channels, many developers won't even bother to build something that only reaches a niche of users. And this is a game where everybody except Google loses.

The older I get, the further Linux seems to stray from the lightโ€ฆ of Unix 

@meena your concerns, from what I understand, are mostly about private companies owning tools that are de facto standards to build infrastructure - and I'm totally with you on that.

But Flatpak/Snap IMHO don't really fall into that category. I mean, they are developed by Red Hat and Canonical respectively, but these are two companies whose core business has been Linux distribution for decades - I'm not a big fan of either, but it's better if it's them to develop a new packaging system rather than Google/Microsoft.

On Docker, I fully agree with you. I don't trust Docker Inc. either. But this doesn't diminish the value of a container-like system as a way to keep the application isolated from the host system. Personally I prefer to use systemd over Docker containers whenever I can, but I'm also aware that this limits the distribution of my software only to UNIX machines that have systemd...

@christianp time to get one of these APL typeballs out of the closet and find a way to connected them to a phone?
dyalog.com/uploads/files/apl50

The older I get, the further Linux seems to stray from the lightโ€ฆ of Unix 

@meena the main problem of (and UNIX in general after the 1980s) has been that of fragmentation.

As a software developer, I need to package and distribute my software to target multiple distributions (as an rpm, deb, xz, etc.). I need to test it on all of those distributions as well. I need to make sure that it's compatible with whatever version of the shared libraries that OS comes with.

The technologies that you mentioned are just possible solutions to this problem.

If I distribute my software as Flat/Snap, then it'll run on whatever OS has Flatpak/Snap installed. If I distribute it as a image, then it'll be like its own tiny OS image that ships all the dependencies it needs to run without touching those of the host OS. And you can destroy it and recreate it easily from a text file, or simultaneously install/run different versions, without breaking the whole host system in the process.

But I also believe that, just because these tools are very good at solving a particular problem, we shouldn't become the guys that have learned how to use a hammer and think that the whole world is nails.

The use-case for bare metal applications is still there. Flat/Snap/Docker all come with a big overhead in terms of storage usage, so running everything on them is just not feasible. Lots of applications still don't mind using the OS shared libs instead of bringing their own in their little backpack. And should be used for what it was supposed to be (e.g. orchestrate complex container architectures in distributed environments), not to run stuff on localhost.

Dear open source users,

If the author of your favorite open source app has announced they stopped developing and supporting the app (because they're frustrated and possibly burned out), please don't suggest they do more free work so that you can continue using the app.
Instead, consider thanking them for their past work and let them know that you enjoyed their app.

Regards,
another open source developer

@molly0xfff this reminds me of the Argentinian bonds crash nearly two decades ago.

Anyone who had a clue of economics knew how risky Argentinian bonds were. We wouldn't invest a single dime in them, unless we had some spare change that we wanted to invest in something that would either burst, or give us crazy high returns in a short time.

But the average housewife didn't know about these things. So when they walked into their bank looking for somewhere to invest their small savings, and the employee would tell them "would you like this Argentinian fund with interests around 20%, while your normal fund would just give you 2%?" (note: no mention about the risk), it was really a no-brainer for them.

Then we all knew what happened with the Argentinian defaults of the 2000s, and many people lost lifelong savings in a moment.

Of course the blame must be laid primarily on those who cheat the public to buy garbage under the promise of high returns, but a little bit of education before investing is always welcome.

If an asset promises returns that are too high to be true, usually there's something wrong with it: high returns are always associated to high risk, and you don't want to put all of your savings in something that is high risk. People should know this simple principle before putting their money into anything.

@blacklight

This is bullshit!
Marcel doesn't deserve this treatment, he is one of the best application developers I know; fairemail, netguard, xprivacy... are indispensable in my day to day life.
It's deplorable, all my contempt to Google.
@EU_Commission @eff @ombudsman

@z428 @EU_Commission @eff @ombudsman having a gatekeeper is surely bad, but it wouldn't be that bad if at least they transparently shared a common set of rules, procedures and remedies.

Take F-Droid for example (an alternative store). In order to submit an app to their store, its source code must be freely available, they would then run a bunch of inspectors on it (including static code analysis and running the app in a sandbox to check which services it connects to), and in case of something that doesn't comply with their rules they tell you exactly what you're supposed to change.

If Google was so transparent, the Play Store wouldn't be such a huge issue. Instead, they don't give you many details on why an app doesn't comply to their rules, they don't give you many actionable points, they are a faceless corporation by design, and when you appeal their decisions or ask for clarification you get an automated answer.

There's nothing intrinsically wrong with being the largest player in a market, but with great powers come great responsibilities. If they can't commit to their duties, then they don't deserve to be the largest player anymore, and the EU should bash them hard until either they learn the lesson, or they stop being the biggest player.

@EU_Commission please do something about this. And @eff please intervene and support the open-source developers who are victims of Google's abuses.

, an open-source email client for that allows you to manage multiple accounts, has been taken down from the Play Store.

Even if its source code is freely available, and both the code and the app's activities have been inspected and audited before being submitted to F-Droid, Google keeps harassing the developer as it considers the app as "spyware", but it refuses to provide ANY details about their findings, or inform the developer on what they are supposed to change to get the app approved again. An appeal request from the developer only resulted in an automatic response from Google.

