fenway

Normally, this would work: SUBSTRING_INDEX( SUBSTRING(post FROM LOCATE('http://',post) ), ' ', 1 ) But in your case, you have a < instead of a space....

I've been programming for almost twenty years (Perl/C/JavaScript/etc.), and I haven't even come close to "that point". *shudder* I use UltraEdit, one of the most powerful editors around; that being said, it's not an IDE, though you can easily compile source code with a bit of magic. Check out v14... incredible. I've been using it forever.

Sounds reasonable... but it sounds like you should be building up an array of where clausing conditions and then joining them with AND at the end.

Otherwise known as a join. SELECT changes.text1,systems.text2 FROM changes LEFT JOIN links ON changes.id = links.changeid LEFT JOIN systems ON links.sysid = systems.id GROUP BY changes.id;

Glad to hear it! Always appreciated when solutions are posted, too.

Agreed; though it seems more relevant here. Check out this post first.

I caution the use of USING... not only is this not sql-standard AFAIK, but SELECT * expansion works a bit differently in v5+:

I use this all the time whenever I'm storing "system" information that is never edited except by hand, and quite infrequently. Name/value pairs are quite easily stored in this fashion, and as long as you read them back all at once, it's fine. Just don't expect the DB to be able to handle this in any meaningful way. Yes, FIND_IN_SET() can be hijacked to deal with very simple boundary cases, but it's bad design.

If used with LIKE, this will return all results that begin with 84.

Forgot about PHP for a minute... if your SELECT query doesn't work from phpmyadmin, you're either connected to the wrong DB, or there are really no records that match your results.

NO! This is a pain to edit, and even more of a pain to search. The only time such a list should ever be used if when you never need to modify individual values.

OK... I will give this another try this evening.

How about an INSERT INTO... SELECT WHERE... using FROM_UNIXTIME?

I finally got to this... but your insert statements don't match your where clause, and don't have any URLs!

So from DB to DB?

1) Bad idea... 2) 100K records isn't big.

But you don't have to any math at all... just read them back in the "new" order, and number them from there.

You can do this with two left joins... but only if you know the fields in advance.

You mean GROUP_CONCAT()? What is the output you'd like to see/

Actually, it's because you haven't quoted your date literails.

You mean "LIKE '84%'"

You shouldn't be storing readable format in the db.

You just have to save the sortorder of all of the items each time... then just enumerate them.

Well, I don't see a ?... post the show create table.

Text? Noooo.... How about 00:20:00?