Now that we know that there is going to be RC6 I think there is no point
in waiting for another week for it just because we have to release
things on Thursday.
I think that the best thing would be to make immediate RC6 and allow
people some time to test it, and if nothing critical comes up, release
the final version in a week or so.
Edin
RC6 will be out on Thursday to give people a chance to identify any
remaining issues with the current state of 5.2.0 as well as allow
existing bug reporters to confirm the critical issues outlined have
been resolved.
Now that we know that there is going to be RC6 I think there is no
point in waiting for another week for it just because we have to
release things on Thursday.I think that the best thing would be to make immediate RC6 and
allow people some time to test it, and if nothing critical comes
up, release the final version in a week or so.Edin
Ilia Alshanetsky
I believe that making a RC now will increase the probability of people
testing out stuff, rather than expecting them to checkout the latest
CVS. I see absolutely no point in waiting to some arbitrary weekday.
Edin
Ilia Alshanetsky wrote:
RC6 will be out on Thursday to give people a chance to identify any
remaining issues with the current state of 5.2.0 as well as allow
existing bug reporters to confirm the critical issues outlined have been
resolved.Now that we know that there is going to be RC6 I think there is no
point in waiting for another week for it just because we have to
release things on Thursday.I think that the best thing would be to make immediate RC6 and allow
people some time to test it, and if nothing critical comes up, release
the final version in a week or so.Edin
Ilia Alshanetsky
I believe that making a RC now will increase the probability of
people testing out stuff, rather than expecting them to checkout
the latest CVS. I see absolutely no point in waiting to some
arbitrary weekday.
Aside from a few issues RC5 is identical to RC6 for testing purposes
it is good enough to allow people to identify major faults. As far as
latest CVS is concerned, there is no need to checkout CVS due to a
wonderful thing called snapshots, which you can usually find at
http://snaps.php.net/.
Ilia Alshanetsky
RC6 will be out on Thursday to give people a chance to identify any
remaining issues with the current state of 5.2.0 as well as allow
existing bug reporters to confirm the critical issues outlined have
been resolved.
If you are referring to bug #39060 then I here by confirm that the bug
outlined in the report is fixed. I doubt however the realpath() bug
in ZTS on BSD will ever be fixed.
-Hannes
Now that we know that there is going to be RC6 I think there is no
point in waiting for another week for it just because we have to
release things on Thursday.I think that the best thing would be to make immediate RC6 and
allow people some time to test it, and if nothing critical comes
up, release the final version in a week or so.Edin
Ilia Alshanetsky
Edin Kadribasic wrote:
Now that we know that there is going to be RC6 I think there is no point
in waiting for another week for it just because we have to release
things on Thursday.I think that the best thing would be to make immediate RC6 and allow
people some time to test it, and if nothing critical comes up, release
the final version in a week or so.
Many people haven't had a chance to look at RC5 yet. I think cycling
through them too fast is going to frustrate people. Give them a couple
of days to report other issues before going to RC6.
-Rasmus
Rasmus Lerdorf wrote:
Edin Kadribasic wrote:
Now that we know that there is going to be RC6 I think there is no
point in waiting for another week for it just because we have to
release things on Thursday.I think that the best thing would be to make immediate RC6 and allow
people some time to test it, and if nothing critical comes up, release
the final version in a week or so.Many people haven't had a chance to look at RC5 yet. I think cycling
through them too fast is going to frustrate people. Give them a couple
of days to report other issues before going to RC6.
Having to wait a week while there is another RC and allowing commits in
the meanwhile, any of which can break things, is not the best release
management from my perspective.
Edin
Rasmus Lerdorf wrote:
Edin Kadribasic wrote:
Now that we know that there is going to be RC6 I think there is no
point in waiting for another week for it just because we have to
release things on Thursday.I think that the best thing would be to make immediate RC6 and allow
people some time to test it, and if nothing critical comes up, release
the final version in a week or so.Many people haven't had a chance to look at RC5 yet. I think cycling
through them too fast is going to frustrate people. Give them a couple
of days to report other issues before going to RC6.Having to wait a week while there is another RC and allowing commits in
the meanwhile, any of which can break things, is not the best release
management from my perspective.
I tend to disagree with you here, Edin.
Releasing one RC just after another - that would be bad management.
RCs are RCs, if something is broken in one of them, let's wait for the next one and see if it ok.
Also, I don't like the rush in the release management which to appears from time to
time, but this time (I mean 5.2) something (or someone) has changed and I really like how it goes,
so let's make it as stable as we can and push it out of the door without any hurry.
--
Wbr,
Antony Dovgal
Antony Dovgal wrote:
I tend to disagree with you here, Edin.
Releasing one RC just after another - that would be bad management.
RCs are RCs, if something is broken in one of them, let's wait for the
next one and see if it ok.Also, I don't like the rush in the release management which to appears
from time to time, but this time (I mean 5.2) something (or someone) has
changed and I really like how it goes, so let's make it as stable as we
can and push it out of the door without any hurry.
I believe that there is very little sense in making release candidates
and continue on committing patches, and in case of PHP major feature
changes while in the release process. I guess changes to the code should
be kept to minimum and reserved for critical bug fixes. Otherwise there
is no telling what broke since the last release candidate and they
become pointless.
Edin