The developer was advised to appeal to the EU, but (maybe rightfully) he said "what's the point? it's going to take them five years anyway just to come with a decision, and in the meantime my app won't be distributed on the major Android channel, and I'll have no incentive to keep working on it". We, as open-source developers, should NOT end up in this situation. We should NOT have the feeling that the institutions are not protecting us because they're just too slow to intervene or even to understand an issue.

This isn't the first time that harasses open-source developers and gets away with it. Email clients alternative to Gmail, as well as any app that accesses what Google deems "sensitive user data" (including emails, calendar, fit data etc.) will now require an expensive (talking of at least $4500 a year) and intentionally cumbersome certification process, and such a certification needs to be renewed on a yearly basis: pmail.com/newsflash.htm. This will mean the end for most of the alternative apps that support Google services.

This isn't about users' security. Most of these apps are open-source, owned by the community, and regularly audited by F-Droid - a store with far higher security standards than the Play Store.

This is just Google declaring their final war against those who dare to access their email, calendar or maps without using Google's apps. This is Google showing the middle finger to the world and saying "the only way to interact with your email and calendar must be through my apps".

More and more open-source developers are being so discouraged by Google's efforts, requests for money and the Kafkaesque labyrinth that they've set up for appeals that they are pulling their apps and services for good.

This shouldn't happen, and the EU has a duty to defend us against this evil corp, because we can't keep defending ourselves. Enough with all the talk about new EU unicorns: if the EU really wants to battle Google, they should do so by defending an enthusiastic community that is already building the alternatives - often without being paid a single dime, while being regularly harassed by big tech.

forum.xda-developers.com/t/app

@kenneth had I worked for them I'd have resigned rather than see their only decent streaming API abandoned and killed like this.

@imacrea that's more or less the way I used to manage my collection before (a self-hosted server with all the mp3s). But that wasn't scalable for a music junkie like me, hence my love/hate relationship with Spotify.

Not all the music I listen to is indie bands that are on , and with an average of 2-3 album discoveries a day I would have to spend my whole salary buying mp3s.

So all-you-can-eat streaming is really the best option, as long as I can integrate it in mopidy, because I've built so much automation around it over more than a decade (from controlling music through IR remotes, to automatically playing my alarm clock music, to the integration with my voice assistant on the RPi, to a multi-room set up with Snapcast, to accessing my collection using any MPD-compatible client) that I *don't want to* use another client/interface, no matter how many bells and whistles they put.

But Spotify is really leaving me with no choice here. Either I work on migrating mopidy-spotify to Librespot (hence continuing with this cat-and-mouse game against a corporation that just doesn't want to provide me with flexibility), or I move all of my music automation to pure Librespot (as well as rewriting the UIs and clients I've built, which is quite a painful option), or I go through the thousands of albums and artists I have saved on Spotify and buy/download them all as mp3s (an even more painful option).

@mihira indeed, I mostly use mopidy-spotify for streaming, which caches a lot, which takes away some of Spotify's abilities to track users and their behaviour.

Luckily nowadays there are also alternatives for streaming like (github.com/librespot-org/libre). But Librespot itself isn't an official API, it's a hackish reverse engineering of their protocols, and as such it still relies on the mercy of Spotify not implementing too many breaking changes.

mopidy-spotify now is broken because of the libspotify deprecation (and it already broke twice in the past 3 years for similar reasons). Like in the previous occurrences, we'll probably work to adapt to the changes, and probably migrate it to Librespot with some GStreamer plugin. But this starts to be a sterile and unproductive exercize that takes away precious time from open-source contributors' time budget just to play a cat-and-mouse game with companies that try their best to break our implementations.

@m2m well this is actually a quite relevant example. If Spotify has a monopolistic position in the market is also because most of the artists publish their music on their platform.

I've also done the same (after all, if you want visibility you need your content to be on the largest platform). I also have a couple of songs published on Spotify. But I'm considering taking them down and just moving to instead. Platforms like Bandcamp also have the advantage of cutting the middleman and bring most of the revenue directly to the artists.

If platforms like Bandcamp had a subscribe-and-eat-all-you-can business model as well, and more indie artists published their content only there, then Spotify would bleed both from the demand and supply side.

@mihira they have invested a lot in the UI. Many of the recommendations they push (as well as Daily Mix playlists etc.) aren't exposed on the API, only on their clients and apps. So if I use e.g. a command-line MPD-based client, I won't see the content that they invested so much in building. Plus, they will only get my search and listening history - not all the goodies and perks that you can get when you track a user directly on a webpage.

@m2m I understand your point (and being a musician myself I'm also sensitive on the issue of fair remuneration for artists), but I also believe that piracy is the only thing that forces innovation when an industry becomes stale.

Without torrents, eMule, Limewire, WinMX and Mega eroding the market share of the music, movies and TV majors, players like Spotify or Netflix wouldn't have been born.

Now that they have stopped innovating and they have become just as stale, expensive, conservative and control-freaks as the businesses they replaced, we need piracy again to erode their market share and let more innovative players enter.

Show older
Mastodon

A platform about automation, software architecture, data science and tech.