I would actually question the effectiveness of using XML or JSON over HTTP as an RPC mechanism. In my experience I get much better performance rolling my own protocol with sockets.
Performance isn't everything. The protocol performance is usually dwarfed by application performance. Using standard data formats over HTTP has some huge advantages: all the pre-existing tools work. For example, you can test your services from the command line using curl, and use other existing tools to parse the data. Debugging a custom protocol and data format over raw sockets is much more difficult. You also get all the existing servers, load balancers, caches, clients, security tools, etc.
We need another layer, between the block layer and the filesystem layer [...] which is simply concerned with allocating some amount of space [...] Filesystems could sit above this layer
FTP is insecure for anything other than retrieval of public information (i.e. anonymous downloads). And for downloads, HTTP is far more efficient as it only requires one request/response. FTP requires five.
They hide the text using JavaScript. View the text-only version and see the replies at the bottom of the page.
You can get the same page as Google by changing your User-Agent to "Googlebot". In other words, they are cloaking. I use Google's Report a Spam Result every time I see Experts Exchange in my search results. Maybe Google will finally put an end to this behavior if enough people report them.
Ever notice that https://amazon.com/ gives a warning because the certificate is for www.amazon.com? Subject Alt Name can fix this. GoDaddy SSL certificates for example.com include www.example.com as a Subject Alt Name (see the cert for https://godaddy.com/).
They also sell certificates for multiple domains called "Multiple Domains Certificate (UCC)" which utilize Subject Alt Name. These certificates work fine in Apache, although you get an incorrect warning in the error log about the CommonName not matching the server name.
(I dislike many of GoDaddy's business practices and would never register a domain with them, but they seem to have the best value for SSL. Google "ssl" and click the GoDaddy advertisement for a discount.)
Why isn't this sort of audit logging (changing, data changing, etc.) built in to the database engine automatically, with a few configuration options available?
many small businesses use their SSN as their tax reporting number rather than an Employer Identification Number, especially for one-person businesses Perhaps they should apply for an EIN online? It is free, and takes about five minutes.
Ec2 is a virtualization platform designed for you to deploy entire server instances. But it's not all that great as a web server because Amazon, along with a total bandwidth charge, charges you per HTTP request.
EC2 charges for instance-hours and bandwidth. S3 charges for bandwidth and per request. You're right that S3 isn't great as a CDN, because that's not what it is designed to do.
From their perspective, either we click now or the ad was useless.
Actually, it depends on if it's CPM or CPC. With CPM, the advertiser wants to maximize clicks (high CTR). With CPC, the advertiser only wants relevant clicks (low CTR) to minimize their CPA.
CPC = cost per click CPM = cost per (thousand) impression CTR = click through rate CPA = cost per action (signup, sale, etc.)
I'm all for the outsourcing of farming jobs, because why should American farmers receive any kind of special privilege?
Because destroying the country's ability to produce food internally is a bad idea. What happens when externally produced food skyrockets in price, or worse, is not available at all?
I recently ordered 3 items with super saving shipping. I received 3 packages, all shipped on different days
Maybe they weren't all in-stock at the same time? It seems odd to do this, especially if they were all under $25, as the shipping costs might exceed the already slim margins. Amazon has a large and complex system that works differently in different cases. My post was meant to give a possible explanation, not to say it always works that way.
Everyone I met at Amazon was smart and wanted to do a good job. Seemly odd behavior is much more likely due to a compex system with insufficient customer transparency rather than intentional anti-customer business rules like "intentionally delay orders with free shipping". Delaying a customer's order to entice paying for shipping on future orders is silly.
If you get the "Super Saving" shipping (i.e. free) when eligible, they'll sit on your order for a several days. If I just buy the regular USPS shipping when he free one isn't available, they ship in a day.
Amazon doesn't intentionally delay your order, but the potential delay is documented. The delay occurs when you order items that come from different fulfillment centers. Amazon aggregates the items by trucking them to a single FC, then ships them to you. (This was explained to me by a co-worker when I worked for Amazon.)
I always order books with Super Saver shipping and they often arrive in three days.
a 0.01% chance of getting caught, calculated with a $10 000 fine/damages payment - so on those completely arbitrary figures an average $100 additional expense
LibTomCrypt provides more generic crypto services (and more of them). You should also note that I don't actively push LTC anymore (I use it, but after what happened to me on usenet I don't give a fuck if anyone else uses it).... LibTomCrypt is the past, not future.
That's too bad. High quality public domain code is rare. Thank you for all your hard work.
Python needs an equivalent to CPAN. I often use Perl instead of Python because of CPAN. PyPI just doesn't cut it. If I'm writing a quick script needing some library, it's quick to find it on CPAN and install with one command. With Python, you have to search PyPI, hope what you want exists on there, hope it works with the version of Python you are using, etc.
Perl packages have a standardized method for doing unit tests, and consequently, many CPAN modules have them. Python does not, so most packages do not have unit tests, or if they do, it isn't as easy to run them as make test.
[delete the existing file and create a new one in its place] is evil and anyone doing it should be KILLed on the spot
This is necessary due to filesystems lacking a transactional API. Suppose you edit a file and insert a line at the top. Your editor has a few choices when saving:
1) Open the existing file for write and rewrite the whole thing.
2) Make a backup of the existing file. Perform #1. Optionally delete the backup.
3) Create a new file, write the whole thing, then rename.
Option #1 has the advantage of preserving metadata, but risks corrupting the original if the machine crashes or loses power, disk fills up, etc.
Option #2 doubles the amount of disk I/O and storage required.
Slashdot Mirror
Performance isn't everything. The protocol performance is usually dwarfed by application performance. Using standard data formats over HTTP has some huge advantages: all the pre-existing tools work. For example, you can test your services from the command line using curl, and use other existing tools to parse the data. Debugging a custom protocol and data format over raw sockets is much more difficult. You also get all the existing servers, load balancers, caches, clients, security tools, etc.
Actually, they did. ZFS has the SPA and DMU layers that sit above the ZPL.
This is exactly how ZFS is designed:
See page 7 of this ZFS presentation for a comparison with the traditional block device model.
The best demo dialog that I've seen is from SHRDLU, written in 1968-1970.
FTP is insecure for anything other than retrieval of public information (i.e. anonymous downloads). And for downloads, HTTP is far more efficient as it only requires one request/response. FTP requires five.
They hide the text using JavaScript. View the text-only version and see the replies at the bottom of the page.
You can get the same page as Google by changing your User-Agent to "Googlebot". In other words, they are cloaking. I use Google's Report a Spam Result every time I see Experts Exchange in my search results. Maybe Google will finally put an end to this behavior if enough people report them.
Any public wifi is an untrusted network. Anyone using that wifi could be an attacker.
Ever notice that https://amazon.com/ gives a warning because the certificate is for www.amazon.com? Subject Alt Name can fix this. GoDaddy SSL certificates for example.com include www.example.com as a Subject Alt Name (see the cert for https://godaddy.com/).
They also sell certificates for multiple domains called "Multiple Domains Certificate (UCC)" which utilize Subject Alt Name. These certificates work fine in Apache, although you get an incorrect warning in the error log about the CommonName not matching the server name.
(I dislike many of GoDaddy's business practices and would never register a domain with them, but they seem to have the best value for SSL. Google "ssl" and click the GoDaddy advertisement for a discount.)
It is in Oracle:
(you might need to create a free login to view these)
Yeah, no one is actually using Amazon's EC2.
Ec2 is a virtualization platform designed for you to deploy entire server instances. But it's not all that great as a web server because Amazon, along with a total bandwidth charge, charges you per HTTP request.
