The Daily WTF: Curious Perversions in Information Technology

2014-04-23 Reply Admin

D-Coder:
anonymous:
golddog:
Speaking of Roman numerals, I read this joke last night and kind of liked it:
A Roman general walks into a bar. Holds up two fingers and says, "I'll have five beers."
Wow, I'm impressed. Next!
Roman walks into a bar. "Gimme a martinus." "You mean martini, don't you?" "If I wanted two, I'd have asked for two!"

Less obvious, but maybe if you watched "Life of Brian", you might know the answer.

2014-04-23 Reply Admin

Anon:
I'd like to know how many records were actually in the number table!

I would presume it is still growing rapidly, thanks to a stored procedure that, when you try to fetch a number that doesn't exist, goes into a loop auto-populating records from select max(id) to N.

Rick · 2014-04-23 Reply Admin

snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?

I haven't touched SqlServer in a while, but IIRC, the default order will be determined by a unique index or primary key.

2014-04-23 Reply Admin

faoileag:
snoofle:
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I really hope you never branch out into publishing a dictionary...

I assume you're not from Germany where Dictionary order != Telephone Directory order

2014-04-23 Reply Admin

Bob:
Remy Porter:
For an article like this,SELECT * FROM comments WHERE comment_text = 'frist' might stick around (personally, I'm a little tired of SQL-based "frists").
How about:
SELECT C.Comment_Text AS [Comment] FROM dbo.stbl_Comments C WHERE C.Comment_ID = 1

Using:

CREATE TABLE [dbo].[stbl_Comments] AS ( Comment_ID INT NOT NULL, -- Number of comment Comment VARCHAR(255) NOT NULL, -- The comment Comment_Text VARCHAR(255) NULL, -- Textual representation of the comment, should always match the value in the Comment field Comment_Irish VARCHAR(255) NULL, -- Irish translation of the comment Comment_GUID UNIQUEIDENTIFIER NOT NULL -- Globally unique identifier for the comment )

TRWTF: No ordinals -- unless they are in a completely different table.

2014-04-23 Reply Admin

Remy Porter:
ubersoldat:
It would be a fun project to create a book titled "How to know your Enterprise Software is crap"

see also: Remy's Law of Software. "If the term 'enterprise' is used anywhere in the description of your software, it is a piece of garbage."

Some words that mean "I have no idea what I'm talking about but you should think I'm brilliant":

enterprise paradigm utilize

These words may have had other meanings at one time but those meanings are long since lost. Like "gay" used to mean happy and carefree, but now means something totally different.

2014-04-23 Reply Admin

Just what do they use the GUIDs in the number table for? Are there really places in the software where someone said, "Ok, at this point we need to get the '7' record from the database. So let's see, the GUID for '7' is 22BFE2A5-33A3-4DA3-9F66-E7BF67276D49, so to get the number 7 all I have to write is select number from number_table where guid='22BFE2A5-33A3-4DA3-9F66-E7BF67276D49'. Then I establish a connect to the database, submit the query, read the result set, and extract that value. Wow, this is so much more enterprisy than writing x=7. Progress!"

2014-04-23 Reply Admin

snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?

The order of the letters in the alphabet was determined by that song.

Nagesh · 2014-04-23 Reply Admin

nobulate:
Nagesh:
Fight club was not bad movie.

You call the rendition of a corporate slave breaking apart psychologically forming multiple personalities as a coping mechanism for reality, "not bad movie"?

You must be one of those people who enjoys Will Smith films too.

Shame.

I have enjoyed much - Pursuit of Happy and Independence Day.

2014-04-23 Reply Admin

Rick:
snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I haven't touched SqlServer in a while, but IIRC, the default order will be determined by a unique index or primary key.

As per the SQL standard, no ORDER BY clause in a query leads to the results being in an undefined order.

SQL Server will MOST LIKELY use the clustered index as a default, but it could also use anything in memory and return a completely different result.

Nagesh · 2014-04-23 Reply Admin

ubersoldat:
Does your software use Struts?

Struts was excellent till spring came here.

Nagesh · 2014-04-23 Reply Admin

Anonymii:
Rick:
snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I haven't touched SqlServer in a while, but IIRC, the default order will be determined by a unique index or primary key.

As per the SQL standard, no ORDER BY clause in a query leads to the results being in an undefined order.

SQL Server will MOST LIKELY use the clustered index as a default, but it could also use anything in memory and return a completely different result.

