HNHacker News
TopNewBestAskShowJobs

romaniv

2,750 karma · joined February 2, 2011

submissionscomments
romaniv··on Parsing JSON is a Minefield
If there are two servers/services that communicate via JSON and they use different parsers, these types of issues can lead to rather nasty problems. Even if both parties fail gracefully on their own.

This gets even worse if your software is an integration layer between two services you do not control.

romaniv··on Microsoft Surface Studio
This seems to compete with Wacom Cintiq more than any Apple product I'm aware of.
romaniv··on DDoS Attack Against Dyn Managed DNS
Withing a small hosting provider this might be pretty simple. Attacker might lease a bunch of new servers and get the IP that was recently released. Then they could launch a DDoS to force address resolution in their favor. It's a bit far-fetched, but a lot of very successful attacks seem that way until someone figures out a way to pull them off.
romaniv··on DDoS Attack Against Dyn Managed DNS
Not sure whether it could be used in a legitimate attack (probably), but it can definitely lead to confusing behavior in some scenarios. You switch servers, your old IP is handed to some random person, your website temporarily goes down - and now your visitors end up at some random website. Would you want that? Especially if you're a business?

Also, "commandeering" an IP of a small hosting might be easier than you think. It depends entirely on how they recycle addresses.

romaniv··on The Neural Network Zoo
Did anyone ever publish an academic paper about a sorting algorithm after merely showing that it beats mergesort time by several percent on couple of synthetic datasets?
romaniv··on JavaScript Fatigue Fatigue
>"web development itself is inherently complicated"

The functionality of 99% of all websites I see these days can be replicated with simple DB calls, string concatenation and a handful of straightforward progressive enhancement libraries on the client side. Some developers simply choose complexity.

The fundamentally complex aspects of web development aren't solved by the stuff people are fatigued by anyway.

romaniv··on Ewww, You Use PHP? (2010)
Syntactically, PHP is atrocious. But it has a lot of very good qualities as a web environment, which has nothing to do with the syntax of the language.

Easy to use documentation. "Batteries included" standard library that specifically caters to web dev. Shared-nothing architecture. No separate compilation step. Easy deployment via source code. Ability to use PHP files as configs and templates. In more recent times, autoloaders and APC cache.

As far the language goes, I also like their approach to dynamic invocations via __call. Very simple and powerful.

romaniv··on OpenAI is Using Reddit to Teach An Artificial Intelligence How to Speak
Might have been an exclusively German language problem, or maybe they cleaned it up since that article was published. I have a phone with T9 and it works remarkably well. Heck, I would prefer it to Swype. (Of course, it's not useful on smartphones, because they no longer have real keys.)
romaniv··on How it feels to learn JavaScript in 2016
The overall complexity was still order of magnitude smaller. There were no transpilers, shims, or complicated multi-step builds, and I knew every single library I was using.
romaniv··on How it feels to learn JavaScript in 2016
And to think that I went into web dev precisely to avoid dealing with complex build chains and horrible UI libraries...
romaniv··on WaveNet: A Generative Model for Raw Audio
Maybe I'm reading this paper incorrectly, but it seems that in this system "voice" is part of the model parameters not inputs. What they did was train the same model with multiple reader voices while using one of the inputs to keep track of which voice the model was currently trained on. So the model can switch between different voices, but only between those which it was trained on.

"The conditioning was applied by feeding the speaker ID to the model in the form of a one-hot vector. The dataset consisted of 44 hours of data from 109 different speakers."

Am I missing something?

romaniv··on Explainable Artificial Intelligence (XAI) Darpa Funding
Until your model starts misdiagnosing people in real life because your training data wasn't 100% perfect (which it never will be for complex real-life problems).

https://vimeo.com/125940125

The problem with unexplainable AI isn't just the lack of explanations. It's the fact that the entire model is a black box, so you cannot somehow tweak a single part of it without retraining.

romaniv··on Are we really so modern?
50 years ago was 1966. Judging by many SF novels from that time, people were actually rather pessimistic about computing power we would have right now. Almost universally, authors predicted some progress, but still thought computers would be expensive, large and not as powerful as what we have now. On the other hand, they were very optimistic about space travel, AI algorithms, robotics, energy prodution/storage tech, personal transportation and the state of industrial engineering.
romaniv··on PHP the Wrong Way
>If you're learning a framework, you have the entire community - which has probably written blog posts, asked and answered StackOverflow questions, and likely has some members who can answer your question at odd hours.

You're assuming that learning the framework will make understanding actual workload code completely effortless. In non-trivial applications this is never the case. You still have to understand... stuff like this, for example:

http://mikehadlow.blogspot.com/2010/04/custom-aspnet-mvc-mod...

When someone writes code from scratch the quality of the code depends on their ability. When someone uses a framework the quality of the code depends on their ability, the quality of the framework and fitness of the framework for a particular problem.

If the developer is not very experienced but sufficiently disciplined, and the framework is well-designed and fits the chosen task well, than yes, you end up with better code. If any of those conditions fail, the code is usually just as bad or worse than vanilla.

romaniv··on PHP the Wrong Way
> The problem is, if I inherit an application created with a well-known framework, I'll have a much easier time figuring out what is what, and looking for answers when something goes awry.

Unless they spend most of their code fighting the framework to accomplish their actual task. Which happens rather often.

Not to mention that learning a new framework is not fundamentally different from learning some person's approach to doing things. Whether or not it's easy depends on the quality of code and documentation. Given equal size of the code that does the job, reasonably written framework-free code is probably going to be easier to understand, because it doesn't have external dependencies.

I never understood why after years of dealing with dozens of competing and rather poorly constructed web frameworks in Java, PHP and now JavaScript people still believe that those things have some kind of inherent value. Remember Struts, anyone? PHP Nuke? WebForms? Would you rather maintain something that used either of those over something that was written in vanilla Java/PHP/C#?

romaniv··on LogicJS adds logic programming to JavaScript
Back in the day the two common use cases were expert systems and natural language processing (which are somewhat related). In this day and age I would most likely use Prolog for rule-based simulations or high-level game AI.
romaniv··on The Age of Absurdity: Why Modern Life Makes It Hard to Be Happy (2010)
If this was an issue with capitalism in general we wouldn't be discussing it this way, since the issue would date back at lease a century.
romaniv··on LogicJS adds logic programming to JavaScript
Well, the classic logic programming language is Prolog, so learning it would be a good start. Here is something fairly lightweight on the subject:

http://www.amzi.com/AdventureInProlog/advtop.php

There was a great PDF that explained how unification and substitution of symbols worked in a concise and readable fashion, but I can't find it right now. Hm, maybe this one:

http://www.cs.ubbcluj.ro/~gabis/plf/Doc/SWI-Prolog/AnIntrodu...

romaniv··on Flip Feng Shui: A new cross-VM exploitation vector
Does anyone know how step #1 works in practice? How can you detect which addresses correspond to same cells in neighboring rows?
romaniv··on On Cybersecurity and Being Targeted
Another interesting thing most people don't realize is that if you switch IP addresses Google demands a phone number for Gmail login, even if the account has no 2FA and no phone number was specified initially. "Give us some phone number to log in". Isn't that strange from security perspective?
romaniv··on On Cybersecurity and Being Targeted
I'm beginning to think that a possible solution is to use an email account entirely different from your primary email for all website registrations.

Which leads me to another idea. If websites supported separate email accounts for messaging and authentiaction/resets and kept the second account name hidden from general public it would (somewhat) help in a lot of cases.

Also, I would be much more likely to do 2FA on more accounts if websites gave me the option to use printed lists of tokens instead of my phone number.

romaniv··on How to use your full brain when writing code
You might listen to Bret Viktor's lecture on Seeing Spaces:

https://vimeo.com/97903574

romaniv··on How to use your full brain when writing code
Computer science has a plethora of studies related to OS process scheduling. All of it is directly relevant to human task management. I've never, ever seen in referenced in any management talks/books/etc.
romaniv··on Gradient Boosting Explained in 3D
Fourier transform is not a classification mechanism. You can transform something into frequency domain, but that doesn't inform you which frequencies are most relevant in classifying samples.
romaniv··on Gradient Boosting Explained in 3D
Can wavelet decompose be trained for classification of feature vectors with many dimensions?
romaniv··on NIST declares the age of SMS-based 2-factor authentication over
I would assume this depends on the nature of your account.

On most of my accounts the probability * cost of getting hacked is much smaller than the probability * cost of loosing access.

romaniv··on Complexities of an infinite scroller
Hm. As far as perceived performance, I recently wrote a proof-of-concept library using a timer and calculating viewport intersection of an element. It feels less annoying than a registered scroll event, although I did not do any instrumented performance comparisons. (The point of the library was progressive enhancement, not performance.)
romaniv··on U.S. Bans Theranos CEO Elizabeth Holmes from Operating Labs for Two Years
What if incumbents provide valid reasons why something wouldn't work?
romaniv··on Refraction – JS library to make modules independent, testable and re-usable
One interesting way of avoiding direct dependencies is to let modules interact with one another via reading/modifying DOM. It probably sounds strange, but the approach works well when you want to create reusable code that also follows the progressive enhancement principle.
romaniv··on AI fighter pilot wins in combat simulation
Comparing comments here to the comments on, say, first Alpha Go post, reveals the amazing amount of AI bias on this website.

When an expert system beats some human in a complex real-life problem the comments are about how it is narrow, boring, not sufficiently tested and ultimately doesn't matter.

When a neural network (with the help of MCTS and an entire data-center full of servers) beats some human in a board game the comments here hype it through the roof and jump to conclusions about the coming dawn AGI.

← PreviousPage 12 of 30Next →