Bugzilla Bug 17172

According to the Gentoo Dev Guide, ebuilds are allowed to have a single letter
in the base version string (i.e. foo-1.1a).  The problem is that portage
complains when one attempts to append a valid suffix (i.e. _pre or _beta) to the
ebuild's name.  It states that a letter cannot precede the " _ ", however the
Dev Guide says that a single letter is allowed.  The suffix doesn't count
towards this single letter, since it isn't part of the base name.  I request
that portage be repaired to allow this aspect of the naming conventions in ebuilds.

*** Bug 27840 has been marked as a duplicate of this bug. ***

I also encountered the same problem with app-arch/lha. The souce
distribution uses "lha-114i-ac20030921" style versioning and I
cannot create ebuild named lha-114i.20030921.ebuild (anyhow, it is not
a valid name according to Gentoo Policy). How should I name it?
Previous release, lha-114i-ac20020903 is in Portage tree but included
as lha-114i-r1.ebuild. I think it is worse (-r? is only meant for
Gentoo internal release and we should bump -ver or _suf) than 
lha-114i.20030921.ebuild. However, it seems adding `-r?' is the only
way to create such an ebuild (I think it is really bad). Should we
allow not only number (and single letter immediately after the last
digit) in -ver but also alphabet ? We might exclude alphabet-only
version in -ver (e.g. lha-114i.ac.20030903 is invalid since it
will be split to `114i', `ac' and `20030903' while we would accept
lha-114i.ac20030903 since `114i' and `ac20030903' would be considered
as valid versioning according to the suggested naming scheme).

Created an attachment (id=21418) [edit]
diff to portage-2.0.49-r15

This bug has been already fixed in 2.0.50_pre*

supposed to be fixed in 2.0.50 which is stable now. If this bug is not fixed
please reopen.

Looks like it's not working yet...

/usr/local/download/portage-cvs/gentoo-x86/net-analyzer/netwatch $ emerge -pv netwatch-1.0b-r3.ebuild 

These are the packages that I would merge, in order:

Calculating dependencies ...done!
[ebuild   R   ] net-analyzer/netwatch-1.0b-r3   [empty/missing/bad digest] [1] 

Total size of downloads: 0 kB
Portage overlays:
 [1] /usr/local/download/portage-cvs/gentoo-x86


/usr/local/download/portage-cvs/gentoo-x86/net-analyzer/netwatch $ emerge -pv netwatch-1.0b_pre3.ebuild 

These are the packages that I would merge, in order:

Calculating dependencies ...done!
Traceback (most recent call last):
  File "/usr/bin/emerge", line 2574, in ?
    mydepgraph.display(mydepgraph.altlist())
  File "/usr/bin/emerge", line 1236, in display
    if portage.pkgcmp(portage.pkgsplit(x[2]), portage.pkgsplit(myoldbest)) < 0:
  File "/usr/lib/portage/pym/portage.py", line 2825, in pkgcmp
    mycmp=vercmp(pkg1[1],pkg2[1])
  File "/usr/lib/portage/pym/portage.py", line 2810, in vercmp
    cmp1=relparse(val1[x])
  File "/usr/lib/portage/pym/portage.py", line 2499, in relparse
    number=string.atof(mynewver[0])
  File "/usr/lib/python2.3/string.py", line 205, in atof
    return _float(s)
ValueError: invalid literal for float(): .0b

/usr/local/download/portage-cvs/gentoo-x86/net-analyzer/netwatch $ diff -Naur netwatch-1.0b-r3.ebuild netwatch-1.0b_pre3.ebuild
--- netwatch-1.0b-r3.ebuild     2004-02-08 19:10:20.000000000 -0800
+++ netwatch-1.0b_pre3.ebuild   2004-02-08 19:06:49.000000000 -0800
@@ -2,7 +2,7 @@
 # Distributed under the terms of the GNU General Public License v2
 # $Header: $
 
-MY_P="${PF/-r/.pre}"
+MY_P="${P/_pre/.pre}"
 SRC_URI="http://www.slctech.org/~mackay/${MY_P}.src.tgz"
 DESCRIPTION="a ncurses based network monitoring program"
 HOMEPAGE="http://www.slctech.org/~mackay/netwatch.html"

Created an attachment (id=25244) [edit]
patch to fix the new error

in portage-2.0.50 this situation creates an Exception as atof() is called on
the part before the underscore (in relparse()).

the patch fixes the problem for me.

*** Bug 41224 has been marked as a duplicate of this bug. ***

also see #37406 for a rewrite of the versioning code that should also fix the
problem (as well as several limitations with the current code).

Is it going to be fixed in the next release?

*** Bug 43580 has been marked as a duplicate of this bug. ***

*** Bug 58973 has been marked as a duplicate of this bug. ***

Bug has been fixed and released in stable portages on or before 2.0.51-r2