Re: Repository 'openSUSE-10.3-Updates' is invalid

houghi wrote:
Kevin Miller wrote:
Sure, but then how do you get notified of updates.
By email once they are done.
Got a link?

A link to what? The email is local and not very interesting to read.
It is the standard output from a cron command.

I was thinking email notifications of updates/patches. So what cron job
are you running to let you know there are patches available for download
and installation? Or is it just the automatic update kicking in?


Some repos are
turned off on my system, such as the Nvidia one. Can't turn them all
off though and still run a reasonably patched system. Or am I missing
something?
Perhaps the fact that you do not do the updates automagicaly.
Guess you could say I'm running semi-automatically. The updater in the
kicker lets me know when there are updates but I install them manually.

I would not call that semi-automatic.

All I gotta do is click on 'install'. Pretty automatic after that.


But even if I did automatic updates, the system would have to query the
repos to realize there were updates to be had. If the repos are turned
off, automatic update would never kick in. Or do I misunderstand you?

The automatic launches and looks for the updates. Just as would, I
imagine, kicker. The advantage is that I do not care wether it takes 5
seconds or 5 hours.

I suppose if I left my system on 24/7 it wouldn't matter. They could
download at midnight. I don't leave it on all the time though. If I'm
using it there are times I don't want to compete for the bandwidth.
When I want to update it's nice to have an efficient mechanism for doing
so.


I don't know whether it's a problem or not. But it is an annoyance and
I know it could be better.
Then by all means give the solution to the developers and/or give us the
bugnumber that you opend and where you are working with the developers
on a solution.
That's a tired old canard. It implies that anybody using open source is
qualified, interested, or has the time/inclination to be a hacker.

No, _YOU_ said that you know that it is possible. That implies that
_YOU_ know how to solve it.

That's a non-sequitur houghi. Knowing something is possible in no way
implies one knows how it's done. I know planes fly but you wouldn't
want to ride in one I designed! I say I know it can be better because
apt does it better. SUSE 10.0 did it better. As Chris has pointed out,
8.something or another did it better. The installation, update and
quality of the .rpms has gone down hill in several cases. That implies
it was better in the past. It doesn't imply *I* know what they were
doing differently.


I mentioned two approaches: apt, which is a proven technology, and
something based on a DNS type approach. I don't know if anybody is
exploring the latter but it does wonders with RBLs, SPF, etc. Not to
mention, um, DNS. <g>

Great and where can I read about your input on bugzilla? Even mentioning
(perhaps a bit more in detail) on how you imagine this to be working.
The APT stuff I get and I am sure that nobody ever thought of it to do
it that way. I just have no idea what you might mean with the DNS stuff.

Research SPF - it uses text entries in DNS to list authorized email
servers. Something similar could possibly be done by having an
equivalent to an A record, or TXT record. The system could do a 'zone
transfer' to pull down the whole list of software and do a diff against
a locally stored list perhaps, or a line by line query maybe. That is,
my system has a list of installed packages and it just does a lookup of
the current revision. If the master file has a serial number like DNS
it could look at that and see whether or not any new updates have even
been released. In an instant one could compare the revision level of
the repos to what I've already downloaded.

I just know on my Debian system I enter 'apt-get update' an 15 to 20
seconds later the list is updated. Refreshing the repos on 10.3 takes
several minutes. 10.0 was slower than Debian, but still leaps and
bounds ahead of 10.3.

And no, you do not have to be a hacker or even able to write code to
give feedback on Bugzilla.

As a matter of fact, last night I reported the timidity problem on
Bugzilla. And while you don't have to be a hacker or coder to do so, it
was a convoluted process. Had to create an ID. OK, no biggie. Entered
a username, and password, then logged in. Got to some bugzilla page
that said I had to validate my new account. EVERY other account I've
ever created via the web emailed me a 'reply to this to activate your
acount' message. But not the Novell/SUSE bugzilla system. I had to go
track down the right web page, and request that the system email me an
invitation to activate the account. That is totally convoluted. Why
don't people participate more? Maybe because of the roadblocks they
encounter. I can't blame them for saying 'screw it' half way through.


This 'good enough for the girls we go with' attitude puts us not far
from the Microsoft camp IMHO. Open source is wide open for innovation
but too often we settle for much less...
What 'we' settle for is what 'we' want. You say you know it could be
better. That means you have the solution. Why do you not share that
solution with the development team.
I've never met anyone one the development team.

I did, but that is irrelevant. Bugzilla is the tool.

In some cases. Timidity for example. But what about my problem with
audacity segfaulting if alsa is running? Nobody else seems to have that
trouble, and there's nothing in the logs. So I should enter a bug that
the developers can't reproduce and that I can't document?


