Michael A. Peters mpeters@mac.com wrote:
I've been lurking in this thread - I'm NOT to get into any arguments - but I do think there could be a better way of _package_ RFE's than the wiki or bugzilla.
Maybe...
A simple page where a user can input the name of the desired package, a url to the homepage, a brief description, and select a category from a drop down menu.
This would be something that extras maintainers looking to contribute could go through.
I'd wagger that extras maintainers have their hands more than full with the packages they do maintain, and would love to have time left over for the /other/ packages they care for...
I.e., if you want a package included/maintained, your best (only?) bet is doing it yourself.
[No, I have no standing to talk officially here. Just from observations elsewhere...]
Horst von Brand wrote:
Michael A. Peters mpeters@mac.com wrote:
What should be available are libraries that are as current as possible.
As an example: FC4 provides /usr/lib/libdb.so.2
Program requires a later version.
libdb.so.3 is needed by gmc-4.5.51-16.i386
Jim
How about a dedicated server to recieve crash reports automatically.
On 6/22/05, Jim Cornette fct-cornette@insight.rr.com wrote:
Horst von Brand wrote:
Michael A. Peters mpeters@mac.com wrote:
What should be available are libraries that are as current as possible.
As an example: FC4 provides /usr/lib/libdb.so.2
Program requires a later version.
libdb.so.3 is needed by gmc-4.5.51-16.i386
Jim
-- Never put off till run-time what you can do at compile-time. -- D. Gries
-- fedora-test-list mailing list fedora-test-list@redhat.com To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-test-list
On Wed, 2005-06-22 at 09:30 +0800, Joel Juliano wrote:
How about a dedicated server to recieve crash reports automatically.
<sarcasm> how about infinite bandwidth and server resources?
and while we're asking for mythical things - how about a unicorn or a pegasus!
Maybe a baby unicorn for Jef. </sarcasm>
A dedicated server to receive crash reports automatically would probably take a lot of abuse and you'd have to sift through the real reports vs the sheer volume of people who just have bad hardware.
-sv
On Tue, 2005-06-21 at 21:36 -0400, seth vidal wrote:
A dedicated server to receive crash reports automatically would probably take a lot of abuse and you'd have to sift through the real reports vs the sheer volume of people who just have bad hardware.
Heh, yeah - I filed several bug reports when RH8 shipped because certain things (xscreensaver) etc had a tendency to crash my machine - all new since symptoms since installing RH8.
Then one day it wouldn't boot. Same ram, video card, hard drive, etc. - only thing I changes was mobo (same chipset) and cpu - and no more crashes ...
On 6/21/05, seth vidal skvidal@phy.duke.edu wrote:
<sarcasm> how about infinite bandwidth and server resources?
and while we're asking for mythical things - how about a unicorn or a pegasus!
Maybe a baby unicorn for Jef.
</sarcasm>
<dead pan> how about http://nadvsh.sourceforge.net/ integration into yum </dead pan>
jef"I won't go till i have oneko I won't go till i have oneko I won't go till i have oneko... in Core"spaleta
<dead pan> how about http://nadvsh.sourceforge.net/ integration into yum </dead pan>
from irc:
yum-nadvsh> you have entered fedora-extras repo... you see abiword.. there are exits to the east and north yum-nadvsh> You have 1200 packages in your pkgSack yum-nadvsh> you have entered fedora-development... dead babies litter the ground yum-nadvsh> there are unresolved dependencies. a grue eats you.
oh the opportunities for hijinks are limitless.
-sv
yes we have so much to show you... seth vidal wrote:
<dead pan> how about http://nadvsh.sourceforge.net/ integration into yum </dead pan>
from irc:
yum-nadvsh> you have entered fedora-extras repo... you see abiword.. there are exits to the east and north yum-nadvsh> You have 1200 packages in your pkgSack yum-nadvsh> you have entered fedora-development... dead babies litter the ground yum-nadvsh> there are unresolved dependencies. a grue eats you.
oh the opportunities for hijinks are limitless.
-sv
On Tue, 21 Jun 2005, Jeff Spaleta wrote:
On 6/21/05, seth vidal skvidal@phy.duke.edu wrote:
<sarcasm> how about infinite bandwidth and server resources?
and while we're asking for mythical things - how about a unicorn or a pegasus!
Maybe a baby unicorn for Jef.
