Re: [Pharo-project] Who is working on which tests? Why are there new, broken tests?
But if the test is revealing a true flaw in the system, better to keep it than to remove it and pat ourselves on the back about "all tests passing." Bill Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254 Email: bschwab@anest.ufl.edu Tel: (352) 846-1285 FAX: (352) 392-7029
oscar@iam.unibe.ch 06/10/08 10:00 AM >>>
Well, I thought we would start to fix the existing broken tests before adding new broken tests. I would also be very happy to see some comments when someone adds a new broken test, so I know what the plan is. Now I added an implementation of Integer>>hex, for example, but I am not so sure why we need this. Ain't too many senders. An easier fix would have been to remove the broken test. Cheers, - on On Jun 10, 2008, at 15:53, Bill Schwab wrote:
Is your concern that the new tests are broken (which seems normal to me that they would start out that way), or are you objecting to their being included in the image?
I can see arguments either way on the latter. If code in the image is wrong and a clearly correct but currently failing test highlights it, I'd rather be alerted to it. You are also correct that having them included draws attention that could easily encourage duplication of effort.
Comments?
Bill
Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254
Email: bschwab@anest.ufl.edu Tel: (352) 846-1285 FAX: (352) 392-7029
oscar@iam.unibe.ch 06/10/08 8:31 AM >>>
I looked at the wiki page http://code.google.com/p/pharo/wiki/CurrentlyWorkingOn but it does not say much.
I would like to work through some tests, but I cannot find an overview.
The tests CharacterTest>>testHex and IntegerTest>>testHex are broken but seem to be very new. Stef, are you working on these? You seem to have written the tests ...
- on
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Hi Bill, That's not what I mean. I agree fully with you that it is good to add tests that show in which way the system is broken. Adding tests for methods that don't even exist doesn't make much sense to me right now. testHex is (was) testing something that doesn't exist. OK if you are doing test-driven development. Not good if you want a system to be coherent and understandable. At the very least such tests need comments to explain why they have been added. Tests that are fragile and meaningless is also bad. The testHash tests that test that a hash returns a specific value are not good. I prefer to get rid of them. On Jun 10, 2008, at 5:34 PM, Bill Schwab wrote:
But if the test is revealing a true flaw in the system, better to keep it than to remove it and pat ourselves on the back about "all tests passing."
Bill
Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254
Email: bschwab@anest.ufl.edu Tel: (352) 846-1285 FAX: (352) 392-7029
oscar@iam.unibe.ch 06/10/08 10:00 AM >>>
Well, I thought we would start to fix the existing broken tests before adding new broken tests.
I would also be very happy to see some comments when someone adds a new broken test, so I know what the plan is.
Now I added an implementation of Integer>>hex, for example, but I am not so sure why we need this. Ain't too many senders. An easier fix would have been to remove the broken test.
Cheers, - on
On Jun 10, 2008, at 15:53, Bill Schwab wrote:
Is your concern that the new tests are broken (which seems normal to me that they would start out that way), or are you objecting to their being included in the image?
I can see arguments either way on the latter. If code in the image is wrong and a clearly correct but currently failing test highlights it, I'd rather be alerted to it. You are also correct that having them included draws attention that could easily encourage duplication of effort.
Comments?
Bill
Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254
Email: bschwab@anest.ufl.edu Tel: (352) 846-1285 FAX: (352) 392-7029
oscar@iam.unibe.ch 06/10/08 8:31 AM >>>
I looked at the wiki page http://code.google.com/p/pharo/wiki/CurrentlyWorkingOn but it does not say much.
I would like to work through some tests, but I cannot find an overview.
The tests CharacterTest>>testHex and IntegerTest>>testHex are broken but seem to be very new. Stef, are you working on these? You seem to have written the tests ...
- on
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
participants (2)
-
Bill Schwab -
Oscar Nierstrasz