pri_id in eventum_issue table is to small: 127 max

[Vlad Safronov]

Hi,

iss_pri_id is tinyint(1) in eventum_issue table and easily runs out out range on 30+ projects.
eventum_project_priority table has pri_id tinyint(1) unsigned and thus is inconsistent with iss_pri_id.

/v

[Elan Ruusamäe (glen666)]

you must of have very old database, it was changed to smallint(3) in 2005

care to post mysql alter statements to make it consistent and bigger?

[Elan Ruusamäe (glen666)]

that old change from 2005:

https://github.com/eventum/eventum/commit/5f8695e3b0afb4e41d9f8da0fb7a8cdb3cea49a4

[Elan Ruusamäe (glen666)]

from that commit, i can extract:

ALTER TABLE eventum_issue modify iss_pri_id smallint(3) NOT NULL default 0;
ALTER TABLE eventum_project_priority modify pri_id smallint(3) unsigned NOT NULL auto_increment;

do you have anything to add?

[Vlad Safronov (vsafronov)]

Thank you. This solved the trouble.
Yes, the database is pretty old :)

[Elan Ruusamäe (glen666)]

i asked you a question:

> you must of have very old database, it was changed to smallint(3) in 2005
> care to post mysql alter statements to make it consistent and bigger?

[Vlad Safronov (vsafronov)]

hm..

Which mysql alter statements to post?

I have applied the suggested above sql:
ALTER TABLE eventum_issue modify iss_pri_id smallint(3) NOT NULL default 0;
ALTER TABLE eventum_project_priority modify pri_id smallint(3) unsigned NOT NULL auto_increment;

Eventum entity is old, running since 2004 or 2005 and I was not one who installed. I upgraded this instance to 2.3.3.
Have I missed something, required database upgrade?

[Elan Ruusamäe (glen666)]

so, still, you found only two fields (the same ones i posted), and the size should be smallint(3)?

[Vlad Safronov (vsafronov)]

yes, I found only two fields,

well, I think SMALLINT is Ok, yet things are not totally consistent due to usage UNSIGNED in only one field:

+ iss_pri_id smallint(3) NOT NULL default 0,

+ pri_id smallint(3) unsigned NOT NULL auto_increment,

this would hardly cause a bug near future though (>32767), yet it is not correct.

[Elan Ruusamäe (glen666)]

On 02.10.2015 15:19, Vlad Safronov wrote:
> yes, I found only two fields,
>
> well, I think SMALLINT is Ok, yet things are not totally consistent due
> to usage UNSIGNED in only one field:
>
> + iss_pri_id smallint(3) NOT NULL default 0,
>
> + pri_id smallint(3) unsigned NOT NULL auto_increment,
>
> this would hardly cause a bug near future though (>32767), yet it is not
> correct.
>

so, lets change to "smallint(3) unsigned" everywhere?

--
glen

[Vlad Safronov (vsafronov)]

Does Brian have any opinion on this? Brian made that change to SMALLINT, was that typo or intentional?
As I said this bug will hardly ever repeat with even SMALLINT..

I would just change booth to just plain INT and stop worring about.

[Vlad Safronov (vsafronov)]

s/booth/both/

[Elan Ruusamäe (glen666)]

committed as f9951348986d182feb2eff7cf4be01e430f98f52

https://github.com/glensc/eventum/commit/f9951348986d182feb2eff7cf4be01e430f98f52

will merge later (depends on prior PR to be merged)