I think Marcus has done more useful harvesting work in the last couple of days than most of us in N weeks.
I'd say that Marcus can and should be trusted to approve stuff, except for one thing - I haven't seen him post critical comments on any of the code he's reviewed, and for anyone that's going to be able to get stuff into the image where we all have to live with it, I want to know that he can be point out where code is lacking.
Except for the selection aspect, this is also important because posting these public comments is the way knowledge is transmitted about good/bad practice, and not just to the author of the specific code.
What does everybody think?
Daniel
Daniel Vainsencher wrote:
I'd say that Marcus can and should be trusted to approve stuff, except
Yup!
for one thing - I haven't seen him post critical comments on any of the code he's reviewed, and for anyone that's going to be able to get stuff
He'll most likely do so if encouraged enough :-)
Michael
Hi Daniel, Marcus and all!
Daniel Vainsencher danielv@netvision.net.il wrote:
I think Marcus has done more useful harvesting work in the last couple of days than most of us in N weeks.
Yup. :-)
I'd say that Marcus can and should be trusted to approve stuff, except
Agree.
for one thing - I haven't seen him post critical comments on any of the code he's reviewed, and for anyone that's going to be able to get stuff into the image where we all have to live with it, I want to know that he can be point out where code is lacking.
I am sure Marcus can do that too! :-) Personally I have found that the code I have reviewed from others generally is good - but class comments and method comments tend to lack IMHO.
I really want the quality of comments to be higher in the official classes. I would personally never let classes in without class comments.
Except for the selection aspect, this is also important because posting these public comments is the way knowledge is transmitted about good/bad practice, and not just to the author of the specific code.
What does everybody think?
Daniel
I definitely think Marcus should be a Harvester - if he likes to be one. :-) Marcus is surely a more competent Squeaker than I am so... ;-)
regards, Göran
On Sun, Jun 08, 2003 at 10:40:28PM +0100, goran.krampe@bluefish.se wrote:
Except for the selection aspect, this is also important because posting these public comments is the way knowledge is transmitted about good/bad practice, and not just to the author of the specific code.
What does everybody think?
Daniel
I definitely think Marcus should be a Harvester - if he likes to be one.
Yes, there's lots of stuff to harvest and I'd like to help with that.
Marcus
goran.krampe@bluefish.se wrote:
Daniel Vainsencher danielv@netvision.net.il wrote:
Except for one thing - I haven't seen him post critical comments on any of the code he's reviewed, and for anyone that's going to be able to get stuff into the image where we all have to live with it, I want to know that he can be point out where code is lacking.
True, although I have to admit that I don't often have comments about the source code that I'm reviewing. More often I have a comment about some aspect of what the change does. (such as my extra comments about the BoundsInHaloFix2 changeset)
I guess I usually just give the source code (diffs) a quick look-over to make sure that nothing jumps out at me as being ugly. As far as judging source code, another factor is who wrote it... typically I will have a level of trust that somone like Ted Kaehler or Ned Konz won't need to have their code heavily scrutinized, but I may look more closely at the code of someone I don't know. (However, external testing is another matter... that's equally important for all submissions.)
Personally I have found that the code I have reviewed from others generally is good - but class comments and method comments tend to lack IMHO.
I really want the quality of comments to be higher in the official classes. I would personally never let classes in without class comments.
This is probably a good rule of thumb. New classes should certainly include class comments before we approve them. It's harder to have any simple rule about method comments... most methods do not need comments and should not have them, but some methods do need comments.
Except for the selection aspect, this is also important because posting these public comments is the way knowledge is transmitted about good/bad practice, and not just to the author of the specific code.
What does everybody think?
I definitely think Marcus should be a Harvester - if he likes to be one. :-)
I agree. (Oh wait, I see he's already approving items anyway... :-) )
- Doug Way
On Mon, Jun 09, 2003 at 12:37:11PM -0400, Doug Way wrote:
I definitely think Marcus should be a Harvester - if he likes to be one. :-)
I agree. (Oh wait, I see he's already approving items anyway... :-) )
Sorry ;-) somewhat premature, I guess. But those were really simple: a one-liner and two documentation-only, if I remember correctly.
Regarding Daniel's suggestion to comment more: Yes, you are right, I'l write more comments about the code in future posts.
goran.krampe@bluefish.se wrote:
I would personally never let classes in without class comments.
There's actually a test in the Tests-package for that: I've done a Abstract testing-class called "ClassTestCase". If you inherit from this one instead of TestCase (and name the test "<name of the class>Test", you get an error if the class-comment is empty.
(and you get 3 more tests: testNew, testUnCategorizedMethods, testRecompileAll)
There are not yet tests for every class in the system, but as soon as the testserver is running, I'l add more.
Marcus
Marcus Denker wrote:
On Mon, Jun 09, 2003 at 12:37:11PM -0400, Doug Way wrote:
I definitely think Marcus should be a Harvester - if he likes to be one. :-)
I agree. (Oh wait, I see he's already approving items anyway... :-) )
Sorry ;-) somewhat premature, I guess. But those were really simple: a one-liner and two documentation-only, if I remember correctly.
I went ahead and added Marcus to the current list of harvesters, to make it official:
http://swiki.squeakfoundation.org/squeakfoundation/45
- Doug
squeakfoundation@lists.squeakfoundation.org
-
Daniel Vainsencher -
Doug Way -
goran.krampe＠bluefish.se -
Marcus Denker -
Michael Rueger