</sarcasm>
<dead pan> how about http://nadvsh.sourceforge.net/ integration into yum </dead pan>
LOL... I once had this idea to create "Apt adventure" with apts Lua-interface where you try to do dist-upgrade from, say, FC1 to FC2. Never got further than the "this would be really sick" thinking stage though :P
- Panu -
<sarcasm>
I think you meant <pissy> :)
A dedicated server to receive crash reports automatically would probably take a lot of abuse and you'd have to sift through the real reports vs the sheer volume of people who just have bad hardware.
You can't mine the data until you have it, and having it is not an obligation to mine it. It's just an opportunity.
There's probably an unbounded amount of work that could go into mining it, but I'd suspect that you could get a pretty good ROI just by going for easy signals-- a rash of crashes on chipset foo on kernel x.y.z, for example. It sounds to me like a pretty decent idea, modulo privacy concerns.
-Ed
On 6/21/05, Edwin Olson eolson@mit.edu wrote:
There's probably an unbounded amount of work that could go into mining it, but I'd suspect that you could get a pretty good ROI just by going for easy signals-- a rash of crashes on chipset foo on kernel x.y.z, for example. It sounds to me like a pretty decent idea, modulo privacy concerns.
well you could go ask upstream gnome about the quality of the data and the amount of mining thats going on with bug-buddy.
-jef
On Tue, Jun 21, 2005 at 07:41:58PM -0400, Jim Cornette wrote:
Horst von Brand wrote:
Michael A. Peters mpeters@mac.com wrote:
What should be available are libraries that are as current as possible.
As an example: FC4 provides /usr/lib/libdb.so.2
Program requires a later version.
libdb.so.3 is needed by gmc-4.5.51-16.i386
This gmc package has a dependency on Berkeley DB 3.x, which hasn't been current since before RHL 8.0 (which is why it's not even in compat-db any more).
Nalin
Nalin Dahyabhai wrote:
On Tue, Jun 21, 2005 at 07:41:58PM -0400, Jim Cornette wrote:
Horst von Brand wrote:
Michael A. Peters mpeters@mac.com wrote:
What should be available are libraries that are as current as possible.
As an example: FC4 provides /usr/lib/libdb.so.2
Program requires a later version.
libdb.so.3 is needed by gmc-4.5.51-16.i386
This gmc package has a dependency on Berkeley DB 3.x, which hasn't been current since before RHL 8.0 (which is why it's not even in compat-db any more).
Nalin
Thanks!
I was surprised that the lib version is one version newer than the one supplied in gnome-libs. Nothing seems to need this older library version either. How did an older version than shipped with RHL 8.0 end up in current gnome-libs?
Jim
rpm -q --whatrequires /usr/lib/libdb.so.2 no package requires /usr/lib/libdb.so.2
rpm -q --whatprovides /usr/lib/libdb.so.2 gnome-libs-1.4.1.2.90-46
Jim Cornette (fct-cornette@insight.rr.com) said:
I was surprised that the lib version is one version newer than the one supplied in gnome-libs. Nothing seems to need this older library version either. How did an older version than shipped with RHL 8.0 end up in current gnome-libs?
That's gnome-libs for GNOME 1, which is really really old. :)
Bill
Bill Nottingham wrote:
Jim Cornette (fct-cornette@insight.rr.com) said:
I was surprised that the lib version is one version newer than the one supplied in gnome-libs. Nothing seems to need this older library version either. How did an older version than shipped with RHL 8.0 end up in current gnome-libs?
That's gnome-libs for GNOME 1, which is really really old. :)
Bill
This is using an FC4T3 to FC4 system. The lib is on my system and gnome-libs-1.4.1.2.90-46 is the only version installed.
Anyway, gmc worked when making a symlink to the below. It was fast, but left extra icons on the desktop. Once started, both gmc and nautilus would respawn if either was killed. Trying gmc with failsafe might be interesting to me.
Jim
ls -la libdb.so.* lrwxrwxrwx 1 root root 11 Jun 7 22:14 libdb.so.2 -> libdb1.so.2 lrwxrwxrwx 1 root root 17 Jun 22 18:15 libdb.so.3 -> /usr/lib/libdb.so (my symlink creation)
rpm -q --whatprovides /usr/lib/libdb1.so.2 gnome-libs-1.4.1.2.90-46
rpm -q --whatprovides /usr/lib/libdb.so db4-devel-4.3.27-4