-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 James Tremblay wrote:
On Friday 04 May 2007 01:29, Pascal Bleser wrote:
James Tremblay wrote:
ok, but historically since 10.0 we have introduced new technology (because we are an engineering distro, not a packaging distro i.e. Ubuntu) that I would not expose the first timer to, in the sense of learning how to overcome the problems that arose from that new technology i.e. the new gnome and kde interfaces, and the updater issues. What "new GNOME and KDE interfaces" ? KDE 3.5.6 ? The new start menus ?
yes, early on, I had mine crash and want to restart more than once.
Ok, well, GNOME... and both menus are WIP. So what, use KDE, switch to the KDE kicker menu.
If opensuse wants to target the soho/homeuser then lets have an LTS version, lets gold master 10.2 without the zenupdater , with gnomebaker, with wireless drivers and their instructions and support it for 3 years like Ubuntu. Other It's not because Ubuntu does an "LTS" that it is the best nor the most viable model. Canonical has still everything to prove wrt their support. Furthermore, LTS is not free, nor is SLED. Actually, if it's a desktop you're after, SLED is less expensive than Ubuntu LTS. And, at the very least, the SUSE engineers have proven their capabilities in the past wrt support, which is not yet the case for the few full-time employees of Canonical.
you make my point the moment you bring SLED into the conversation. SLE is rock solid , I use it for many things at my school and I love it. Finding answers to problems is easy with support.novell.com and as long as you don't want some package thats not included, all the applications that a person needs to work in the enterprise are all at your finger tips and since an enterprise techie is involved,wrt modifications, adding outside packages isn't really a question, is it?
I don't see what you mean with your last line. You can add packages built for openSUSE 10.1 from 3rd party repositories. But that pretty much defeats the idea behind SLED. [...]
Wireless configuration support and the network manager are still unstable especially when registering to the customer center during installation and lots of people don't rewire their homes, they buy a wireless router and wireless cards and pay someone to set it up or call Dell to walk them through it. Are we ready to compete with that ? No , i don't think so , we would need
Sure but that's not any different from other Linux distributions. openSUSE is pretty good wrt hardware detection and easy of installation IMO.
Blame the wireless card vendors, they're making an awful mess out of chipsets, incompatible devices with the same model number, closed specs, etc...
I'm not saying Ubuntu is any better in this regard.
OK, and I understand your point of view from an end-user perspective. But there's no magic wand in this case. We really have to blame it on the wireless chipset/card vendors.
that 900 system in place, we would need the update.opensuse.org to be set up as the default update channel in yast( even if registering fails) and have it redirect to other mirrors, we would need software.opensuse.org to be the default application source for add-ons (like the multiverse) and a The build service repositories (software.opensuse.org) just contain too much software -- having it split up into many repositories as now is the right model. You wouldn't want experimental xorg73 packages to show up in the same repository as the latest Firefox package.
Then there needs to be some work done to present software in the correct usage categories. i.e. experimental , stable , games , productivity , hobbies, etc. and of course searchable without running of to someone else's web page , i.e. webpin, which isn't an obvious description of what you get there and should be a drop down choice of search.opensuse.org
Again, it's just not as simple as it might look at a first glance. Those "usage categories" you're talking about are actually "stability levels" (uh, bad name but well..). That's really tough to do, and Debian/Ubuntu pretty much fail at it. What you end up with on Debian is having a stone age old "stable" distribution, an outdated "unstable" and a broken "testing". Point is, what is the process for deciding whether a package is "tagged" as "stable", "unstable" or "experimental" (aka Debian's "testing") ? Bug reports ? Success reports ? That's pretty much the only way to measure the stableness of a package. The recent Novell survey shows that 85% of the people who took the survey (count bleeding noobs out, they didn't even know about the survey) don't contribute to openSUSE in any way. What would make those people contribute at least their experience wrt stability of individual packages (or groups of related packages) ? A good, integrated web UI would definitely help (one has to say that Canonical's proprietary Launchpad is a really nice tool for that purpose), but still... Whining won't cut, people have to contribute, even if it's "just" bug reports, success reports, etc... I would agree that a less challenging user interface would be needed for that though, Bugzilla isn't for noobs.
update.opensuse.org makes sense though.
support.opensuse.org forum\KB to even exist. These things are all being
http://suseforums.net/ http://suselinuxsupport.de/ http://opensuse.us/
not located as a click from opensuse.org , remember these people that we are trying to gather have never heard of SuSE. They only know , if at all , opensuse. They want a onestop shop.
