Forum Replies Created
-
AuthorPosts
-
August 26, 2014 at 3:48 pm in reply to: Feature Request: Intelligent Defaults For Database Copy Dialog #35024MitchParticipant
Peter,
I understand your perspective. I certainly don’t expect that everyone works the way that I do, but I’m equally certain that I’m not the only one who would benefit from such a change. A user option that enabled the behavior I described would allow people like me to be more productive while not negatively impacting the others.
As far as having many connections open, I always have at least 4 connections open, sometimes more. The changes I described would work well for me.
MitchParticipantI do have the same problem in the Object Browser “change object color” dialog.
The problem persists across machine restarts. It is very reproducible.
MitchParticipantI do not get the pop-up dialog after clicking on other colors. I have not been able to do anything to make it appear.
How do you change color from the Object browser?
MitchParticipantThank you.
In the XML job file, when
is specified, what is the encoding scheme for the passwords?
MitchParticipantA STOP button on the tab would be acceptable.
In this instance, I personally would prefer the behavior I originally described, but controlled by an option, the default value of which was the current behavior, so as not to anger existing users. Or the default being the new behvaior, but with clear and noticable documentation indicating how to change the option.
MitchParticipantI ran SQLyog.exe from :~/.wine/drive_c/Program Files (x86)/SQLyog and still got only the wine icon.
MitchParticipantYes, please add it to the list.
Food for thought: in most environments, mine included, it is impractical to rename database tables, especially for a reason like this. I am dealing with a system that has been in place for years. References to some tables are scattered all over the codebase. It would require hundreds of hours of work to rename some of our tables.
I'm willing to accept the fact that I may be the only one who wants or needs such a feature, but it would certainly make my work more productive.
MitchParticipantSpeaking for myself, the convenience of the shortcut would far outweigh the occasional confusion of not being able to set to null when the column is defined as not null.
As for “set to default” and “set to empty string” I have NEVER used either of those, so I would not mind no shortcuts for them.
What would be even better (in my opinion) would be user-definable keyboard shortcuts – then each user could tailor them to his/her particular needs.
MitchParticipantYes. That would be very helpful.
MitchParticipantI certainly think this behavior would be a new option, defaulting to the current method.
The frequency that this happens varies – probably several times a week on a normal basis, but sometimes many times a day depending on what is going on.
MitchParticipantI accidentally discovered a partial work-around for this.
In the form view, if I select the text in the field that I want to watch (in this case the seconds behind master from show slave status), then whenever I click the refresh button, the query re-executes then automatically scrolls the form view output until the selected field is visible. This is a reasonable workaround, as it also makes the field of interest easy to spot since it is in reverse color font. It does take a moment to become visible, because the display starts at the top and then scrolls down, but it is better than manually scrolling each time.
MitchParticipantI didn't see where in the Database Synchronization Wizard to save a job file either.
Can you post a sample job file, so that I can create one manually to use until this issue is addressed?
MitchParticipantThat does work for setting the initial database, but it doesn't select the database in the list of databases near the top of the window, which still reads “No database selected”. I will use this as a workaround for now, but I think it would be better to have an explicit option for this.
MitchParticipantI've played with some options trying to figure out some more clues. It only crashes when the “EXPLAIN EXTENDED info” box is checked under “Enable Query Profiler”. When “EXPLAIN EXTENDED info” is not checked, the query runs fine. (And it runs fine with “Enable Query Profiler” unchecked.)
Trace from SQL server is shown below.
I can run “EXPLAIN EXTENDED SELECT * FROM mlc2 WHERE webobjectreviewid=14 LIMIT 0, 1000;” by itself OK, both in SQLyog and from the command line.
You might examine the code that parses the result of the EXPLAIN command. The only difference between EXPLAIN and EXPLAIN EXTENDED seems to be the inclusion of the “filtered” column, so you might check the handling of that value. Here is what I get from EXPLAIN EXTENDED on the command line:
mysql> EXPLAIN EXTENDED SELECT * FROM mlc2 WHERE webobjectreviewid=14 LIMIT 0, 1000;
+—-+
+
+
+
+
+
+
+
+
+
+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+—-+
+
+
+
+
+
+
+
+
+
+| 1 | SIMPLE | mlc2 | const | PRIMARY | PRIMARY | 4 | const | 1 | 100.00 | |
+—-+
+
+
+
+
+
+
+
+
+
+Here is the explain for the query that does not crash;
mysql> EXPLAIN EXTENDED SELECT * FROM mlc2 LIMIT 0, 1000 ;
+—-+
+
+
+
+
+
+
+
+
+
+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
+—-+
+
+
+
+
+
+
+
+
+
+| 1 | SIMPLE | mlc2 | ALL | NULL | NULL | NULL | NULL | 1 | 100.00 | |
+—-+
+
+
+
+
+
+
+
+
+
+SQL trace from crash:
120403 8:52:17 729 Query SET PROFILING = 1
729 Query SET profiling_history_size = 0
729 Query SET profiling_history_size = 15
729 Query SHOW STATUS
729 Query SELECT * FROM mlc2 WHERE webobjectreviewid=14 LIMIT 0, 1000
729 Query SHOW STATUS
729 Query SHOW PROFILES
729 Query select state, round(sum(duration),5) as `duration (summed) in sec` from information_schema.profiling where query_id = 16 group by state order by `duration (summed) in sec` desc
729 Query SET PROFILING = 0
729 Query EXPLAIN EXTENDED
SELECT * FROM mlc2 WHERE webobjectreviewid=14 LIMIT 0, 1000
729 Query SHOW WARNINGS
MitchParticipantAttached is dump of a table that exhibits the same problem. It is a copy of the original table containing only the single row that causes the problem.
This statement works fine:
SELECT * FROM mlc2;
This one causes the crash:
SELECT * FROM mlc2 WHERE webobjectreviewid=14;
-
AuthorPosts