It's not granular enough for small scale/personal use. If someone were considering using S3 for a small project, or even just personal backup, this wouldn't really help them.
The sliders should offer some way to get more granularity. Maybe scale the sliders the higher they go, or have a separate option to select the scale you want to work on.
I was thinking the same thing. I've just started using S3 on a rails project for learning purposes and (I think) it's free for the amount of data I'm using.
Good point - I thought about this a bit and figured I could take one of a couple of approaches 1) log scale (probably a bit confusing / harder to implement) 2) a switch between personal and business (could do similar switches for other settings like region
Given Heroku's popularity and the simplicity of its sliders, why has Amazon not put in a few days to build this on S3? I'm asking genuinely as I'm wondering if it really is enterprisey oversight or it really gets them more signups or usage without providing this kind of tool.
As a historical sideline, Amazon is no stranger to sliders. Its diamond search (of all things) was one of the first uses of sliders afaik to narrow down a search, in the early days of Ajax (http://ajaxian.com/archives/ajax-showcase-amazoncom-diamond-... - mentioned in May, 2005)
Also surprising there's no affiliate programme for S3, coming from the mother of all affiliate programmes.
Doesn't work on my phone. I am used to seeing slider controls that don't support dragging on touchscreens, but these don't appear to support the fallback of tapping a point on the scale to move the slide to that point.
For me complex part is S3 pricing is IO requests. I have no clue how many requests expect. If you could find approximate values and make a list like: Webpage (1000 Read requests and 10 Write requests), Web Application (10000 read and 10000 write), etc. Or some other way to get an idea of what it should be approximately...
That's my biggest problem with pricing S3.. May be I am alone in that.
Compared to Hetzner. Not a cloud provider, but a root server provider. It's also more expensive than Strato - also a root server provider. Of course, not with the same availability, but for most startups it's more than enough.
You compare a hoster for servers with a storage solution. Don't get me wrong, I know that hetzner offers cheap iron and it's really amazing, but all this comes at the expense of the network and the support.
RE: S3 – it scales any time and on demand. I can upload 2 TB starting now and don't have to order a server which may or may not be around on Monday (tomorrow). You could compare S3 to something like Rackspace Cloudfiles or Nirvanix. Your Hetzner argument makes little sense – not even as far as TCO is concerned.
Also, good luck recovering your servers and storage array on a public holiday in Bavaria. ;)
Yes, you are right: if you need all those things you mention, than hetzner is not an alternative.
However most startups an small busnisses don't, and maybe never will. I mentioned Hetzner because it's the most known here, but there are many other similar providers.
I'm using several Strato root servers since almost 10 years and never had an incident so far, even if it's "low" availability.
Whatever stack or solution you want. IMHO what matters is the monthly bill: everytime I tried to use something from Amazon, I ended up paying too much. Of course, this is all relative and very subjective :).
Love the idea and implementation, thanks a lot for this!
Though more granularity, as it has been said, would be nice, this is good enough for me and lots of people as it is right now.
What I REALLY would like to see (and I know this is not really a feature request for this, but just an idea I'm throwing out there) is this made for other Amazon AWS services as well.
Common scenarios, like "I would like to run a wordpress site on an EC2 instance serving a couple thousand hits per month", or "I'd like to store a 1 TB backup of my personal computer on Glacier and retrieve it when my computer crashes (say in 1 year)".
The point is, like you said yourself, if you're spending a lot of money, you shouldn't trust anyone else, but having these "ballpark figures" would be really useful for people that are considering these services.
To that point, it's hard to get a handle on how many GET, LIST, POST, etc. operations one would routinely do in the course of a blog or something else. While that can get pretty complex (especially if you end up throwing varnish in to minimize calls to your datastore), a calculator including various rules of thumb would be great.
If you pay for Dropbox what exactly are you paying for? Software to communicate with S3?
I'll accept that the software (python+rsync+an icon) is arguably "value added", but for the paying Dropbox user, what they are getting is still S3 storage.
You're paying for a service that auto syncs your data across multiple devices efficiently and keeps versioned backups of everything.
The fact that they currently use s3 for storage is an implementation detail.
So basically it's the same as rsync and a little scripting, but without commands and scriptability? I don't use Dropbox so I'm curious. I'd like to know if they're offering something to paying customers that S3 does not, apart from their particular combination of python and rsync. If we just focus on storage is Dropbox more expensive? If yes, by how much? I think this is a valid question.
If you just focus on storage, yes, Dropbox is more expensive.
But this is kind of like saying iOS is just an expensive closed variant of FreeBSD.
Since I'm not drawing a salary from Dropbox's marketing division, I'm not going to go in depth here but I'll just say that Dropbox can be used easily and efficiently by normal people who have never seen a command line. Whereas your home brew python+rsync scripts can only be used by nerds like us.
It would be interesting to see how your homegrown sync solution works on various operating systems (desktop + mobile) and offers a web-based UI to do the same on the go.
my dad loves dropbox. he would be lost with s3. for most people it's not "arguably value added"; it's the difference between possible and not.
(another vote for a log scale (or something equivalent, like a separate "units" that sets the scale to 1s, 10s, 100s...). also, the tick marks should be on reasonable, round values)
then maybe you should have said something along those lines earlier, instead of repeatedly asserting that s3 and dropbox are equivalent (and then asking if you'd missed anything).
Do you read carefully? When I asked if I missed anything I was referring to anything besides the arguable "value add". My apologies if that was not clear. But the value add is precisely what you focused on, and argued (predictably): the python, rsync and the icon. I'm asking about the S3 component.
The sliders should offer some way to get more granularity. Maybe scale the sliders the higher they go, or have a separate option to select the scale you want to work on.