I guess you do not work with large databases and trying to find a person the first time NO MATTER how the name is spelled. It is just good customer service. with as your example an Irish person entering a Scottish name and vis-a-vers-a easily making a mistake.

Now add in German names with a beta symbol instead of a double s. The Greek with AE combo letter vs AE. And all the other punctuations from the French, German, and ... That are NOT normally not on capitalized words.

I did hotel software for 20yrs, these are common problems for cross boarder travel. I wrote custom search and translation software to "fix" these issues so the guest service can shine even with errors. But the cost was greater disk size and two files to read for each sorted name.

Most developers, do not get that the world is made up multiple standards and refuse to consider local vs database relationships like:

Boolean: Yes/Si/... No/No/..

Amounts: ,. ., 0,2,3 places from right for display, 0,2,3 places from right for value (USD: ,.22, JPY: .,20)
Dates:: ./ 0 suppressed ISO, JPN, USA

How do you thing the do working with phone@, addresses and names?

Database engines fail with these simple-complex constructs because sorting and matching tests are still left hand and character set driven. A database MUST treat all of these names the same: McClean, MacClean, MCLean, Mc Clean, Mac Clean. McCleen, ...

You need to do a little research...

1) USPS is responsible for address validation. It is their system that will add a new address to the database. Right now about 98% of all addresses will validate correctly(unknown). Specially with 5 different types of postal codes that affects how an address is validated. When it validates incorrectly it is normally an error in the USPS system. Nothing can fix that expect them.

2) GIS information normally runs 3-6 months behind USPS.

3) Tax rates are currently gathered monthly and supplied to clients by multiple accounting companies just for this need. Must all tax changes occur 3-4 months before they go into effect. Even if you miss the date, there is normally a grace period in which you implement the change.

YES all these services today you pay. If you use the USPS will be free, but do not over use it.

That is why I would like USPS to be the tax expert, you validate an address for shipping and they supply you the tax rate. Since it is USPS, they would be "always right", and indemnify the user of service. If an address does not validate, that too will be reported *AND* a tax rate will returned to as a default. We used the closest spelling match, and if more than 1 was found, the lowest rate. This logic passed all states, since it was the best that could be used at the time.

If the USPS steps up, the cost should be no more than what it costs today to validate an address. They are already passing back route information and other unneeded information for normal processing, the tax rate(s) that are in effect would be just a little more.

Burden??? What Burden???

It is database issue with GIS lookup.

If Amazon is already doing this Washington State, then the hard part is over, since the tax districts in Washing State follow the Elementary School Districts as part of the boundaries.

Personally, the USPS (or even FedEx and UPS) should step up and be the "tax experts". Since they validate the address (hence GIS) they can determine the correct rate from the over 160k districts in US.

GIS is important since a given address maybe on City, County, State, Federal land and different rates apply.

I did this software for a large bug company. Had to get the correct rate for almost 10 million addresses, since the "cash register" is at the service location. It is not hard, but once a month we had a service supply us new tables.

Access Control - using VLANs and "firewalls" can be broken with errors
Business Continuity - system is unavailable for *any* work
Damage - system needing to be reload to return to working status w/ loss of intermediate work.
Notification - Is there some thing that maybe wrong
Missed Posting - due to todays separated servers one being "down" could cause *lost* postings and revenue.

That is just a short list. SOX opens a whole array issues, Best to let the SOX team to worry.

The network is now the system!

1) You should not be making any direct changes to the network with out correct design, test and sign off.

2) You should already have a redundant network structure, so "half" can be loss without any loss to network operations. This way the change can be tested in parallel.

3) You should always report to SOX officer when a request outside correct operations and management is made. It makes it their responsibility to solve the legal issues, for not following their written standards, before you began.

The Wiki article would translate to what you said. But there is another view that needs to be thought about. That is "how it is connected?".

ROM and RAM are normally connected directly to address and data buses of CPU and thought today more as memory. The disk are normally connected though an interface (some times another computer/controller) and thought as storage. This all about the PHYSICAL connection. It is also why other connections to "disks" can be created like RAM disks, iSCSI and tape-drive-disks.

I work on a machine that OS "looks" to machine via LOGICAL view. So the disk is just one large slow virtual memory space and the RAM is just faster / closer cache of that virtual space, and cache is faster still and closer. Makes nice to add new drives since it just one big flat model. It even makes journals easy since you can have it stripe the journals over 100 drives with single option. My "current" machine is 32 processors, 640GB of memory, and 900+ disks in RAID5 clusters for 35TB of storage.

