Ikizukuri, or, That Kinda Sick Frog Video

I’m sure at least some of you have seen the video now making the rounds of the Japanese woman consuming the bottom half of a frog as sashimi while the top half blinks at her from her dish. I’m not going to post it here as watching it has made me feel a bit queasy, but it’s fairly easy to find, Google “live frog sashimi” if you really want to see it. (Not for the faint of heart or those lacking a strong stomach.)

I’m pro-cultural understanding here, so let me clarify a few important points before we all go on a rampage about Japanese culture:

Ikizukuri (生き作り) is the name of the practice, and literally translates to “prepared alive”.

Yes, the animal – usually a sea creature of some persuasion, frogs, fish, octopodes, shrimp – is not killed and prepared, but just prepared. I am pro-cultural understanding, but I am not a cultural relativist, and in most cases I firmly believe that this constitutes vicious animal cruelty. Wanting fresh fish is one thing, but there’s something a little bit sick about arranging the meat around the still-moving corpse. I’ve had carpaccio prepared from a freshly caught tuna, but that tuna was pretty dead before it got sliced up. Fresh food is great; still-moving food not so much. Some say the movement is unconscious nerve twitches and that the animal is dead when it’s sliced open, but I am not a neuroscientist and I can’t confirm that. (I tried Googling around for it and couldn’t find anything that met my criteria of confirming or refuting – sorry, guys.) I mean, I hope it dies quickly, but I don’t know.

This is not common Japanese cuisine.

The restaurant it was at is called Asadachi. Asadachi (朝立ち) literally means “morning stand-up” or, in our English slang, “morning wood”. The restaurant name is indeed the same and it is intentional. It’s a fly-by-night operation in Tokyo’s lovingly named Piss Alley. Piss Alley .. well, let me defer to the experts here: http://www.tofugu.com/2012/11/23/tokyos-infamous-piss-alley/ (if you like reading about Japanese culture, Tofugu is great, by the way – the author, Koichi, writes in a fun and accessible manner).

The food Asadachi serves is more “virility food”, or folk remedies to make you hale and hearty. It’s actually fairly well-known for bizarre food, and I mean bizarre by their own standards, not ours.

Japanese cuisine does often include very fresh fish, but when I visited Japan, and in all the Japanese cooking I’ve witnessed, they buy their fish dead. My wonderful Japanese host mother, Momoko, made us a big plate of hand-rolled sushi the Friday before I left, and it was all pretty dead (and she knew I didn’t want the watered-down version – if they ate it, I wanted to eat it, no holding back).

This isn’t commonly eaten – it would be like claiming all Koreans eat dog, when in fact most Koreans (62% under the age of 30 according to Wikipedia) are mildly disgusted by the idea (and the dogs are livestock dogs, not commonly kept as pets). So, you know. Choose your words.

It’s not sushi, by the way.

It’s sashimi. To be sushi, it has to be served with shari, or sushi rice, which this is not. Sashimi is raw seafood served alone.

In conclusion: it’s sick, but no more sick than some of the stuff we do.

Oh yes, slaughtering animals and eating them semi-alive (or sometimes even just straight up alive – see odori-ebi [踊り海老] for details) is, if you’ll pardon my crassness, pretty fucked up. But we keep pigs in tiny cages and then let them bleed out when we slaughter them. We clip chickens’ beaks and let them live in their own filth. We drug cows to make them lactate constantly.  This is less a question of what we’re eating, but how we’re eating it. Most Hindus, for example, find the idea of eating a cow blasphemous. Yet we think nothing, most of us don’t, of ordering a burger and happily devouring it. We Westerners tend to think poorly of those who eat horse. (Fact: horse is nothing special. They sell it ground in some Québec supermarkets. Kangaroo is also not great.) Yet I feel ultimately that it’s the right of everyone to, on the whole, eat what they want. It’s just prudent and worth encouraging that it be done prudently, with care to the environment, and if you’re going to slaughter animals for food, just be decent. Don’t let them suffer.

Do You Really Need a Degree?

DISCLAIMER: MASSIVE BIAS

Lately I’ve been a little pokey, seeing what exactly it is is required for a technically competent QA person of my experience to get a job. I mean, I’m not necessarily going anywhere, just curious, largely about what’s required for a QA person vs. software engineer in test vs. software engineer, what folks are looking for, etc.

I’ve found a prodigious number of postings that on the first line demand a degree in computer science or similar, or equivalent.

Two things stick out to me here that I want to go into.

What is “equivalent”?

