Adriano dos Santos Fernandes
Adriano dos Santos Fernandes
@hvlad look for email sent to us by Pavel Zotov: `Problem "string right truncation / expected length 32765, actual 65533" after fix gh-7592`
Since in some conditions Firebird allows parameters with length 65533, let's wait for a better moment (v6) and maybe increase it in all conditions.
A test case would make easy to identify the problem and would be fundamental to validate a fix.
How can both threads of the same attachment be active in the engine at the same time? Shouldn't one be blocked in the attachment lock of `EngineContextHolder`?
> The most suspicious as for me is the call of MET_index_pagespace() (that run some query) from Statement::release(). And AFAIK, it's not valid to do it from AST handler.
This problem has been resolved in stdout around the 1900 years.
> I would just make ISQL correctly handle RESULT_SEGMENT. If you looked at details of the problem, you would saw that ISQL does not receives `RESULT_SEGMENT` at all.
The Firebird team thanks you for effort duplication. And I'd say this is a problem here too, the way it does is inefficient and more complex than necessary.
> Today we have the following artifact listed on the Wiki homepage as the download for Windows: > > https://github.com/FirebirdSQL/firebird-odbc-driver/actions/runs/11053399614/artifacts/1982531858 This is problematic. Artifacts of actions may expire. > Perhaps...
> Yes, actions can be triggered by a tag and a release can be made (with the corresponding arfifacts), but I'll repeat myself - this is not suitable for a...