HNHacker News
TopNewBestAskShowJobs

kolektiv

1,625 karma · joined December 25, 2008

London/Horsham (and anywhere else interesting).

Tech Theory + Practice. I care most about changing the world in meaningful ways through technology. I care very little about yet another X.

Contact me at andrew at xyncro dot com.

irc:kolektiv:irc

[ my public key: https://keybase.io/kolektiv; my proof: https://keybase.io/kolektiv/sigs/0d5gliRyPmqTUqjf5C4XNh5y1sry6und7d3z4LYWCyo ]

submissionscomments
kolektiv··on Docker - the Linux container runtime
kolektiv on GH too.
kolektiv··on Geoffrey Hinton has joined Google
Sigh. I don't know whether I'm alone in being saddened that many of the excellent thinkers of our time, who once might have spent their lives probing the the very limits of our comprehension in their field, now end up comfortably turning their minds to the services of (in the long term) rather short sighted commercial ends.

I can imagine that this sentiment is probably against the prevailing HN mood, but I've been thinking a lot lately about how certain kinds of thought and investigation are only enabled and supported by certain structures. The point of a university used to be being the highest pinnacle of thought. A place where people could have everything not related to intellectual pursuit taken care of so they could devote themselves to the pursuit of greater knowledge. Now that seems to be replaced by a corporate campus.

It's a cliche, of course, but how many breakthroughs of meaning are we missing out on because the brightest and best of our generation are now no longer seeking after truth, but instead seeking after a way to make more people click on an ad? Ah well, it's late, and as ever I'm post-sober. Best of luck to him and his team - I've enjoyed his lectures and papers. I hope they may continue, but I sadly suspect not.

kolektiv··on Spidering the web with CasperJS
Ah! Thanks for clarifying, that makes significantly more sense. I hope you didn't take my querying as too unkind!
kolektiv··on Spidering the web with CasperJS
I'm trying to work out what you'd want this for. I suppose as a quick "yes this does actually work" but I can't think of many less efficient ways of spidering things - whole headless browsers? Ouch.

If you needed to do this, you'd use something that already exists - if you just wanted to do it in Node, for some particular reason, why not just use plain requests and an HTTP parser?

I don't mean this to be overly critical, but just in case anyone's thinking this would be a good idea - there are better ones. Just to learn from, fair enough.

kolektiv··on B- environment merits B- effort
Laziness does not have to equate to being a slacker. Lazily (or inventively) pursuing your employers goals is not a problem - generally if you find a way to satisfy them with less work, everyone's happy. Slacking would be following your own goals (loafing around, hiding) when they conflict with your employer.

As the article says, if your people aren't bothered or motivated by your goals, they probably shouldn't be there (at least in skilled positions - of course there are plenty of jobs which are purely for the paycheck). It's not worth changing your company based on the attitudes and motivations of people who shouldn't be there anyway.

It is easier to hide in (many) big companies, but that doesn't mean that the solution should be to remove all potential hiding places - it should be to remove (or re-motivate) those who are hiding.

kolektiv··on Marissa Mayer is killing telecommuting, and that’s a good thing
Quite. "I'm giving up everything in exchange for huge stock grants, a 7-figure base salary and industry wide recognition" does not allow you to say "you should give up everything in exchange for 100k and very little job security".

If they want to, sure, but nobody has a right to expect that from others, in the same way as founders have no right to expect the same level of dedication from employees without the same potential for reward.

[edited for minor punctuation]

kolektiv··on Marissa Mayer is killing telecommuting, and that’s a good thing
> Marissa's gesture applies to Yahoos who been doing lots of personal and "side projects" away from premises at Yahoo's expense and not necessarily for Yahoo's benefits

Does it? Because from all the information provided publicly it seems to apply to all remote workers - are you sure they're all lazy moonlighters? Doesn't seem probable.

And don't apply your own experience to everyone else where it doesn't apply - I work from home a lot too - I have no dogs, no kids, no smells - in fact I have a very nice office, with better equipment and environment than any client I've ever worked with. But that's just me, of course.

kolektiv··on Resurgence of Neural Networks
Encouraging to hear, it turned up from Amazon earlier! Coincidental. Also arrived is Connectome by Seung - anyone read it yet? Opinions?
kolektiv··on Sexism is not funny, let's stop laughing
For the same reason that there should be jokes about pretty much everything. They're a form of human communication and expression which can be very valuable to group dynamics.

