Hacker Newsnew | past | comments | ask | show | jobs | submit | routelastresort's commentslogin

Containerization is so widespread... from almost every single programming book and beginner class to the foundation for most of the Internet running today. As a technology, it's very easy to teach. I suggest keeping an open mind with relevant and ubiquitous technology and at least coming up with a compelling alternative for any of the use wildly popular use cases that have made it commodity tech at this point.

Lack of real moderation and reddit tier opinions like this are why I no longer visit this site on a daily or even regular basis.


90% of my timeline is on Threads is all stuff I don't even follow, while Bluesky is exactly what I ask for. Decentralized platforms are the only way to go, even if it's the centralized version and vision. The Zucc way of ramming (overly) suggested content and ads down your throat will die if there are alternatives.


> Decentralized platforms are the way to go even if it is centralized

Wat.


Probably as in you can move your account to another server seamlessly. Mastodon also relies on "centralised" servers; none of them are p2p


Isn’t that federation?


Just federated identity, like OpenID. It doesn't necessitate distributed systems with communicating backends, like Mastodon or Usenet, just that it allows the user to bring their own IdP and therefore to have "sovereignty" to verify their own identity across platforms. The alternate route is to have users cryptographically certify their registrations post-hoc, e.g. Keybase


You cropped the key part

> ... centralized version and vision


> while Bluesky is exactly what I ask for

Not exactly. They can't even show you full timelines: https://bsky.app/profile/dholms.xyz/post/3jzsdcorxx42n


Trivial to code up a custom feed to show however long you want.


wat?

Also: I'm definitely not going to code anything?


I thought Bluesky was only decentralized in theory. Have alternative open source servers been released?


They've launched the 'sandbox' which is the federated/decentralised dev/staging environment. So yes, you can run your own PDS (server) and have it federate with others, but it's still very much in development and 'here be dragons'

https://github.com/bluesky-social/pds/blob/main/SANDBOX.md


Super cool!


They announced that more feed options, including followers only are coming.

They rushed to market.


Considering that "followed only" is far simpler to build, that's a pretty lame excuse and doesn't do anything to change my opinion about whether they're priorities are healthy.


I’m not a fan of that decision either. But I can understand why they added it first. Otherwise, when you sign up, you have empty feed and it’ll take a long time for you to get enough of content, as others subscribe, you follow them and they start posting.

With this decision they ensured you have content from the day one. Often irrelevant (which is risky), but TBH, I’ve found their algorithm to be ok. I’ll drop it as soon as possible, but for v0, I’m ok with it.


> 90% of my timeline is on Threads is all stuff I don't even follow

Most people want this though. They don't just want to see people they follow, they want to be served an endless stream of takes and fight videos


[flagged]


A less uncharitable interpretation of the parent comment is that decentralized platforms are the only way to guarantee personal freedom, save a saintly BDFL or government regulation. Not likely to happen.

The “even if it’s the centralized version” is an acknowledgement that some users may prefer centralization, or the platform is still too young to usefully decentralize. As long as there is a way out, or a credible plan, freedom is preserved.

The third Matrix succeeded due to choice—even at the subconscious level.


Some of it absolutely seems like fluff or stating the obvious. That being said, the majority of the organizations out there do not walk the walk. Especially in product management, the actual product or output lacks a soul or purpose. I'd say that the majority of traditional companies and startups all suffer from this problem.

I've even worked at places with successful products where this mindset is missing, and the success is by accident or proxy to more solid endeavors.

Alignment with vision matters too. Too many companies are like many-headed hydras, where a role like Engineering SREs understand the value proposition and missing pieces of a product or offering better than the actual product management team and executive sponsor.

It's easy to say it's remedial, but it's more difficult for companies to actually live it and be successful because of it. Source: I seen't it.


Ah, Anandtech 20 page reviews... like it was 1999.


If you're a company that supports the EFF (even through individual donations, corporate matching, etc.), I'd love to hear about you, and consider supporting you as well!


As would I!

Mostly, I'm EFF's International Director, but as of yesterday, I've taken a temporary secondment to work for six months with our development (read: fundraising) team.

We're facing a number of challenges on many fronts, and like many of our colleagues in the non-profit world right now, we're having to ramp up quickly to face them.

