Re: a fallacy (was Re: when can a package be made architecture-dependent?)
On 18 Jan 2003, Adam DiCarlo wrote:
> > If every Debian developer refused to support > architectures he/she
> > didn't have immediate access to, non-x86 Debian would > disappear
> > pretty quickly.
>
> If the upstream software maintainers state they don't want to support
> certain architectures, what the hell, isn't that their perogative?
>
> Are we in some sort of inverted fascist world where just because an
> author gives away some software, she's suddenly required to commit to
> supporting whatever we say? Remember, supporting different
> architectures can be a bear of a job.
There's no need to go tossing terms like "fascist" around, and no one is
suggesting that the upstream authors be "required" to do anything. The
question is what standards Debian developers should set for
themselves...what does it mean for *Debian* to support 11 architectures
unless the packagers make some effort in this regard, even though any
individual packager is unlikely to have access to all 11.
Nor am I suggesting heroic porting efforts, like rewriting large sections
of assembly code. Just a minimal effort---at the least, not tagging a
package as architecture-specific if it should be reasonably possible to
port, even if it fails to build immediately.
That's all this thread is about: tagging. Please don't start talking
about Nazis and turn it into a flamewar, if it's not too late already.
Cordially,
Steven G. Johnson
Reply to: