HN user

endriju

120 karma

We need to save life on our planet. If we act now, it will still take up to 75 years for our current actions to fully show their impact.

Posts0
Comments42
View on HN
No posts found.
[GET] "/api/user/endriju/stories?hitsPerPage=30&page=0": 500 Failed to fetch user stories

I think about this quite often, as 2 of my kids were born in Ireland, but grew up elsewhere (Slavic country). I speak English with them, but so do some of my friends/neighbors with their kids. The other kids would seldom speak English though, while my kids seem to default to English somehow. Btw. they left Ireland when they were 2y and 9months old (now 9 and 7), never attended daycare there, were only exposed to English at grocery stores, doctors etc. My point is, the early exposure to the language seem to make huge difference. I'm sure it's a combination of other things too but I'd say it is definitely a strong factor. [edits: grammar]

I think the problem with these general-purpose PaaS on top of Cloud is exactly that they're general-purpose. You will run into limitations and issues as soon as you start doing things only 5-10% of their customers need, like for example running privileged docker containers in kubernetes or wanting to manage parts of your infra with terraform. I'd much rather see tools helping me deploy one particular niche use case in Cloud really well. BTW I'm trying to follow this mantra while building https://staticbot.dev

I'm building StaticBot.dev. I was surprised how tedious the manual setup for hosting static websites on AWS infra still was after 2 decades in the industry, and as I wanted to put a few websites out there to test waters for some ideas I had recently, decided to tackle it myself. So basically scratching very own specific itch: deploying and managing a fleet of static project websites in AWS infrastructure with IaC and nice UI. I wouldn't myself use something with (hidden) vendor lock-in, so opted for "hybrid" approach where user can deploy conveniently using the tool but has up-to-date terraform code avaiable in s3 so can take over project deployment anytime. Not much of a business mind so I might open source the whole thing later on (though the value of code kind of plumetted lately as AI can generate it so well).

'"Third option is to launch a clone of the product with a different name, use it for capturing the ad traffic and then direct it towards the original product. This seems workable, but also a bit too experimental for comfort."'

You might also have the option to twist the above - similar to what I did with my product http://exmerg.com. I created a copy of it at http://gridoc.com and directed the traffic from the old product to the new one. Users gradually started going directly to the new one (gridoc) and the original web is now almost forgotten.

If you think that you are lucky to have good friends and helpful colleagues, maybe turn it around and think to yourself - they are lucky to have me!

Personally, I consider people who are humble or even low self-esteem as more interesting. Maybe it even correlates with intelligence? Idk, but the fact that you might feel like you lack confidence can actually make you look more interesting to people.

I know many people (including me) who try to minimize contact with with over-confident people when possible.

Running a 1-man SaaS as a side-project has been one of the best things career wise. It helped me gain a different perspective on software (design, devops, sales, support etc.) in general. Being able to bring some of that experience into my day-time jobs was really rewarding.

The Saas i'm running is Gridoc (https://gridoc.com) - it pays for the car and a few bills. I'm pretty sure it could be doing better but it has been on auto-pilot for a couple years now (started a family).

Soundex 8 years ago

A pretty easy way to discover how Soundex works can be playing with http://gridoc.com/fuzzy-matching/ - a tool for fuzzy record matching that supports Soundex and Levenshtein Distance.

Disclaimer: I'm the author of the tool.

"Vienna was the heart of an Austro-Hungarian empire of about 53m people that stretched from Innsbruck in the west almost as far as the Black Sea in the east. After 1867 the empire was divided into two: a Magyar-dominated Hungary, ruled from Budapest, and a heterogeneous, multi-ethnic, multilingual other half, ruled from Vienna."

In all fairness the Magyar-dominated Hungary could be attributed the same attributes, it was multi-ethnic and multilingual, especially in the suburban areas of the kingdom.

Ethnic map from 1910 [1] shows it pretty well (Croatia and Slavonia were left out from the picture).

[1] https://en.wikipedia.org/wiki/Kingdom_of_Hungary#/media/File...