We get a huge amount of expert advice and moral support from technologists here on Hacker News, but we only really ever chip in with the occasional technical comment.

For the next six months at least I'd be delighted to talk with anyone who is as interested in supporting us financially, or understanding better how that works. You can mail me on danny@eff.org or comment here.

This is a bit of an experiment for me (I'm more comfortable talking about global censorship and surveillance than asking people for a few bitcoins), as well as for EFF. To keep with the mod rules here, it's probably worth trying to keep on-topic to this post, but happy to chat offline or arrange or more general conversation.


Hi, I'd be happy to chat. We (Aptible) have donated as a company in the past and will in the future. Chas at aptible.com


Hey chasb, email sent!


One thing that has always kept me from donating to the EFF is the apparent lack of on-the-ground lobbying operations; I contribute to Public Knowledge instead, because they (seem to be) much more involved in the policy process.

Am I reading the situation correctly? If so, do you guys have plans to shift strategy given the direction of Congress today?


PK are great people, and definitely a good port of call.

For our strategy shift, I'd read our 100 day plan here: https://www.eff.org/deeplinks/2017/01/our-100-day-plan , especially point 7, which is about our D.C. ground game.

We have a domestic legislative team (led by Ernesto Falcon, PK's former VP of government affairs https://www.eff.org/about/staff/ernesto-omar-falcon ). It's not something that we talk about a lot, because the work tends to be in the weeds, doing line-by-line critiques of new laws, or talking to dedicated staffers. We try and do deep dives about our legislative work in EFFector when we can, but mostly people here about our big activism or tech pushes.

One of the aspects of how Congress is directed at the moment is that direct lobbying for civil liberties within Congress may be a bit of a limited strategy (something we're familiar with from previous times in our existence). In those situations, we think it's equally important to equally emphasize the alternative strategies -- fight executive overreach and unconstitutional law in the courts, create technology to protect users, and work on legal protections at the state and international level. D.C. lobbying is part of that (and we've been thinking more seriously about other ways to build that up which I'd be happy to brainstorm with you), but we think these days it works best if you integrate it with other strategies. The power of being a D.C. insider isn't as great as it has been these days, and the strength of advocacy, court challenges, and tech is showing itself right now. - d.



Hi there. We (Aptible) are big fans.


Likewise and god damn I respect you if you do.


PolitiFact: "completely unbiased" goalpost mover. His password was an iteration of 'password' - e.g. no real difference.


Um, his win8 login password being a variation of password at some point doesn't automatically imply that his gmail password was "password".

I thought it was widely accepted that Podesta was the victim of a spearphishing attack (coupled with bad advice from IT), rather than just "hacked" via password-guessing.


A few points:

* Yes, lots of evidence says that Podesta did give his password to a hacker the spear phishing email you mention. The email dump cuts off soon after this event, the phishing bit.ly link was visited then according to the stats page, and we can see the phishing email here: https://wikileaks.org/podesta-emails/emailid/34899

* This appears to be a quote of someone speaking, I think it's pretty reasonable not to pronounce the @. All common variations on password are equally vulnerable to password guessing programs. There's a rule to use l33t speak in JTR and many other common programs.

* You can make a GMail account with the password p@ssword, something else they don't mention, choosing instead to go with a literal interpretation.

* It doesn't seem to mention that he also lost his phone in a DC cab, which is another possible source of leaks.

* They never appear consider whether or not the person who gave the emails to Wikileaks told them Podesta's password, they just say there's no evidence for them to verify, other than Gmail rejecting a password of 'password' (but not p@ssword), after which they rate that claim as false.


I don't disagree with any of those points. I simply disagree with the leap to the conclusion by several posters that because Podesta used "p@ssw0rd" in one place (based on the context, I'd guess that was presumably the initial password as set by some staffer), he definitely reused that same password for a different account.

I will also agree that the Politifact article's conclusion is hasty, and their selection of evidence is questionable at points (such as gmail account creation).


I think that's a fair assessment.


> His password was an iteration of 'password'

What the email indicated was that someone temporarily set a Windows 8 password to a variant of password.

