Re: Created branch 1.0

2015-02-09 Thread Vikram Dixit K
The build check in HIVE-8933 fixed in HIVE-8845. On Mon, Feb 9, 2015 at 11:32 AM, Vikram Dixit K wrote: > Hi Ed, > > This was the case with 0.14. It was fixed before 1.0 went out in HIVE-8933. > > Thanks > Vikram. > > On Mon, Feb 9, 2015 at 9:08 AM, Alan Gates wrote: > >> That's fixed, correct?

Re: Created branch 1.0

2015-02-09 Thread Vikram Dixit K
Hi Ed, This was the case with 0.14. It was fixed before 1.0 went out in HIVE-8933. Thanks Vikram. On Mon, Feb 9, 2015 at 9:08 AM, Alan Gates wrote: > That's fixed, correct? I do not believe there were any SNAPSHOT > dependencies in 1.0. > > Alan. > > Edward Capriolo > February 9, 2015 at

Re: Created branch 1.0

2015-02-09 Thread Alan Gates
That's fixed, correct? I do not believe there were any SNAPSHOT dependencies in 1.0. Alan. Edward Capriolo February 9, 2015 at 8:40 Because we can not really have a stable api if by definition we build around snapshot dependencies. On Mon, Feb 9, 2015 at 11:38

Re: Created branch 1.0

2015-02-09 Thread Edward Capriolo
Because we can not really have a stable api if by definition we build around snapshot dependencies. On Mon, Feb 9, 2015 at 11:38 AM, Edward Capriolo wrote: > Question. > > https://issues.apache.org/jira/browse/HIVE-8614 > > Did we not just agree in this thread that hive will no long have > depen

Re: Created branch 1.0

2015-02-09 Thread Edward Capriolo
Question. https://issues.apache.org/jira/browse/HIVE-8614 Did we not just agree in this thread that hive will no long have dependency that are SNAPSHOT? On Fri, Jan 23, 2015 at 1:06 AM, Brock Noland wrote: > Hi Alan, > > I agree with Xuefu and what was suggested in your statement. I was > thin

Re: Created branch 1.0

2015-01-27 Thread Brock Noland
Hi, Yes as my main concern was that the 1.0 release would delay the subsequent release. If we are ready to do RC's now, I find this plan acceptable. Thank you for working hard to get this rc out. Brock On Tue, Jan 27, 2015 at 12:28 PM, Vikram Dixit K wrote: > Hi Folks, > > It has been a few d

Re: Created branch 1.0

2015-01-27 Thread Vikram Dixit K
Hi Folks, It has been a few days and I have done all the work needed to produce a 1.0 RC and think it is better to have a vote on it. I still hope that we can have this release as 1.0 and Brock's release as 1.1. By the end of the day I think having more releases is a good thing for the community a

Re: Created branch 1.0

2015-01-23 Thread Sergey Shelukhin
I think the way it is done in Hadoop space is better for Hadoop space (and better wrt consistency, us being in the Hadoop space). Because no single company or QA process controls or covers all the changes to the product, and some changes go unseen by every actor, stabilization period is a must...

Re: Created branch 1.0

2015-01-23 Thread Sergey Shelukhin
We really didn't define 1.0 very well. The way HBase did this (correct me if I'm wrong) is that 1.0 was the stable release (stable APIs and code). Then 1.1, 1.2 etc. become patch releases, and 2.0 becomes the release for all the new crazy stuff... On Fri, Jan 23, 2015 at 4:01 PM, Lefty Leverenz w

Re: Created branch 1.0

2015-01-23 Thread Lefty Leverenz
Setting aside branching specifics for a moment, what are the pros & cons of doing release 1.0 now or later? - From a documentation point of view, planning a later 1.0 release could give us time to catch up on the backlog of doc tasks, or at least prioritize the backlog and address the mos

Re: Created branch 1.0

2015-01-23 Thread Szehon Ho
Wherever I've seen in enterprise software, the trunk-based development model has been the standard where all release branches are cut from trunk and short-lived. I've never heard of a case where a branch originally designated for 0.14 (minor release) is cut again to become 1.0 (major release), and

