[xmonad] Call for xmonad.hs

[Gwern Branwen]

On 2008.10.19 15:08:01 -0700, Andrew Sackville-West <andrew at swclan.homelinux.org> scribbled 2.5K characters:
> On Sun, Oct 19, 2008 at 03:59:07PM -0400, Gwern Branwen wrote:
> > I'm not sure whether it's worthwhile to
> > update them - if they're not being updated by their user, it's
> > reasonable to infer the user is no longer using XMonad
>
> I disagree. I use xmonad everyday on three different machines and
> wouldn't trade it for anything. But my config in the archive is
> woefully out of date. I'll take this as a prompt to jump over and
> update mine...

Heh heh, yes, one of the reasons I want everyone to update their wiki config is so that I can make such inferences. If configs aren't very up to date in general, then that constitutes weak Bayesian evidence of use.

> > (in which
> > case they aren't useful in considering how to make XMonad users'
> > lives easier)
>
> that follows.
>
> > or are using a different config (and then obviously we
> > should care less about a configuration even its creator doesn't want
> > to use).
>
> the assumption that the user doesn't use the config is just wrong,
> IMVHO. It turns out for me, really, the system is so stable and
> usable, once tweaked to a reasonably pleasant default, that there is
> no need to change. I only started using AvoidStruts (or whatever it's
> called) the just the other day because of an unrelated change to my
> config which broke my xmonad.hs because defaultGaps wasn't around
> anymore. That config hadn't been rebuilt in *months* at least. I
> missed a couple of upgrades, at least, because what I had _just
> worked_.

This is a problem I didn't really expect (although I should've). Besides simply keeping this in mind ('for every user, expect there to be more than a few shadow users'), how does this affect us? Do we seriously need to worry about installations/users going back to, say, Debian stable? Or can we just focus on 0.8?

> I think one of the signs of the success of a product is that certain
> subsets of the user population *stop* upgrading. Not because they
> aren't using the product, but because the product fits their needs to
> a level where it's just a tool and no longer a new thing that needs to
> be maintained. tribute to xmonad ;)

Well, I think XMonad itself is pretty stable alright. There doesn't seem to be much left to do - I've seen things like using XCB suggested, but that seems both a) likely to never happen*, and b) only require installation of libraries and recompilation etc., with no config changes (I guess? I'm not sure about this). XMC, though, I consider to be far from done. We don't even have compositing support, and that's an entire rich area on its own**.

* since it hasn't happened yet and quite a bit of time has passed since the first discussions and attempts. Doomsday Principle and all that.
** I demand an XMC module which hooks into my household alarm system so that when my house is on fire, I can have cool 3D flames race across my screen when switching windows!

--
gwern
SALDV Aum enigma remailers embassy NATIA in Armani Sayeret
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
Url : http://www.haskell.org/pipermail/xmonad/attachments/20081020/f7529d7e/attachment.bin