What the email did not indicate that his gmail account itself used a variant of password as the password. It is disingenuous to conflate the two and claim his email password was password as Assange did in numerous interviews.


His Gmail password was probably runner4567, at least his icloud password was and he emailed it. Also not changed when that email leaked.

Just stop with the nonsense that his email password was "p@ssword", as the truth is just as stupid.


The good thing that can come of this would be the resurgence of the Linux desktop. Apple just crossed the line for me (<32GB, no SSD upgrade, messing with the keyboard). I'll miss the almost perfect trackpad, but there's no way I'll support these choices.


Historically, the side effect this tends to have is that a lot of programmers flock to a new platform but bring the way they use and write programs with them. Fifteen years ago we ended up with CORBA- and Window Registry-like components bolted on top of Linux desktops. On the other hand, seeing how nowadays the Linux "community" is mostly a bunch of big cloud, automotive and IoT/vaporware companies, this could be the closest the Linux desktop has to salvation.

So far, the attempts to bring elements of the OS X experience to the Linux desktop have been very... cargo cult, in the absence of a more forgiving word. Fetishizing design choices and simplicity has made it very unpleasant to deal with a modern Linux system. Many users of tiling-wm-and-terminal-only desktops don't do so "just" because it's the most efficient option, they do it because the alternatives are horrifyingly bad.


I wasn't aware the Linux desktop needed salvation. Maybe it needs salvation for non-technical end users but it's good to excellent for end-users whose primary use involve actually using Linux for what it's good at which is basically being a fully open OS and IDE for software and other engineers.


It has been exactly two days since the last time I had to dig down and figure out what broke in PulseAudio this time. Things are certainly bad for non-technical users, but those of us who write Linux software for a living don't have it much better, either.


Do you know why it breaks? Because Ubuntu 16.04 just works on my laptop. Things that require one-time technical setup: Function key (screen brightness) and Optimus/bumblebee. I also set up touchegg because Unity hates multi-touch trackpad. But since initial setup I never have to touch anything again, and can upgrade everything via Software Update.

I would say it is no better than Windows at all (since you also need to install correct driver on Windows).


Cynically, I'd say it breaks because the only mistakes that we live with, try to work around or fix are the ones made no later than the mid-90s. Everything else, we live with for a while, then boldly proclaim that they're outdated, junk from another age, no longer appropriate for a modern system and then promptly rewrite -- so unsurprisingly, a lot of components of a major Linux desktop are basically beta-quality and/or in a continuous state of flux.

Look at the Gnome 3.22 changelog. It lists things like support for multiple renaming, being able to set alarms for events in Calendar and seamless photo sharing via Google Photos and email. I remember being excited about all these features (except Google Photos, for obvious reasons) when I was using Windows. 2000. I was using Windows 2000.

That's why we're rejoicing that we'll soon have ASLR in all mainstream distros and support for a display server/compositor where windows can't snoop on other windows is just around the corner, while Microsoft is perfecting call-flow integrity and has had a proper compositor since the days when we were barely able to bolt our applications on top of an X11 compositor.

Edit: Non-cynically: most of the breakage happens because the level of complexity involved in a most modern technologies is way over the level that can be meaningfully managed by a community. systemd, xdg-everything, they're all very useful tools, but only a handful of people can properly use them, and it doesn't help that so many of them work for Red Hat and aren't exactly transparent about a lot of things. This breeds mistrust and brings about a lot of unjustified criticism along with the justified one.

As for why PulseAudio in particular keeps breaking, I'm not familiar enough with its source code to say. My problems revolve around things like randomly deciding to use another output device. I work around it by not using it, really. Every couple of months I take KDE and Gnome for a ride, they keep breaking, I open up the page of my local Apple dealer, I gaze incredulously at how much money they want for that hardware, close the page, pacman -Rcs plasma-meta and get back to WindowMaker.


