GigaSchema extends the startup() function of Schema shell like this:

function startup()
{
Utility::cleanDir(TMP.’cache’.DS.’models’.DS); // delete BS
parent::startup();
}

And here’s the code of cleanDir() in Utility class

function cleanDir($path)
{
$folder=& new Folder($path);
$tree=$folder->tree($path, false);
foreach ($tree as $files) {
foreach ($files as $file) {
if (!is_dir($file)) {
$file=& new File($file);
$file->delete();
}
}
}
return;
}

Hope it helps.

like ignore cache could be i
…cake schema generate -fi

or delete cache could be d
…cake schema generate -fd

Or just make it so that -f blows out the cache. That would like make sense instead of searching and searching and searching and searching because the generate command doesn’t seem to see whats in the database ever.

Or i could just svn:ignore my tmp/cache.

While the shell clearly states: “Comparing Database to Schema…”, it actually compares the model cache to the schema. By removing the model cache, it will be regenerated before the comparison takes place.

Here’s what you’ll do:

(I’m not seeing anything at this point in the article, where I assume that there was at one time an explanation about using the schema exported earlier in the post?)

And done, our database is now synced.