Hi Saptarshi,

In principle I'm fine with these, but let's get at the specifics of
datatypes.

We probably use the mysql text, mediumtext, and longtext types a lot,
because they're really usefully-sized datatypes. What will we replace those
with?

What other datatypes are you expecting to change?

-Darius

On Wed, Aug 17, 2011 at 11:25 AM, Saptarshi Purkayastha <[email protected]>wrote:

> Hi,
>
> While we were recently discussing that OpenMRS should run on multiple
> database servers, I started working on Support Multiple Databases in
> OpenMRS Installation and Update<https://tickets.openmrs.org/browse/TRUNK-1925>
> .
> There are some of the following broad changes that need to be made the the
> liquibase xml so that we can install OpenMRS on different databases (target
> MySQL, Postgres, MsSQL... and may be later Oracle).
>
> 1.) Change datatypes and create tables that are compatible with all the
> databases (directly for compatible types or database specific for
> non-compatible types)
> 2.) Remove precision from many columns which are not compatible across all
> those database servers
> 3.) Divide a larger changeset into smaller changeset so that they can be
> done commonly (by changing syntax) across multiple database servers.
> 4.) Due to move to liquibase 2.0, all checksums for changesets have been
> NULL'd and then changed to the new format. This should happen automatically,
> but if anyone depends on these checksums, then you should reply to this
> email :)
>
> We may also have to change if anywhere non-ANSI syntax or MySQL-syntax has
> been used in the DAO. I haven't reached that far to tell how many such
> instances exist, but I hope there aren't many such places
>
> What do other developers think about these changes? Any suggestions on the
> way?? Anything that u think should be avoided or done?? These are fairly
> large changes and may result in incompatible checksums for already run
> changesets.
>
>
> PS: On a sidenote, I would like to highlight that for new installations it
> is silly that we are doing changesets based installations. While the world
> has moved to image based deployments for OS and large-software packages, we
> did the reverse and moved to changesets based installations. Upgrades are
> easier through changesets, but for new installations they are lengthy and
> boring. I would like to propose the for new installations, just an sqldump
> deploy is easy and fast, while keeping to changesets for upgrades.
>
> ---
> Regards,
> Saptarshi PURKAYASTHA
>
> My Tech Blog:  http://sunnytalkstech.blogspot.com
> You Live by CHOICE, Not by CHANCE
>  ------------------------------
> Click here to 
> unsubscribe<[email protected]?body=SIGNOFF%20openmrs-devel-l>from 
> OpenMRS Developers' mailing list

_________________________________________

To unsubscribe from OpenMRS Developers' mailing list, send an e-mail to 
[email protected] with "SIGNOFF openmrs-devel-l" in the  body (not 
the subject) of your e-mail.

[mailto:[email protected]?body=SIGNOFF%20openmrs-devel-l]

Reply via email to