Hacker News new | past | comments | ask | show | jobs | submit | incision's comments login

> How does this address either of my points?

It addresses it as a human being discussing the meaning of the article as opposed to a pedant-bot dicing words to 'win'.

> I find it weird that guides like this still suggest seperate root, home, boot and swap partitions.

The classic reason is preventing space / inode exhaustion of / by user accounts and preventing time of check, time of use vulns. Tough the latter have since been addressed in other ways [1]. There's also a certain amount of convenience and cleanliness to separating system and data that way.

1: https://wiki.archlinux.org/index.php/Security#Preventing_lin...

Thanks for this.

Lots of great sequences from the 80s presented there along with on my of my all-time favorite openings - Ghost in the Shell [0].

I've always felt failing to put together a compelling opening is a missed opportunity, particularly for a series. It's a hook, a primer to put the audience in mind of the best eras and episodes every time.

0: http://www.artofthetitle.com/title/ghost-in-the-shell/

Wow, it has probably been 15 years since I read something from Raph Koster. So long that I had to think for a bit to recall why that name didn't quite feel right - until Designer Dragon popped into my head.

I can't disagree with a single thing said here.

I tend to think of WoW as the Buy-N-Large generation ship of gaming.

It created an entire word that was in some sense a paradise, but at the exclusion of all the awesome, frustrating, exciting, tedious richness that came before it.

The episode of This American Life referenced is well worth a listen if you've ever been curious about what trying to do good work inside a crushing bureaucracy sounds like.

1: http://www.thisamericanlife.org/radio-archives/episode/536/t...

See also: on how they pick the bosses of career regulators


What are there, 1B Android users? I'm sure Google would like that to be 7B.

I'd wager the complaints of the sort of people who frequent lobste.rs or HN are not representative of the majority of current users and almost certainly not of the next 6B potential users.

Overall, it's everything Google said they would do with material [1].

It's fine to dislike these things, I don't necessarily like them myself, but claiming they're simply wrong or "for no reason" is somewhere between angry and lazy.

Just a glance at the new vs old keyboard or gmail shows a direction - moving / duplicating interactions toward the bottom right (logical as screens get bigger), breaking out punctuation into discrete keys (presumably logical given the way most people actually type).

The overall restyling is adding consistency between apps and particularly between mobile and the desktop.

Animations are really valuable anytime someone doesn't come to the table with a complete mental model of how something fits together already. Basically, they're not for you Mr. Programmer.

1: http://www.google.com/design/

When I help people out with their computers and I run updates (pretty much everyone I go to stay with!) they always complain if there's been a change in the UI at all - "why's that button moved?" "where's my favourites?" "how do I access my inbox now?" "why does it keep popping up that thing?" ...

I'd really like to see some respect for this from people like Firefox - "we won't move any major UI elements or update the default skin more than once every 2 years". Of course the marketing is all about the novelty at the moment - steadfastness doesn't count for much in tech circles it seems. With stable UI you can have a less rapidly changing system for users to comprehend; running updates for security fixes shouldn't mean you have to face a new menu paradigm or new default screen or new tab shape or new whatever every few weeks.

Seriously. I don't enable auto-update on supported users apps for this reason.

I couldn't disagree more. Of course there no sense in moving things around just for the sake of changing something, and implementing those design changes is always prone to making bad decisions. However I pretty much want the applications I use to move forward and if this means iterative design adjustments, then by all means, do it.

The people complaining about any kind of change are usually the ones who learn using an application by trying to exactly memorize navigation paths and word-for-word expressions in menus, which feels absolutely wrong to me. It's like learning stuff in school by memorizing the content word for word, without understanding any of them or the context in general. It may work for while, but not for long.

I know that the average person is very different from the tech-savvy crowd around here, in fact I work with non-technical people everyday, yet this helplessness once a button or a feature has been moved is still very surprising to me and I think it's a better idea to try to educate people not to hang on to memorized paths, but instead look for plausible context, really consider where you'd expect a certain feature to be located. That's exactly what the developer did for his app - well, in most cases. Avoiding change in software design is not the solution.

No, it's not a better idea to "educate users" it is rather to respect the users. Imagine when your car would change the controls on random mornings. People want the work done like you just want to drive. Don't project your untypical preferences to the others.

I'm almost sure that if you're a serious programmer you use a command line. You know, where you type in one line, the output is written in the following lines, emulating the electrically-controlled typewriter which prints the letter one after the another on the roll of the paper from the sixties. Think about that. "Carriage return" once actually caused the physical carriage to return and the line feed actually activated the paper feeder to roll the paper one line up.

A lot of changes in the UI are often driven more by the changes in the politics inside of the company (who is going to be seen for his visible contribution to the "change") than any real need. Think about that.

No, what we have to do is educate you on why you cannot expect everyone's brains to work the way you want them to.