Honestly, I work with Oracle database every day and what I see is records are always returned in the order they are entered if there are no indexes or primary keys on the table.

Yes. you can take that to bank with you.

chubertdev · 2014-04-23 Reply Admin

Anonymii:
Rick:
snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I haven't touched SqlServer in a while, but IIRC, the default order will be determined by a unique index or primary key.

As per the SQL standard, no ORDER BY clause in a query leads to the results being in an undefined order.

SQL Server will MOST LIKELY use the clustered index as a default, but it could also use anything in memory and return a completely different result.

Correct, there is no guaranteed order.

Fun/stupid SQL trick (semi-related):

select 4
union
select 1
union
select 3
union
select 2

chubertdev · 2014-04-23 Reply Admin

Jay:
snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?

The order of the letters in the alphabet was determined by that song.

s/ASC/DESC

I'm lysdexic.

chubertdev · 2014-04-23 Reply Admin

QJo:
Bob:
Remy Porter:
For an article like this,SELECT * FROM comments WHERE comment_text = 'frist' might stick around (personally, I'm a little tired of SQL-based "frists").
How about:
SELECT C.Comment_Text AS [Comment] FROM dbo.stbl_Comments C WHERE C.Comment_ID = 1

Using:

CREATE TABLE [dbo].[stbl_Comments] AS ( Comment_ID INT NOT NULL, -- Number of comment Comment VARCHAR(255) NOT NULL, -- The comment Comment_Text VARCHAR(255) NULL, -- Textual representation of the comment, should always match the value in the Comment field Comment_Irish VARCHAR(255) NULL, -- Irish translation of the comment Comment_GUID UNIQUEIDENTIFIER NOT NULL -- Globally unique identifier for the comment )

TRWTF: No ordinals -- unless they are in a completely different table.

Different database, actually, on its own server.

chubertdev · 2014-04-23 Reply Admin

Jay:
Remy Porter:
ubersoldat:
It would be a fun project to create a book titled "How to know your Enterprise Software is crap"

see also: Remy's Law of Software. "If the term 'enterprise' is used anywhere in the description of your software, it is a piece of garbage."

Some words that mean "I have no idea what I'm talking about but you should think I'm brilliant":

enterprise paradigm utilize

These words may have had other meanings at one time but those meanings are long since lost. Like "gay" used to mean happy and carefree, but now means something totally different.

Those words are pretty robust.

2014-04-23 Reply Admin

I particularly liked the bit where people discussed the mods discussing the frist posts that no longer existed.

chubertdev · 2014-04-23 Reply Admin

The Balance:
I particularly liked the bit where people discussed the mods discussing the frist posts that no longer existed.

The frist rule of frist posts is that you should expect them to get recursively baleted.

2014-04-23 Reply Admin

Remy Porter:
A simple "frist" is going to get deleted. For an article like this,SELECT * FROM comments WHERE comment_text = 'frist' might stick around (personally, I'm a little tired of SQL-based "frists"). The key point is- if it makes us laugh, we're less likely to delete it.

Thus, "Frist Club was not bad movie"?

(I would never do that, but it's amusing here.)

2014-04-23 Reply Admin

Jay:
Just what do they use the GUIDs in the number table for? Are there really places in the software where someone said, "Ok, at this point we need to get the '7' record from the database. So let's see, the GUID for '7' is 22BFE2A5-33A3-4DA3-9F66-E7BF67276D49, so to get the number 7 all I have to write is select number from number_table where guid='22BFE2A5-33A3-4DA3-9F66-E7BF67276D49'. Then I establish a connect to the database, submit the query, read the result set, and extract that value. Wow, this is so much more enterprisy than writing x=7. Progress!"

For RI of course. You might have a one-to-many record relationship between, say "7" and um... well I don't work in an enterprise so I can't say just what, but I'm sure there's something. And the "7" might need to be changed to an "8" without affecting all the other tables. Or if you delete "7" (reminds me of a Stephen Wright joke) you want all the related records deleted too.

2014-04-23 Reply Admin

n9ds:
Or if you delete "7" (reminds me of a Stephen Wright joke) you want all the related records deleted too.

"I bought a telephone that didn't have any 5s. So I ran into a friend and he asked, 'Hey, why haven't you called me?' I said, 'My phone doesn't have any 5's.' He asked, 'How long have you had it?' 'I don't know, my calendar doesn't have any 7's.'"

