[ts-gen] dbase question...

Bill Pippin pippin at owlriver.net
Mon Jul 19 10:31:01 EDT 2010


Thanks much for taking care of Vince's question about mysql; from

> ... After a new install [the shim exits with] ...
> Problem: 516 db record uid -- vector offset mismatch ...
> stemming from MySQL 5.1.47 on Fedora 12 and have tried
> ... innodb_autoinc_lock_mode = 2 ...

and Jordi:

> ... Fedora 12 system for testing purposes and, after adding 
> the following parameters in '/etc/my.cnf' it worked finely
> with its default MySQL version 5.1.47:

> transaction-isolation = SERIALIZABLE
> innodb_autoinc_lock_mode = 0

Vince, about your question with respect to Postgres:

> Has any further thought been given to switching backends to
> Postgresql ...

In a recent post, from last Thurday:


I wrote:

> The following is ... interesting primarily for the roadmap
> information about planned features ...


> Although the lack of a windows port is, absent someone else
> porting the code, unlikely to change in the near future, addition
> of Postgres support is likely soon, a complete revamping of symbols
> handling to dynamically build the symbols database from contract
> data queries will be ready sooner than that, and broadly increased
> api level support over a range of api levels is next on our roadmap.

In the above, "Postgres ... soon" should be read to mean, *after*
both increased api level support (versioning), and the dynamic
symbols database creation (contract fill) tasks.  I've just started
the versioning task this morning.

As always with free software, these plans reflect current intentions,
and are subject to change.  Those needing a more certain roadmap are
free to contact our support email address, or contribute code.  Either
approach would be welcome.

For those wishing to contribute to a Postgres port, please realize
that the C++ coding, for the file fill.c, is not a problem, and I'll
enjoy adding that when the time comes; the real headache is outside
the shim, with the database setup sql scripts.  That work includes
adding sequence number tables and triggers for all table inserts, to
replace the mysql autoenum declarations; as well as revising the sql
to eliminate all other mysql-isms.  Boring, time consuming, but not
that hard.  Anyone wishing to start in on shim development could do
a lot worse than starting with the database, and with the postgres
port task.  That being said, the database will change radically
with dynamic contract fill, one reason the fill task comes first, so
anyone wanting to work on the Postgres port should contact me first.



More information about the ts-general mailing list