For many people, computers are just a thing they use, whether for work or for keeping in touch with friends or whatever. They aren't something that they dedicate very much of their brain to. Computer interfaces have lots of text, lots of images -- some of which are clickable, some of which aren't -- and the rules aren't very consistent from device to device. On top of all that, there are some genuine dark UI patterns, like software from CNet that tries to trick users into downloading useless crap.

What you're basically expecting is for people to dedicate more of their attention to computers. And guess what? Mechanics wish people would dedicate more attention to their cars. Contractors wish people would dedicate more attention to their homes. Landscapers wish people would dedicate more attention to their yards. And in each of their professional forums, you'd hear some of them arguing with equal veracity that people need to be more educated about all of these things.

As a user who learns apps by concept, not memorization, it's still incredibly frustrating when the UI paradigm changes (such as the old switch from organized menus to chaotic ribbons, and from useful text to ambiguous icons). But even a conceptually well-versed user of software starts to develop a spatial/muscle memory for working significantly faster; experienced, intelligent users can still be thrown completely by minor UI changes that take them out of their working flow state.

As a front-end engineer who works on desktop Firefox for Mozilla, I think we have actually held to your "no major redesigns more than once every 2 years" request ;)

Australis was a UI change that we don't anticipate doing more than once every two years. We recognize the costs of doing so, for users, add-on devs, and Mozilla devs.

I've recently updated to 33 (on Linux and Windows) and had my startpage altered - hid a lot of the site links, added an unwanted search form. It did add a new cog, which usually means settings but in this case is a toggle for the pinned sites (completely breaking the expected standard-through-use, offering no affordance as to it's action, contradicting the use of the cog to access preferences).

Since the Firebird days I've been using FF (on and off, I had a dalliance with Opera, Chrome, some other variants) and even I had to first search what the newtab page was even called (I forget now, it doesn't tell me anywhere in the UI) then having read a few posts I noticed that some people had different size site links and tried scroll-zoom and so returned to my default of 3 x 5 sites shown.

It's tiny things like this - you open your browser and newtab has been altered (to make way for adverts in this case) and smack there you are. It's not "your" browser it's the company that offered the upgrade, it's theirs now .. until you can claw it back again. Then the next upgrade comes, some security fix and you apply it with trepidation.

Is it really 2 years since the addon bar was removed, the big orange menu button was removed and the new "burger" menu link was added, the new config for the menu link, the streamlined tabs, the removal of the menubar, tabs moved above the address bar. Time flies?¹

[Looking it seems the major Australis changes (removal of addon bar for example) came in 2013, this thread is full of rants against changing the UI and moving towards Chrome https://blog.mozilla.org/addons/2013/05/27/major-compatibili...]

¹ - I genuinely don't remember. I also used to use fixes/addons to have a compact menu with chrome removed (using KDE settings) so I'm not clear on default UI.

This is why I won't recommend Android to any friends or family. Being the most technically literate person most of them know, they all come to me for help fixing problems with their computers or phones.

The UI changes in iOS7 caused a whole bunch of questions and problems with my "support network", but with Android it would be like that every year with every OS update.

The reason I don't recommend iOS is because Apple keeps changing their policy on apps and their stuff works only with products they make. I like my friends and family to not live in Apple's world view / walled garden.

So, it's not about Apple designing great products; i recommend android because it's a better choice for freedom (which usually always ends up having to make more effort).

Non-techy people aren't good at tech. They don't grok it. Usually they operate it by learning it by rote. "Click the button there, then click/tap this". To them working a piece of tech is like learning off spells. When you change the UI, it's like someone changed the language you're trying to work in. They have to relearn from scratch.

Same here. The technical folks will immediately update, discover the UI annoyances and find means to overcome them. But I notice that non-technical family and friends become very much annoyed when they are continuously pushed to update their software and almost everything changes within the UI.

Companies have switched from introducing new features and backend improvements (what the technical folks used to care about) to what they feel everyone can see, the UI. As a result, the software has not fundamentally changed, minus a few small new features, but it looks like a brand new product.

One solution might be to keep supporting the previous two or three layouts, and provide an easy way to switch back and forward between them.

> Animations are really valuable anytime someone doesn't come to the table with a complete mental model of how something fits together already. Basically, they're not for you Mr. Programmer.

Yeah, whenever I see sentence like "as a user, I don't care about animations" I just stop reading the rest. I appreciate the fact that you, the article/comment writer, have good mental model about what is going on and you want to be as fast as possible, but that comment also shows that you aren't a typical user and you don't have a clue how common people thinks.

Well thought out animation is also something that should improve usability. It directs user focus and can give hints as to what just happened.

I actually think most of the animation in Lollipop is rather well thought out from a UX perspective. It might be a teensy bit showy but I rather like it.

Exactly. Not to mention, these changes make many users feel like they are getting something new (for free even). Even if it just a style change, my girlfriend really enjoys seeing the new look of the phone. I actually purposely let her do the updates on my android phone because I'm not overly excited about the changes but I know she is (she is an apple user so can't experience new android stuff otherwise).

