List:MySQL++« Previous MessageNext Message »
From:Warren Young Date:January 24 2006 12:13am
Subject:Re: Borland C++ Builder Interest
View as plain text  
Frederic Laruelle wrote:
> 
> Could you provide a little more details on the expected workload?

First, obviously, you have to fix all the problems that you reported in 
your other email.  I didn't look at it carefully, but it all looks to be 
fairly trivial stuff.  The hardest part of making these changes is doing 
so in such a way that you don't break other platforms, and that 
maintains the current code style.  Read the HACKERS file at the top of 
the MySQL++ distribution, especially the "Adding Support for a Different 
Compiler" section.  But don't take it too literally because it is based 
on the current build scheme.  Instead of "makemake", think "Bakefile".

After that, you should port your changes to the v2.1 branch.  This 
should be trivial, as the Bakefile changes are the only thing on that 
branch right now.  You should be able to make a patch file with diff and 
apply it directly.

Finally, you need to start playing with Bakefile to get it to create the 
necessary project files automatically.

> How many releases do you have per year (on average)?

Since v2.0 settled down back in September, we've only made one release. 
  That was about two months ago.  Releases tend to come in spurts, 
though.  I expect several soon after v2.1, just because of the way 
people are.

So depending on your perspective, it's one release every few months, or 
one release about every two weeks.

> What is your best guesstimate on the effort needed (in man hours) to
> get it to work today?

The first two items above you should probably be able to do in an hour 
or two.  The last one could be easy, or it could be a major time sink. 
You won't know until you try.

If you don't yet understand Subversion or Bakefile, budget some time for 
learning those tools.

> I think my threshold to consider your offer lies around a couple hours
> per month maximum...

I hope you can put more time in at the start than later on.  Once the 
port is working, you should only need to put in time to answer 
Borland-specific questions on the mailing list, and to make patches to 
fix up anything that breaks Borland support in the future.  By far the 
biggest part of the work is up front.
Thread
Borland C++ Builder InterestFrederic Laruelle21 Jan
  • Re: Borland C++ Builder InterestRichard Ulrich23 Jan
  • Re: Borland C++ Builder InterestWarren Young23 Jan
Re: Borland C++ Builder InterestWarren Young24 Jan