Doug Gregor wrote:
Frankly, I think this whole approach of "fixing the process" is wrongheaded. We're in this mess because our *tools* are broken, not our *process*.
That's the issue. its the process that's broken. No amount of improvement in the tools will fix it.
What doesn't work? Regression testing.
This is true. And it can never be made to work and be useful under the current procedures.
- Doug
P.S. Here are some of the many things I could have done that would have been more productive than writing the message above: 1) Made regression.py work with Subversion, so that we would be performing regression testing on the trunk.
wouldn't be necessary under the new proposal.
2) Looked at the changes made on the RC_1_34_0 branch to determine which ones can be merged back to the trunk.
wouldn't be necessary under the new proposal.
3) Fixed some of the current failures on the trunk.
wouldn't be necessary under the new proposal.
4) Setup a new nightly regression tester.
wouldn't be necessary under the new proposal.
5) Studied Dart2 to see how we can make it work for Boost
Hmm I don't know what this is. But another tool won't make a difference.
6) Investigated the problems with incremental testing.
of course any improvement is necessary. But current tools work well enough. They are not the bottleneck.
7) Improved the existing test reporting system to track Subversion revisions associated with test runs, and link to those revisions in the Trac.
wouldn't be necessary under the new proposal.
8) Improved the existing test reporting system to track changes from day to day
wouldn't be necessary under the new proposal.
Before I reply to any messages in this thread, I'll be thinking about that list. Will you? P.P.S. I know I sound grumpy, because I am. The amount of time we have collectively used discussing policies would have used far more wisely to improve the tools we have.
You guys have been making a heroic effort and in no way to I want to denigrate your efforts. But I think you're shoveling against the tide. We have more problems now than before not because someone isn't working hard enough or putting enough time. The current process is fundamentally flawed in that it doesn't scale. Working harder with better tools isn't going to produce the improvements that re-thinking the process will. Of course the beauty of this is we really don't all have to agree. You're free to improve the tools for trunk testing and the like and those of us who want to are free to use branches for development. If we can't agree to move to smaller incremental releases - well, we can do what Joaquin has done - post incremental improvements on a library by library basis. We'll see how it works out. Robert Ramey