2014-04-23 Reply Admin

Remy Porter:
Nice. You left out, however:
Comment_GUID_Roman VARCHAR(255) NULL --the GUID represented as a series of Roman numerals

Don't forget:

Comment_GUID_Hex VARCHAR(255) NULL --the GUID represented in hexadecimal

Comment_GUID_Oct VARCHAR(255) NULL --the GUID represented in octal

Comment_GUID_Bin VARCHAR(255) NULL --the GUID represented in binary

chubertdev · 2014-04-23 Reply Admin

n_slash_a:

Comment_GUID_Hex VARCHAR(255) NULL --the GUID represented in hexadecimal

Way too confusing.

2014-04-23 Reply Admin

Beve Stallmer:
Now that's ENTERPRISE QUALITY!

/turbo package/

2014-04-23 Reply Admin

Jay:
snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?

The order of the letters in the alphabet was determined by that song.

Do Re Mi Fa So La Ti Do

2014-04-23 Reply Admin

John:
faoileag:
snoofle:
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I really hope you never branch out into publishing a dictionary...
I assume you're not from Germany where Dictionary order != Telephone Directory order

Japan too. Phone book order: 神戸 (Kanbe) 木下 (Kinoshita) 神戸 (Kobe) Dictionary order depends on whether it's a character dictionary (in which case there are only two entries for the above examples) or something ordered by pronunciation (which would yield the same as the phone book order i nthis case).

I'm in the phone book too. Last time I noticed, I was the only one on the page that was printed in Roman letters, surrounded on all sides by a bunch of normal letters. I was in the correct place by pronunciation.

2014-04-23 Reply Admin

Coyne:
I bet they have an incident when the database switches to Unicode.

Actually the procedure avoids problems with most ANSI code pages as well as Unicode. Consider ISO-8859-1. Instead of putting ñ where it belongs right after n in alphabetical order, it gets shoved off to some numerical values higher than punctuation marks (not those punctuation marks, those other punctuation marks). So if you're taking airline reservations or credit card applications, you have to do the kind of test that this procedure does and kick out anyone who has an ñ in their name.

NTFS has something like that too. When formatting an NTFS volume, one of the special files tells how to uppercase a character. So if you're running with a Turkish locale at the time you format a volume, i should upcase to İ and ı should upcase to I. But for some reason I think Windows forgets to check the upcasing table at the time of actually creating or opening a file. Due to some bug, Windows uses the locale that the user's process is running under at the time of creating or opening -- no wait, Windows uses the system-wide locale for non-Unicode programs because conversion between ANSI and OEM character sets occurs in kernel mode which doesn't know about user mode locales -- well anyway, the various upcasing tables stored in various partitions get ignored. You can end up with interesting random collections of filenames that are sometimes equal but sometimes different, sometimes openable but sometimes not openable.

chubertdev · 2014-04-23 Reply Admin

Norman Diamond:
John:
faoileag:
snoofle:
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I really hope you never branch out into publishing a dictionary...
I assume you're not from Germany where Dictionary order != Telephone Directory order
Japan too. Phone book order: 神戸 (Kanbe) 木下 (Kinoshita) 神戸 (Kobe) Dictionary order depends on whether it's a character dictionary (in which case there are only two entries for the above examples) or something ordered by pronunciation (which would yield the same as the phone book order i nthis case).
I'm in the phone book too. Last time I noticed, I was the only one on the page that was printed in Roman letters, surrounded on all sides by a bunch of normal letters. I was in the correct place by pronunciation.

I don't get that language.

select 'Kanbe', ASCII(N'神' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS), ASCII(N'戸' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS)
select 'Kinoshita', ASCII(N'木' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS), ASCII(N'下' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS)
select 'Kobe', ASCII(N'神' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS), ASCII(N'戸' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS)

What am I missing?

2014-04-23 Reply Admin

snoofle:
Les:
TRWTF is that there's no "Order By" in the query. That means that the alphabet is not guaranteed to be in alphabetical order.
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?

Some research has shown that there is an alphabetical bias.

For example, in science papers, when multiple authors are listed in alphabetical order, readers nevertheless tend to give more credit (probably unconsciously) to the names that appear first.

Random order may work in some cases , but if the list is long enough, you need rules so that you can quickly find an individual item.

