@ypercube wow. Blatant omission. Don't you hate it when the answer is just staring you in the face and you don't recognize it. Post that as an answer ant I'll accept.
at a guess, they didn't populate the phone number field for that sipermarket in their database, and their presentation layer code doesn't check for null values
because every supermarket will have a phone number in the database, right guys?
@ConcernedOfTunbridgeWells Strange I could swear I've seen a few questions that would fit quite nicely under that tag
(albeit a different usage of the word than we use in here...)
Ahaha I hate it when I spot typo errors in what I write approximately 2 minutes and one second after they're posted :v
@JackDouglas I don't know what the world's coming to. First @gbn has no mullet, now bollocks isn't professional enough. Anybody would think this was a serious forum.
@JackDouglas 100% serious. What are the benefits of creating 100 different tables for 100 different types of lookup strings? The table(s) will be in memory in either case, so the only real difference I see is programmability. I'd rather key into a single table than have to specify the table name depending on the lookup type. The latter is harder to do without dynamic SQL.
@ConcernedOfTunbridgeWells I did that once and I'm still regretting it. Also I'm still the one doing the CRUD for various reasons so lots of pain for no gain in this particular case :-(
@AaronBertrand I've seen it done on systems before - Agresso, Oracle Financials to name a couple. The internal MDS repository seems to be an EAV structure. Kalido make a suite of tooling including a MDM tool based on an EAV model
That makes more sense. If your lookups are static, separate tables is easier (IMHO). If it needs to be user-configurable, then that way thar be dragons
@AaronBertrand Blog makes sense. "EAVs are evil" is a useful rule of thumb, but like all rules of thumb, if you are really sure you know what you are doing you should ignore it. I don't think I'd ever advise someone to go down that route unless I knew a whole lot of detail about their requirements though.
Right, you captured my point: I wasn't trying to suggest that everyone should use EAV, but rather that it shouldn't automatically be on the reject list.
Hard to determine exactly what you're trying to find, but if you're trying to distinguish rows that have 1, 2, or n / characters:
DECLARE @x TABLE(y VARCHAR(32));
INSERT @x SELECT 'Foo bar0'
UNION ALL SELECT 'Foo/bar1'
UNION ALL SELECT 'Fo/o/bar2'
UNION ALL SELECT 'Fo/obar1'
UNION ALL SELECT 'F...
The downside is I've only ever used mis-applied EAV, I've never used it where it'd be recommended (which as far as I can tell is if you're supporting user customisability beyond the bounds of all reason, or if you have to support sparse data sets).
The right answer is to correct table1 so that it is using the right data types. In the meantime, assuming you need to match both date and time, you can try this:
and CONVERT(DATETIME, t.notes_datetime, 101) = s.notes_datetime
I wonder if one could make up a NoSQL drinking game, e.g. Take a sip when someone mentions scalability Take a sip when someone says ACID isn't necessary Finish your glass when the term 'web scale' gets used in a non-ironic context ... and so on.
Does anybody knows how Google or Yahoo perform searches for keywords against very very huge amounts of data? What sort of database or technologies do they employ for this?
It takes few milliseconds, but they have more than a billion pages indexed.
@AaronBertrand In any public forum the proportion of bumping into crazies is a linear function of the number of punters on the forum. Given your posting volumes and the size of Stackoverflow you've got some non-zero chance of running into a dickhead of some description.
You don't quite have SQL Server's proprietary UPDATE FROM syntax down. Also not sure why you needed to join on the CommonField and also filter on it afterward. Try this:
UPDATE a
SET a.CalculatedColumn = b.[Calculated Column]
FROM Table1 AS a
INNER JOIN Table2 AS b
ON a.CommonField = b...
@AaronBertrand In the days of bulletin boards, people complained about arseholes and flame wars. In the days of usenet they complained even more about arseholes and flame wars. Stackoverflow goes to a bit of trouble to make flame wars difficult, although it doesn't totally succeed and peiople manage to have flame wars in comments. However, they haven't managed to prevent arseholes.
