move to use v5.36;
tested by me over the last few weeks, and tb@
also fixed a "manual install" bug properly reported by tb@

aside that there should be *no functional change*.
If you see any message like "hey, the number of params is wrong"
it is a fringe case I didn't run into and should be easy to fix.

document (hopefully) near future directions

allow search objects to say "keep_all" and make PackageRepositoryLists
aware of that for match_locations

we might not get update_info if unsigned!

new syntax: pkgname%branch to select a branch wherever it makes sense.
see pkg_add(1) for details.

Search framework makes that surprisingly simple.

handle conflict with partial in a less hackish ways, by passing an
extra "with_partial" parameter to specs.
This also removes quite a lot extraneous objects.

allows .libs* packages to match correctly in the presence of @conflict markers

handle failures from ports tree with more care.
this does avoid infinite loops in case of wrong plists.
necessary since print-plist-with-depends CAN fail...

initial scaffolding for @rcscript: for now, acts like normal files,
but allow absolute pathnames (treat that as an implicit @cwd).

allow say and errsay to work without parameters, as it's ways common.
create verbose_system up in state, because it makes sense without verbose.

whitespace cleanup

whitespace fixes

better spec handling: instead of dying, create badspec objects that never
match anything. Add method is_valid (and propagators) to know whether a
spec is valid (to be used in pkg_create in a systematic way).

use Auto::cache

fix partial stems, as noticed by numerous people

change the internal representation of Search::Stem to allow the addition
of supplementary stems (useful for quirks)

tweak the rules a bit: the search filter for "more recent than" ought to
notice it sees older stuff. Even if they get filtered later on, this would
turn into a "no need to update".

(in actual use, I probably will need to open those packages and look at
the pkgpaths to make sure they are not valid candidates, and to compare
full signatures... oh well)

kill more unused code

remove dead code: as far as possible, we use only
match_locations/filter_locations since we deal with location.

external API only uses spec->filter(), so we keep that one.
We still need the generic match/filter adapters for Stem searches.

bad espie: use strict/warnings consistently, and fix two nits and two
actual errors !

fix, need to keep =version for "no need to update" to occur...

okay, so updates become "true" updates: unless you say -F downgrade
then packages with an older version will be filtered out.

maybe I'll change the code later (two XXX), but allows flavor after stem,
so that e.g.
pkg_add star--
pkg_add star--static
are now not ambiguous.

remove class FatPackageLocation. Instead make PackageLocation a hierarchy
so that I can handle installed packages without playing ping-pong with
the PackageRepository hierarchy (mostly)

synch with my new style search/location changes
It's likely there will be some fallout, but it's getting a bit too large
to keep around.
This does kill a few very old oddities as well.

tweak the semantics so that no-default-conflict works.

requested by mbalmer@,sturm@. okay winiger@

I am on drugs, I don't know where I found split can take strings as
arguments. Repair...

some minor systematic changes.
- mark all regexps I can with /o if they can be compiled once.
- turn $o->method() into $o->method
- remove unneeded prototypes
- reduce split /re/ into split "string" where possible.

add match_locations/filter_locations methods to search stuff that returns
full pkg locations instead of package names.

Add pkgpath hints to pkgspec. Make PackageRepository/Source functional:
if a pkgpath is hinted at, then we build that package, and we delegate
to the corresponding File: repository.

Some XXX: we need to unlock the database temporarily, as the build
*requires* shared access to stuff that is going on (and possibly write
access for installing subpackages, so this cannot work 100% yet).

Still, this allows the transparent build of one single package, at least.

for completion, allow Stem object to filter as well.

redo search filters as real objects. Accordingly, rename match_list to
filter.

put Search objects into their own file, finally

document (hopefully) near future directions

allow search objects to say "keep_all" and make PackageRepositoryLists
aware of that for match_locations

we might not get update_info if unsigned!

new syntax: pkgname%branch to select a branch wherever it makes sense.
see pkg_add(1) for details.