Really, though – given the sometimes-dramatic differences between the way schools teach computer science, what is “equivalent”? Do you just want 4 years’ experience? Experience in what? I did 3 years of my degree before I got a job – is that equivalent? I’ve got the fundamentals of CS for sure and stacked some pretty good testing experience on top of it. Also, I know a number of great developers, engineers, and general computer scientists who couldn’t test their way out of a paper bag. Additionally, some folks have this bizarro tendency to undervalue their QA people, pay them less, treat them as second-class citizens — no one who’s spent hundreds of thousands on a piece of paper is going to go for a QA position like that unless you make it better. If you want a technical QA person, you need to search for a technical QA person — not a developer.

Furthermore, let me tell you about how many times I’ve used things like Big O in my testing work.

0.

Now, don’t get me wrong — I can do it. I can glance over an algorithm and make a fairly good estimate of how well it’s going to run given the data structures and design. But asking specific Big O questions in an interview for a tester reveals that you’re a bit out of touch with what someone in testing actually does, unless you’re looking for someone who is developing internal tools rather than automated test suites. I have tests running for a SaaS suite whose player has been loaded 9,000 times since you began reading this sentence.

I’ve never needed Big O. I’ve needed “this is kind of slow, oh it has three nested for-loops, that’s why, let’s make it faster” or “are we really loading the ENTIRE JVM for this?”. O(JVM) = slow; that’s more valuable than Θ(log x). In practice, we need the difference between “satisfactorily speedy” and “not good enough”, and the ability to pick out where a bottleneck might be in a system rather than scritching some numbers on paper and leaving ourselves more confused than we started. Plus – in a job environment, I can have the CLRS textbook on my desk and peruse it at my leisure. No, I don’t remember Dijkstra’s shortest-path algorithm by heart. I know it exists and I know what it does. I could pick it out if I saw it. But I can’t write it down from memory, and I’m not sure why that’s not okay. Rote memorisation does not an engineer make.

So what do I need that degree for?

Let me put it to you in the simplest terms:

Considering the exorbitant price tag on a university education, my return on investment both financially and in terms of mental health is now negative. I watched friends be driven close to suicide by the stresses of college, and you know what? It happened to me too. That is not worth the time or money. I would rather get in another way, read books, learn on my own, gain experience organically, and have the same job, but I am not going to push myself literally to the brink of madness so someone can pat me on the back and give me a salary. I care more about being happy and healthy and maybe making a little less money than being miserable, indebted, but having some cushy job.

I can test software. I know my fundamentals, and you know what? I’m decent at it given the personal feedback I’ve received. I can write software, too, and if I need a nudge in the right direction, I can ask colleagues and friends if an algorithm exists for what I’m trying to do. I can search documentation. I can Google or StackOverflow for information (Joel Spolsky, thou hast been verbed).

You don’t need a degree to have a job.

Now, to be clear – to be something like a doctor, you do need some formal study or some way to prove that you aren’t a hack with a bone saw. But for software? Nah. Not really.

But Alana! How do we weed people out?

Read their resumés.

But I get eleventy bajillion of them!

There are other ways to weed someone out.

Did they write you a cover letter? Do they sound interested in the position, or are they throwing the same document at a few walls hoping it sticks? Do they have anything remotely relevant in the way of on-the-job experience?

Do they contribute to open-source? Trust: if there is one way to both make and break someone in terms of being a good tester or software engineer, it’s making them contribute to an open-source project of a reasonable size. There are people there who will have no qualms tearing their code apart over and over and over (and I do mean absolutely savaging it), and an engineer worth their salt will learn from their mistakes. You do not want to work with an engineer who gets emotionally attached to their code.

If there’s a genuine interest and at least some experience, talk to them on the phone! You will learn very quickly whether or not they have the chops.

But a degree means chops, Alana!

Let me tell you a story.

Recently I interviewed two people for a QA intern position. Mind you, this is how I started my career, and I had not a single whit of experience in QA. I didn’t know what an assert was, nor a positive or negative test. Nada. Zilch. I could code, sure, but QA? Nothing. I could (and did) go into the interview and think of test cases for a simple scenario, and actually had to be told “okay, that’s good, let’s move on”.

One was a grad student, the other an undergrad, and the latter was very, very green. Of course it would be inappropriate for me to detail any more about them or tell you specifically what was asked, but the point here:

The grad student could not think of a single test case or assertion when prompted.

Not one.

I’m not exaggerating.