I believe the solution is to use a random alphabetical order. That is, the order is derived randomly, ONCE, then used every time an arbitrary order is called for in the entire paper/project/etc., or even for the entire life of an organization. The order has to be published to be any use, of course, probably included with every publication from the organization.

2014-04-23 Reply Admin

chubertdev:
Norman Diamond:
John:
faoileag:
snoofle:
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I really hope you never branch out into publishing a dictionary...
I assume you're not from Germany where Dictionary order != Telephone Directory order
Japan too. Phone book order: 神戸 (Kanbe) 木下 (Kinoshita) 神戸 (Kobe) Dictionary order depends on whether it's a character dictionary (in which case there are only two entries for the above examples) or something ordered by pronunciation (which would yield the same as the phone book order i nthis case).
I'm in the phone book too. Last time I noticed, I was the only one on the page that was printed in Roman letters, surrounded on all sides by a bunch of normal letters. I was in the correct place by pronunciation.
I don't get that language.
select 'Kanbe', ASCII(N'神' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS), ASCII(N'戸' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS)
select 'Kinoshita', ASCII(N'木' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS), ASCII(N'下' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS)
select 'Kobe', ASCII(N'神' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS), ASCII(N'戸' collate Japanese_Bushu_Kakusu_100_CS_AS_KS_WS)
What am I missing?

You're missing phone book order. You did dictionary order (character dictionary).

2014-04-23 Reply Admin

"The best thing about this solution is how easily it can be modified to support the exciting new lowercase letters that are starting to gain traction in some forward-looking businesses."

Had me right there! :D

2014-04-24 Reply Admin

Norman Diamond:
John:
faoileag:
snoofle:
The concept of an ordered alphabet is kind of arbitrary (is there really any good reason why A comes before B, other than your mother told you so?), so why insist that the letters be returned without some entropy?
I really hope you never branch out into publishing a dictionary...
I assume you're not from Germany where Dictionary order != Telephone Directory order
Japan too. Phone book order: 神戸 (Kanbe) 木下 (Kinoshita) 神戸 (Kobe) Dictionary order depends on whether it's a character dictionary (in which case there are only two entries for the above examples) or something ordered by pronunciation (which would yield the same as the phone book order i nthis case).
I'm in the phone book too. Last time I noticed, I was the only one on the page that was printed in Roman letters, surrounded on all sides by a bunch of normal letters. I was in the correct place by pronunciation.

By "ordered by pronunciation", do you mean "ordered by the name when written phonetically (in English, whatever that means in this context), in standard English alphabetical order"?

2014-04-24 Reply Admin

Does your software store bitmap image data in a database, pixelwise?

Does your software store all other data in a database as 100MB XML files?

2014-04-24 Reply Admin

QJo:
By "ordered by pronunciation", do you mean "ordered by the name when written phonetically (in English, whatever that means in this context), in standard English alphabetical order"?

My guess is "ordered by pronunciation" means ordered according to the kana writing.

2014-04-24 Reply Admin

yep - it's certainly in the Financial Times style guide. Here's one from a UK university:

https://www.york.ac.uk/communications/publications/writing/style-guide/#N

"Spell out numbers up to and including ten. After ten use figures except for exact measurements and charts where figures can also be used for numbers below ten"

Bonus question.... Should one write nought, one, 10, 11 when writing in binary?

The Great Lobachevsky · 2014-04-24 Reply Admin

Jay:
Remy Porter:
ubersoldat:
It would be a fun project to create a book titled "How to know your Enterprise Software is crap"

see also: Remy's Law of Software. "If the term 'enterprise' is used anywhere in the description of your software, it is a piece of garbage."

Some words that mean "I have no idea what I'm talking about but you should think I'm brilliant":

enterprise paradigm utilize

These words may have had other meanings at one time but those meanings are long since lost. Like "gay" used to mean happy and carefree, but now means something totally different.

I had a manager who couldn't get through a meeting without using the word synergy. I propose adding to the list.

2014-04-24 Reply Admin

Anon:
Mark:
'the Number_Text field sometimes contained a value like "nine", and other times contained a value like "17"' may not be such a bonus WTF.
My employer's corporate style/writing guide mandates numbers <=nine written in letters, and >=10 in numerals.

Pretty sure I read in some general "writing style" guide somewhere that that is how you should represent numbers when writing. It just looks better. Apparently.

That was the style I was taught growing up. One of the many style points I always thought was idiotic. "Yesterday I had nine widgets, today I have 10" is officially considered correct, but I think it is wrong to mix forms in a sentence (or paragraph or document).

