On Wed, Apr 14, 2004 at 06:47:42PM -0400, Chris Metzler wrote: > This thread got started because people were frustrated about having > to explain stable vs. testing vs. unstable to new users trying > Debian. But it appears to me that a lot of people with strong > ideas on how to fix that don't understand the differences themselves.
this is good evidence that there must be a better approach than the one we're currently using. john doe will read "stable" and might think it means that "it's got all the current upstream bug fixes" when what we mean by it is "we stopped adding new stuff to this one a long time ago, and haven't found any serious conflicts in quite a long time". john doe will read "unstable" and may think it means "not stable", whereas what we mean is "probably stable, and we're working on making it more stable". john doe will read "testing" and may think it means "experimental" when we intend it to mean "whatever isn't stable here will be really soon, and it'll be the new stable version". there's a discrepancy between what a newbie it likely to infer, and what the old hands have learned to interpret. > The web page http://www.debian.org/devel/testing explains what testing > is. It isn't what many people in this thread seem to be suggesting. exactly this confusion could be alleviated by a better naming scheme. but perhaps we should examine AVOIDING descriptive names altogether... as Tom Massey also has a point: On Thu, Apr 15, 2004 at 12:14:05AM +1000, Tom Massey wrote: > I vaguely suspect that renaming the releases won't actually > solve the problem that it's meant to - reducing confusion > among new Debian users. You're likely to just end up with a > new set of labels to explain. Any name you come up with is > going to be too short to fully explain the situation: call > stable 'server', testing 'desktop' for example, and you still > have to explain that the server release is good for desktops > if you prefer stability over new stuff, and the desktop > release might be good for a server if you need more recent > packages and don't want to search for backports. You can't fit > all that info into a short name. I run unstable on my desktop > machine, stable on my mail server because I know what the > names mean. Education as to what goes in to the various Debian > releases is the key, and changing the release names doesn't do > much for that. perhaps instead of trying for descriptive-but-too-short a name for each layer (stable/testing/unstable) of release, we should stick to ONLY the colorful code names which will give the newbie NOTHING to assume about current-vs-stable, and then they're more likely to traipse over to the description page to learn what's what. i.e. instead of stable / server / rocksolid testing / workstation / almost unstable / cuttingedge / future maybe we should avoid the descriptive names and use only ...slink ...potato woody sarge sid ? -- I use Debian/GNU Linux version 3.0; Linux boss 2.4.18-bf2.4 #1 Son Apr 14 09:53:28 CEST 2002 i586 unknown DEBIAN NEWBIE TIP #101 from Joost Kooij <[EMAIL PROTECTED]> : Looking for a way to CREATE A PAGE OF LINKS to all the */index.html that already exist in your /usr/share/doc tree? apt-get install dwww then point your browser to: http://localhost/dwww Also see http://newbieDoc.sourceForge.net/ ... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