It is perfectly possible to joke about horrific things without intending to condone or support those things. That said, the place for these jokes is unlikely to be in groups of people you don't know at tech conferences. That's insensitive, rude and often intimidating - so don't.

kolektiv··on I found your mitten
That's very cute. But the pedant in me will not let this pass without noting that many of what you've found are in fact gloves. Mittens do not have fingers. I'm sorry to have to point that out :) On the other hand (groan) you are helping the semi-gloveless too, so well done for that.
kolektiv··on Political failure modes and the beige dictatorship
Just to point out that the House of Commons voted that way - that's the combined vote of all parties eligible. The Conservative party was rather split on the issue, I think only a (largish) minority of the Conservative party itself voted in favour (although I can't remember the split offhand). That's still rather better than you would have got a few years ago, but then the Conservative party is now basically split down young/urban/secular-ish and older/rural/religious-ish lines and votes on these issues tend to be somewhat predictable.
kolektiv··on In praise of failure: the key ingredient to children’s success
I'm partially only writing this reply as a slightly more "+1" than a simple upvote, but from my experience employing and leading people (programmers, generally) this certainly applies in the UK as well. Not only at elementary school age either, but right the way through education. More and more in the past few years I've come across people who've achieved degrees in subjects without ever really coming across a problem without a potential solution discussed in the near vicinity - perhaps an adjacent chapter.

When presented with just a problem, and no hint of where to find a solution, the end result has often been a mix of horror/terror/anger/paralysis. They have all of the knowledge required in the toolbox, but no practice at applying it in non-directed ways or of having to try (many) failing ideas before one works.

Very good luck with your teaching - I wish we had a lot of teachers over here who thought the same way.

kolektiv··on Show HN: Greater Peace of Mind Guaranteed, or Your Money Isn't Refunded
You have a typo in your subline: "Pledge money and vow to mediate for 30 days straight."

I am assuming mediate should be meditate, unless you really want my help in a peace process (not wise).

kolektiv··on How do you manage a team of geeks?
Couldn't agree more. There's this odd implication, after listing a load of usually pointless make-work tasks and tedious process, that geeks are somehow odd for disliking this. In reality, you're just dealing with a group of people who, because of the nature of what they do with a lot of their life, are perhaps better than average at seeing and analysing that state of affairs. The rest of your company probably doesn't like it much either, but they're perhaps not quite as sensitive to pointlessness as people who deal with it much more.

What you should be learning from the things that "your geeks" (blech) dislike, is what the rest of your company also dislikes but isn't telling you. The fact that you can fill in the timesheets for 15 people should show you exactly how pointless they are. You already know what they're doing! What are you trying to track exactly? I'm presuming you're not paying them based on the timesheets, otherwise they probably wouldn't let you fill them in.

That's just a small example. You don't have some weird sub-species. You have people who are attuned to certain things as a side effect of their job/life choices. Now work out what the rest of your company are attuned to, and go and listen to them as well.

kolektiv··on What one book could give me a new, useful superpower?
I should never comment after a fine and bibulous meal, but... I'm not sure how appealing I find the idea of a society where people thought in that way. Perhaps Bacchus is distracting me from the nuance of that point however.
kolektiv··on Deep Thoughts on the Food Tech Revolution by Captain Obvious McClure
Heh, well, I've just got back from a lovely restaurant, and I have to admit that they had a wonderful cellar and a well-matched sommelier, so this will be brief, as I'm having to re-read quite diligently!

I don't really think I'm being elitist - but I will admit that I'm talking about a sector within a market. Clearly I'm not talking about fast food, I'm talking about a specific type of dining experience. If this technology existed within a KFC (and was executed well) I'd like it (yes, I'm that weird guy that likes good dining and the occasional meal of absolute junk...)

I don't think that most restaurants currently piss off customers though - if they did, they'd generally be broke I think. I think you're assuming that the changes you suggest would make the best even better - and I genuinely don't think that would be the case. I think that character, personality, idiosyncrasy etc. are the makers of a great restaurant as much as the food, and I didn't feel like these ideas fitted that - at least within certain markets.

