RE: [cdt-dev] Progressing towards 2.0 GA

[Sebastien Marineau <sebastien@xxxxxxx>]

Hi Doug,


> On the subject of the 2.0 branch, I guess we should do it as 
> soon as the 
> RC build is done on Monday. As we have mentioned, we will be doing 
> performance work on the parser and that is likely to blow 
> everything up 
> right away. The sooner we get started on this the more time 
> we will have 
> to recover. We will likely do this work on the 2.0 branch and 
> merge it 
> back to head once we're stable again. That'll let the 2.1 guys march 
> ahead.

On the subject of parser work, should we not take the reverse approach,
e.g. do the work on the head and move it to the release branch (2.0.x)
after?
This way the 2.0 branch remains stble and shippable?

The second question is with the branch itself -- do you want to do the 
branch, or shall I? Either way works for me...

> 
> An interesting discussion also revolves around the 3.0 work 
> and where that 
> would go. Having too many streams will kill us, but I would 
> love to see 
> the managed build work march ahead as soon as we can. I'm not 
> sure we can 
> get that all done and stable by the end of October, tough.

Right. Any way we could stage it such that 2.1 becomes a mini-release
along the way to 3.0 with a subset of the features? This also has the
benefit of
getting the new features "out there" and getting some feedback on them
before
3.0.


Sebastien

> 
> Doug Schaefer, IBM's Eclipse CDT Architect
> Ottawa (Palladium), Ontario, Canada
> 
> 
> 
> Sebastien Marineau <sebastien@xxxxxxx> 
> Sent by: cdt-dev-admin@xxxxxxxxxxx
> 06/23/2004 12:04 PM
> Please respond to
> cdt-dev
> 
> 
> To
> "'cdt-dev@xxxxxxxxxxx'" <cdt-dev@xxxxxxxxxxx>
> cc
> 
> Subject
> RE: [cdt-dev] Progressing towards 2.0 GA
> 
> 
> 
> 
> 
> 
> Hi Kleo,
>  
> thanks for driving this -- I absolutely agree with your 
> suggestion.The 
> proposal would be to:
>  
> - Target the first CDT release candidate for Monday, June 
> 28th. Between 
> now and then, all 
> CVS commits would have to be for identified 2.0-gating 
> defects and would 
> have to be reported to the CDT 
> dev list. As of Monday, all commits would stop along and 
> everyone would 
> hammer away on the CDT RC.
> One open question is whether we should branch the CDT tree at this 
> point.My suggestion would be to
> do so.
>  
> - After Monday, only gating critical defects would be fixed. We would 
> follow the same process as
> Eclipse for code checkins - Any checkin would have to be 
> submitted to the 
> CDT dev list (as a patch)
> for review, along with rationale, criticality, and bug ID. 
> Commiters could 
> veto the fixes if we deem them too
> risky or inappropriate.
>  
> - We spin a second release candidate (if needed) on 
> Wednesday, soak in it, 
> and declare it GA if
> no stoppers are found, else we go through the RC cycle again.
>  
> How would this work for everyone? There is definitely a lot 
> of interest 
> and desire from everyone to have CDT
> GA at the same time as Eclipse 3.0, and this would put us 
> within a few 
> days. While CDT 2.0 will not be
> perfect, we can document known issues and limitations and 
> address them in 
> upcoming releases.
>  
> As a follow-up, we can have a separate discussion on the 
> definition and 
> timing of what will go into 2.0.1 and 2.1.
> I'll send a follow-up to your email, Kleo, with some 
> proposals and we can 
> discuss from there.
>  
> Thanks,
>  
> Sebastien
> -----Original Message-----
> From: Kleanthis Hapitas [mailto:khapitas@xxxxxxxxxx]
> Sent: Tuesday, June 22, 2004 12:02 PM
> To: cdt-dev@xxxxxxxxxxx
> Subject: [cdt-dev] Progressing towards 2.0 GA
> 
> 
> Hi all, 
> 
> I thought we should go around and see how we are progressing 
> towards the 
> GA milestone targeted for the end of the month. 
> 
> The Eclipse platform declared RC3 just this past weekend, and 
> they have 
> identified about 25 defects that they will still want to fix 
> before going 
> GA. So, they are planning to have RC4 built by the middle of 
> this week, 
> and once those 25 fixes have been tested and verified, and 
> assuming no 
> further regressions, they will announce GA likely before the 
> end of this 
> week (June 25). 
> 
> Given that Eclipse will be done by June 25, we should think 
> about drawing 
> a line for when the CDT could build our (first) release 
> candidate. Once we 
> do that, the goal would be that everyone takes a good look at 
> the build to 
> make sure we have no further show stoppers. If we don't we 
> could declare 
> GA after a few days of soak. If we do, we would do another release 
> candidate build, and so on... 
> 
> So, what say you? What's a reasonable target for an initial release 
> candidate? End of the week? Early next week? 
> 
> Cheers, 
> 
> Kleo Hapitas 
> 
> 
> _______________________________________________
> cdt-dev mailing list
> cdt-dev@xxxxxxxxxxx
> http://dev.eclipse.org/mailman/listinfo/cdt-dev
>