Hacker News Reader

Comment by sunchild | Hacker News Reader

You realize how poorly your comments reflect on your product, right? I'll certainly never support Tarsnap after reading this garbage.

View on HN

I'd rather be poor, friendless, and honest than be rich and have friends only because I refused to air controversial opinions.

I'm glad I don't have any investors or employees who could be hurt, though.

You seem to think you are being brutally honest. You aren't, you're being painfully arrogant. There is a massive difference.

I expressed an honestly held opinion. It is possible that opinion was the result of painful arrogance. Those are not contradictory.

Typically people understand being brutally honest as providing a painful to hear but helpful or contributory opinion.

When your opinion comes from nothing but arrogance, ignorance, and a 4chan style need to draw attention to yourself by being deliberately insensitive, then your opinion isn't brutally honest. It's worthless, and you should keep it to yourself.

Taking the opportunity of a man's death to call him overrated is not speaking truth to power: it's tacky and tasteless by any measure.

Don't pride yourself too much on being "honest". It is far easier to go around always telling the truth than to apply discretion.

I don't agree with cpercival's comments, and I am quite saddened by the loss of Steve - more so than I would have expected.

But it should be pointed out that the comments in no way reflect poorly on Tarsnap. It's a fantastic product.

(other than the fact that the account balance alerts only consider storage used, not traffic charges. Grr.)

edit: Downvoted already? For being a satisfied customer? Really?

other than the fact that the account balance alerts only consider storage used, not traffic charges. Grr.

I wish I had a good solution to this -- but it's impossible to predict when someone is going to be using lots of bandwidth, and alerting based on the current total spending rate had a very high false positive rate since bandwidth usage usually spikes for a single day and then goes down to a small fraction of the storage cost.

The downvoting is probably for being off-topic.

Any discussion of the technology of someone who is running a small business is certainly on-topic.

So, what about using exponential smoothing or 95% peak estimation to get a better "current" spending rate? There is also the possibility to do some data mining on the actual usage (assuming you have collect that data and are willing to use it in aggregate to improve service) — i.e. customers who spend like you do and have a balance like yours usually run out of money after x days.

Another option I just thought of: try and estimate usage several different ways, and then check the variance — if its low, then issue an alert. I wouldn't be surprised if there is some research on how to do this, so it should be cook-up-able relatively easily if its a feature you want to add.

Exponential smoothing and peak estimation won't help with the fundamental problem, which is that warning based on storage costs alone produces the correct result -- in the sense that funds run out after exactly 7 days -- 90% of the time. Any non-trivial addition of bandwidth costs into the equation is going to hurt far more often than it helps.

Well, it doesn't produce the correct result for us, where bandwidth costs are more than 7 times storage costs, so we get our warning far too late.

Agreed.

cperciva has demonstrated a complete lack of tact or class, but neither of these changes the fact that he does make an excellent product.

Reply on news.ycombinator.com