Search framework makes that surprisingly simple.

handle conflict with partial in a less hackish ways, by passing an
extra "with_partial" parameter to specs.
This also removes quite a lot extraneous objects.

allows .libs* packages to match correctly in the presence of @conflict markers

handle failures from ports tree with more care.
this does avoid infinite loops in case of wrong plists.
necessary since print-plist-with-depends CAN fail...

initial scaffolding for @rcscript: for now, acts like normal files,
but allow absolute pathnames (treat that as an implicit @cwd).

allow say and errsay to work without parameters, as it's ways common.
create verbose_system up in state, because it makes sense without verbose.

whitespace cleanup

whitespace fixes

better spec handling: instead of dying, create badspec objects that never
match anything. Add method is_valid (and propagators) to know whether a
spec is valid (to be used in pkg_create in a systematic way).

use Auto::cache

fix partial stems, as noticed by numerous people

change the internal representation of Search::Stem to allow the addition
of supplementary stems (useful for quirks)

tweak the rules a bit: the search filter for "more recent than" ought to
notice it sees older stuff. Even if they get filtered later on, this would
turn into a "no need to update".

(in actual use, I probably will need to open those packages and look at
the pkgpaths to make sure they are not valid candidates, and to compare
full signatures... oh well)

kill more unused code

remove dead code: as far as possible, we use only
match_locations/filter_locations since we deal with location.

external API only uses spec->filter(), so we keep that one.
We still need the generic match/filter adapters for Stem searches.

bad espie: use strict/warnings consistently, and fix two nits and two
actual errors !

fix, need to keep =version for "no need to update" to occur...

okay, so updates become "true" updates: unless you say -F downgrade
then packages with an older version will be filtered out.

maybe I'll change the code later (two XXX), but allows flavor after stem,
so that e.g.
pkg_add star--
pkg_add star--static
are now not ambiguous.

remove class FatPackageLocation. Instead make PackageLocation a hierarchy
so that I can handle installed packages without playing ping-pong with
the PackageRepository hierarchy (mostly)

synch with my new style search/location changes
It's likely there will be some fallout, but it's getting a bit too large
to keep around.
This does kill a few very old oddities as well.

tweak the semantics so that no-default-conflict works.

requested by mbalmer@,sturm@. okay winiger@

I am on drugs, I don't know where I found split can take strings as
arguments. Repair...

some minor systematic changes.
- mark all regexps I can with /o if they can be compiled once.
- turn $o->method() into $o->method
- remove unneeded prototypes
- reduce split /re/ into split "string" where possible.

add match_locations/filter_locations methods to search stuff that returns
full pkg locations instead of package names.

Add pkgpath hints to pkgspec. Make PackageRepository/Source functional:
if a pkgpath is hinted at, then we build that package, and we delegate
to the corresponding File: repository.

Some XXX: we need to unlock the database temporarily, as the build
*requires* shared access to stuff that is going on (and possibly write
access for installing subpackages, so this cannot work 100% yet).

Still, this allows the transparent build of one single package, at least.

for completion, allow Stem object to filter as well.

redo search filters as real objects. Accordingly, rename match_list to
filter.

put Search objects into their own file, finally

allow search objects to say "keep_all" and make PackageRepositoryLists
aware of that for match_locations

we might not get update_info if unsigned!

new syntax: pkgname%branch to select a branch wherever it makes sense.
see pkg_add(1) for details.

Search framework makes that surprisingly simple.

handle conflict with partial in a less hackish ways, by passing an
extra "with_partial" parameter to specs.
This also removes quite a lot extraneous objects.

allows .libs* packages to match correctly in the presence of @conflict markers

handle failures from ports tree with more care.
this does avoid infinite loops in case of wrong plists.
necessary since print-plist-with-depends CAN fail...

initial scaffolding for @rcscript: for now, acts like normal files,
but allow absolute pathnames (treat that as an implicit @cwd).

allow say and errsay to work without parameters, as it's ways common.
create verbose_system up in state, because it makes sense without verbose.

whitespace cleanup

whitespace fixes

better spec handling: instead of dying, create badspec objects that never
match anything. Add method is_valid (and propagators) to know whether a
spec is valid (to be used in pkg_create in a systematic way).

use Auto::cache

fix partial stems, as noticed by numerous people

change the internal representation of Search::Stem to allow the addition
of supplementary stems (useful for quirks)

tweak the rules a bit: the search filter for "more recent than" ought to
notice it sees older stuff. Even if they get filtered later on, this would
turn into a "no need to update".