I'm in desperate need of scalable Apache Spark cluster available through API that would make it easy to submit jobs that could process arbitrary size datasets but would let me abstract away the scaling part of the problem. I don't understand how there's nothing like that already considering popularity of Spark.

Basic income introduces a few drawbacks:

- talented people will not need to do consulting or daytime jobs to support their startups or side-project, which can lead to them living in a bubble

- there will always be people who spend all their income on gambling, drugs, sex, whatever - with basic income they just have more cash to waste (although this could be controlled somehow)

- low-skilled jobs aren't going anywhere soon - how do we want to motivate people to do jobs like garbage collection or road maintenance - this is already a problem in countries with exuberant social systems like France

I don't think WebSockets belong on a website (even an interactive one as discourse...

From the article "Discourse does not use WebSockets. Discourse docker ships with HTTP/2 templates."

The whole article is worth reading. It looks like the author is well aware the WebSockets limitations.

In 2014 I have worked on http://exmerg.com but it wasn't until recently for it to start generating some passive income. Still, the infrastructure costs (AWS) are higher than the revenue, so money wise the balance is passive (not counting the hours spent building the service which I could use for client work).

The positive side effect of building it was having additional source of motivation to learn new things in various fields like UI/UX design, cloud services, business operations and other SaaS related stuff.

The main lesson it taught me was how much hidden work needs to be done to attract users and monetize a web service. I think a start-up that would make it easier to deploy, maintain and monetize web services would be an interesting one.

Felix Dennis (†2014)

What’s It Like Then, Being Rich?

What’s it like then, being rich,

Knitting gold to warm an itch?

        Very much like being poor;

        Wealth is just a key — no more.

Why not share this magic key

To luxury — and start with me!

        Surely better that you earn it;

        Could I trust you to return it?

How does one become your ilk?

Is it bred in mothers’ milk?

        Many paths can lead to riches,

        Few in sunlight, most in ditches.

Wherein lies the difference

From us — this odd ambivalence?

        Envy, malice, obligations;

        Toadying from poor relations.

Grown far richer than his neighbour,

Why would any rich man labour?

        Wealth is salt in wine immersed,

        Quaffing but excites the thirst.

Salt my arse! It’s filthy greed —

How many homes does one man need?

        For some the trick’s in trading it;

        For others, in parading it.

I’ve seen. But surely, comfort pales

Perched on padded Chippendales?

        When affluence holds no surprise,

        Wonders come in other guise.

Aye, the eyes of tart and whore!

What might you miss, if you were poor?

        Time.  The luxury of choices.

        First editions.  Old Rolls-Royces.


Except his poems, I also recommend his great book on becoming rich http://www.felixdennis.com/books/how-to-get-rich/

Regarding the switching between different subjects or projects, what helped me was to quickly write down TODO notes of actions I should immediately take when picking up the subject again. For example "Finish exercise X", "Watch lecture Y" or "Run unit tests of project Z".

I found my self dive in to the subject much faster when there was no need to think about where to start.

Thanks everyone for the valuable feedback, i appreciate it! The option that i am most likely to go with is to open source and document the interesting parts of the code (parsing sheets, merging data) and keep the site as an example use case for the OS projects.

What I miss on the site is current offer of tours. Sometimes you have to "fake it before you make it". If that doesn't work out you can always switch to maintenance mode and come back to the project once you figure out what to do next.

When looking at your site, i realized how important the video is - it provides comprehensive explanation, especially when considering the product complexity. Anyway that's from June so I guess you've done a lot of progress since then.

Just a quick answer - I am actively developing with ExtJS for about 3 years now. It depends on what product are going to build. I always say that ExtJS is great for 'excel on the web' kind of applications - desktop like apps. The grid component is unbeatable, so is the layout system (which limits the CSS nightmare pretty much).

However there are caveats of which mostly discussed is steep learning curve. It's true that it takes some time till one uses ExtJS in the right way. Blog posts like this one are a good read before starting new project http://www.sencha.com/blog/top-10-ext-js-development-practic...

I have no experience with KendoUI, but from what I saw it is the closest match to ExtJS in terms of developing desktop-like JS apps. And it's free.

Edit: looks like KendoUI is not free anymore.