Re: Old libgmp2 problems (was Re: Why did you break my package!)
On Sun, 7 Nov 1999, Ben Collins wrote:
> Ok, first of all, I would like to point out that you have reduced this to
> a really ugly situation. Your initial email to me was condescending and
> completely out of line. That is the reason my replies have been somewhat
> cold in nature. That does not excuse your attitude, vulgar language, and
> downright shitty displacement.
The whole thrust of my communication with you was that communicating with
the package maintainer is a must if the confusion around this particular
bug is to be avoided in the future.
The frustration you say in "vulgar" language was generated by your false
assumptions and your refusal to admit that it would have been a better
deal if you had talked with me before doing an NMU. I have never objected
to the technical content of the NMU.
>
> Since I am Cc'ing this to debian-sparc, I will not include any of the
> previous correspondence between us. I do want to include that list, since
> you will here from others who were affected by this problem, not once, but
> twice, leaving their systems unaccessible.
>
> To start from the begining...
>
> You uploaded a new libgmp2 2.0.2-1 that at first would not even work
> because of the silly soname problems, thus it was never compiled for
> sparc. Then you uploaded a 2.0.2-2 that fixed this. Now the problem arose
> after it was compiled for sparc. Soon people started to restart their sshd
> servers. Upon trying to start them back up, they noticed it hung trying to
> generate part of the key. This was attributed to libgmp2.
>
> What followed were serveral posts. I believe Steve Dunham was involved,
> and several others. A patch was produced to fix the problem (I believe it
> came from Dave Miller). I was told that a bug was filed on libgmp2 with
> this patch included. A binary _only_ NMU was made for sparc to get it
> working again, until you, the maintainer, could include the patch into the
> main debian source.
To my knowledge no bug report was ever submitted. It certainly isn't in
the BTS at this time, and no one contacted me to suggest that there was a
problem.
>
> Several weeks later, you uploaded a 2.0.2-3, that did NOT include this
> patch. People started emailing this list _again_ complaining and wondering
It did not include the patch because I had no knowledge of its existance.
This is why I have been suggesting (strongly) that contacting me IS
important to the process.
> why we had such a messed up libgmp2 for sparc (rather irrate that they
> couldn't access their remote systems anymore). So, I made an NMU this time
> (I did not make the first), using the patch and made sure to do a diff
> upload so that it would be more noticable. I did not check the BTS, since
> I was under the impression that it was already there. I also did not
> contact you because I have way too much to do as it is.
So you just violated all principles of maintainership and took a shot in
the dark. If someone else hadn't submitted a bug against the NMU package I
would never have realized it was there, and most probably made another
release over top of it as well. You have just outlined two major mistakes
that only got caught because of fortuitous circumstances. Stop blaming me
for you unwillingness to follow normal established proceedure!
>
> Now the fact that you did not notice this for some time, tells me that a)
> it wouldn't have made a difference to email you then ask for a new
> version, b) we would have been left all this time without a working
> libgmp2 for sparc, and an NMU would have been needed anyway.
>
If that is what it tells you, then you don't know me at all, and just want
to be pissy. Well, I don't buy it. Anyone who has contacted me about bugs
in my packages has always been paid attention to. Your suggestions are
simply bogus crap.
> Now note something:
>
> [bmc@blimpo(10:25pm)-~]%dpkg -l libgmp2
> Desired=Unknown/Install/Remove/Purge
> | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed
> |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err:
> uppercase=bad)
> ||/ Name Version Description
> +++-===============-==============-============================================
> ii libgmp2 2.0.2-3.1 Multiprecision arithmetic library
> [bmc@blimpo(10:25pm)-~]%l /usr/lib/libgmp*
> lrwxrwxrwx 1 root root 15 Aug 14 19:26 /usr/lib/libgmp.so.2 -> libgmp.so.2.0.2
> -rw-r--r-- 1 root root 125236 Aug 14 19:25 /usr/lib/libgmp.so.2.0.2
> lrwxrwxrwx 1 root root 15 Aug 14 19:26 /usr/lib/libgmp2.so.2 -> libgmp.so.2.0.2
>
> This is on my sparc, using the package I NMU'd. The links are fine
> (although I think the nameing of the library is really fscked up, but
> that's beside the point). The bug report you refer to is not of the sparc
> arch upload I made. Now, after looking at the bug report, and your rules
> file, there is no way this was because of the current package. My only
> conclusion is a) ldconfig, or b) bad upgrade, from an older package.
>
As I told you earlier, the bug report is bogus, as the reporter failed to
install the libgmp2-dev package which supplies the links he was "missing".
The only purpose served by this report was to let me know that there was
an NMU out there.
> One of these would be a system problem, the other would be _your_ problem.
> None of them are from my NMU.
>
Agreeed.
> My final word on this, take your shit talking crap someplace else. I have
> an arch to get release ready, and you are taking up my time.
>
Same to you Ben ;-)
Luck,
Dwarf
--
_-_-_-_-_- Author of "The Debian Linux User's Guide" _-_-_-_-_-_-
aka Dale Scheetz Phone: 1 (850) 656-9769
Flexible Software 11000 McCrackin Road
e-mail: dwarf@polaris.net Tallahassee, FL 32308
_-_-_-_-_-_- See www.linuxpress.com for more details _-_-_-_-_-_-_-
Reply to: