My first guess would be that the domain uses a custom nameserver that logs all dns queries and serves back a different IP to each user. Then when the http request comes in to that specific IP, it can check the dns logs to identify the resolver.
But surely.. that would either require (a) a massive IP space (i guess that works if the site is IPv6 only), or (b) assuming a pretty short time between dns lookup and http request, and hence possibly false results if the dns lookup is cached somewhere and the IP has since been reused
I'd love to know if there's a better way of implementing this.
A better way would be to hand each client a unique hostname (sessionidhere.whatsmyresolver.stdlib.net) and match that dns query to the http client? For example as an ajax call from the main website.
SSL connection error
Unable to make a secure connection to the server. This may be a problem with the server, or it may be requiring a client authentication certificate that you don't have.
It's not a certificate problem, I think somebody's misconfigured the server. It's listening on port 443, but not actually responding appropriately to TLS traffic.
There are certain modes where it simply cuts the connection when the server is done transmitting. There's no length header or end-of-transmission character or anything like that.
Once I did a little research and wrote it all out, Gopher actually came out worse than I thought. Even if it were brought up to Gopher+ specifications (which haven't been implemented by anyone in 20 years), it would still be way behind HTTP/1.1 in terms of fixing design flaws.
You are criticizing gopher from the point of view of "it must have the same features as HTTP". Some of the things you call "problems" are just different ways of doing it.
And if the TCP sliding window is an issue for you, why not fix TCP?
No, these are design flaws that anything over TCP/IP needs to deal with. In particular, closing the connection to indicate the end of a file transfer is really bad (what if somebody pulls an ethernet cord?)
Fix sliding window? I don't think you understand TCP. This is an optimization to TCP to prevent the need to ACK every packet. It means fewer bytes on the wire and less CPU load for TCP processing, while also maintaining a respectable degree of reliability.
TCP checksums are not sufficient for data streams of more than a few kB.
> Let's be honest, you just need to know how are you going to deal with the file, not what kind of file it is.
How you are going to deal with it is a function of what kind of file it is, so to know the former you need to know the latter. The internet Media Type system is far from perfect, but much better.
> You need something along the lines of - this resource is readable text - this resource is binary data - this resource is an image
If that was really all you needed, you wouldn't need the subtypes in the internet Media Type system, while experience in the real world has shown that you do need that information.
Because I want to differentiate between JPEG, GIF, and PNG images? The how is going to be different for each one. Without the protocol telling me, I can only make a reasonable guess on how to process it.
Practical question - let's say you spin new EC2 on AWS. This results in VM generating new host sshd key for this instance as part of bootstrapping new instance process.
How do you place host sshd key in DNS records without first connecting to this instance and getting the key, action which requires you to check/verify host sshd key (which is NOT in DNS yet)?
The way I can think of is using IAM credentials to allow modification of a Route53 record using the AWS CLI from the newly spun server. You will need to write a short script that you could push to the new instance either through EC2 MetaData or UserData fields.
If there is a demand I could try whipping up a CloudFormation template to do it.
Either create the ssh host key outside and set it inside the image or integrate the new host into your configuration management and report the sshd host key back and export it to the zone database. Of course your AMI could also just automatically populate the DNS entry via some shared secret or API ...
>> Invalid point: Formulae are easy to follow and audience can easily verify/reproduce your work.
Actually this is another valid point of Excel.
Unless the financial model was done incorrectly (using VBA, complicated formula chains, unclear logic), it is very easy to verify financial model in Excel, which is a big plus.
Excel has been partly to blame for many bugs in mathematical models, in part because of its datatypes and in part because of its builtins. Some of these bugs are not at all easy to spot, depending on how you go about the building and the verifying.
What mathematical models? For your company financial assumptions and forecasts you generally need, in essence, the sum(...) function, percentages and multiplication, repeated ad nauseam, and if any of your formulas even have space for 'not easy to spot' bugs then the bug is in thinking of making such formulas in the first place.
Your spreadsheets would be big because of quantity of different things included, not because of complexity of those things