Nor do I have the development background to seriously contribute to
it.

Sure you have. You use openSUSE and that is qualification enough. If
only developers would ask for things or even test the software, you
would get something completely different.

I presume that the development teams participates in the 'greater SUSE
community' to some degree, i.e., they keep their ears to the ground to
get a sense of how things are working out. But maybe they don't.

Sure they do. They look at what that community brings into Bugzilla. If
it isn't in Bugzalla it isn't a problem. Then there are many, many
openSUSE mailinglists that they will read.

Bugzilla is hardly the haunt of normal users. It's a geek tool and
if you're basing the experience of the average joe on what's in it,
you're only scratching the tip of the iceberg. As for mailing lists,
"Many many" is the operative word, and spare time is a rare commodity.
Here's a quote from http://en.opensuse.org/Communicate: "Please note
that the support mailing list is very active and can send you over 100
emails per day!" Sorry. I'm just not up to wading through 100 more
emails a day.


Maybe they have their own little list server or forums and don't care
what the experience is like for us mere mortals. That would explain a
lot.

That is just stupid and almost trolling and you know it. If you, as mere
mortal, would just step down your ivory tower and just look for a few
moments on the openSUSE mailinglists what is going on, join an IRC
meeting (perhaps just lurking) and what not, you would know that you are
absolutely and completely misinformed.

Ivory tower? Hardly. More like a broke hitchhiker on the information
superhighway.

I've always liked SUSE. For the most part it's what I want. There are
enough things goofy in 10.3 however that make me inclined to try other
distros. Maybe they do it better.

Great. That is what choice is about

So far I'm not wild about Debian (except for apt - it works a treat).
But maybe Ubuntu will work, I dunno. They seem to be listening to
consumer feedback at any rate, judging from their widespread adoption.

And again I am asking you what customer feedback have you given to
openSUSE that they have ignored? And no, telling the people here is not
giving customer feedback.

Maybe if they made it easier to give feedback more folks would. Most of
us want to use our computers, not build them.

Bugzilla has other advantages. You can see how far they are with
solving the problems and enhancements.

Yes, that is an advantage of using it.


If you say that you do not like what they are doing, that is fine with
me. However if you say (or imply) that they do nothing then I know that
you are lying and can't even bother to do do any investigation yourself.
No problem with that either, except for the part where you insult hard
working people.

The former. I'm saying that they did better in the past and I don't
like what they've done lately. I stayed on 10.0 until it was out of
support because of the pain they introduced with red carpet stuff. I
know they work hard - I never said they do nothing. And I don't lie. I
may make mistakes. But that's different.


Have you even tried looking up what the development team is doing? And
no, I am not going to point you in the right direction. I want to see
wether you are seriously interested or just do the easy thing and blame
everything on other people and keep telling that the grapes are to sour.

My computer is just a tool. When I go out in my woodshop, I expect my
tools to perform a particular task. I don't re-engineer my tablesaw or
router, or whatever. In the past I've bought cheap tools. I don't
hesitate to say they were junk. I just replace them with better tools
the next time. I expect a certain level of quality in the software I
use. If it's not there I'll say so. But it's just software. I'm not
going to make it my life's crusade to fix it. My hats of to those that
do, but I have other interests...

....Kevin
--
Kevin Miller
http://www.alaska.net/~atftb
Juneau, Alaska
Registered Linux User No: 307357
.

Relevant Pages

  • RE: From release notes for FC5T3 (web)
    ... I used battle on public lists over MS vs. Linux and ... Unless we were to redesign the installer ourselves, ... I don't what you definition of productive is but this bugzilla was ... What you don't get if you don't read the developers list is ...
    (Fedora)
  • Re: Linux 2.6.21
    ... Mailing lists don't track bugs. ... The _only_ reason why I originally started regression lists was because ... So what's the difference between bugzilla and lkml.org? ... In Bugzilla we are able to see how bad our bug handling is... ...
    (Linux-Kernel)
  • Re: 2.6.25-rc2: Reported regressions from 2.6.24
    ... BTW it would be nice to receive CC in the bugzilla, ... directly not to put them in the CC lists in there, so I'm a bit cautious ... ... even though the bugzillas probably have the correct addressess ...
    (Linux-Kernel)
  • Re: [SLE] Bugreports?
    ... The advantage on the other side is that our bugzilla ... >> ensures that our developers are spending more time with reports that are ... > It would be on the public relations interest to give feedback to feedback, ...
    (SuSE)
  • Re: Reporting bugs. (Re: FC6 through the rear view mirror)
    ... suggestion made by the Fedora list community accepted by Fedora ... If you don't monitor bugzilla and you don't monitor the developers list, ... The developers have mostly left the user lists so as not to ...
    (Fedora)