The undergrad wasn’t a knockout, but did better thinking of cases to test the particular scenario (by which I mean, thought of a few rather than looking confused).

A degree proves nothing unless you talk to them. It can be literally a five minute phone call to gauge them, their interest, their personality, and if you’re unimpressed, end it there. Otherwise, it can be a full phone interview.

Screening someone via their degree .. it’s not the same as, say, Cisco credentials for sysadminning. Degrees vary wildly and I would opine that it’s best to make a note of it and then let the person themselves with their experience and ability prove their worth.

Job Interviews and Tech Companies

I interviewed an intern candidate the other day that reminded me of a friend of mine.

The guy was clearly terrified; I’d make a guess that it was probably his first interview for a job in his desired industry. He was trying to be laid-back but the sweating bullets and nervous laughs gave him away. Being an analytical gal and also a newly-minted interviewer, I was considering this and wondering why it was I liked him, and I ended up pinning down a few things he did very right that I suspect will be valuable for both me and you in the future, as well as a couple of general Things To Know.

1. Write a thank-you note.

I really can’t stress enough how valuable thanking your interviewer(s) for their time is. It’s a clichéd phrase but I absolutely mean it — particularly at a company with a laid-back vibe or even just a small operation, your interviewer is pretty likely to be someone you’d be working with. We’re people. After your interview, we go back to our desks and pick up where we left off in our code.

Hard at work. Right where we left off. Yup.

It’s an incredible testament to your social and communication skills to indicate that you are vocally grateful for someone’s time, especially if those people are people who will later be helping you out on a project (should you get the job, a-course). It also makes someone feel good to be recognised in that way and is a wonderful method of getting the people you talked to in your corner.

I’ve done multiple interviews for one position before, and a lot of times everyone has different strengths and weaknesses. A particular candidate being visibly interested in the job and gracious enough to even send a quick note thanking everyone for their time is a pretty solid way to edge out other candidates who might have been stronger in a particular area.

2. Talk about your extracurriculars and personal projects.

Particularly for companies that place a lot of value on culture, having a personality and talking about what makes you you is a better idea than you might have been told. I don’t really care much about what you did in school — I care if you liked it, of course, but what I really want to know is what really pumps you up. If you did a project in Haskell for school but like Node.js so much that you redid it just for fun, the one I want to hear about is the Node project because it gives me a lot more information about not only what you know, but what you want to know.

People who are really passionate about a thing will absorb knowledge about it at breakneck speed. If two people come in for a Ruby on Rails job and candidate A has more experience on paper but candidate B really loves Ruby on Rails to the point that she’s stayed up nights fiddling around with it, I’d give candidate B strong consideration in most cases.

When I ask if you have any cool projects outside of school or work, I want to know about projects outside of school or work, even if they’re not 100% relevant to the task at hand. I mean, maybe don’t tell me about your programmatic categorisation of your medieval torture devices, but other than that, if I ask, I really do want to know! Even having a side project speaks well of you. If you’re going for a Java gig but you built an analog synthesizer from start to finish, that shows me that you have the initiative to get things done and pursue knowledge of something that interests you.

3. Don’t bullshit your experience.

I don’t know about others but it really sets my teeth on edge when I know someone is trying to talk themselves up to a level they’re not at. Listen – I’m not stupid, okay? I know you’re out to impress me because I have a strong influence over whether or not we hire you, but remember – I am evaluating you for fitness in a lot of areas. I am also a master of bullshit myself, having taken AP exams in high school.

Good tip for life in general, actually.

Do not feed me bullshit on your resumé. Do not feed me bullshit in your interview. It’s really unimpressive when I can tell you’re trying to do everything short of straight up lying. It is incredibly refreshing when someone says “I’ve never worked with that technology but I’m excited to start learning it” or “I’ve never played around with x library, but I’ve used y package, which is for a similar purpose”. It is okay not to know something. Admitting you don’t know it rather than talking around it does two things:

1) Shows me you’re honest.

2) Lets me form a more accurate picture of whether or not you’re a fit for the position.

#2 is actually more important and I’ll explain why in a minute.

4. Evaluate me.

An interview is not just for me to evaluate you — it is also for you to evaluate me, so do it. Come with questions designed to determine whether or not you really want to be at the company, not just questions you think will impress me. Most people at a company want you to be happy working there, because at its most economical, a happy employee is a productive employee. Conversely, we want to be happy working with you.