I also prefer the Oxford comma and logical punctuation, despite what I was taught. Style guides are just stupid.

chubertdev · 2014-04-24 Reply Admin

Norman Diamond:
You're missing phone book order. You did dictionary order (character dictionary).

The thing that I don't get is that both Kanbe and Kobe come up as 144/140. How can you tell that they're different?

chubertdev · 2014-04-24 Reply Admin

Dan:
I also prefer the Oxford comma and logical punctuation, despite what I was taught. Style guides are just stupid.

I don't prefer the Oxford comma, I demand it.

2014-04-24 Reply Admin

ubersoldat:
It would be a fun project to create a book titled "How to know your Enterprise Software is crap" with stuff like:
Does your software goes to the database to perform stupid ops?

Does your sofware has 5 or more layers of abstraction?

Does your software use Struts?

A very enterprisey book, probably about 700+ pages, bound in a nice photocopy of a wooden table.

More realistically, you could do that in two questions. Basic set theory precludes the necessity of the thrid.

2014-04-24 Reply Admin

chubertdev:
Norman Diamond:
You're missing phone book order. You did dictionary order (character dictionary).

The thing that I don't get is that both Kanbe and Kobe come up as 144/140. How can you tell that they're different?

When registering for a phone line, bank account, employment contract, etc., whatever, you write your name as usually written (usually using Chinese characters) and ALSO write the pronunciation (using kana). In my case sometimes I write my name as written in some foreign countries (using Italian characters) and also write the pronunciation (using kana), but sometimes write in kana in both places.

However, the phone book itself doesn't include furigana, so if you're reading a phone book instead of writing your own name to register for a phone line, you don't always know the pronunciation of the names you're looking at. On business cards and some other kinds of printed materials, people with ambiguous names sometimes include furigana so the reader will know the pronunciation instead of having to be told.

2014-04-25 Reply Admin

Perhaps they are holding out juuuust in case EDBIC (nee "the first enterprise substituion cipher") comes back en vogue.

2014-04-25 Reply Admin

Remy Porter:
Nice. You left out, however:
Comment_GUID_Roman VARCHAR(255) NULL --the GUID represented as a series of Roman numerals

Ahh yes, HexaRoman digits. I think Cicero talked about these.

chubertdev · 2014-04-25 Reply Admin

Norman Diamond:
chubertdev:
Norman Diamond:
You're missing phone book order. You did dictionary order (character dictionary).

The thing that I don't get is that both Kanbe and Kobe come up as 144/140. How can you tell that they're different?
When registering for a phone line, bank account, employment contract, etc., whatever, you write your name as usually written (usually using Chinese characters) and ALSO write the pronunciation (using kana). In my case sometimes I write my name as written in some foreign countries (using Italian characters) and also write the pronunciation (using kana), but sometimes write in kana in both places.

However, the phone book itself doesn't include furigana, so if you're reading a phone book instead of writing your own name to register for a phone line, you don't always know the pronunciation of the names you're looking at. On business cards and some other kinds of printed materials, people with ambiguous names sometimes include furigana so the reader will know the pronunciation instead of having to be told.

Reminds me of the usage of "Coke" in the south.

"I'll have a Coke." "What kind?" "Iced tea."

oheso · 2014-04-25 Reply Admin

Norman Diamond:
Phone book order: 神戸 (Kanbe) 木下 (Kinoshita) 神戸 (Kobe)

Furigana for the win!

Tka, Japan has phone books? Oh, yeah, those tomes that get delivered every year that sit collecting dust in the entryway until we decide to throw them away. Come to think of it, not sure when we last received a set.

oheso · 2014-04-25 Reply Admin

QJo:
written phonetically (in English, whatever that means in this context), in standard English alphabetical order"?

See Norman's answer about entering your name in phonetic kana. Then things are sorted in kana order:

http://en.wikipedia.org/wiki/Goj%C5%ABon

2014-04-26 Reply Admin

It might have been too subtle for you to understand

2014-04-28 Reply Admin

roman numerals?

2014-04-29 Reply Admin

Dan:
I also prefer the Oxford comma, and logical punctuation, despite what I was taught.

FTFY

2014-04-29 Reply Admin

The concept of assigning a GUID to every number made my head go all explodey.

Your Letters are Numbered

Leave a comment on “Your Letters are Numbered”