(in actual use, I probably will need to open those packages and look at
the pkgpaths to make sure they are not valid candidates, and to compare
full signatures... oh well)

kill more unused code

remove dead code: as far as possible, we use only
match_locations/filter_locations since we deal with location.

external API only uses spec->filter(), so we keep that one.
We still need the generic match/filter adapters for Stem searches.

bad espie: use strict/warnings consistently, and fix two nits and two
actual errors !

fix, need to keep =version for "no need to update" to occur...

okay, so updates become "true" updates: unless you say -F downgrade
then packages with an older version will be filtered out.

maybe I'll change the code later (two XXX), but allows flavor after stem,
so that e.g.
pkg_add star--
pkg_add star--static
are now not ambiguous.

remove class FatPackageLocation. Instead make PackageLocation a hierarchy
so that I can handle installed packages without playing ping-pong with
the PackageRepository hierarchy (mostly)

synch with my new style search/location changes
It's likely there will be some fallout, but it's getting a bit too large
to keep around.
This does kill a few very old oddities as well.

tweak the semantics so that no-default-conflict works.

requested by mbalmer@,sturm@. okay winiger@

I am on drugs, I don't know where I found split can take strings as
arguments. Repair...

some minor systematic changes.
- mark all regexps I can with /o if they can be compiled once.
- turn $o->method() into $o->method
- remove unneeded prototypes
- reduce split /re/ into split "string" where possible.

add match_locations/filter_locations methods to search stuff that returns
full pkg locations instead of package names.

Add pkgpath hints to pkgspec. Make PackageRepository/Source functional:
if a pkgpath is hinted at, then we build that package, and we delegate
to the corresponding File: repository.

Some XXX: we need to unlock the database temporarily, as the build
*requires* shared access to stuff that is going on (and possibly write
access for installing subpackages, so this cannot work 100% yet).

Still, this allows the transparent build of one single package, at least.

for completion, allow Stem object to filter as well.

redo search filters as real objects. Accordingly, rename match_list to
filter.

put Search objects into their own file, finally

we might not get update_info if unsigned!

new syntax: pkgname%branch to select a branch wherever it makes sense.
see pkg_add(1) for details.

Search framework makes that surprisingly simple.

handle conflict with partial in a less hackish ways, by passing an
extra "with_partial" parameter to specs.
This also removes quite a lot extraneous objects.

allows .libs* packages to match correctly in the presence of @conflict markers

handle failures from ports tree with more care.
this does avoid infinite loops in case of wrong plists.
necessary since print-plist-with-depends CAN fail...

initial scaffolding for @rcscript: for now, acts like normal files,
but allow absolute pathnames (treat that as an implicit @cwd).

allow say and errsay to work without parameters, as it's ways common.
create verbose_system up in state, because it makes sense without verbose.

whitespace cleanup

whitespace fixes

better spec handling: instead of dying, create badspec objects that never
match anything. Add method is_valid (and propagators) to know whether a
spec is valid (to be used in pkg_create in a systematic way).

use Auto::cache

fix partial stems, as noticed by numerous people

change the internal representation of Search::Stem to allow the addition
of supplementary stems (useful for quirks)

tweak the rules a bit: the search filter for "more recent than" ought to
notice it sees older stuff. Even if they get filtered later on, this would
turn into a "no need to update".

(in actual use, I probably will need to open those packages and look at
the pkgpaths to make sure they are not valid candidates, and to compare
full signatures... oh well)

kill more unused code

remove dead code: as far as possible, we use only
match_locations/filter_locations since we deal with location.

external API only uses spec->filter(), so we keep that one.
We still need the generic match/filter adapters for Stem searches.

bad espie: use strict/warnings consistently, and fix two nits and two
actual errors !

fix, need to keep =version for "no need to update" to occur...

okay, so updates become "true" updates: unless you say -F downgrade
then packages with an older version will be filtered out.

maybe I'll change the code later (two XXX), but allows flavor after stem,
so that e.g.
pkg_add star--
pkg_add star--static
are now not ambiguous.

remove class FatPackageLocation. Instead make PackageLocation a hierarchy
so that I can handle installed packages without playing ping-pong with
the PackageRepository hierarchy (mostly)

synch with my new style search/location changes
It's likely there will be some fallout, but it's getting a bit too large
to keep around.
This does kill a few very old oddities as well.

tweak the semantics so that no-default-conflict works.

requested by mbalmer@,sturm@. okay winiger@

I am on drugs, I don't know where I found split can take strings as
arguments. Repair...

some minor systematic changes.
- mark all regexps I can with /o if they can be compiled once.
- turn $o->method() into $o->method
- remove unneeded prototypes
- reduce split /re/ into split "string" where possible.

add match_locations/filter_locations methods to search stuff that returns
full pkg locations instead of package names.

