forums › forums › SQLyog › SQLyog: Bugs / Feature Requests › Nasty Bug Found In Text Entry: V5.14 And V5.15
- This topic is empty.
-
AuthorPosts
-
-
July 19, 2006 at 12:34 pm #21960
peterlaursen
Participantit is all explained here:
http://www.webyog.com/faq/8_99_en.html
I think your column is defined as DEFAULT NULL ?
For those few words that are KEYWORDS, you must enclose the string in backquotes like `default` .
-
July 19, 2006 at 10:58 pm #21961
mattRobertson
Memberpeterlaursen wrote on Jul 19 2006, 05:34 AM:it is all explained here:http://www.webyog.com/faq/8_99_en.html
I think your column is defined as DEFAULT NULL ?
For those few words that are KEYWORDS, you must enclose the string in backquotes like `default` .
Yes that fixed it. Thank you!
So if you ever edit a record with one of these values in it, SQLYog will assume you want its SQLYog-specific behavior and — if that assumption is incorrect — corrupt it. And this is by design? I have a GUI so I can edit the data myself, not so something can decide to edit it for me. If there is no way to turn off this behavior I could wind up corrupting any database record I touch. I go into client systems every day and many times I haven't got a clue whats inside until I look. I can't just go in and change a field that needs it anymore as SQLYog could break the record? Thats not something I should ever have to worry about from a DB GUI.
In the case of the cms I am editing, thankfully I don't use this so much that I can't provide some instruction to work around it. But this is a built-in way for your product to crash the client's entire application if they forget that instruction.
I can see where the words other than “default” are highly unlikely to see themselves placed as standalone values, but “default” itself seems mighty common, and its of course not a reserved word in mySQL itself.
How can this be disabled? Honestly I have never heard of a db fornt end that will decide for itself what it will do with existing data that it wasn't expressly asked to alter by the user, and am a little dumbfounded to find out the behavior was deliberate.
At least put in a switch somewhere in preferences!
-
July 19, 2006 at 11:10 pm #21962
peterlaursen
ParticipantQuote:At least put in a switch somewhere in preferences!We will discuss that.
BTW: Please observe too that also functions are evalutated.
-
July 20, 2006 at 12:08 am #21963
mattRobertson
Memberpeterlaursen wrote on Jul 19 2006, 04:10 PM:We will discuss that.BTW: Please observe too that also functions are evalutated.
Hopefully nobody has something like that in a data field. It is unlikely but nothing is impossible.
I would suggest the default behavior is to shut this feature off, since it can — and in my case did — result in data alterations without any direct action on the part of the developer. If I hadn't had a freebie 5.02 sitting on another machine I (or more accurately my client) would have been out of business until I wrote and ran manual SQL queries to replace the destroyed data. It doesnt' take long to write a query but it took a couple of hours to figure out which of the dozens of tables was throwing an error and why.
-
July 20, 2006 at 7:22 am #21964
Ritesh
MemberWe will put that as a preference in the next release.
-
-
AuthorPosts
- You must be logged in to reply to this topic.