Overheard in a cubicle farm

rusyn

"We've hardcoded the IP address in the database."

RogerWilco

@Brian Fernandes said:

"We've hardcoded the IP address in the database."

And? Depending on what problem you're solving, I can see this as a valid solution. This can be a WTF, but then you need to provide information why.

Unless you mean the word "hardcoded" refering to something stored in a database? I would not consider something in a database "hardcoded" usually, but it might depend on the kind of person saying it and their background.

rusyn

@RogerWilco said:

Unless you mean the word "hardcoded" refering to something stored in a database? I would not consider something in a database "hardcoded" usually

Ting

@RogerWilco said:

This can be a WTF, but then you need to provide
information why.

More after a chuckle than a WTF

@RogerWilco said:

Unless you mean the word "hardcoded" refering to something stored in a database? I would not consider something in a database "hardcoded" usually, but it might depend on the kind of person saying it and their background.

.

create table iptable (
ipAddress varchar(15) not null
)
go
insert into iptable (ipAddress) values ('127.0.0.1')
revoke all on iptable to public
grant select on iptable to public

There, its now hardcoded in a database.

FTFY

Evo

Maybe they meant they hardcoded it in a stored procedure? I can imagine that being said to be "hardcoded in the database".

tafinucane

Or the db normally stores a fqdn, then an application tries to use gethostbyname or similar to get the ip. This wasn't working for whatever reason, so they hardcoded the ip in the database.

rudraigh

I chuckled.

MiffTheFox

Pfft, everybody knows that hardcoding means using a VARCHAR or an UNSIGNED INT or something resonable like that. Obviously, the non-hardcoding, and therefore correct, way to do so would be to have four fields for octets as foriegn keys to an `ip_numbers` table. That way, if they ever extend IP beyond 255, we can just add more rows to the numbers table. It's future-proof!

morbiuswilters

@MiffTheFox said:

Pfft, everybody knows that hardcoding means using a VARCHAR or an UNSIGNED INT or something resonable like that. Obviously, the non-hardcoding, and therefore correct, way to do so would be to have four fields for octets as foriegn keys to an `ip_numbers` table. That way, if they ever extend IP beyond 255, we can just add more rows to the numbers table. It's future-proof!

Well, it's not that future proof since you would need 16 columns to support IPv6. Unless you are proposing that the "octets" be 32-bit. Then again, a 15-char varchar could be just as "future-proof" if you use Unicode.

belgariontheking

This thread is now about over-normalization and ipv1000.

Discuss.

morbiuswilters

@belgariontheking said:

This thread is now about over-normalization and ipv1000.
Discuss.

IPv1k? Hell, they already jumped to 128-bit addresses, where else can they go? "We will provide enough addresses for every single atom in the universe to have its own, private, 128-bit address space."

Jaime

@morbiuswilters said:

@belgariontheking said:

This thread is now about over-normalization and ipv1000.

Discuss.

IPv1k? Hell, they already jumped to 128-bit addresses, where else can they go? "We will provide enough addresses for every single atom in the universe to have its own, private, 128-bit address space."

Actually, that sounds like IPv7. All it would take is 208 bits to give each atom a 128-bit space. Currently, each atom can have its own 48-bit address space. Enough to allow every atom to have every legal ethernet MAC address simultaneously.

morbiuswilters

@Jaime said:

@morbiuswilters said:
@belgariontheking said:

This thread is now about over-normalization and ipv1000.

Discuss.

IPv1k? Hell, they already jumped to 128-bit addresses, where else can they go? "We will provide enough addresses for every single atom in the universe to have its own, private, 128-bit address space."

Actually, that sounds like IPv7. All it would take is 208 bits to give each atom a 128-bit space. Currently, each atom can have its own 48-bit address space. Enough to allow every atom to have every legal ethernet MAC address simultaneously.

2^80 != 10^80. The number of atoms in the universe is at least 2^266. Still, your basic point is valid: IPv4 to IPv6 quadrupled the address space. To be on the safe side, we'd have to quadruple it again to give every atom his or her own IPv6 address space.

@morbiuswilters said:

To be on the safe side, we'd have to quadruple it again to give every atom his or her own IPv6 address space.

For a reasonable price, of course.

morbiuswilters

@BC_Programmer said:

@morbiuswilters said:
To be on the safe side, we'd have to quadruple it again to give every atom his or her own IPv6 address space.

For a reasonable price, of course.

Hey, bloated, over-engineered standards don't write and toothlessly police themselves!

ammoQ

Still not enough. You guys always forget quantum physics. We need an IP address for every particle that might exist, even if it's just for a tiny fraction of a second and nobody notices it. Otherwise, sooner or later we'll run out of address space again.

dhromed

I am reminded of Problem Sleuth, in which Future Pickle, Past Pickle, Future-Future Pickle and Past-Future Pickle are summoned by Godhead Pickle to reform the cosmos by replacing every particle.

belgariontheking

@morbiuswilters said:

IPv1k? Hell, they already jumped to 128-bit addresses, where else can they go? "We will provide enough addresses for every single atom in the universe to have its own, private, 128-bit address space."

Well, to be fair, every update after ipv46 just provided faster communication among the atoms so they could friend each other on facebook more efficiently.

"Oh hey, I'm an atom in a molecule of shit." "NO WAI ME TOO!!! BE MAI FRIEND!"

Then, after about v172 or so, predictive technology was built in. So now, every host already knows what packet you're about to send it, and you don't have to send it at all. Also, they're all really excited about the horse porn that's coming their way.

dhromed

@belgariontheking said:

Then, after about v172 or so, predictive technology was built in. So now, every host already knows what packet you're about to send it, and you don't have to send it at all. Also, they're all really excited about the horse porn that's coming their way.

I am not a doctor, but I play ~~one~~ it on the internet, and I endorse this message.

Jaime

@morbiuswilters said:

@Jaime said:

@morbiuswilters said:

@belgariontheking said:

This thread is now about over-normalization and ipv1000.

Discuss.

IPv1k? Hell, they already jumped to 128-bit addresses, where else can they go? "We will provide enough addresses for every single atom in the universe to have its own, private, 128-bit address space."

Actually, that sounds like IPv7. All it would take is 208 bits to give each atom a 128-bit space. Currently, each atom can have its own 48-bit address space. Enough to allow every atom to have every legal ethernet MAC address simultaneously.

2^80 != 10^80. The number of atoms in the universe is at least 2^266. Still, your basic point is valid: IPv4 to IPv6 quadrupled the address space. To be on the safe side, we'd have to quadruple it again to give every atom his or her own IPv6 address space.

I was only off by a factor of 100,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000. Not too bad.

bstorer

@belgariontheking said:

Well, to be fair, every update after ipv46 just provided faster communication among the atoms so they could friend each other on facebook more efficiently.

My atoms keep poking each other. It tingles.