We’re also usually a bit nervous. I’m a bit anxious by nature and also want my interviewee to have a good experience and leave wanting to come back and work for us. Having a conversation and asking me questions that you have as well as letting me go through the questions I have for you makes me leave with a considerably warmer impression than I would have if it were a robotic question and answer session. At the very bottom-of-the-barrel minimum, it shows me you’re interested in working for the company, and if you turn out to be a nice and congenial person, all right, double prizes!

Interviewing for Pizza Planet!

This is why #2 up in my last point is so important, by the way — we both need to be a good fit for each other if it’s going to be a happy relationship (and it is a relationship).

—

Anyway.

I should probably summarise here, but the only thing I can come up with is the ever-futzy “be yourself”. Throw some glitter if you want the full glurgy cliché experience. In the most practical way possible, I do mean it, though — be professional, but also remember that no one wants to work with a drone with no personality. The people I spend 40+ hours a week in the company of? Shit, I should hope we like each other.

Party on, Wayne.

Don’t Starve: The Little Sandbox that Could

If you’re one of those people who just adopted Steam for Linux and are practically chewing your desk for good games, I have got news for you.

Actually, it might just be me who’s chewing the desk, but, ya know. News.

Anyway, after S4L came out, I saw that one of the bestselling games was Don’t Starve. I’m persnickety about descriptions and it initially didn’t seem particularly interesting to me, but on a second go-round I figured it wasn’t gonna hurt to shell out $15 and if I didn’t like it, well, I could always trade it away for Half-Life 2.

I’m happy to report that I do like it, as a matter of fact. As someone who can be a wee bit tricky to please, it’s a fun game that leaves a lot for you to find without having the devil’s learning curve. I’m going to break it down into a few categories and rate each out of 5, as I’ll do with future game reviews: Gameplay, Eyes & Ears, Performance, and Overall. I also have a coupla neat tips I’ve found helpful that I’ll throw in at the end.

—

Gameplay: 4/5

I’m going to tell you right now that if you don’t like sandbox games, particularly ones with a strategy element, you are going to hate this game. The official genre is “open-world wilderness survival”, which is just a nice way of saying “sandbox that tries to kill you a lot”.

Opening up Don’t Starve and starting a new game plunks you in the middle of ostensibly nowhere with some wonk in a suit saying you don’t look so good. The objective of the game being not to starve, you begin collecting such things as grass, twigs, carrots, berries, and other items to help you achieve your objective, crafting weapons, and finding items. You need to be standing near a light source (typically fire) when night comes, lest you be eaten by the Grue. And no, it’s not me making the Zork joke there; it’s really in the game.

The game gets steadily more complex from there and has an almost dry humour in making ways for you to die. Your hunger getting too low means you starve. Your health getting too low means, well, duh. Your sanity getting too low means your nightmares start to attack you. I think we’ve covered the basics. Any player of games like NetHack or Battletoads will be unfazed by this, but one frustrating aspect for those not adjusted to such dynamics is that when you die, that’s it. You’re done. Everything’s gone. Game over, back to the beginning to pick more grass, I don’t care if you had an ice staff and were about to win the game; you die, you’re done.

The character selection is nice – you begin by only being able to play Wilson, but as you survive and die and survive (or move on to the next world), your XP goes into unlocking new characters with different strengths and weaknesses.

The day-to-day gameplay itself is straightforward. There’s a lot of grinding in the beginning, e.g. you’ll spend a considerable amount of time chopping down trees for logs to feed your fire. The object of the game is to find the five Things (one of them is the Box Thing), assemble them, and move on to the next world, but the map is large and this takes a considerable amount of time unless you’re lucky with the random generator. How well you get settled in the beginning is also a function of how lucky you got with where the game put you in the beginning, though it will generally put you in a spot that has a robust food source nearby. It’s challenging but not mean.

That said, you will die a lot. I got screwed over by bees, lightning, winter, the Father and Son Walrus Hunting Party, futzing around with spiders (admittedly that was my fault), and tentacles, just to offer a few dumb ways to die. There are a few ways to get around dying which I won’t spoil, but they leave you at a pretty sturdy disadvantage.

Overall, good. The frustration in sometimes getting started in a crap spot and dying whenever the game feels like reminding you who’s boss knocks it down a star, but 5/5 means perfect. This is, however, engaging and fun.

Eyes & Ears: 4/5

The game’s art style is borderline steampunk, quasi-macabre, but still manages to be cute in a kind of backwards way. There are some things in it that are straight up creepy (hounds come to mind), but on the whole it’s a cute 2-D cartoony-sketchy style that is well-rounded without being overcomplicated.

