Re: tla and keeping code

[Richard Stallman]

    What I'm thinking of doing is to rewamp the whole
    "fetch-extract-patch" bit to use TLA archvies.  So instead of fetching
    any code using ftp/http/etc, extracting it and then patching it before
    building, we would just grab the TLA archive for the package.

    This is quite a big change since not many GNU projects use TLA to
    handle source code managment.  So one would have to mirror a bunch of
    GNU projects and non-GNU projects manually for some time.

The meat of your proposal is that instead of fetching the code,
extracting it, and patching it each time you make a release, you will
do that and save the results.

Using TLA seems to be a side issue, orthogonal to that.
I don't see how it helps or relates to the rest.