Message79271
| Author |
belopolsky |
| Recipients |
belopolsky, lemburg, loewis |
| Date |
2009年01月06日.16:27:08 |
| SpamBayes Score |
0.0 |
| Marked as misclassified |
No |
| Message-id |
<d38f5330901060826n204906b9lc98cdf8f53e7f994@mail.gmail.com> |
| In-reply-to |
<49637E8A.6000905@egenix.com> |
| Content |
On Tue, Jan 6, 2009 at 10:53 AM, Marc-Andre Lemburg
<report@bugs.python.org> wrote:
..
> I don't follow you: those symbols are not meant for public use anyway,
> so we can easily change them without any "costs". The same goes for any
> of the private, but still exported symbols in the API, ie. all _Py*
> APIs.
This is the same that I said: "renaming is cost free" if it is done
together with the type change. Martin was right that there is cost
associated with any change. The benefit of "completeness" for a
specialized build would not justify even having this discussion IMO.
On the other hand since the type change needs to be done in order to
make the code correct on a 64 bit platform and compile without
warnings with gcc, a patch is needed anyways. Since we are already
paying the cost of change, renaming is free. There is no disagreement
between you and me here. |
|
History
|
|---|
| Date |
User |
Action |
Args |
| 2009年01月17日 12:14:32 | loewis | set | spambayes_score: 0.830059 -> 0.0 |
| 2009年01月06日 16:27:09 | belopolsky | set | recipients:
+ belopolsky, lemburg, loewis |
| 2009年01月06日 16:27:08 | belopolsky | link | issue4850 messages |
| 2009年01月06日 16:27:08 | belopolsky | create |
|