Re: Created branch 1.0

2015-01-23 Thread Gopal V
On 1/23/15, 6:59 AM, Xuefu Zhang wrote: While it's true that a release isn't going to include everything from trunk, proposed 1.0 release is branched off 0.14, which was again branched from trunk long time ago. If you compare the code base, you will see the huge difference. From the stability a

Re: Created branch 1.0

2015-01-23 Thread Sergey Shelukhin
I don't think that 1.0 release is expected to have major increase in functionality, in fact I strongly disagree with this. Major release is supposed to be stable. With 0.14, we have a good evidence of use both in the community, and releases within the platforms by major vendors that caused bugs (i

Re: Created branch 1.0

2015-01-23 Thread Xuefu Zhang
> Every release of hive has been a subset of tip of the trunk (we branch > for release while trunk moves ahead), and super set of changes of > every previous release. So every release so far has had a subset of > functionality of hive trunk branch (if that is what you are referring > to). While it

Re: Created branch 1.0

2015-01-22 Thread Brock Noland
Hi Alan, I agree with Xuefu and what was suggested in your statement. I was thinking we'd release the next release as 0.15 and then later there would be 1.0 off trunk (e.g. what would have been 0.16) and thus be superset (minus anything we intentionally remove). As I have said several times, I'd

Re: Created branch 1.0

2015-01-22 Thread Thejas Nair
On Thu, Jan 22, 2015 at 2:49 PM, Edward Capriolo wrote: > If we do a 1.0.0 release there is no problem with us later releasing 14.1 > or a 14.2. I think everyone would understand that the 14.2 being released > after 1.0.0 would likely have back ported features. Releasing a 15.0 after > 1.0 would n

Re: Created branch 1.0

2015-01-22 Thread Thejas Nair
On Thu, Jan 22, 2015 at 12:31 PM, Xuefu Zhang wrote: > Hi Thejas/Alan, > > From all the argument, I think there was an assumption that the proposed > 1.0 release will be imminent and 0.15 will happen far after that. Based on > that assumption, 0.15 will become 1.1, which is greater in scope than 1

Re: Created branch 1.0

2015-01-22 Thread Edward Capriolo
If we do a 1.0.0 release there is no problem with us later releasing 14.1 or a 14.2. I think everyone would understand that the 14.2 being released after 1.0.0 would likely have back ported features. Releasing a 15.0 after 1.0 would not make as much sense as we probably do not want two active branc

Re: Created branch 1.0

2015-01-22 Thread Xuefu Zhang
Hi Thejas/Alan, >From all the argument, I think there was an assumption that the proposed 1.0 release will be imminent and 0.15 will happen far after that. Based on that assumption, 0.15 will become 1.1, which is greater in scope than 1.0. However, this assumption may not be true. The confusion wi

Re: Created branch 1.0

2015-01-22 Thread Alan Gates
I had one clarifying question for Brock and Xuefu. Was your proposal to still call the branch from trunk you are planning in a few days 0.15 (and hence release it as 0.15) and have 1.0 be a later release? Or did you want to call what is now 0.15 1.0? If you wanted 1.0 to be post 0.15, are yo

Re: Created branch 1.0

2015-01-22 Thread Thejas Nair
Brock, Xuefu, We seem to have trouble reaching to a consensus here. (Please see my arguments why I don't see this causing confusions, and let me know if it changes your opinion). How should we move forward ? Do you think we need to go through a formal vote regarding the release plan as per hive by

Re: Created branch 1.0

2015-01-22 Thread Thejas Nair
I don't see any reasons for confusion. >From a user perspective, 1.0 is going to have a super set of changes of 0.14. 1.1 (based on planned 0.15 release) will have a super set of changes in 1.0 . On Wed, Jan 21, 2015 at 10:47 PM, Xuefu Zhang wrote: > I strongly believe that the concept of 1.0 o

Re: Created branch 1.0

2015-01-22 Thread Thejas Nair
On Wed, Jan 21, 2015 at 7:09 PM, Brock Noland wrote: > Sorry about the confusion. That link misses the content to which I was > replying. I agree that 1.0 should be stable, but that does not mean that I > agree that 1.0 should come from 0.14. For a stable release, we would need to go through an

Re: Created branch 1.0

2015-01-21 Thread Xuefu Zhang
I strongly believe that the concept of 1.0 out of a branch as proposed is creating the greatest confusion in the community. If for any reason that 1.0 cannot be cut from the trunk, that means that we are not ready and so shall wait until so before considering such a release. Thus, I'd -1 on this p

Re: Created branch 1.0

2015-01-21 Thread Gopal V
On 1/21/15, 7:09 PM, Brock Noland wrote: Too be clear I strongly feel creating 1.0 from 0.14 will be confusing. In fact it's already crrated confusion amongst folks on this list. Furthermore 1.0 should be created from trunk and be a superset of previous releases. I don't think there is any con

Re: Created branch 1.0

2015-01-21 Thread Brock Noland
Vikram, Sorry about the confusion. That link misses the content to which I was replying. I agree that 1.0 should be stable, but that does not mean that I agree that 1.0 should come from 0.14. Here is one with the content to which I was replying: http://mail-archives.apache.org/mod_mbox/hive-dev/2

Re: Created branch 1.0

2015-01-21 Thread Vikram Dixit K
@Brock, I created this branch from 0.14. I created this branch based on the email thread discussing 1.0, http://search-hadoop.com/m/8er9YGX8g2 where you had said you agreed with the suggestion from Enis from HBase who said that we should base 1.0 on a stable version rather than making it a featu

Re: Created branch 1.0

2015-01-21 Thread Thejas Nair
Hi Lefty, Yes, you are right. Anything that is not fixed in 0.14 and is fixed in 1.0 would have 1.0 as the fixed version. Yes, 0.15.0 would then become 1.1.0 . Yes, it is a good idea to document this translation somewhere. On Wed, Jan 21, 2015 at 3:28 PM, Lefty Leverenz wrote: > So my initial i

Re: Created branch 1.0

2015-01-21 Thread Thejas Nair
As discussed in the previous thread [1] I don't see why it would be horribly confusing to have- 1.0.0 = 0.14 + changes to delete things due to deprecation + additional API documentation. 0.15 would then become 1.1.0 . I should be able to submit some patches for documenting public APIs by tomorrow.

Re: Created branch 1.0

2015-01-21 Thread Brock Noland
Hi, I should have just waited to check where this branch came from, but I was quite surprised when I received this mail so I too quickly fired off a response. I apologize to everyone for the spam. The reason I was surprised is that I don't feel there was a consenus comming out of the 1.0 discussio

Re: Created branch 1.0

2015-01-21 Thread Lefty Leverenz
So my initial impression was correct -- instead of calling it release 0.14.1, we're calling it 1.0.0. Or am I hopelessly confused? Will 0.15.0 be 1.1.0? (If so, I'll need to edit a dozen wikidocs.) Will release numbers get changed in JIRA issues? Presumably that's not possible in old comments,

Re: Created branch 1.0

2015-01-21 Thread Eugene Koifman
could we include HIVE-9390 & HIVE-9404? This has been committed to trunk. They add useful retry logic to support insert/update/delete functionality. On Wed, Jan 21, 2015 at 1:06 PM, Vikram Dixit K wrote: > Hi Folks, > > I have created branch 1.0 as discussed earlier. All the jiras that have > 0

Re: Created branch 1.0

2015-01-21 Thread Brock Noland
Vikram, Which branch was the 1.0 branch created from? Brock On Jan 21, 2015 1:07 PM, "Vikram Dixit K" wrote: > Hi Folks, > > I have created branch 1.0 as discussed earlier. All the jiras that have > 0.14 as the fix version should be committed to 1.0 branch instead. The list > of jiras that are