incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Stahl <...@openoffice.org>
Subject Re: Code covered by the Oracle grant
Date Tue, 07 Jun 2011 12:06:33 GMT
On 07/06/11 11:42, Christian Lippka wrote:
> Hi Thorsten,
>
> Am 07.06.2011 11:09, schrieb Thorsten Behrens:
>> Simon Brouwer wrote:
>>> The real question is whether anything essential is missing that Oracle
>>> can't supply and that is very difficult to replace.
>>>
>> If you re-read Christian's mail, the answer to both is "yes". And
>> another remark: given the overall state of the code (~20 years of
>> sedimentation), the full project history is of great value, when one
>> tries to figure out how one specific piece of code came to pass.

yes, the history is definitely valuable; in fact i sometimes am 
frustrated that it only starts in 2000 and misses out the first 10 years...

>> All of that makes starting off from the hg repo appear desirable ...
> While I fully agree that the commit history is of value, I do not see
> the need to
> include them when switching to AL. IMHO it is perfectly legal for anyone
> to clone
> the currently available repositories and archive them and also make them
> available
> publicly. So those information will not be lost, this is the internet :-)

certainly a valid point.
but not having to go to "some other place" to access the history is 
convenient.

> This is not an argument against having the history, I'm perfectly fine
> with that
> solution also. But in this case my personal preference would be to start
> clean.

i don't agree with this.

not using the current HG repository will create a lot of pain when 
integrating the currently outstanding CWSes (assuming those will be 
donated as well, of course).

for example, take a look at the contents of CWS aw080, which replaces 
the tools RTTI with C++ standard RTTI, with changes all over the place, 
and as a patch is >10MB, and tell me that you want to apply that patch 
manually against a code base that has diverged (losing the CWS internal 
history in the process).

besides, as you already noted, there are the scalability problems of 
SVN's merge tracking, that made it unusable with OOo's branch-heavy 
development model after half a year.

having used HG for years now, i see the proposal to use SVN as a big 
regression for OOo development.

regards,
  michael

Disclaimer: These are my personal opinions as an individual interested 
in the future of an open source office suite. I do not speak for my 
current employer.


---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org


Mime
View raw message