Subject: Re: making 'make replace' safer
To: <>
From: Jan Danielsson <>
List: tech-pkg
Date: 07/16/2006 23:25:56
This is an OpenPGP/MIME signed message (RFC 2440 and 3156)
Content-Type: text/plain; charset=ISO-8859-15
Content-Transfer-Encoding: quoted-printable

Peter Schuller wrote:
>>> How does that differ from what wip/pkgmanager already does?
>> I looked at pkgmanager, and discounted it because it depends on common=

>> lisp.  I remember also thinking that it didn't do what I want, but now=

>> I forget - I should look again.
> I gotta ask - are a lot of people annoyed by this? I mean from a user's=
> perspective. Will many people refuse to use a tool because it's written=
 in a=20
> language that they for some reason don't want to use?

   "refuse" is such a strong word.

   Speaking in general: I don't *want* to run tools which require me to
install a complete runtime for a new language which I will never use for
anything but that application. Also, one of the strengths of open source
is that you can fix bugs yourself. If a program is written in C (or C++)
I can fix some bugs myself. If it is written in a somewhat more "exotic"
language, I can't.

> I have to admit I am leaning towards possibly re-implementing pkgmanage=
r in=20
> some other language, mostly due to get support for threads, better=20
> portability and better POSIX integration.
> But then the question becomes what language people consider acceptable.=
> it's REALLY REALLY a huge sticking point I *might* be convinced to do i=
t in=20
> C/C++, but I would *so much* prefer not to. The prime candidate at the =
> would be Ruby, which seems fairly accepted over in the FreeBSD camp (wh=
> with portupgrade and such).

   If you write it in C/C++, then people who don't want to have Ruby
installed won't have to install Ruby.

> Will many people have objections to that? Any opinions? Are there any=20
> (relevant to this situation) portability issues with Ruby that I am not=
> of?
> If C is an absolute requirement for a tool to be generally accepted - w=
> the use of boehm-gc kill off that advantage? Opinions?

   I have written several scripts in Python which perform maintenance on
my systems. I chose Python because I know it, and it allows for pretty
rapid development. But truth be told, I don't want all these Python
scripts. It's a *pain* when I Python happens to get updated during a
recursive "make update" operation. The real fun starts when the build
fails for Python.

   I understand if you choose something other than C/C++, since you'd be
doing so for the same reason I chose Python for some of my scripts. But
if the issue where up for voting, I would vote for C/C++.

   (Yes, Some Day, I will convert my Python scripts to a combination of
C/C++ programs and shell scripts).

Kind Regards,
Jan Danielsson
Te audire non possum. Musa sapientum fixa est in aure.

Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"

Version: GnuPG v1.4.4 (MingW32)