EC2 charges for instance-hours and bandwidth. S3 charges for bandwidth and per request. You're right that S3 isn't great as a CDN, because that's not what it is designed to do.
when I asked [a Chinese kid] about the massacre at Tienamen Square, he said "What massacre?"
What happens when you ask Americans about the massacre at Kent State?
From their perspective, either we click now or the ad was useless.
Actually, it depends on if it's CPM or CPC. With CPM, the advertiser wants to maximize clicks (high CTR). With CPC, the advertiser only wants relevant clicks (low CTR) to minimize their CPA.
CPC = cost per click
CPM = cost per (thousand) impression
CTR = click through rate
CPA = cost per action (signup, sale, etc.)
I'm all for the outsourcing of farming jobs, because why should American farmers receive any kind of special privilege?
Because destroying the country's ability to produce food internally is a bad idea. What happens when externally produced food skyrockets in price, or worse, is not available at all?
3 bottles that look like liquor bottles
Those are Torani syrups (used to flavor coffee).
Plus, in the event of a sudden power off or crash, software raid can corrupt your disk if you're running with a parity disk.
RAID-Z is designed to prevent this.
One word: Costco
I recently ordered 3 items with super saving shipping. I received 3 packages, all shipped on different days
Maybe they weren't all in-stock at the same time? It seems odd to do this, especially if they were all under $25, as the shipping costs might exceed the already slim margins. Amazon has a large and complex system that works differently in different cases. My post was meant to give a possible explanation, not to say it always works that way.
Everyone I met at Amazon was smart and wanted to do a good job. Seemly odd behavior is much more likely due to a compex system with insufficient customer transparency rather than intentional anti-customer business rules like "intentionally delay orders with free shipping". Delaying a customer's order to entice paying for shipping on future orders is silly.
If you get the "Super Saving" shipping (i.e. free) when eligible, they'll sit on your order for a several days. If I just buy the regular USPS shipping when he free one isn't available, they ship in a day.
Amazon doesn't intentionally delay your order, but the potential delay is documented. The delay occurs when you order items that come from different fulfillment centers. Amazon aggregates the items by trucking them to a single FC, then ships them to you. (This was explained to me by a co-worker when I worked for Amazon.)
I always order books with Super Saver shipping and they often arrive in three days.
I know people who would commit felonies to have one of those for their page file.
Or their database log disk.
a 0.01% chance of getting caught, calculated with a $10 000 fine/damages payment - so on those completely arbitrary figures an average $100 additional expense
You did the math wrong:
0.01% * 10000 = $1
LibTomCrypt provides more generic crypto services (and more of them). You should also note that I don't actively push LTC anymore (I use it, but after what happened to me on usenet I don't give a fuck if anyone else uses it). ... LibTomCrypt is the past, not future.
That's too bad. High quality public domain code is rare. Thank you for all your hard work.
Python needs an equivalent to CPAN. I often use Perl instead of Python because of CPAN. PyPI just doesn't cut it. If I'm writing a quick script needing some library, it's quick to find it on CPAN and install with one command. With Python, you have to search PyPI, hope what you want exists on there, hope it works with the version of Python you are using, etc.
Perl packages have a standardized method for doing unit tests, and consequently, many CPAN modules have them. Python does not, so most packages do not have unit tests, or if they do, it isn't as easy to run them as make test.
Python really needs to copy CPAN.
[delete the existing file and create a new one in its place] is evil and anyone doing it should be KILLed on the spot
This is necessary due to filesystems lacking a transactional API. Suppose you edit a file and insert a line at the top. Your editor has a few choices when saving:
1) Open the existing file for write and rewrite the whole thing.
2) Make a backup of the existing file. Perform #1. Optionally delete the backup.
3) Create a new file, write the whole thing, then rename.
Option #1 has the advantage of preserving metadata, but risks corrupting the original if the machine crashes or loses power, disk fills up, etc.
Option #2 doubles the amount of disk I/O and storage required.