The soundtrack is repetitive without being annoying. It’s actually pretty good in terms of soundtrack; the music and sounds fit the game and, at least for me, induced the emotion appropriate to the situation. There are a few pretty good “OH GOD WHAT IS THAT” moments. Also, bunnies scream.

A few bugs seem to be here, i.e. the combat music will sometimes start playing when you’ve either done something like chopped a tree near a spider but neither aggro’d nor engaged it yourself or just not started combat, but they’re uncommon enough as to be minor curiosities rather than annoyances.

Performance: 4.5/5

Don’t Starve runs pretty well on my modest laptop setup. Ubuntu 12.04 on a ThinkPad Edge E430, 4GB RAM, Core i3, integrated gfx. It’s not strictly necessary to have a high-end mouse as there aren’t really any twitch skills required; my Logitech USB mouse does the job just fine.

There are a couple of blips that occur even when no other windows are open besides Steam, but I’m genuinely unable to determine if that’s me or the game. I don’t see it with other games that have more demanding graphics, though, so I can’t give it 5 stars.

Overall: 4.17/5

Don’t Starve is a cute and fun sandbox game. It doesn’t break any new ground and can be repetitive and frustrating for those unprepared for the challenge of starting anew every time you die, but if you get past that, it’s gleeful entertainment and a good way to lose track of time.

Fun Tips from Your Friendly Neighbourhood Casual Gamer

A few things I’ve found it advisable to do from the get-go:

1. Make it your business to find a pig village.

Pigs are friends and food! A pig village has a number of advantages and I’d advise finding one and setting up camp there. Pigs are neutral mobs, though hitting one will aggro all the pigs in the area, which could leave you very dead in very short order. You can also befriend them temporarily by giving them meat. They’ll follow you around and kill things attacking you. A group of 5 pigs makes you borderline invincible.

There’ll also be a Pig King. Things you either find lying around or dig up from graves can be traded with the Pig King for gold nuggets, which you need for things like the Science and Alchemy machines.

Also, most of the time, where there are pigs, there are spiders, and aggroing a few spiders and kiting the pigs into them is a pretty good way to farm meat, silk, and Spider Glands. Three spiders will make short work of a pig, and conversely, two pigs will make short work of a good number of spiders. Also, if the spiders kill the pig, get in there and kill the spiders — they’ll eat any meat drops. Spiders sometimes drop Monster Meat on death, but seriously, don’t eat that. Don’t feed it to the pigs either unless you want Werepigs (hint: you don’t). Save it – it’s an ingredient for some item recipes later on.

2. Ignore swamps unless you have friendly pigs or there are Merms nearby

Tentacles will beat you like a rented mule. However, much like pigs, Merms can be pretty easily kited into tentacles, and depending on who wins, you will find some combination of fish, frog legs, monster meat, tentacle spots, or a tentacle spike (which is the second-strongest weapon in the game). Tentacles also only attack one target at a time, so having a group of pigs or Merms go to town will take it out in pretty short order.

3. Digging up graves is worth the sanity hit most of the time

Unless you’re really straight up losing it, digging up graves is a valuable source of gold nuggets and gems. The items you get are largely useless except for trading with the Pig King for gold. A Hardened Rubber Bung is worth 8 gold nuggets, and they’re relatively common if you go digging.

You can recover sanity by having friendly pigs around or picking flowers if you don’t have any sanity-improving items on.

4. Leaving a cave open near your campsite could prove more than you bargained for. Also, don’t go spelunking unless you’re really for-sure prepared

If you leave an open cave near your campsite, you will be subject to Batilisk attacks every single dusk. Considering the fact that they swarm, they will make short work of you. Fortunately, they are deterred by fire.

—

Until next time, party on, Wayne.

3 Language Learning Resources I Like and So Should You

Anyone who knows me knows I nerd out hard over language. I’d give my metaphorical left nut to sit around and learn languages all day for a living if that were a thing anyone got paid to do.

But nobody does, so I am reduced to slinging JavaScript for my supper and indulging my quasi-healthy obsessions in my off time. I’ve found that some resources are vastly more effective than others for me, and I would reasonably suspect from both numerous anecdotal testimonials in some cases and unfounded speculation in others that it’d be more effective for others, too.

1. Duolingo

Duolingo is a free website that has succeeded at gamifying learning just enough to keep users interested without sacrificing elements of teaching. Its business model is such that learners can use it for free while doing lessons and translating documents, while individuals or entities wanting a document translated pay for the service. Also, the owl mascot is freakin’ adorable.

