8 ms·

Show HN: I made a simple persistent counting API

54 points by dustyreagan 13y ago

dustyreagan 13y ago
I was doing some usability testing in one of my webapps and needed a simple way to count conversions. There are lots of A/B testing solutions out there, but I just wanted a persistent count I could increment on specific user actions. So I created the ArbitraryCounter.com. I find it useful, but I'm wondering if anyone else might. What do y'all think? PS. This is WAY beta, and mainly a proof-of-concept. But I'll continue developing it if there's an interest it.
- PeterisP 13y ago
  Perhaps POST requests would be more appropriate for incrementing/decrementing? GET's may be either cached(skipped) or duplicated for various reasons, resulting in incorrect counts.
- ankurpatel 13y ago
  Can someone explain how counters can be so hard to implement ourselves that we need to use this?
  - terabytest 13y ago
    It's probably not that they're hard, but rather that they might require some boilerplate code and bloat that somebody might want to offload to an external service like this one.
  - dustyreagan 13y ago
    The idea was to make something easier than booting up a database. Personally, I just needed a semi-persistent count that I could discard later. This fit the bill for me, and being able to retrieve the counts via JSON, I can easily add them to my Geckoboard dashboard.
Jake232 13y ago
I think you should be able to incr/decr by more than 1. Maybe allow http://arbitrarycounter.com/vb/fruit/apples+N http://arbitrarycounter.com/vb/fruit/apples+N where N can be any integer > 0
- pkulak 13y ago
  Yeah, batching would be good to encourage.
- angersock 13y ago
  X-Increment-Size: 10 Eh, eh?
- dustyreagan 13y ago
  I added an IncrementBy and DecrementBy variable you can pass in your POST request, and update the documentation. How's that work for ya'?
pwf 13y ago
GET should never modify data. I would recommend changing that to a POST before something indexes it and decides to crawl it every 5 minutes. See here for more reasons: http://stackoverflow.com/questions/705782/why-shouldnt-data-be-modified-on-an-http-get-request http://stackoverflow.com/questions/705782/why-shouldnt-data-... I realize this isn't supposed to be highly accurate, but it's something to keep in mind if it continues to grow.
- dustyreagan 13y ago
  Agreed. I updated the site and docs. Thanks! :)
geuis 13y ago
I love simple api's like this. Its kind of the same mental model I keep around http://jsonip.com http://jsonip.com. A simple service that does one thing really well. Its been working well and has grown to millions of requests a day. Keep working on this, I can see several uses already.
- ancarda 13y ago
  >millions of requests a day. I'm interested to know how much it takes to run jsonip? Does it cost a lot? Edit: I've just seen the Pro version. Are you turning jsonip into a freemium model?
  - geuis 13y ago
    It's a node.js app that has until recently been a single-process app. I just integrated clustering to take advantage of the extra cores Linode gave me. It's a simple but high traffic service and node has handled it like a champ. I have a relatively inexpensive account with Linode that so far has suited the site's traffic just fine. I'm exploring the idea of offering a higher-tier service. The general feature list is outlined on getjsonip.com but pricing isn't finalized. I'm still collecting signups and feedback right now.
- dustyreagan 13y ago
  Thanks! I appreciate that! :)
egonschiele 13y ago
> All URLs are public. We recommend using a unique group name to avoid collisions with other users. So anyone can increment your counter without you knowing? Seems like this means you can never rely on the count being right. As a side note, the link on the bottom is broken.
- terabytest 13y ago
  Unless they know what counter you're using or you make it public, it's unreasonable that anyone could mess with your counter.
  - egonschiele 13y ago
    This would be an example of security through obscurity.
    - cobbal 13y ago
      Yes and no. You could argue that use of passwords or private keys is security through obscurity, but that definition is pretty meaningless. Security through obscurity doesn't really apply when you can make the search space arbitrarily (often exponentially) big very cheaply, as is the case of URLs here.
  - dragonwriter 13y ago
    > Unless they know what counter you're using or you make it public, it's unreasonable that anyone could mess with your counter. ...or you choose convenient, meaningful $group and $id names, which makes even unintentional collisions not at all unlikely.
- dustyreagan 13y ago
  The idea was to launch the proof of concept with public URLs, then if I discovered people were interested, create private URLs. Thing is private URLs require an authentication system, which is fine, but I didn't want to overdevelop the proof-of-concept. I fixed the link. Thanks! :)
dragonwriter 13y ago
Incrementing or decrementing a counter is neither side-effect-free nor idempotent, and GET should generally be both, so you really shouldn't use GET for those operations. [1] POST would probably be the usual choice for these operations. (Though, since you aren't really creating a subresource but are instead performing a defined transformation on an existing resource, you could probably make a case for PATCH.) Web APIs are protocols implemented on top of HTTP, and should respect the semantics of HTTP unless there is a clearly-identified compelling reason not to. [1] http://tools.ietf.org/html/rfc2616#section-9.1 http://tools.ietf.org/html/rfc2616#section-9.1
- hnriot 13y ago
  why should they respect the semantics? Those were arbitrary and out dated. I don't see any need for that.
  - wcarss 13y ago
    Web crawlers and web page views index using GET; if GET affects your server state, so do page views and indexing. GET is idempotent. That's not outdated, and it's not arbitrary: it is literally stated in the protocol specification. It's a useful distinction created and used by professionals to maintain sanity in complex schemes of communication. Please do not ignore technical details because you think you're smarter than everyone else on the planet, or that the problems they solved a decade ago somehow went away in the face of node.js. edit: Just in case you (or anyone reading this) is unaware of it, http://www.ietf.org/rfc/rfc2616.txt http://www.ietf.org/rfc/rfc2616.txt (or pretty: http://pretty-rfc.herokuapp.com/RFC2616 http://pretty-rfc.herokuapp.com/RFC2616) Find Section '9.1 Safe and Idempotent Methods'.
    - mnutt 13y ago
      Naturally, it is not possible to ensure that the server does not generate side-effects as a result of performing a GET request; in fact, some dynamic resources consider that a feature. The important distinction here is that the user did not request the side-effects, so therefore cannot be held accountable for them. I think that, depending on what his API is trying to do, GETing a counter URL to increment it could be acceptable if his service's main purpose is to be accessible via client-side javascript. That being said, once you go down that road you have to begin taking into account everything that may depend on GET idempotence: send back proper caching headers as well as busting caches from the client side, decide whether or not you want to block javascript-parsing web crawlers, etc.
frankcaron 13y ago
"There's an API for that" is the new "there's an app for that".
- meowface 13y ago
  Yep. I understand this is just a little toy thing, but people are really trying to start businesses off of "APIs" that do not much more than OP's. The whole point of an API is that you defer a significant amount of serverside processing and logic to a remote host, and receive a nicely formatted and machine-readable response. You have to deal with the overhead of a TCP connection (handshaking / RTT), an HTTP request, and an HTTP response. If you're doing that every time a user visits your web page (or app view or whatever), especially multiple times, then you better have a good reason for doing so.