@AaronBertrand I tend to view them as more indicative of the voter than the answer they voted on. If the answer is wrong then downvote for sure, but at least try to explain what's wrong with it.
@AaronBertrand I've downvoted whithout leaving a comment before, but mostly when there already is a comment explaining why (and I happen to agree with that comment), otherwise I try to explain my vote
Hi, I just got an answer of mine converted to a comment. Which is totally correct. It was a comment. (well a request for clarification). However, can somebody explain to this newbie why I can not add comments to 1) initial questions and 2) Answers added by other people?
@JackDouglas It's completely anonymous, so it doesn't customise results for you. OTOH the context in the google search does a better job of picking up relevant maternial IMO.
Depends on the situation. If your search bubble is useful to you then google will probably get better results.
I remember seeing a description of the big alpha box they ran it on at DEC in 1995-ish. It had 32 CPUs and 1GB of RAM. I remember being quite impressed with the scale of the machine.
Not sure I particularly lament OSF/1 but HP should never have canned the Alpha.
@gbn DEC was still making PDP-11's until 2000 or so. RSX/11 had a fairly loyal customer base in the industrial control space and DEC made ruggedised PDP/11s.
@gbn I used one as a pleb for a while at AK uni. There was an Alpha at UOC running VMS, but people mainly used it for running legacy SAS code.
The engineering school was quite big on DEC kit, but I did computer science and they were basically a Solaris shop. I learned CS on Intel Solaris on a really high spec (for the day) PC that was quite a bit cheaper than the cheapest entry level SPARC box that Sun made.
@EdwardDortland Didn't realise you were in here... just responded to your comment on the FACT table Q. Feel free to throw any Qs back in here if you prefer
@Phil lol, and of course that's the one piece of functionality that doesn't confirm 5 ways to sunday that you actually want to DELETE YOUR ENTIRE DATASET!!!
what am I talking about. I was adding a table the other day, and when I executed the CREATE TABLE statement, sql told me I already had an object named "blah", so I typed in "DROP TABLE blah" hit F5 and then said out loud, "oh sh*t"
time to go back to the latest backup. luckily it was a table that hadn't been updated in months.
@swasheck ahhh so much fun. I created a dedicated database along with a set of stored procs to handle all our maintenance plan needs. Now I just have a single SQL Agent job that runs a master stored proc that automatically rebuilds/recreates fragmented indexes, updates statistics for tables that have a changed row count, and does backups automatically. The database also has a DDL trigger that propagates design changes to the other servers on our network.
@MaxVernon I was completely thinking that I was going to drop an unused monitoring job ... BOOM! ... at least i've got recent experience restoring system dbs
@jcolebrand It is just like most of IT, hurry up and wait. At least the kids aren't here - would have been stir crazy and then escalated once power and Internet go down. They were due to fly out to Florida tomorrow for a vacation, so switched them to a flight yesterday. Hopefully can just write their trip off as evacuation expenses.
@jcolebrand I'm not sure we're going to see sustained hurricane force winds - the intensity forecasts have been dropping - maxing out at 70 on the models right now.
I am trying to dump a large table (40GB, 600,000 rows) using a SELECT ... INTO OUTFILE query. I want to use that because I want my results sorted in a particular way (so mysqldump is not good enough).
The problem is, I am getting an error code 28. I am pretty sure it is because it is running ou...
I have a query that returns a single row of data. Let's say it looks like this (with column headers).
Fruit | Veg | Meats | Nuts |
----------------------------------------
Apple Lettuce Veal Almond
Instead of this, I'd like to return this single row as a two column tab...
@swasheck I guess I just haven't had much direct need to do that. Although yesterday an msdb almost consumed all disk space. Turned out it was having an issue with queue messages not getting deleted.