You can do as many or as few lessons as you want in a sitting. It’ll send you reminder emails to do a few lessons each day, and holy crap does it guilt you when you miss a day. “You made the owl cry” on Day 2 of missed lessons tugs on the heartstrings of anyone with a soft spot for sobbing green owls. Seriously, I was afraid that at day 7 I’d get “the owl is on a hunger strike. You monster.”

At any rate, it’s a decent system, though it doesn’t have much in the way of encouraging retention other than pressing you to practise daily. A convenient feature is being able to automatically practise whatever your weakest words are. You can also test out of lower levels if you’re already proficient in a language and just want to do more advanced lessons.

You begin with four hearts per lesson session (three in later levels) and lose one for each wrong answer. If you lose all of them and then get another question wrong, you have to start the lesson over. It has a feature where you can speak into a microphone and it will evaluate your pronunciation, but I’ve never been able to get it to work with Ubuntu (or perhaps my microphone isn’t good enough – either way). You can turn this off, though, and only receive translation tasks. The downside here is that to get anything really beneficial, you either need headphones or to be somewhere where voices from the magic box speaking in strange tongues won’t cause nationalistic hysteria.

It currently has Spanish, French, German, Italian, and Brazilian Portuguese for English speakers, and each of those languages’ native speakers can learn English (except, oddly enough, German). Rumours abound about which language they plan to add next, some plausible, some less. (Polish: plausible. Mandarin: less plausible. Anything with a non-Latin alphabet would be tricky, and a language written in logographs would be halfway impossible under Duolingo’s system.)

Considering it’s free and has an unbelievably user-friendly interface, it’s definitely worth a go.

Duolingo (free)

2. WaniKani & Textfugu

Okay, so this is kinda cheating, but the two go hand in hand far too much to ignore one for the other. These are specifically for Japanese.

Both products are the brainchild of the masterminds at Tofugu, a Japanese culture blog. WaniKani is a genius kanji-learning application, while Textfugu is an interactive textbook that fills in the blanks of the Japanese language (i.e., grammar, some extra vocab). WaniKani will eventually attempt to get through all the Jouyou kanji, or, the standard set of 1700 or so kanji you need to comfortably read a Japanese newspaper. Textfugu follows along.

WaniKani uses a spaced repetition system (or, SRS) to instill memorisation and recall, as well as a bunch of totally ridiculous mnemonics (said affectionately – ridiculous though they might be, they work). The first review of a learned item is 4 hours after the lesson. A correct answer pushes the next review to 8 hours, then a few days, then a few weeks, then a month, so on and so forth. It tries to present the item to you right when you’re about to forget it.

WaniKani is currently in private beta, though requesting an invitation will usually procure you one pretty quickly. Textfugu is available generally. Both give a taster of the material to see if the user likes it (and it’s a decent taster, not just a honeypot to steal money) and then have a subscription service. Textfugu costs $20/mo or $120 for forever, whereas WaniKani is currently $8/mo or $80/yr, which will go up to $10/mo or $100/yr once it’s in general availability. Users who bought in at the $8 level are locked into those prices.

I have to say that having paid for both has done me invaluable amounts of good in terms of actually remembering the kanji in front of me, and is worth every penny. I will also say that, for a native English speaker, Japanese is a major time commitment, especially if you want to get the kanji into your long-term memory. Try them, and if you’re willing to commit to really learning the language, they are superb.

WaniKani & Textfugu (freemium; $8/mo and $20/mo respectively or $80/yr / $120/lifetime).

3. Lang-8

Lang-8 is a resource for those who already have some language proficiency and are looking to continue.

The way it works is that you post journal entries in the language(s) you’re learning. Native speakers then correct those journal entries (they can be about anything; I like to talk about what I had for lunch or what I’m doing right then for the languages I’m a beginner in). On the other side of it, you correct posts made by users learning your native language.

It’s a wonderful way to learn to speak and write naturally — many users are willing to participate in group Skype chats for practice speaking.

A free account lets you state you’re learning up to two languages (you can change them at any time but may only have two at any given time). Paying for an account (about $8 a month) gives you unlimited languages for learning and more visible journal entries for correction. Free accounts still get plenty of attention on their posts, though, so unless you want to learn 3 or 4 languages at once, like I do, a free account is just fine.

It can be a bit frustrating in that some people doing corrections are learners posing as native speakers, but these are few and far between and most of the corrections are valuable.

Lang-8 (freemium; free accounts allow up to 2 languages, premium unlimited languages for $8/mo)