> Animations are really valuable anytime someone doesn't come to the table with a complete mental model of how something fits together already.

This argument conflates the problem (lacking a mental model) with a solution (animations). This Mr Programmer is confused by elements flying around in front of them. Maybe regular users get confused too?

> This Mr Programmer is confused by elements flying around in front of them.

Try it out for yourself and see if you find it confusing. In many places it's designed to enhance ease of use and does a fairly good job of it.

I agree with you, but it seems the train has already left the station.

I don't use my tablet all that much, but I don't find myself questioning how something is done (that stuff didn't really change), or why things look different (I don't really care). It's just... different, nothing else.

>"The new VLT results indicate that the rotation axes of the quasars tend to be parallel to the large-scale structures in which they find themselves. So, if the quasars are in a long filament then the spins of the central black holes will point along the filament."

Though I have no real idea what I'm talking about...

This feels intuitive to my mental picture of the universe.

The description of this large scale structure and the expansion of the universe has always put me in mind of watching the patterns form and reform from drips in a soapy sink or an elastic fabric being pulled apart.

In both cases, you end up with these big expanses bordered by dense stringy areas. That the motion of the stuff that snaps / shears / collapses or whatever into these strings and knots would be aligned seems perfectly logical.

As the article points out, Woz was Chief Scientist at Fusion-IO [1] as well. That company struggled [2] before being acquired by SanDisk for less than the IPO price.

It was a different story in 2011 when Woz went on CNBC and said that the company "has grown as fast as Apple so far" [3] pushing the stock to the highest point it would ever reach just a month before the share lock-up expiration [4].

1: http://en.wikipedia.org/wiki/Fusion-io

2: http://www.theregister.co.uk/2013/10/25/fusionio_flasher_fla...

3: http://m.cnbc.com/us_news/45074931

4: http://www.forbes.com/sites/ericsavitz/2011/12/06/fusion-io-...

Yeah, those are layers in the image. That image [1] (727MB) is built on buildpack-deps [2] (695MB) which is itself built on the base jessie (154MB).

It would be pretty straightforward to adapt those Dockerfiles to create an image which includes only the dependencies for building node. It would end up looking a lot like 'node:slim' [3] (288MB).

Ideally, Docker will eventually have the functionality to more easily strip out transient requirements like build dependencies from the final images.

1: https://github.com/docker-library/node/blob/013858ac35afb9ca...

2: https://github.com/docker-library/buildpack-deps/blob/a201b1...

3: https://github.com/docker-library/node/blob/013858ac35afb9ca...

>"We don't need more complex provisioning tools. We have plenty of provisioning tools."


Thankfully, the Docker team seems in agreement with this based statements about avoiding making Dockerfiles "too clever" and the response to various proposals.

As you point out, most of the "issues" here are really misconceptions.

I expect it's a tough balance for any new(er) project. Maximizing exposure and adoption, but avoiding negative perceptions from being applied in ways aren't optimal.

Dockerfiles are deliberately dumb to let other tools take over as necessary, is my understanding.

My experience (over the last year) is that they're so limited as to be pretty useless. They don't even do what they're advertised to do, ie give you a reliable way to reproduce a build, and they're inflexible for my idea of real-world work with Docker. Where they're good is in giving everyone a point of reference.

I had a discussion with the maintainers last year about this:


I have a problem with most CM tools in that they're for moving target systems, not immutable ones. Ansible is the closest, but our experience has been that development on it is slow relative to the tool we use (see below). It's saved us a ton of money.

I blog on this and similar topics here:


The "tool for building and maintaining complex Docker deployments" is here:

http://ianmiell.github.io/shutit/ https://github.com/ianmiell/shutit

I also talk about this here:


>"They don't even do what they're advertised to do, ie give you a reliable way to reproduce a build, and they're inflexible for my idea of real-world work with Docker."

Not exactly, as the thread you link points out you can reference an image ID in FROM rather than the name:tag which has potential to change silently.

It's the equivalent of using a package manager against a repo you don't own without pinning - expect problems.

This can be mitigated by FROM'ing via ID or avoided entirely by running your registry where tags are reliable.

Admittedly, these things are not necessarily obvious, but I think it's a bit disingenuous to paint Dockerfiles as worthless or broken.

That said, ShutIt looks very cool and seems to address exactly some of my concerns / desires about working with Docker.

I just don't agree with framing it in opposition to and at the expense of what exists.

There's value in a container description that is fully self-contained, transferable and 'dumb' enough to be transparent.

Hi, yes you're right - you can reference an image ID. However, as soon as you go to the network you're lost - any apt-get/yum update or install could break your system in surprising ways.

Having done _lots_ of builds lately I can vouch for that (see my blog for some examples).

In the end the image ID _is_ useful, but the dockerfile itself has limitations.

I agree with your last point as well - my evangelism comes from solving problems at my company in this way (which I know are not uncommon problems) rather than any belief that it beats others objectively.

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