Re: [Mingw-users] Environment variables

[Jared M. <abs...@gm...>]

> Date: Fri, 04 Jan 2013 19:18:05 +0000
> From: Keith Marshall <kei...@nt...>
> Subject: Re: [Mingw-users] Environment variables
> To: min...@li...
> Message-ID: <50E...@nt...>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> On 04/01/13 04:23, Jared Maddox wrote:
>> I really don't want to deal with writing gnu configure code if I can
>> avoid it. As for this program, if I had to learn that particular
>> system in order to port it then I WOULDN'T have started this project.
>
> This attitude simply defies logic.  So, you will commit WEEKS to the
> development of crufty, fragile, difficult-to-maintain makefile logic,
> (which surely will NOT cover all the bases), to save the few HOURS it
> would take to learn the rudiments of autoconf -- rudiments which, once
> learned, will likely allow you to complete the entire task in a matter
> of DAYS, and the result will be orders of magnitude more robust.
>

Weeks? Days? I successfully compiled DGD BEFORE I sent the original
post in this thread, and did it in maybe an hour. It would have taken
me longer if I hadn't done it before (a few months ago, I believe),
but that's equally true for a configure-based build approach.

Before you sent this email that I'm replying to, on the 4th, I spent a
few hours congealing the suggestions in this thread and reading the
documentation on GNU's make to create a revision of my original patch.
That compiled too, by the way.

I'd say that I've spent maybe five or six hours on this. Further,
since the project doesn't already use autoconf, I would have to rework
the entire build system to do what you advise. Do you think that I'm
getting paid for this? This is for me, not for someone with a
checkbook in hand. I've made my port simple and easy to both extend to
other platforms, extend to other sub-platforms, and override to force
the makefile to work. As long as someone can run the original
makefile, they should be able to run my revision of it.

On top of all of that, this is not the primary project of it's type
for it's primary maintainer. I have no intention of supporting this
myself, so I want him to accept this patch series into HIS version.
Replacing the build system that he already knows with a new one
strikes me as likely to make that difficult.

If you think that it's reasonable to expect my to use autoconf to
compile DGD in this situation, then you and I are using different
interpretations of "reasonable".

>> Various *nix platforms (Darwin, Linux, the big three BSDs, and Sun)
>> are already supported, and I'm not trying to add support for them
>> anyways,  ...
>
> If this were true, then you would not be asking the questions you have.
>

If it were false, I wouldn't have been able to compile the program
before I made the starting post of this thread. In fact, I possibly
wouldn't have posted the original message in the first place, since I
could have still been working out how everything's INTENDED to work.
The reason why I was asking these questions is because I was running
the initial test without knowing what was actually available to test.
I'd actually looked for that info before, but I'd never found it.



At any rate, I've already gotten the info that I needed, so thank you all.