Add pkgpath hints to pkgspec. Make PackageRepository/Source functional:
if a pkgpath is hinted at, then we build that package, and we delegate
to the corresponding File: repository.

Some XXX: we need to unlock the database temporarily, as the build
*requires* shared access to stuff that is going on (and possibly write
access for installing subpackages, so this cannot work 100% yet).

Still, this allows the transparent build of one single package, at least.

for completion, allow Stem object to filter as well.

redo search filters as real objects. Accordingly, rename match_list to
filter.

put Search objects into their own file, finally

new syntax: pkgname%branch to select a branch wherever it makes sense.
see pkg_add(1) for details.

Search framework makes that surprisingly simple.

handle conflict with partial in a less hackish ways, by passing an
extra "with_partial" parameter to specs.
This also removes quite a lot extraneous objects.

allows .libs* packages to match correctly in the presence of @conflict markers

handle failures from ports tree with more care.
this does avoid infinite loops in case of wrong plists.
necessary since print-plist-with-depends CAN fail...

initial scaffolding for @rcscript: for now, acts like normal files,
but allow absolute pathnames (treat that as an implicit @cwd).

allow say and errsay to work without parameters, as it's ways common.
create verbose_system up in state, because it makes sense without verbose.

whitespace cleanup

whitespace fixes

better spec handling: instead of dying, create badspec objects that never
match anything. Add method is_valid (and propagators) to know whether a
spec is valid (to be used in pkg_create in a systematic way).

use Auto::cache

fix partial stems, as noticed by numerous people

change the internal representation of Search::Stem to allow the addition
of supplementary stems (useful for quirks)

tweak the rules a bit: the search filter for "more recent than" ought to
notice it sees older stuff. Even if they get filtered later on, this would
turn into a "no need to update".

(in actual use, I probably will need to open those packages and look at
the pkgpaths to make sure they are not valid candidates, and to compare
full signatures... oh well)

kill more unused code

remove dead code: as far as possible, we use only
match_locations/filter_locations since we deal with location.

external API only uses spec->filter(), so we keep that one.
We still need the generic match/filter adapters for Stem searches.

bad espie: use strict/warnings consistently, and fix two nits and two
actual errors !

fix, need to keep =version for "no need to update" to occur...

okay, so updates become "true" updates: unless you say -F downgrade
then packages with an older version will be filtered out.

maybe I'll change the code later (two XXX), but allows flavor after stem,
so that e.g.
pkg_add star--
pkg_add star--static
are now not ambiguous.

remove class FatPackageLocation. Instead make PackageLocation a hierarchy
so that I can handle installed packages without playing ping-pong with
the PackageRepository hierarchy (mostly)

synch with my new style search/location changes
It's likely there will be some fallout, but it's getting a bit too large
to keep around.
This does kill a few very old oddities as well.

tweak the semantics so that no-default-conflict works.

requested by mbalmer@,sturm@. okay winiger@

I am on drugs, I don't know where I found split can take strings as
arguments. Repair...

some minor systematic changes.
- mark all regexps I can with /o if they can be compiled once.
- turn $o->method() into $o->method
- remove unneeded prototypes
- reduce split /re/ into split "string" where possible.

add match_locations/filter_locations methods to search stuff that returns
full pkg locations instead of package names.

Add pkgpath hints to pkgspec. Make PackageRepository/Source functional:
if a pkgpath is hinted at, then we build that package, and we delegate
to the corresponding File: repository.

Some XXX: we need to unlock the database temporarily, as the build
*requires* shared access to stuff that is going on (and possibly write
access for installing subpackages, so this cannot work 100% yet).

Still, this allows the transparent build of one single package, at least.

for completion, allow Stem object to filter as well.

redo search filters as real objects. Accordingly, rename match_list to
filter.

put Search objects into their own file, finally