54 points by dustyreagan 13y ago

I was doing some usability testing in one of my webapps and needed a simple way to count conversions. There are lots of A/B testing solutions out there, but I just wanted a persistent count I could increment on specific user actions. So I created the ArbitraryCounter.com. I find it useful, but I'm wondering if anyone else might. What do y'all think? PS. This is WAY beta, and mainly a proof-of-concept. But I'll continue developing it if there's an interest it.

PeterisP 13y ago
Perhaps POST requests would be more appropriate for incrementing/decrementing? GET's may be either cached(skipped) or duplicated for various reasons, resulting in incorrect counts.
ankurpatel 13y ago
Can someone explain how counters can be so hard to implement ourselves that we need to use this?
- terabytest 13y ago
  It's probably not that they're hard, but rather that they might require some boilerplate code and bloat that somebody might want to offload to an external service like this one.
- dustyreagan 13y ago
  The idea was to make something easier than booting up a database. Personally, I just needed a semi-persistent count that I could discard later. This fit the bill for me, and being able to retrieve the counts via JSON, I can easily add them to my Geckoboard dashboard.

Jake232 13y ago

I think you should be able to incr/decr by more than 1. Maybe allow http://arbitrarycounter.com/vb/fruit/apples+N http://arbitrarycounter.com/vb/fruit/apples+N where N can be any integer > 0

pkulak 13y ago
Yeah, batching would be good to encourage.
angersock 13y ago
X-Increment-Size: 10 Eh, eh?
dustyreagan 13y ago
I added an IncrementBy and DecrementBy variable you can pass in your POST request, and update the documentation. How's that work for ya'?

pwf 13y ago

GET should never modify data. I would recommend changing that to a POST before something indexes it and decides to crawl it every 5 minutes. See here for more reasons: http://stackoverflow.com/questions/705782/why-shouldnt-data-be-modified-on-an-http-get-request http://stackoverflow.com/questions/705782/why-shouldnt-data-... I realize this isn't supposed to be highly accurate, but it's something to keep in mind if it continues to grow.

dustyreagan 13y ago
Agreed. I updated the site and docs. Thanks! :)

geuis 13y ago

I love simple api's like this. Its kind of the same mental model I keep around http://jsonip.com http://jsonip.com. A simple service that does one thing really well. Its been working well and has grown to millions of requests a day. Keep working on this, I can see several uses already.

ancarda 13y ago
>millions of requests a day. I'm interested to know how much it takes to run jsonip? Does it cost a lot? Edit: I've just seen the Pro version. Are you turning jsonip into a freemium model?
- geuis 13y ago
  It's a node.js app that has until recently been a single-process app. I just integrated clustering to take advantage of the extra cores Linode gave me. It's a simple but high traffic service and node has handled it like a champ. I have a relatively inexpensive account with Linode that so far has suited the site's traffic just fine. I'm exploring the idea of offering a higher-tier service. The general feature list is outlined on getjsonip.com but pricing isn't finalized. I'm still collecting signups and feedback right now.
dustyreagan 13y ago
Thanks! I appreciate that! :)

egonschiele 13y ago

> All URLs are public. We recommend using a unique group name to avoid collisions with other users. So anyone can increment your counter without you knowing? Seems like this means you can never rely on the count being right. As a side note, the link on the bottom is broken.

terabytest 13y ago
Unless they know what counter you're using or you make it public, it's unreasonable that anyone could mess with your counter.
- egonschiele 13y ago
  This would be an example of security through obscurity.
  - cobbal 13y ago
    Yes and no. You could argue that use of passwords or private keys is security through obscurity, but that definition is pretty meaningless. Security through obscurity doesn't really apply when you can make the search space arbitrarily (often exponentially) big very cheaply, as is the case of URLs here.
- dragonwriter 13y ago
  > Unless they know what counter you're using or you make it public, it's unreasonable that anyone could mess with your counter. ...or you choose convenient, meaningful $group and $id names, which makes even unintentional collisions not at all unlikely.
dustyreagan 13y ago
The idea was to launch the proof of concept with public URLs, then if I discovered people were interested, create private URLs. Thing is private URLs require an authentication system, which is fine, but I didn't want to overdevelop the proof-of-concept. I fixed the link. Thanks! :)

dragonwriter 13y ago

Incrementing or decrementing a counter is neither side-effect-free nor idempotent, and GET should generally be both, so you really shouldn't use GET for those operations. [1] POST would probably be the usual choice for these operations. (Though, since you aren't really creating a subresource but are instead performing a defined transformation on an existing resource, you could probably make a case for PATCH.) Web APIs are protocols implemented on top of HTTP, and should respect the semantics of HTTP unless there is a clearly-identified compelling reason not to. [1] http://tools.ietf.org/html/rfc2616#section-9.1 http://tools.ietf.org/html/rfc2616#section-9.1

hnriot 13y ago
why should they respect the semantics? Those were arbitrary and out dated. I don't see any need for that.
- wcarss 13y ago
  Web crawlers and web page views index using GET; if GET affects your server state, so do page views and indexing. GET is idempotent. That's not outdated, and it's not arbitrary: it is literally stated in the protocol specification. It's a useful distinction created and used by professionals to maintain sanity in complex schemes of communication. Please do not ignore technical details because you think you're smarter than everyone else on the planet, or that the problems they solved a decade ago somehow went away in the face of node.js. edit: Just in case you (or anyone reading this) is unaware of it, http://www.ietf.org/rfc/rfc2616.txt http://www.ietf.org/rfc/rfc2616.txt (or pretty: http://pretty-rfc.herokuapp.com/RFC2616 http://pretty-rfc.herokuapp.com/RFC2616) Find Section '9.1 Safe and Idempotent Methods'.
  - mnutt 13y ago
    Naturally, it is not possible to ensure that the server does not generate side-effects as a result of performing a GET request; in fact, some dynamic resources consider that a feature. The important distinction here is that the user did not request the side-effects, so therefore cannot be held accountable for them. I think that, depending on what his API is trying to do, GETing a counter URL to increment it could be acceptable if his service's main purpose is to be accessible via client-side javascript. That being said, once you go down that road you have to begin taking into account everything that may depend on GET idempotence: send back proper caching headers as well as busting caches from the client side, decide whether or not you want to block javascript-parsing web crawlers, etc.

frankcaron 13y ago

"There's an API for that" is the new "there's an app for that".

meowface 13y ago
Yep. I understand this is just a little toy thing, but people are really trying to start businesses off of "APIs" that do not much more than OP's. The whole point of an API is that you defer a significant amount of serverside processing and logic to a remote host, and receive a nicely formatted and machine-readable response. You have to deal with the overhead of a TCP connection (handshaking / RTT), an HTTP request, and an HTTP response. If you're doing that every time a user visits your web page (or app view or whatever), especially multiple times, then you better have a good reason for doing so.