Re: Fair large change to contributors - Mailing list pgsql-www

Stefan Kaltenbrunner wrote:
> Joshua D. Drake wrote:
>> Magnus Hagander wrote:

>> Consider... I have a table called people but fill it full of companies.
>> That is wrong. Consistency is extremely important to the maintenance of
>> any project and any code base.
> 
> I disagree here - I think renaming is simply a useless act and I'm firm
> with magnus here that playing this game will hardly gain us much but
> having to deal with annoyances here and there :-(

It isn't just a rename, it is a move to the appropriate organized 
structure that is represented correctly except where people were too 
lazy to do it correctly before. Remember the content has moved from 
developers to community.

Both your and magnus's arguments boil down to two things:

1. We want the diff correct. This has some legitimacy and I can see 
where this would be annoying. I proposed a couple of solution. None are 
acceptable to Magnus so I don't know what to do. Frankly I don't think 
we are missing a lot, especially if we use svn mv. We can always 
manually diff if we absolutely have to.

2. We designed it so we don't have to be consistent. This with respect 
is easily the dumbest thing being argued in this thread. Considering we 
are only arguing two things, and not actually arguing the merit of the 
patch itself that says a lot. If I were to submit something inconsistent 
into -hackers I would be beaten until I was bloody and begging for 
someone to teach me the ways of correctness.

> 
>> Your suggest opens the doors for a horrible pile of spaghetti over the
>> years. Consider... we have developers in contributors, sponsors in
>> donors, resources in links on and on and on.
>>
>> It is a mistake, flat out.
>>
>>> For example, hosting pages are served by professional.php. Techdocs are
>>> served by communitydocs.php. That is *intended*.
>>>
>> If it is intended it is a mistake. Techdocs should be served by techdocs
>> and now my point above is made.
> 
> why ?

Please read the thread. I have made it very clear. This is a management 
nightmare in the making and frankly the fact that I know the above now, 
I am going to submit a patch to fix that laziness too.

> I don't see the point in renaming files for no real purpose (other
> than "it looks nicer to the five odd people browsing the source" and
> having to deal with annoyances as a result)
> 

I propose we rename all files based on the timestamp plus milliseconds 
of their first commit. The dispatcher is there, it can take care of 
everything. Barring that, because we want svn diff capability... we 
leave all old files as they are but all new files must be named by the 
EPOCH.

Joshua D. Drake



> 
> Stefan
> 
> ---------------------------(end of broadcast)---------------------------
> TIP 9: In versions below 8.0, the planner will ignore your desire to
>        choose an index scan if your joining column's datatypes do not
>        match
>