This has been common place on IBM equipment for years.

That last machine had batteries on each of i/o controllers (RAID) with dual memory cards so that card system can take a "hit" and even half the i/o controller could go, and you can still change the I/O controller and KEEP the data in transaction.

The cabinet itself had dual power supplies with batteries, each board used dual power from dual supplies. The cabinet took two 30A 480V 3-phase feeds, and supplied 400V DC to the boards, which then dropped the power down need rates.

Europe is not the easiest...

Italy had 3 VAT taxes that the rate was 0.0%. They could not be mixed and had to reported each month to the state and displayed on the bill.

Europe also has tax on tax rules. One tax has a sub class of tax, because the base charge was not an option to customer.

Lastly, taking the Bar example. Vienna had three different VAT rates for milk. If you had a restaurant and bar together...
    If you drank the milk in restaurant, it was at one rate. (drink with food)
    If you drank the milk in bar area (next table over), it was another rate. (bar drink)
    If you pored it on you breakfast ceral, it was a third rate. (food)
And all would have to be shown on the bill.

But the other-side, VAT is VAT once you get it setup, you list amount of tax you paid for the items, and you paid, you send in the difference.

There are over 160,000 different sales tax rates in the US. Then there are exceptions like to a maximum item value of $2000 in Memphis or which item is charged which tax.

Also the tax rates are normally aligned with boundaries that postal code, city name or county do not follow, so just checking what mailing address does not help. In Washington State the tax rate follow elementary school boundaries for MTA additional amounts.

There are exception areas that like county land inside of incorporated city boarders. Arkansas has alot of towns like this one block in the center of town has a different rate.

Lastly, there is Texas that bases the tax based on location of the business.

All of this do able - I did it for nationwide service company (kill bugs). Personally, I would place the tax calculation as part of shipping requirement and look to Fedex and UPS to supply the tax rate to use.

You are missing his point...

Why does the content owner have to block Google? Google has to ask permission to index books, why not webpages?

This is the edge that missing in the conversation.

robots.txt does not block Google or anyone else, it just a form of a "site use contract" that had come into play help manage the relationship between two computers. Reading it still consumes resources (yes very little), why should the contain owner pay for something he does not want to happen in the first place?

Yes, there is more here, like he really would like a cut of Goggles pie. But the question of "Why must I pay to prevent you from using what is mine?", will also need to answered.

IT was data room not a data center and it was 25yrs ago...

On the third floor was data room with computer and phone switch next too the main A/C flume going from basement to 20flr (walking closet without floor or ceiling. The external auditors came though and wrote up the location that there was no raised floor in case of flooding.

Site's response: HA HA HA - Flood - 3rd floor?

Next week the water main at the 5th floor brpke and came down the A/C flume and flooded the third floor in 2 feet of water!

Opps

It runs better than 9.04 on this machine that I am using. This is a K6-3D/400 with 256M and 10G drive. It was upgraded from 7.04 - 16 hours per release.

Issues since 9.10...

Failure during boot get Xwindows/gnome to start. On new log on screen is now a choice of gnome and safe gnome. Just change to the other one and boots OK.

During first boot Netscape kept kicking errors about xorg. Those when a way on second full boot.

Do not like new update apt just showing up with a click. Liked better the icon in tool bar.

Now, that was funny!

I too have old hardware:
486sx16 with 12MB and 275MB drive, 2 ISA 10baseT - use for testing of IPCop.org linux firewall. Works great.

K6-3D 400 with 256mb and 10G drive - Had to get Ubuntu 7.04 because new versions would not find IDE drvies correctly and had problems with ACPI (or lack of). Then upgrade upgrade upgrade not to 9.04. Took awhile. It is what I am writing on now.

My next projects are Dual PPro 200 with 256MB and 2 2G drives. It was running Linux before with NT3.51 in a VMWare session, but that been a few years. 2 SMALL form P6-500 with 256MB and 10GB.

I also run Core 2 QUAD with 8GB and 256G and 1TB with VMWare 5 sessions. So I do not just have old hardware.

I am finding Linux has loss some of its roots for being able to bring more work out of older equipment. But it does not stop us from trying, does it. Good luck with your project.

I used an old redhat diskette a few months ago to netload from their servers, was fun and slow but did work. Other information for doing these types of installs.
http://www.debian.org/distrib/netinst
http://marc.herbert.free.fr/linux/win2linstall.html -helpful.

Also remember the Async ports on these old machines and be connected to another machine and used for network loading too. ;-) Just takes longer be works very well.

Thank you