Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Three Months to Scale NewsBlur

blog.newsblur.com

Post reply on HN
Print view View original

71–80 of 146 posts

Re: Three Months to Scale NewsBlur

#71

Post by rssident »

conesus wrote: post #20
mynegation wrote: post #16

Earlier quoted context omitted.

I tried to install NewsBlur, but instructions were not very helpful. I tried to get as far as I can, editing fabfile and installing gajillions of new python modules, but still failed. Don't get me wrong, it is wonderful and very commendable of you to open source NewsBlur! Nor do I feel entitled to the source code or free ride (I will probably end up paying for the hosted solution). Just providing some feedback, given…

This is something I'll be working on soon enough. There are many, many dependencies (not only for the web app, but there are three DBs that you have to have installed as prerequisites - mongo, postgres/mysql, and redis). The real problem with setting up your own instance of NewsBlur is that you'll have to do your own feed fetching. This is effectively what you're paying for when you pay for NewsBlur. Let me break it…

You should make an intelligent feed updating algorithm.

If you have at least fifty previous entries from a feed it's easy to predict fairly accurately when that feed will need to be updated again.

That's how the indexer for http://rssident.com works. Saves you tons of cpu cycles.

Re: Three Months to Scale NewsBlur

#72

Post by conesus »

your_perception wrote: post #67
conesus wrote: post #31

Earlier quoted context omitted.

Originally price, now price + API for deploying new instances. If you check out my fabfile.py script you'll see that I have EC2 setup and ready but I don't use it because it costs more, runs slower, and is a pain in the butt to maintain.

I am guessing you're using DO VPS for very specific purposes. I have a VPS on it myself, but they do not provide any way for VPSes to network with each other, which is annoying for multi-host setups. If you have them communicating with each other, I'd appreciate it if you shared your approach. And congrats on your increased traffic.

It's the one manual step I have to do. After I provision a server, I copy the IP and stick it in a few places (my fabfile, my hosts file that I then re-copy to every other machine, and my HAProxy config).

While I do provision new servers regularly, I don't auto-scale, so I don't have a need for pure automation. It would be nice, but that's something I look into when I learn about new deployment procedures.

Re: Three Months to Scale NewsBlur

#73

Post by UberMouse »

conesus wrote: post #68
UberMouse wrote: post #55

Earlier quoted context omitted.

Why can't there be an option in the cast of 1. for it to automatically open the first unread story when I click "All stories"? Basically what I am doing by pressing the first unread button. I can understand not doing that by default since people may want to see their unread feeds, but I always just go to the first unread one anyway.

In that case, there's a preference waiting for you that you're going to just love...

Haha, excellent, thank you for helping me. Also "Text" view seems to be missing from the Default View preferences, is this intentional?

Re: Three Months to Scale NewsBlur

#74

Post by korg250 »

lazugod wrote: post #41
Roritharr wrote: post #7

Earlier quoted context omitted.

Could you atleast describe what is better about Feedly, so not only the creator of NewsBlur, but also anyone looking to spend money on an RSS Solution gets your informed input?

I briefly tried both. * Feedly has a simple interface which is much more like Google Reader. But it depends on a plugin (or an app) rather than being an allwhere accessible website. Their Kindle Fire app seemed to load content extremely slowly, and its interface wasn't a simple list. And it's free, so I can't be a customer and I don't know how they earn money. * NewsBlur is complicated and unresponsive. There are a l…

"and I don't understand why an arrow follows the mouse."

Really? It is a feature for better reading, so you can can easily see which line you are.

Re: Three Months to Scale NewsBlur

#75

Post by rssident »

lazugod wrote: post #41
Roritharr wrote: post #7

Earlier quoted context omitted.

Could you atleast describe what is better about Feedly, so not only the creator of NewsBlur, but also anyone looking to spend money on an RSS Solution gets your informed input?

I briefly tried both. * Feedly has a simple interface which is much more like Google Reader. But it depends on a plugin (or an app) rather than being an allwhere accessible website. Their Kindle Fire app seemed to load content extremely slowly, and its interface wasn't a simple list. And it's free, so I can't be a customer and I don't know how they earn money. * NewsBlur is complicated and unresponsive. There are a l…

Try http://rssident.com

I am working on making the UI more user friendly and adding OPML import/export.

It's going to be supported by advertising with a subscription service for premium features and no ads.

Re: Three Months to Scale NewsBlur

#76

Post by btipling »

mynegation wrote: post #16
conesus wrote: post #8

And I just want to remind the technical crowd here that NewsBlur is 100% open-source: http://github.com/samuelclay I've been asked quite a few times why I open-source the code. The answer is simple. Let me use an example from Joel Spolsky. (From http://www.joelonsoftware.com/articles/fog0000000052.html ) The dominant spreadsheet, with 100% market share, is Lotus 123. You're the product manager for Microsoft Excel. As…

I tried to install NewsBlur, but instructions were not very helpful. I tried to get as far as I can, editing fabfile and installing gajillions of new python modules, but still failed. Don't get me wrong, it is wonderful and very commendable of you to open source NewsBlur! Nor do I feel entitled to the source code or free ride (I will probably end up paying for the hosted solution). Just providing some feedback, given…

> I am not a web developer so that can partially explain my struggle with fabric, Django and multitude of webdev and devops modules that you use.

GitHub is not an app store. You can't expect to "install" a software as a service's code base. Open sourcing a project is opening it up to engineers, providing a self-hosted option on the other hand should be an easy to install alternative that probably requires another full-time engineer to handle as it needs to come with ability to upgrade, get logs, support and documents...

Re: Three Months to Scale NewsBlur

#77

Post by dangrossman »

sdfjkl wrote: post #44

That was surprisingly clueful of PayPal. Usually that part of the story is way more dramatic and ends up with alternate payment options.

Only in the stories. The majority of situations where PayPal calls probably end just like this. I've taken that call myself and didn't have a problem. The stories we read are almost all about people doing things they shouldn't with any payment processor; that's why it only gets worse when the processor learns the details.

Re: Three Months to Scale NewsBlur

#78

Post by thefuturewasnow »

recuter wrote: post #15
mynegation wrote: post #9

I wonder what Samuel's attitude towards taking external financing is. Because NewsBlur is having a hell of a momentum on Google Reader news (pun not intended).

Well now that he has a graph pointing up and to the right he can certainly raise money and collect developers as if they were Pokemon and play games with the valuation in the hopes that somebody Acquihires Newsblur or buys them for some opaque strategic reasons. Maybe Google will buy them! :) Edit: I kid, but that's actually a possibility in my mind and would be an ironic outcome. Maybe Facebook or Yahoo or somesuch…

it would be better if Clay keeps it and doesn't sell it. read earlier comment regarding one employee and little overhead. kudos to open sourcing the project and being able to support it. i'm pitching in even if it doesn't become my new go-to rss reader.

Re: Three Months to Scale NewsBlur

#79

Post by rodgerd »

The inevitable file descriptor limits on Linux means that for every database connection you make, you use up one of the 1,024 file descriptors that are allocated to your process by default. Changing these limits is not only non-trivial, but they don’t tend to stick.

You're clearly a talented programmer, but perhaps employee #2 should be a sysadmin.

Re: Three Months to Scale NewsBlur

#80

Post by jcampbell1 »

rdl wrote: post #34
conesus wrote: post #31

Earlier quoted context omitted.

Originally price, now price + API for deploying new instances. If you check out my fabfile.py script you'll see that I have EC2 setup and ready but I don't use it because it costs more, runs slower, and is a pain in the butt to maintain.

Oh, I meant vs. spending $20k on hardware, or partnering with someone like Softlayer/Rackspace as a promotional thing (I suspect either would be willing to give you a great deal to use as a marketing case)

My guess... DO offers hardware in 60 seconds. Also cheaper/faster than EC2.
Post reply on HN