131839 - @::fetchahead not undefed in SendSQL. This can/has caused problems with errors in reports of "Mid-air collision detected!"

Status: RESOLVED WONTFIX

(Bugzilla :: Query/Bug List, defect, P2)

Description

[Quentin Arce]

This causes a problem in FetchSQLData as it does not fetch a new row array from
the database if it still contains data.

This was not a problem until I performed updates and selects in the code prior
to this error check.

This has not been fixed in the CVS 2.15 code base.

Add 
 undef @::fetchahead;

at line 214 of globals.pl

This solved the problem.

Thanks,

Quentin Arce

Comment 1

[Quentin Arce]

Just thought I should update the version since it is still a problem in the 2.15
source base.

Comment 2

[Joel Peshkin]

[e-mail] seems to care about these so they wont make 2.18.  If
someone adopts them, they can move them back. Retargetting to 2.20

Comment 3

[Max Kanat-Alexander]

I'm not even sure this is an issue still, but SendSQL is going away anyhow. I
think it's already gone from process_bug.

Comment 4

[Frédéric Buclin]

Is that a problem on the 2.16 branch? If yes, this should be fixed. I think
2.17.4 is not affected due to the rewrite of SendSQL & co, see bug 163290.

Comment 5

[Max Kanat-Alexander]

This is a very minor problem which would require tremendous invasive effort to
fix. It causes very few visible issues. 2.16 is very locked-down.

The blocker will fix this for the tip; I don't want to fix it for 2.16.

Comment 6

[Frédéric Buclin]

(In reply to comment #5)
> The blocker will fix this for the tip; I don't want to fix it for 2.16.

I agree about 2.16. Only security bugs should land on this branch. Moreover, I
think that the 2.18 branch and the tip are already fixed per bug 163290, see my
previous comment. Then we should wontfix this bug.

Comment 7

[Max Kanat-Alexander]

Yep. Agreed. If this problem doesn't show up in 2.18, users can upgrade to 2.18
if they want to have it fixed. And I'm fairly certain that by 2.22 we won't even
be using SendSQL anymore, anyhow.