regarding footprint-mismatch error message
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate. -- Fredrik Rinnestam
Fredrik Rinnestam wrote:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
I agree, it wouldn't be the first time I had to rebuild a port just because I forgot -uf. Glenn
On Sat, Nov 25, 2006 at 11:29:10PM +0100, RedShift wrote:
Fredrik Rinnestam wrote:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
I agree, it wouldn't be the first time I had to rebuild a port just because I forgot -uf.
[ though some are just mocking around without bothering to think about *why* users often confused by software and don't bother to fix the cause of confuse or at least give some advice, I will do lost part, an advice and uncover the cause ] You don't have to rebuild package if footprint mismatched, built .pkg.tar.gz file cooked fine, but not installed or upgraded if -i or -u options given respectively. The cause of confusion is: pkgmk prints nothing but package *building* error: =======> ERROR: Footprint mismatch found: NEW drwxrwxrwt root/root var/spool/mail/ NEW drwxrwxrwt root/root var/tmp/ =======> ERROR: Building 'abc#1.2.3.pkg.tar.gz' failed. Not install/upgrade failed, but building. And that's is why it's not obvious that package really built. [ Now it's not so funny, right? ] Thus, all you have to do is pkgadd manually. The second thing you could do is to look into /etc/pkgmk.conf for the ignore flag, which is is not an option in some cases (if you really give attention to mismatched packages, and want to force installation/upgrade of mismatched packages manually).
Glenn
Good day, -- Anton (irc: bd2)
Fredrik Rinnestam wrote:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
Use the pkgutils in contrib. # Han -- http://www.xs4all.nl/~hanb/software/crux/
On Sun, 2006-11-26 at 13:29 +0100, Han Boetes wrote:
Fredrik Rinnestam wrote:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
Use the pkgutils in contrib.
That's not supposed to be there [1]. [1] http://lists.crux.nu/pipermail/crux-contrib/2006-November/000764.html
Mark Rosenstand [2006-11-26 18:13]:
On Sun, 2006-11-26 at 13:29 +0100, Han Boetes wrote:
Fredrik Rinnestam wrote:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
Use the pkgutils in contrib.
That's not supposed to be there [1].
[1] http://lists.crux.nu/pipermail/crux-contrib/2006-November/000764.html
I have a feeling he doesn't care that much: https://crux.nu/bugs/?do=details&id=138 Regards, Tilman -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
On Sun, 2006-11-26 at 18:46 +0100, Tilman Sauerbeck wrote:
Mark Rosenstand [2006-11-26 18:13]:
On Sun, 2006-11-26 at 13:29 +0100, Han Boetes wrote:
Fredrik Rinnestam wrote:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
Use the pkgutils in contrib.
That's not supposed to be there [1].
[1] http://lists.crux.nu/pipermail/crux-contrib/2006-November/000764.html
I have a feeling he doesn't care that much: https://crux.nu/bugs/?do=details&id=138
... about anything: http://lists.crux.nu/pipermail/crux-contrib/2006-November/000756.html http://lists.crux.nu/pipermail/crux-contrib/2006-November/000759.html http://lists.crux.nu/pipermail/crux-contrib/2006-November/000761.html
Use the pkgutils in contrib.
That's not supposed to be there [1].
[1] http://lists.crux.nu/pipermail/crux-contrib/2006-November/000764.html
I have a feeling he doesn't care that much: https://crux.nu/bugs/?do=details&id=138
... about anything:
http://lists.crux.nu/pipermail/crux-contrib/2006-November/000756.html http://lists.crux.nu/pipermail/crux-contrib/2006-November/000759.html http://lists.crux.nu/pipermail/crux-contrib/2006-November/000761.html
Han. Since contrib does have standards and rules, ignoring this (pretty serious) dispute will not make it go away - only create irritation. Please join the discussion and atleast make your case for these dupes and maybe it can be resolved swiftly. -- Fredrik Rinnestam
Fredrik Rinnestam wrote:
Since contrib does have standards and rules, ignoring this (pretty serious) dispute will not make it go away - only create irritation. Please join the discussion and atleast make your case for these dupes and maybe it can be resolved swiftly.
Which discussion? I keep providing a solution and they keep getting ignored. That's not a discussion. That's a one way direction of telling people to RTFM and shut up while there is a technical solution. A technical solution which is staring in your face, yet you simply join the guys that prefer to tell RTFM. Odd... # Han -- http://www.xs4all.nl/~hanb/software/crux/
Fredrik Rinnestam wrote:
But to create a dupe in contrib as a way to merge your ideas with mainline is not in anyones interest, IMO.
It's not a dupe. And I want to provide this port in contrib because so many want to use it, so they won't have to install my repo to get it. # Han -- http://www.xs4all.nl/~hanb/software/crux/
On Thu, Nov 30, 2006 at 09:03:19PM +0100, Han Boetes wrote:
Fredrik Rinnestam wrote:
But to create a dupe in contrib as a way to merge your ideas with mainline is not in anyones interest, IMO.
It's not a dupe.
I see the port as derivative work, however the name is what causes the problem in my opinion. Maybe it should be renamed to pkgutils-han to emphasize it is different from the original.
And I want to provide this port in contrib because so many want to use it, so they won't have to install my repo to get it.
Not like it is difficult to get ports from other repos, but I think that, for simplicity, it should be made available from contrib.
Jesse Kokkarinen wrote:
Han Boetes wrote:
Fredrik Rinnestam wrote:
But to create a dupe in contrib as a way to merge your ideas with mainline is not in anyones interest, IMO.
It's not a dupe.
I see the port as derivative work, however the name is what causes the problem in my opinion. Maybe it should be renamed to pkgutils-han to emphasize it is different from the original.
I'll rename the directory. # Han -- http://www.xs4all.nl/~hanb/software/crux/
Han Boetes wrote:
Jesse Kokkarinen wrote:
Han Boetes wrote:
Fredrik Rinnestam wrote:
But to create a dupe in contrib as a way to merge your ideas with mainline is not in anyones interest, IMO.
It's not a dupe.
I see the port as derivative work, however the name is what causes the problem in my opinion. Maybe it should be renamed to pkgutils-han to emphasize it is different from the original.
I'll rename the directory.
Yay, a peaceful resolution :) -- Jay Dolan jdolan.dyndns.org A: Because it's annoying to read. Q: Why is top-posting bad? A: Top posting. Q: What's the most annoying thing about usenet?
Fredrik Rinnestam [2006-11-25 23:18]:
pkgmk should print out a warning instead of an error on footprint mismatch. Most "errors" are trivial and perfectly normal. I feel that a warning would be more appropriate.
I think this is a good idea. What's more important though is the exit code that's used for this condition. If it stays at non-zero, calling apps will treat it as a fatal error. If you install a bunch of ports at once via prt-get, it's rather important, since it will decide whether prt-get stops to tell you about the footprint mismatch or whether it will just go on... Mmh. Regards, Tilman -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
participants (8)
-
Anton Vorontsov -
Fredrik Rinnestam -
Han Boetes -
Jay Dolan -
Jesse Kokkarinen -
Mark Rosenstand -
RedShift -
Tilman Sauerbeck