Actually two clicks away (if you don't count the welcome page): http://en.opensuse.org/Communicate#Web_Forums But indeed, it's ridiculously well hidden. Note that a useful web forum will never, ever be hosted on opensuse.org as long as the legal situation in the US (and some other countries, most notably Germany (see the Heise linking case)) remains as it is now. If hosted on opensuse.org, Novell is liable for the content, and linking to package repositories like Packman or Guru (or others), that contain packages that might infringe patents in some countries (e.g. USA) is a no-no for Novell, which is quite understandable. And you will certainly agree that not being allowed to link to such packages or repositories would seriously limit the usefulness of a web forum ;)
discussed and worked on but we can't put the cart before the horse and start extolling the virtues of our distro to the homeuser, when in reality we the enthusiast are the only ones that can make it all work with a little guidance from friends in obscure places like the irc forums. Some one needs to start IRC "forums", "obscure" ?
yes, obscure to the masses , take a group of windows power users with no linux knowledge hand them a laptop and a copy of opensuse , tell them they get to keep the laptop if they can recreate their home computers applications and functionality , without any help from us. Bet they don't even think about an irc forum, and if they do, get frustrated trying to get the system to get them there. We are trying to sell computers to the single mom who's kids need a computer but the one with the Microsoft tax is 200.00 dollars more, as well as the power user.
Personally, I think that "windows" and "power users" is an oxymoron. But yeah, that's not the point ;) I would also like to see a desktop shortcut that is installed by default on KDE and GNOME which would start an IRC client on irc://freenode.net/#suse (and a shell script that wraps irssi for the console, just in case you can't start the X server any more) Name it "online help" and obscurity is gone ;)
consolidating the places we get help from on an intuitive location like "support.opensuse.org" because the masses are used to the host domain having these kinds of pages. our front page is difficult to navigate because the things homeusers are looking to find aren't what we look for when we go
File bugs for those suggestions, give feedback to the web designers of the site when they make a proposal on the list (but not hidden in a large post on this list ;))
start a mailing list "opensuse-website" I'll join
It's opensuse-wiki@opensuse.org
there, even software.opensuse.org has changed recently and it is obscure even to me how to search for packages in the different repos. I was looking to see
True, it's still lacking a lot on the end-user interface, but efforts have been concentrated on the packager side of things up to now (which is even more important in the first place). Hopefully there will be some sort of interface for users soon. As for finding packages, Benjamin Weber's webpin is currently the best option.
is that mature enough to sell to the single mom?
Probably not, you'd have to understand what "package" and "package repositories" are. But tell me, what would be "mature enough to sell to the single mom" ? Googling for a crack for commercial software, get a copy full of trojans and virii, scan it with 1 or 2 cracked antivirus suites, run the .exe installer, etc... ? Ultimately, the only viable solution is probably a web interface where you can browse software packages by category, by tags, have a search function, where each item shows the website, screenshots, a more or less noob-safe description and a single-click option to add the containing repository (if needed) + install the package with its dependencies. But that would be a *lot* of work. And with more than 85% of openSUSE users not getting involved into the effort...
if we had a 10.2 package for NBD and had to go to google to find it on opensuse.org, this is unacceptable to the novice. So that novice goes looking for a click and run distro like Ubuntu.
Ubuntu is no "click and run distro", it's by no means easier to set up and use than openSUSE. The only thing that's better with Ubuntu is the package manager. Idealizing Ubuntu as being perfect and not having any issues is ridiculous.
you confuse installation with adding software , which is probably why Mr. Dell is looking at Ubuntu for his end user PC's he knows that not only techies will be asking for them and he has looked at what offerings are made to that curios windows power user after the install.
That's an extreme simplification of a situation that involves lobbying, pressuring, lots of money, etc... (yes, I mean between Microsoft and Dell, in both directions).
No our structure is still to immature to sell it to the homeuser masses, therefore I agree with Justin and Bruce and Ted, opensuse is the techies version.
Not to say there aren't a lot of things to improve, but it isn't a "techies version" more than SLED nor Ubuntu nor Fedora nor ...
As far as I'm concerned they are all techies versions and Linux suppliers, not Linux itself, are the ones that need improvement. We need to hire people in marketing to layout our sites we need to do more user testing like Novell
Marketing people have nothing to do with anything productive or efficient, and definitely not with usable websites (except the very, very few good marketing people out there). We would rather need web designers, usability experts, ... umm.. no, wait. What we would need in the first place is 50% of the >85% joining the effort, even with modest contributions. Maybe the real question is: how to bring those people into the active community ? Better site ? Better tools ? Probably. A kick in the ass ? Possibly ^^
did to find out what SLED needed and how people used it. Only now we are working on delivery not functionality so we need to do usability testing on our websites. I stand by my assertion that our structure is still to immature to offer to the masses.
Define "our structure". Ok, that probably answers the question.
cheers
- --
-o) Pascal Bleser http://linux01.gwdg.de/~pbleser/
/\\