phpgroupware-developers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: SV: [phpGroupWare-developers] Nested db-objects


From: Dave Hall
Subject: Re: SV: [phpGroupWare-developers] Nested db-objects
Date: Mon, 26 May 2008 11:13:01 +0000

On Mon, 2008-05-26 at 10:52 +0200, Maât wrote:

> We need to move, we need to produce an alpha and then betas and 
> hopefully a stable release

Agreed.  This is why I am reluctant to entertain the notion of delaying
the release any longer than is absolutely necessary to get a stable,
secure and maintainable, release in the hands of our broad user base.

> We need to attract new contributors, and for that we also need to set up 
> a fun-to-code atmosphere

I am all for fun, but at the same time there are some realities which
nee to be faced.  Maintaining old, hack ridden, buggy, insecure code
isn't fun - I have spent 100s of unpaid hours doing just that.

To get a security release out the door costs around 2 days work.  

Tracing stupid bugs in poorly written code can waste hours.

If people don't like the requirements of being an official phpgw app, no
one is forcing them to have that status.  I think we are well within our
rights to mandate a minimum level of quality within the project.  I also
think that those who put the many hours into maintaining the broad code
base are more qualified to propose such standards.

> Could you stop just five minutes this childish game and try to find a 
> win-win solution for everybody ?

I think you are looking for common ground where there is none.

> What's wrong with the idea of accepting TEMPORARILY sigurd patch for 
> branching and alpha tag then working all together to have this feature 
> cleaned for first  beta tag ?

The clone behaviour is not negotiable.  If you need an independent db
object, then use the clone, that is what it is for.  It will also clean
up after itself.  If you are trying to use transactions and use multiple
db objects, then you must be putting business logic into your storage
layer and so you are doing something wrong.  Easy to use, hard to abuse,
that should be the mantra, and the current db class does exactly that.

> What's wrong with the idea of letting Sigurd give an estimation of the 
> time he would need instead of deciding for him ?

I don't think we should be allowing one person to dictate the release
schedule.  When announced, the community supported the release schedule.
Given Sigurd's track record we will be having the same debate over issue
after issue which he doesn't like in trunk and having to go through this
same process again and again.  It is a pointless exercise.

> Dave : Could you try to soften a little bit your team management 
> communication and give people a chance to prove their good will ?

I think you need to look at the last few months.  Everytime Sigurd is
unhappy with something within the ReSight project he comes on the list
to try to get his own way.  This usually involves distorting the facts.
When that doesn't work he may resort to threats.

> How can you all hope to see phpgw gain new contributors and new audience 
> if nobody want to make an effort to improve all that need to be improved ?
> 
> We can see things change and improve... but the first requirement to see 
> that happen is to want it !
> 
> sorry regards,

I appreciate what you are trying to do here Pascal.  I have tried to
resolve my differences with Sigurd in various forums, including spending
over 2weeks meeting in Norway last month, and it seems to go nowhere,  I
am sorry that the project has to endure this constant bad blood on the
lists, but I am at a loss to how else this can be handled.  Sigurd is
unwilling to accept that there is problems with his code and that a hack
can't be fixed by another hack.  The tree is full of "temporary fixes",
TODOs and FIXMEs.  I think others have expressed a desire to move
forward and build a quality groupware platform.

A release always bring in new developers, but few stick around.  Over
the years I have witnessed the people coming to the project, checking it
out, then realising the state of the code base and justifiably running a
mile.  We need to take the issue of code quality and standards seriously
if we are to build a sustainable project.

Cheers

Dave





reply via email to

[Prev in Thread] Current Thread [Next in Thread]