I couldn't have said it better. The truth is, all my life, I wanted a Linux desktop that looks beautiful and just works. Simply, stupid, works. No magic involved. I work with Linux remotely almost everyday. I love to tinker, I love to learn - but it must be my choice what to tinker with and what I am going to learn next. I want to play with things that make me happy and give me satisfaction. And constantly maintaining my system and googling around for solution is not one of them. I tried... Crom help me, I've really tried. After yet another failure I just felt extremely disappointed and moved back to Windows at home. At work I am using MacBook Pro (previous one) and I am very happy about it - like it better than Windows. Honestly, OS X desktop is the best what happened to me so far. My last Linux desktop adventure happened few years ago, though, so maybe, just maybe, it got better recently... I still have hope, maybe there is some windows manager out there, but I'd need to thoroughly evaluate it on some spare machine before moving forward with it as my main desktop system.


This is probably why i'm going to be replacing my 2014 with a 2017 MBP (hopefully they're not worse...), despite wanting to go with Linux.

I can't stomach dealing with driver/etc issues. I can't handle paying $2k only to deal with debugging things that i want to "just work".

Sure, it might work perfectly, or it might be a nightmare. I've had both experiences on desktop linux.


I reluctantly decided that my MacBook days were over with these 2016 machines. But moving to linux was a huge step. So I bought a $360 2011 ThinkPad W520. A quad core laptop (with 32gb ram capacity). To experiment. Well, it was a dream machine, so now I'm using a quadore ThinkPad p50 (xubuntu) (with NVIDIA graphics). This is a great experience. It runs VMware way better than macos, and the keyboard ....


I've read about i3 several times over the past few years but finally checked it out a few weeks ago, and it fits exactly what I've been looking for in a Linux desktop for so long.

All I've ever wanted from Ubuntu/Gnome/KDE/XFCE (at least while developing) was the ability to launch programs easily from the keyboard, to tile windows in various configurations, and to manage virtual desktops.

i3 makes all of this easy with an incredibly simple and logical set of keyboard shortcuts. To anyone thinking about trying out i3, the learning curve isn't as bad as it seems, especially if you use a distro like Manjaro-i3 that can take care of some of the harder parts for you.


It's funny how much I think about the trackpad when I consider leaving apple. Seems like a a stupid reason when you tell non-apple users, but seriously, everything else feels so primitive in comparison.


Resurgence? When was it ever popular? Iirc, market share right now is under 5%.

I say this as pretty much a lifelong Linux desktop user. I've been waiting and hoping for it to take off, but.....


Macs took over some of the developer space that Linux laptops used to be in. As a usable Unix environment with excellent hardware, it's a solid investment.


Me too. I wanted to move to a quad core. I was ready to get the new 15" but 16gb? My son has a hand-me-down 2010 MacBook Pro with 16gb. Anyway now I have a P50 quad core, with a 95whr battery, 7w idle power consumption, expandable to 3 HDD, two of which are as fast as the Mac, and a keyboard which is ridiculously good. It runs xubuntu. And I got change after selling my 18month old 13" MacBook pro. reality distortion bubble has been popped a bit.


The Linux desktop seems to be getting ever more dragged in crap like Unity, systemd, and broken ideas on top of broken ideas

Though you can replace those components, the time being wasted there makes the Touchbar and other Apple crap seem the worse of two evils


I can't say there is any usability difference between systemd and upstart on my Ubuntu 16.10 machine.

Which brokenness are you referring to?



Even the XPS 13 (supposed to be built for Linux support) seems to have a lot of basic hardware/software problems.

I was looking for an excuse to get one recently but found enough testimony of breakage to not make the switch.


I tried to buy one of these recently. Dell won't sell them in my country. I looked Too bad I can't buy it with Linux (or nothing!) pre-installed.


XPS's trackpad is pretty good, not MacBook good, but almost there.


I tried using a really fast and expensive 128GB card as extended storage. They're pretty much unusable, slower than a fast USB3 stick, and less practical than a Samsung T1 SSD or something like that.

I don't mind the lack of an SD slot, but I won't buy another Macbook Pro until there's 32GB of memory. I have a late-2013 Haswell MBP with an OWC-upgraded SSD, and I prefer it over my work computer, a 32GB maxed-out Thinkpad running Linux... in pretty much every aspect.


Well, it's safe to say you dodged a bullet then :)


I can no longer drink, due to medical issues. I'd still go out with my team for the bonding, and probably just leave early every time. I really like how the Chef Community Summit has drinking and dry events, like Game Night. Making inclusivity a goal is fun for everyone!


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: