I worked with someone that developed an online museum designed roughly to answer this problem. I'm afraid it was a long time ago and I've forgotten the URL, but it was in Scotland if that helps...
The basic idea is each exhibit had various 'tell me more' links that took you off in different directions, and the description of each exhibit was dynamically generated based on the other exhibits you had seen so if you start at say the Mona Lisa the description might simply say it was painted by the famous artist Da Vinci and it took him two years because he is such a perfectionist, and if you then go to say Cezanne it will talk about painting the abstract rather than realism, etc. Seemed a good idea to me anyway.
There are different sorts of senior developers. I know people with that as their title because they've been a developer at a company for twenty years (and have been slowly rotting for 19 of those). At a company like that I can easily imagine a capable young graduate earning more than the old guard - all it would take is for someone in HR to realise the problem and be quietly addressing it.
Monte-carlo will pick the path which leads to a win most often and this tends to produce the behaviour you've described. They still struggle when they're ahead in terms of guessing the most likely catastrophe and preventing it, I suspect that is the advance that will come next
Hardly sounds hard - the developer is encoded into the app. Do you really want me to post an algorithm? how about hash(UDID + developer name) as a first shot?
I'm not sure, the statistics would be pretty meaningless if it were opt-in.
Overall though, I'm not convinced it's a good idea. If I were looking for a calculator on my phone / tablet, I would want to know that it has a heap of capacity which I am not using. You'd probably pick up the day to day frustrations that using your app has by talking to people more effectively than gathering usage statistics.
For instance, imagine your app has lousy graphing capability (I don't know your app, so treat this as hypothetical). According to the statistics, you'll find your users don't use the graphing capability. If you log more than you claimed above, you might have enough to know that every user used the graphics once before giving up on it, or you might not. Either way, I'd be impressed if you could infer the root cause (wanted functionality, but what's provided sucks). A user study would pick this up quickly and accurately.
They are a bit random, you might get a few features which just one person wants - and you'll have to be careful about feature creep, leading questions etc, but those are all surmountable problems.
Somebody else noted that the standard deviation of ability of male chess players is higher but the mean is roughly equal, and that a similar property exists in mathematics, go and quite a few other pursuits. So I would posit the theory that males are more inclined to obsessive behaviour while females tend to have more balanced lives, but neither has inherently more aptitude for chess.
Two new business examples choosing Oracle... (not really disagreeing, more FYI)
1) Startup that is bankrolled by VC funding. The VC bring in experienced managers, used to working in a big corporate environment and they immediately start trying to recreate a similar environment in the startup. That's across the whole board, DR/BCP, HR forms, etc. not just Oracle.
2) Small company starts to move from data is not important to data is important and the IT guy has no open source experience. He compares the price of SQL Server and Oracle, and at the standard edition level Oracle is cheaper. Later the company grows and needs partitioning - they're not going to migrate to SQL Server enterprise and they're unlikely to have run across EnterpriseDB.
Oh, and one point I would disagree on. I think MySQL's existence is good for Postgres. MySQL strongly focuses on ease of setup / use by beginners and I think that that competition is good for encouraging Postgres to cater for beginners.
Right - there's not much point in #4 from a business perspective.
1) It costs money 2) Being caught is unlikely but the cost if you're caught is huge 3) They'd do it anyway without the bribe.
About the only real benefit of 4 I can see is the increased control - you could specify how large and when, which would help ensure nobody else steals the market before you
Right, and if you read the CNET article he mentions that he's already disposed of all the checked hardware.
He also mentioned that the extra cost of hardware + embarrassment of missing meetings due to being detained and missing flights means his business is losing contracts and money, and he's thinking of refusing international clients. Maybe that's the government's goal.
Well yes, but you only have to look at other countries to see an alternative - e.g. Japan a few years ago or China now.
In the US contracts are long and complicated because the cost of a mistake is high. The other side is likely to also go through them with a fine toothed comb for loopholes, and legal recourse is expensive, slow and unpredictable. In other countries, businesses will not do business with you if you have a bad reputation so there is a strong disincentive from exploiting loopholes - meaning that the contracts can usually be reduced to a few pages or less and what's written there does not need to be triple checked for loopholes. Similarly other countries court systems are much cheaper so that a day in court will cost you hundreds of dollars.
So no, I don't think you could eliminate the profession, but I think the US has elevated it to a place that is unhealthy for the rest of the economy, and lowering them down a few pegs would have many more benefits than costs.
Instead of someone breaking your 20 character password, all they have to do is find a password that hashes to the same as your SHA1 hash. Because of weaknesses in the SHA1 algorithm, any password contains only approximately 8 bytes (8 characters) of data. Put another way, until we improve off SHA1 it is not particularly useful to have a password over 8 characters because it's cheaper to crack the hash than the password anyway.
Roughly correct, one of the status codes returned means it is on fire.
This was a real issue, not an Easter egg. The old line printers were very fast and if they got jammed they got hot quickly and really did burst into flames, so if you get that error then you had better run into the printer room.
While I'm not familiar with the pricing of tickets at the globe, at the time there was a much bigger division between first class and third class tickets generally. So pandering to the masses was probably less effective.
Mine was in statistical learning and while on paper you could say the job I got when I left (data miner) was related, the reality is different. At uni I was developing new algorithms and pushing them to their limits, while in the workplace I was being contracted out at an hourly rate so it was much more about getting reasonable results quickly and so I didn't use any techniques beyond undergrad - clients simply wouldn't be willing to pay for the time it would take.
There were benefits to my work from staying at uni longer, such as better writing skills and more experience applying techniques, but I could've left after undergrad and got those just as easily in five years on the job.
Where I went if you drop out of a PhD then you will usually be given a MS if you have got to about half way, but you are not given one if you complete the doctorate.
(from a 1Password blog post, sorry about the poor English)
The 1Password file is very encrypted, yes. So any additional encryption is unnecessary. Opening it on your Mac doesn’t decrypt the file so others can see it — the decryption happens on your Mac, not on the network. So the file sitting there is still encrypted.
(from the 1Password FAQ)
The slightly longer answer is that your data is encrypted using AES, the same state-of-the-art encryption algorithm used as the national standard in the United States. 1Password uses 128-bit keys for encryption, which means that it would take millions of years for a criminal to decrypt your data using a brute force attack.
--
The point is that as long as your master password is decent, you can give the password file to anyone (or have DropBox compromised) and not have to worry about the passwords getting out.
If say 1% of spammers get hung out to dry like this and 99% get rich then many people would be willing to take those odds. Increasing the punishment of those 1% wouldn't help, you need to increase it from 1%. I think that would have a cumulative effect since it would start scaring people off and thereby increase your percentage faster.
It's also the reason that the RIAA concentrates on making its cases high profile - it's a reasonable substitute for more prosecutions.
He was fined roughly $100 per spam sent. If you say his spam messages wasted on average 10 seconds then he wasted 40,000,000 seconds or roughly one year FTE. So if you value the time of the average recipient at say $40k then he caused about $40k worth of actual damage.
I believe tripling it is the standard for punitive damage which would make a total fine of $120k. I personally think that's reasonable for a single small spam campaign but I can understanding people wanting to bankrupt him for live in order to send a clearer message.
Slashdot Mirror
I worked with someone that developed an online museum designed roughly to answer this problem. I'm afraid it was a long time ago and I've forgotten the URL, but it was in Scotland if that helps...
The basic idea is each exhibit had various 'tell me more' links that took you off in different directions, and the description of each exhibit was dynamically generated based on the other exhibits you had seen so if you start at say the Mona Lisa the description might simply say it was painted by the famous artist Da Vinci and it took him two years because he is such a perfectionist, and if you then go to say Cezanne it will talk about painting the abstract rather than realism, etc. Seemed a good idea to me anyway.
There are different sorts of senior developers. I know people with that as their title because they've been a developer at a company for twenty years (and have been slowly rotting for 19 of those). At a company like that I can easily imagine a capable young graduate earning more than the old guard - all it would take is for someone in HR to realise the problem and be quietly addressing it.
Similar.
Monte-carlo will pick the path which leads to a win most often and this tends to produce the behaviour you've described. They still struggle when they're ahead in terms of guessing the most likely catastrophe and preventing it, I suspect that is the advance that will come next
Hardly sounds hard - the developer is encoded into the app. Do you really want me to post an algorithm? how about hash(UDID + developer name) as a first shot?
I'm not sure, the statistics would be pretty meaningless if it were opt-in.
Overall though, I'm not convinced it's a good idea. If I were looking for a calculator on my phone / tablet, I would want to know that it has a heap of capacity which I am not using. You'd probably pick up
the day to day frustrations that using your app has by talking to people more effectively than gathering usage statistics.
For instance, imagine your app has lousy graphing capability (I don't know your app, so treat this as hypothetical). According to the statistics, you'll find your users don't use the graphing capability. If you log more than you claimed above, you might have enough to know that every user used the graphics once before giving up on it, or you might not. Either way, I'd be impressed if you could infer the root cause (wanted functionality, but what's provided sucks). A user study would pick this up quickly and accurately.
They are a bit random, you might get a few features which just one person wants - and you'll have to be careful about feature creep, leading questions etc, but those are all surmountable problems.
Somebody else noted that the standard deviation of ability of male chess players is higher but the mean is roughly equal, and that a similar property exists in mathematics, go and quite a few other pursuits. So I would posit the theory that males are more inclined to obsessive behaviour while females tend to have more balanced lives, but neither has inherently more aptitude for chess.
Two new business examples choosing Oracle... (not really disagreeing, more FYI)
1) Startup that is bankrolled by VC funding. The VC bring in experienced managers, used to working in a big corporate environment and they immediately start trying to recreate a similar environment in the startup. That's across the whole board, DR/BCP, HR forms, etc. not just Oracle.
2) Small company starts to move from data is not important to data is important and the IT guy has no open source experience. He compares the price of SQL Server and Oracle, and at the standard edition level Oracle is cheaper. Later the company grows and needs partitioning - they're not going to migrate to SQL Server enterprise and they're unlikely to have run across EnterpriseDB.