If you're ever in London and you fancy a discussion about what would really make for an incredible restaurant, and you fancy discussing it over wonderful food, give me a shout...

(Dinner's on me, of course)

kolektiv··on Deep Thoughts on the Food Tech Revolution by Captain Obvious McClure
God I'd hate this. Restaurants are places of relaxation, company, discovery, beautiful foods. I want to order something which sounds good and then see whether it is - I want it to come to me as a surprise - "Oh Tuscan Rabbit? I wonder how the chef will do it confit... Oh! Wow, let's try this then!"

It's not about efficiency, it's about atmosphere. I don't want a glowing screen anywhere near my dining table. I don't want anything which helps me order "what I had last time". I don't want them to ask me about the last time I was here 3 years ago either.

Can we invent and work on technology which adds to the beauty and experience of life next please? Either that or get the brain in a jar thing working.

kolektiv··on What Compsci textbooks don't tell you: Real world code sucks
I know what you mean, and if I thought this article was actually targeted at students, and with any intention of enlightenment, I would agree with you. I just don't feel it deserves that credit. It's wandering, poorly researched hearsay.

I would call it a hatchet job if I could tell by reading it what the target is, but it isn't even good enough for that.

kolektiv··on What Compsci textbooks don't tell you: Real world code sucks
Good grief, what a vast amount of supposition backed up by very little evidence or insight. "People often write bad code." Yes, we know that. "There's more than one reason for this." Yes, we know that too.

The rest of the article is just random presumptions thrown together, for what reason I couldn't say. The best developers work in finance because the money's highest? Yeah, OK then... (To be fair, I think there is a qualifying "considered to be" in there somewhere) K is impenetrable? Hmmm. Haskell is an offshoot of ML and unreadable? Umm, factually wrong.

There's lots wrong with software at scale, and the pressures of financial institutions along with high budgets etc do produce an interesting set of problems, but this article offers nothing new about anything at all in the field.

As an aside, I already don't trust an author when they start off with a claim that textbook code is good code! It's almost universally accepted to be trivialised and simplified to the point of "don't do this in the real world" simply because it has to be. Some of the worst code I've seen has been in textbooks (and that's worst as in "dangerous" rather than as in "badly written and in need of refactoring/replacement/enclosing in concrete).

Edit to note: I think it's actually beyond incompetent journalism and in to plain lying to present a piece of code as evidence of Haskell unreadability when the link to that code on SO refers to it clearly as "obfuscated Haskell code".

kolektiv··on Ask HN: Salary Increases Over Time
It seems like a horrible mangling of syntax. I don't see why you wouldn't simply say "...another whole year of experience..."
kolektiv··on How do we read code?
Seconded. Surprising how distracting something so trivial is given recent (unrelated) context. My mind immediately sprang to the conclusion that perhaps this sentence is partially applied - a legacy of an evening spent writing some interesting functional code for treating MIDI as observable sequences.
kolektiv··on Grammy for MIDI Creators Dave Smith, Ikutaro Kakehashi
Haha, seconded. That list got expanded as well to allow for multiple byte (segment) IDs, and the expanded list still feels like a blast from the past! The only thing about MIDI is that the official standard doc is still only officially available as dead tree. While there's plenty of good (and sometimes better) treatments of it online, I wish they'd just catch up a little - I don't really want to send a cheque from the UK to Cali' for it!

Of course, that's rather a digression - it's a surprisingly decent standard, and what it's enabled is startling. Well deserved, and congratulations to them!

kolektiv··on Uses For A Message Queue
I'll go for qualified agreement to some extent :)

You're right in saying that, for example, exactly once delivery is a symptom of a business requirement. I do agree, but even something like message de-duplication has issues with scaling in a genuinely distributed system. Admittedly, whether you choose to look at that as part of messaging (a term which I admit to possibly over-broadening sometimes, as I tend to look on it as more of an approach than a technology) is open to debate, but I do get where you're coming from much better now.

On the second point, I wouldn't describe my scenario as a distributed transaction - there's also very possibly no single server on which to do a classic database transaction - as soon as you do there's a scalability barrier. Anyway - it's more a difference of scenarios I think, and I'd agree that in the ones you've described the solutions sit outside of a messaging layer/domain.