@JackDouglas i'm slowly disengaging bh.se but i'm trying to periodically lurk. usually the "accepted answer" is of poor quality or is something that does not fit (my) standards of bh in general. the
@swasheck I wouldn't say we're all experts. Some of the stuff I answer I didn't know before I answered. Just did a bit of research & tested myself before answering. The site's a good way of learning random things
Is it possible to set a limit on the amount of physical disk space a user (or group of users) can use within a single database? For example is there some method of associating a schema with a file group and then setting a maximum size for the file group? The actual database would have other table...
@JackDouglas - indeed. and while i'm normally fine letting people be wrong about humanities-oriented things ... but some of the things are just so egregiously wrong and misleading that i feel like it's irresponsible to not raise an objection
I'm in a situation where I would like to update the date of all the values in the table to have the same day but persist the time of each record.
So these three dates...
8/28/2012 14:00:00
8/28/2010 12:00:00
8/28/2008 10:00:00
Should be turned into...
8/28/2012 14:00:00
8/28/2012 12:00:00
8/28/...
DECLARE @Target DATETIME = '20120828';
UPDATE dbo.MySpecialTable
SET DateField = DATEADD(DAY, DATEDIFF(DAY, DateField, @Target), DateField);
This will work if the target date is before or after the value in the column.
@AaronBertrand it also seems exclusive to newer versions. i was going to answer with something similar to yours, but tried to get information about the requirements. your answer deals with those requirements anyway
When a question is migrated, it actually gets duplicated to the new site. That includes duplicating your "favorite" selection on it. You can remove that one.
It does make it somewhat more difficult to remove it from favorites on the originally posted site, though; when you click most links to th...
I have a table (in SQL Server 2008 R2, in case that matters), defined with the following rows. The DateAdded column is a SmallDateTime data type.
ID DateAdded
1 2012-08-01 12:34:02
2 2012-08-01 12:48:25
3 2012-08-05 08:50:22
4 2012-08-05 11:32:01
5 2012-08-26 09:22:1...
I am Pavithra from RDBMS team and your SR is assigned to me. Can you upload the screenshot/spool output of the error which is occurring when you run the query?
Hi Pavithra Do you have a working instance of Oracle 11.2 to try my minimal working example on? If you can reproduce it you can get a trace file easily enough. If you want me to, I can talk you through the process of getting a trace file (but I expect there are some experts there on support who know?) Kind regards Jack
select id, (select count(distinct column_value) from table(ids)) from (select id, collect(id) over (order by id) as ids from (select to_number(trunc(level/4)) as id from dual connect by level<20));
Does anybody knows how Google or Yahoo perform searches for keywords against very very huge amounts of data? What sort of database or technologies do they employ for this?
It takes few milliseconds, but they have more than a billion pages indexed.
I accidentally dropped the mysql database which was in my online phpmyadmin?
I am trying to dump a large table (40GB, 600,000 rows) using a SELECT ... INTO OUTFILE query. I want to use that because I want my results sorted in a particular way (so mysqldump is not good enough).
The problem is, I am getting an error code 28. I am pretty sure it is because it is running ou...
I have a query that returns a single row of data. Let's say it looks like this (with column headers).
Fruit | Veg | Meats | Nuts |
----------------------------------------
Apple Lettuce Veal Almond
Instead of this, I'd like to return this single row as a two column tab...
I'm in a situation where I would like to update the date of all the values in the table to have the same day but persist the time of each record.
So these three dates...
8/28/2012 14:00:00
8/28/2010 12:00:00
8/28/2008 10:00:00
Should be turned into...
8/28/2012 14:00:00
8/28/2012 12:00:00
8/28/...
I have a table (in SQL Server 2008 R2, in case that matters), defined with the following rows. The DateAdded column is a SmallDateTime data type.
ID DateAdded
1 2012-08-01 12:34:02
2 2012-08-01 12:48:25
3 2012-08-05 08:50:22
4 2012-08-05 11:32:01
5 2012-08-26 09:22:1...
