Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 19 Apr 2001 13:53:51 -0500
From:      Glenn Johnson <gjohnson@srrc.ars.usda.gov>
To:        ports@freebsd.org
Subject:   Re: Handling tarballs with no version
Message-ID:  <20010419135351.A60770@node7.cluster.srrc.usda.gov>
In-Reply-To: <20010419205924.G1527@ringworld.oblivion.bg>; from roam@orbitel.bg on Thu, Apr 19, 2001 at 08:59:24PM %2B0300
References:  <20010419125305.A48871@node7.cluster.srrc.usda.gov> <20010419205924.G1527@ringworld.oblivion.bg>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Apr 19, 2001 at 08:59:24PM +0300, Peter Pentchev wrote:

> On Thu, Apr 19, 2001 at 12:53:05PM -0500, Glenn Johnson wrote:
>
> > I have a port that I maintain that does not use a version number in
> > the tar file.  If I update the port and bump either the PORTVERSION
> > or PORTREVISION variable, then users who have the port installed
> > will get a CHECKSUM mismatch when doing an upgrade.  How can I
> > force the update of the distfile when one installs the port?  I was
> > thinking that running the distclean target prior to fetch would work
> > but what is the cleanest way to implement that?
>
> Absolutely bump the PORTREVISION; then, leave it up to the users
> to make sure they have the latest.  Yes, this will break automated
> builds, but such is life :(
>
> Invoking distclean before fetch would be a *bad* idea - it would force
> a re-fetch every time somebody tries to build this port, or some port
> dependent on it.

Well, what I had in mind was to only run the distclean target if the tar
file was out of date.  I thought maybe an md5 comparison could be done
and if that fails then run the distclean target to purge the old tar
file, followed by a fetch of the new tar file.

The problem I am envisioning with the current arrangement is that the
user could set NO_CHECKSUM=yes after it fails the first time.  The port
will then be built using the old distfile but labeled as whatever the
current PORTREVISION is set to.  There would be a mismatch here.  The
other possibility of course is that the user gets a failure someplace
else, likely a patch, but this at least avoids the problem of the
PORTREVISION not matching reality.

-- 
Glenn Johnson
USDA, ARS, SRRC			 Phone: (504) 286-4252
New Orleans, LA 70124		e-mail: gjohnson@srrc.ars.usda.gov

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-ports" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010419135351.A60770>