Always interesting to hear, thanks for clarifying!

kolektiv··on Uses For A Message Queue
My commiserations on the train journey, I don't miss the Brighton <> London commute one bit (assuming it's the same Clapham, of course).

Guaranteed delivery is indeed easy if your messages are idempotent, but if they're not then life is significantly more difficult. Most messaging systems have an "at least once" guarantee, but building in an "exactly once" guarantee, especially with a system with any significant distribution doesn't seem easy at all (tell me if I've missed something though, but event ordering in distributed systems has a reasonably interesting history).

The same obviously applies to message ordering. These (presumably incrementing/timestamped?) ordering identifiers are fine if they don't have to cross a sender/batch boundary, but plenty of situations are more complex than that. How do you "simply" avoid the problem of either centralised synchronisation or time slip in ordering, as well as out of order delivery with major partition events?

Just curious as to whether you're saying that these are all easy problems, or that the domain is easy if you have enough constraints!

kolektiv··on Why I'm No Metrosexual
A little OT, but it does make me chuckle when you read this on a blog with little to no visual affordances. The title and the date are both permalinks, but there's no indication until you mouse over to check. That's fine - for the sake of cleanliness it's an acceptable trade off. And also, importantly, we've learnt that they probably are. Consistency is also important. Whether people will learn Metro successfully (statistically) etc. is yet to be seen I guess, but it's more complex than this.
kolektiv··on Marijuana Majority Website Highlights Celebs And Tech Leaders Who Support Pot
I genuinely don't know much about this question. What makes them more capricious than DUI limits for alcohol for example? People vary in their metabolism and intoxication levels there - is there much greater or less predictable variance with marijuana?

I also have to admit that trying to in effect argue that people should not be prevented from "driving stoned" (yes, I know that's not really what they're saying - I hope!) is a peculiar strategy to get the majority of people onside. I rather like a drink. But I wouldn't argue in favour of drink driving.

kolektiv··on You keep using that word "meritocracy"
"If we met the utopian ideal we toss around in blog posts, we’d still have lots of middle-aged women in this field. We’d have black people. We’d have Asian people – not a smattering, but a majority, cause the world is mostly Asian people."

Just had to pick up on this, purely because it's symptomatic of another form of blindness. We DO have Asian people. Go and look in Asia. There aren't many Asian programmers in companies in my current town because the demographic is not a small scale mirror of the world demographic. I wouldn't expect to find even 8% white guys in a programming office in Manila.

That's not to ignore the fact that certain sections of populations are, for whatever reason, dramatically underrepresented in companies in various prominent countries, but throwing around random meaningless soundbites like that doesn't help anyone's argument.

kolektiv··on Introducing Go by Example
Looks like an excellent project, with a really clear and simple approach. I'd love to see more of these approaches for other languages, but I'm looking forward to working through this one.
kolektiv··on Twitter responds to IFFFT
It feels like a fairly passive aggressive message. Kicking off with a "blame the victim" jibe in the homework line (I'm not commenting on whether they are actually the victims here, merely the phrasing), it strikes an initial tone of "this is their fault, not ours".

Moving on to an entirely disingenuous non-apology ("i'm sorry you were offended" vs. "i'm sorry I offended you"), it doesn't get much better. Then on to an appeal to authority - we tried it with other developers and they didn't care at all.

The thing is, I can't criticise from a business level. I have no idea what their internal world looks like (we can all have our own guesses). But the message is inept at best. The upshot is that it's hard to imagine anyone building something (innovative or not) on the Twitter platform. If it's within the TOCs today, would this message give you confidence that it will be next week?

By attempting to kick out half of your hangers on, you end up kicking out all of your hangers on. Seems like a lesson worth heeding for giants and hangers on alike.

kolektiv··on QRCodesDoneRight.com - let's teach people how to use QR codes right
The question isn't "is there utility" but "is there sufficiently more utility than alternatives to justify their use". I don't think I've come across a situation where the answer to that is yes. The barriers are high.

Most people a) don't have a reader b) don't really know what they are. Compare this to something like a URL/domain which people do recognise the format of, or a discount code ("use SUBWAY10 for 10% off!") to achieve the same goals.

Seriously, I would love to hear of any use cases where a QR Code was the absolute best approach.

← PreviousPage 7 of 11Next →