Oh, and one point I would disagree on. I think MySQL's existence is good for Postgres. MySQL strongly focuses on ease of setup / use by beginners and I think that that competition is good for encouraging Postgres to cater for beginners.
Which is a shame, because from a technical perspective I really like .NET.
Right - there's not much point in #4 from a business perspective.
1) It costs money
2) Being caught is unlikely but the cost if you're caught is huge
3) They'd do it anyway without the bribe.
About the only real benefit of 4 I can see is the increased control - you could specify how large and when, which would help ensure nobody else steals the market before you
Right, and if you read the CNET article he mentions that he's already disposed of all the checked hardware.
He also mentioned that the extra cost of hardware + embarrassment of missing meetings due to being detained and missing flights means his business is losing contracts and money, and he's thinking of refusing international clients. Maybe that's the government's goal.
Well yes, but you only have to look at other countries to see an alternative - e.g. Japan a few years ago or China now.
In the US contracts are long and complicated because the cost of a mistake is high. The other side is likely to also go through them with a fine toothed comb for loopholes, and legal recourse is expensive, slow and unpredictable. In other countries, businesses will not do business with you if you have a bad reputation so there is a strong disincentive from exploiting loopholes - meaning that the contracts can usually be reduced to a few pages or less and what's written there does not need to be triple checked for loopholes. Similarly other countries court systems are much cheaper so that a day in court will cost you hundreds of dollars.
So no, I don't think you could eliminate the profession, but I think the US has elevated it to a place that is unhealthy for the rest of the economy, and lowering them down a few pegs would have many more benefits than costs.
So does ever other database nowadays, even MySQL.
Interesting that you list only commercial DBs, do you have any trouble using postgres on very large databases?
Toyota Auris / Corolla Hybrid due 2011 at £18,950 (about $30k).
In general computations yes, but not in anything that is most naturally expressed as vector computations.
Instead of someone breaking your 20 character password, all they have to do is find a password that hashes to the same as your SHA1 hash. Because of weaknesses in the SHA1 algorithm, any password contains only approximately 8 bytes (8 characters) of data. Put another way, until we improve off SHA1 it is not particularly useful to have a password over 8 characters because it's cheaper to crack the hash than the password anyway.
Roughly correct, one of the status codes returned means it is on fire.
This was a real issue, not an Easter egg. The old line printers were very fast and if they got jammed they got hot quickly and really did burst into flames, so if you get that error then you had better run into the printer room.
Lets look:
List 1: Get the basics, hard working, motivated
List 2: Creatively brilliant
Except in very few cases, I know which I'd rather employ. I think you just demonstrated the value of big name Ivy League schools.
Yes, this and functional programming were the two subjects at uni which I do not use directly but significantly changed how I work.
While I'm not familiar with the pricing of tickets at the globe, at the time there was a much bigger division between first class and third class tickets generally. So pandering to the masses was probably less effective.
Mine was in statistical learning and while on paper you could say the job I got when I left (data miner) was related, the reality is different. At uni I was developing new algorithms and pushing them to their limits, while in the workplace I was being contracted out at an hourly rate so it was much more about getting reasonable results quickly and so I didn't use any techniques beyond undergrad - clients simply wouldn't be willing to pay for the time it would take.
There were benefits to my work from staying at uni longer, such as better writing skills and more experience applying techniques, but I could've left after undergrad and got those just as easily in five years on the job.
Where I went if you drop out of a PhD then you will usually be given a MS if you have got to about half way, but you are not given one if you complete the doctorate.
(from a 1Password blog post, sorry about the poor English)
The 1Password file is very encrypted, yes. So any additional encryption is unnecessary. Opening it on your Mac doesn’t decrypt the file so others can see it — the decryption happens on your Mac, not on the network. So the file sitting there is still encrypted.
(from the 1Password FAQ)
The slightly longer answer is that your data is encrypted using AES, the same state-of-the-art encryption algorithm used as the national standard in the United States. 1Password uses 128-bit keys for encryption, which means that it would take millions of years for a criminal to decrypt your data using a brute force attack.
--
The point is that as long as your master password is decent, you can give the password file to anyone (or have DropBox compromised) and not have to worry about the passwords getting out.
I don't think it works though.
If say 1% of spammers get hung out to dry like this and 99% get rich then many people would be willing to take those odds. Increasing the punishment of those 1% wouldn't help, you need to increase it from 1%. I think that would have a cumulative effect since it would start scaring people off and thereby increase your percentage faster.
It's also the reason that the RIAA concentrates on making its cases high profile - it's a reasonable substitute for more prosecutions.
He was fined roughly $100 per spam sent. If you say his spam messages wasted on average 10 seconds then he wasted 40,000,000 seconds or roughly one year FTE. So if you value the time of the average recipient at say $40k then he caused about $40k worth of actual damage.
I believe tripling it is the standard for punitive damage which would make a total fine of $120k. I personally think that's reasonable for a single small spam campaign but I can understanding people wanting to bankrupt him for live in order to send a clearer message.