Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!

Bug 476982

Summary: default src_install installs README.html into docdir instead of htmldir
Product: Gentoo Hosted Projects Reporter: Julian Ospald <hasufell>
Component: PMS/EAPIAssignee: PMS/EAPI <pms>
Status: RESOLVED WONTFIX    
Severity: normal    
Priority: Normal    
Version: unspecified   
Hardware: All   
OS: Linux   
URL: https://dev.gentoo.org/~ulm/pms/head/pms.html#x1-1030009.1.9
Whiteboard:
Package list:
Runtime testing required: ---
Attachments: fplll-4.0.4.ebuild

Description Julian Ospald 2013-07-15 22:02:38 UTC
Created attachment 353400 [details]
fplll-4.0.4.ebuild

example ebuild attached

caused by the wildcard
Comment 1 Julian Ospald 2013-07-15 22:30:53 UTC
proposal:


for d in README* ChangeLog AUTHORS NEWS TODO CHANGES \
        THANKS BUGS FAQ CREDITS CHANGELOG ; do
    if [[ -s "${d}" ]] ; then
        if [[ "${d}" == *.html ]] ; then
            dohtml "${d}"
        else
            dodoc "${d}"
        fi
    fi
done
Comment 2 Ulrich Müller gentoo-dev 2013-08-19 21:39:08 UTC
(In reply to Julian Ospald (hasufell) from comment #1)
> for d in README* ChangeLog AUTHORS NEWS TODO CHANGES \
>         THANKS BUGS FAQ CREDITS CHANGELOG ; do
>     if [[ -s "${d}" ]] ; then
>         if [[ "${d}" == *.html ]] ; then
>             dohtml "${d}"
>         else
>             dodoc "${d}"
>         fi
>     fi
> done

TBH, I don't like this. It adds some clutter to the default function, and I wonder how many ebuilds could actually profit from such special treatment of README.html.

If we are going to change something, I'd rather drop the wildcard and install plain README only.
Comment 3 Julian Ospald 2013-08-19 21:40:35 UTC
(In reply to Ulrich Müller from comment #2)
> (In reply to Julian Ospald (hasufell) from comment #1)
> > for d in README* ChangeLog AUTHORS NEWS TODO CHANGES \
> >         THANKS BUGS FAQ CREDITS CHANGELOG ; do
> >     if [[ -s "${d}" ]] ; then
> >         if [[ "${d}" == *.html ]] ; then
> >             dohtml "${d}"
> >         else
> >             dodoc "${d}"
> >         fi
> >     fi
> > done
> 
> TBH, I don't like this. It adds some clutter to the default function, and I
> wonder how many ebuilds could actually profit from such special treatment of
> README.html.
> 
> If we are going to change something, I'd rather drop the wildcard and
> install plain README only.

clutter?
Comment 4 Ulrich Müller gentoo-dev 2013-08-19 21:47:14 UTC
Mainly the dohtml, which I expect to be useful in rare cases only. This isn't appropriate for a default function.
Comment 5 Julian Ospald 2013-08-19 21:48:35 UTC
I cannot follow that logic. The default function misbehaves and needs to be fixed.
Comment 6 Ulrich Müller gentoo-dev 2013-08-19 22:04:56 UTC
It's a *default* function, so it is expected that it cannot handle all cases. There's always the possibility to write your own src_install() for the few ebuilds that need it.
Comment 7 Julian Ospald 2013-08-19 23:15:45 UTC
(In reply to Ulrich Müller from comment #6)
> It's a *default* function, so it is expected that it cannot handle all
> cases. There's always the possibility to write your own src_install() for
> the few ebuilds that need it.

handling more cases by default reduces the need of writing your own functions, so it is an improvement
Comment 8 Ulrich Müller gentoo-dev 2013-08-20 09:43:46 UTC
No need for a function, defining DOCS is enough.

Anyway, I suggest that you get consensus in the -dev ML if you want this feature.
Comment 9 Michał Górny archtester Gentoo Infrastructure gentoo-dev Security 2013-08-20 09:45:58 UTC
Two points:

1. we don't support any kind of explicit excludes. This one feels like implicit exclude, and is inconsistent with this. It's going to be confusing.

2. I feel like html/ fits better big docs with index.html and so on. Creating the directory for single README file is just wrong.
Comment 10 Ulrich Müller gentoo-dev 2014-08-28 11:28:07 UTC
(In reply to Ulrich Müller from comment #8)
> Anyway, I suggest that you get consensus in the -dev ML if you want this
> feature.

AFAICS, no such discussion has taken place.
No progress since one year, therefore closing.