[#287] - Clean up SQL files including deleting unused fields and updating sample data
- Closed
- 26 Jul 2012
- Medium
- Build: master
- # 287
- Diff
- elinw:sampledatenew
User tests: Successful: Unsuccessful:
This pull request updates the core sql files for mysql including deleting fields that have not been used at all in the 1.6-2.5 series. It also removes references to those fields in the small number of places that they were present.
This also updates the sample data files to reflect these changes and to fix a few other details.
Please note the testing sample data is to be deleted prior to GA release.
Also please note that there will be additional changes, this is just round one.
Will do the schema changes and updates next, but I'd be happy to have
people do some testing as is and confirm it's good. Although I"m happy to
share the updates that I've done.
About recreating the tables if they don't exist ... two things.
1. sometimes people may have messed with their core tables.
2. We have a goal of letting people reset their site and install a new set
of sample data.
ALTER TABLE #__contact_details DROP imagepos;
ALTER TABLE #__content DROP title_alias, DROP sectionid, DROP mask,
DROP parentid;
ALTER TABLE #__newsfeeds DROP filename;
ALTER TABLE #__weblinks DROP sid, DROP date, DROP archived, DROPapproved;
ALTER TABLE #__menu DROP ordering;
ALTER TABLE #__weblinks ADD version int(10) unsigned NOT NULL
DEFAULT '1';
ALTER TABLE #__weblinks ADD images text NOT NULL;
ALTER TABLE #__newsfeeds ADD description text NOT NULL;
ALTER TABLE #__newsfeeds ADD version int(10) unsigned NOT NULL
DEFAULT '1';
ALTER TABLE #__newsfeeds ADD hits int(10) unsigned NOT NULL DEFAULT
'0';
ALTER TABLE #__newsfeeds ADD images text NOT NULL;
ALTER TABLE #__contact_details ADD version int(10) unsigned NOT
NULL DEFAULT '1';
ALTER TABLE #__contact_details ADD hits int(10) unsigned NOT NULL
DEFAULT '0';
ALTER TABLE #__banners ADD created_by int(10) unsigned NOT NULL
DEFAULT '0';
ALTER TABLE #__banners ADD created_by_alias varchar(255) NOT NULL
DEFAULT '';
ALTER TABLE #__banners ADD modified datetime NOT NULL DEFAULT
'0000-00-00 00:00:00';
ALTER TABLE #__banners ADD modified_by int(10) unsigned NOT NULL
DEFAULT '0';
ALTER TABLE #__banners ADD version int(10) unsigned NOT NULL
DEFAULT '1';
ALTER TABLE #__categories ADD version int(10) unsigned NOT NULL
DEFAULT '1';
On Fri, Jul 20, 2012 at 5:30 PM, Michael Babker <
reply@reply.github.com
wrote:
Curious as to why you've got the sample_*.sql files now recreating the
tables. Also, do you have an update SQL for the schema changes that are
being made?
Reply to this email directly or view it on GitHub:
#287 (comment)
Why not now ..
I'm not a big fan of adding table fields that are not set up in code. Could we do that in two steps? I'm afraid we might end up with more orphan fields that way.
No we're not going to and it is extremely wrong on many levels to do complex sql fixes and code same time, at least more than one or two lines. As much as possible we should try to isolate changes. This is a very complex commit as it is. I promise if we don't get other changes to standardize the code in we can go back and take them out. There is no way ever to move forward if you are not willing to take the first step.
Curious as to why you've got the sample_*.sql files now recreating the tables. Also, do you have an update SQL for the schema changes that are being made?