Adds support for reverse addressing relative to the current step, as well as calculating the path to the special device node obtained in the previous step. Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com> --- features/bootchain-core/README.md | 6 +++ .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++--- 2 files changed, 38 insertions(+), 5 deletions(-) diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md index 8d341e0..ba66259 100644 --- a/features/bootchain-core/README.md +++ b/features/bootchain-core/README.md @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need. - Modularity: loading methods are initially separated from the common code and daemon. +- Via resolve_target() supports not only forward, but also reverse addressing, + relative to the current step. For example, a record like `step-3/dir1/dev` + will process the result of `dir1/dev`, made in the third step from the current + one. Together with the overload of the chain of steps, direct addressing is safe + only when storing the numbers of the completed steps in files, whereas reverse + relative addressing it is safe in any case and can often be more convenient. - Allows you to work with shorter and more familiar paths to special files devices thanks to the use of `DEVNAME` along with `dev`. diff --git a/features/bootchain-core/data/bin/bootchain-sh-functions b/features/bootchain-core/data/bin/bootchain-sh-functions index e73fd35..0eebbaf 100644 --- a/features/bootchain-core/data/bin/bootchain-sh-functions +++ b/features/bootchain-core/data/bin/bootchain-sh-functions @@ -27,19 +27,46 @@ get_parameter() resolve_target() { local target="$1" + local where="${target%%/*}" + local n="${#where}" - case "${target%%/*}" in + target="${target:$n}" + + case "$where" in '') ;; pipe[0-9]|pipe[0-9][0-9]|pipe[0-9][0-9][0-9]) - target="$mntdir/dst/step${target:4}" + [ -d "$mntdir/dst/step${where:4}" ] || + return 0 + target="$(readlink-e "$mntdir/dst/step${where:4}")${target}" + ;; + step-[1-9]|step-[1-9][0-9]|step-[1-9][0-9][0-9]) + case "$destdir" in + "$mntdir"/dst/step[0-9]*) + n="${destdir##*/}" + n="$(( ${n:4} - ${where:5} ))" + ;; + *) + return 0 + ;; + esac + [ "$n" -ge 0 ] && [ -d "$mntdir/dst/step$n" ] || + return 0 + target="$(readlink-e "$mntdir/dst/step$n")${target}" ;; *) if [ -z "${prevdir-}" ]; then - message "no previous stop to use" - return + message "no previous step results to use with $name" + return 0 + fi + if [ -z "$target" ] && + [ "$where" = DEVNAME ] && + [ -s "$prevdir"/DEVNAME ] + then + read -r target <"$prevdir"/DEVNAME ||: + else + target="$prevdir/${where#/}${target}" fi - target="$prevdir/${target#/}" ;; esac -- 2.24.1
On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote: > Adds support for reverse addressing relative to the current step, > as well as calculating the path to the special device node obtained > in the previous step. > > Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com> > --- > features/bootchain-core/README.md | 6 +++ > .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++--- > 2 files changed, 38 insertions(+), 5 deletions(-) > > diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md > index 8d341e0..ba66259 100644 > --- a/features/bootchain-core/README.md > +++ b/features/bootchain-core/README.md > @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need. > > - Modularity: loading methods are initially separated from the common > code and daemon. > +- Via resolve_target() supports not only forward, but also reverse addressing, > + relative to the current step. For example, a record like `step-3/dir1/dev` > + will process the result of `dir1/dev`, made in the third step from the current > + one. Together with the overload of the chain of steps, direct addressing is safe > + only when storing the numbers of the completed steps in files, whereas reverse > + relative addressing it is safe in any case and can often be more convenient. На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется одинаковым. Хотя бы current-1 > - Allows you to work with shorter and more familiar paths to special files > devices thanks to the use of `DEVNAME` along with `dev`. > > diff --git a/features/bootchain-core/data/bin/bootchain-sh-functions b/features/bootchain-core/data/bin/bootchain-sh-functions > index e73fd35..0eebbaf 100644 > --- a/features/bootchain-core/data/bin/bootchain-sh-functions > +++ b/features/bootchain-core/data/bin/bootchain-sh-functions > @@ -27,19 +27,46 @@ get_parameter() > resolve_target() > { > local target="$1" > + local where="${target%%/*}" > + local n="${#where}" > > - case "${target%%/*}" in > + target="${target:$n}" > + > + case "$where" in > '') > ;; > pipe[0-9]|pipe[0-9][0-9]|pipe[0-9][0-9][0-9]) > - target="$mntdir/dst/step${target:4}" > + [ -d "$mntdir/dst/step${where:4}" ] || > + return 0 > + target="$(readlink-e "$mntdir/dst/step${where:4}")${target}" > + ;; > + step-[1-9]|step-[1-9][0-9]|step-[1-9][0-9][0-9]) > + case "$destdir" in > + "$mntdir"/dst/step[0-9]*) > + n="${destdir##*/}" > + n="$(( ${n:4} - ${where:5} ))" > + ;; > + *) > + return 0 > + ;; > + esac > + [ "$n" -ge 0 ] && [ -d "$mntdir/dst/step$n" ] || > + return 0 > + target="$(readlink-e "$mntdir/dst/step$n")${target}" > ;; > *) > if [ -z "${prevdir-}" ]; then > - message "no previous stop to use" > - return > + message "no previous step results to use with $name" > + return 0 > + fi > + if [ -z "$target" ] && > + [ "$where" = DEVNAME ] && > + [ -s "$prevdir"/DEVNAME ] > + then > + read -r target <"$prevdir"/DEVNAME ||: > + else > + target="$prevdir/${where#/}${target}" > fi > - target="$prevdir/${target#/}" > ;; > esac > > -- > 2.24.1 > > _______________________________________________ > Make-initrd mailing list > Make-initrd@lists.altlinux.org > https://lists.altlinux.org/mailman/listinfo/make-initrd > -- Rgrds, legion
26.10.2021 14:12, Alexey Gladkov пишет: > On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote: >> Adds support for reverse addressing relative to the current step, >> as well as calculating the path to the special device node obtained >> in the previous step. >> >> Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com> >> --- >> features/bootchain-core/README.md | 6 +++ >> .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++--- >> 2 files changed, 38 insertions(+), 5 deletions(-) >> >> diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md >> index 8d341e0..ba66259 100644 >> --- a/features/bootchain-core/README.md >> +++ b/features/bootchain-core/README.md >> @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need. >> >> - Modularity: loading methods are initially separated from the common >> code and daemon. >> +- Via resolve_target() supports not only forward, but also reverse addressing, >> + relative to the current step. For example, a record like `step-3/dir1/dev` >> + will process the result of `dir1/dev`, made in the third step from the current >> + one. Together with the overload of the chain of steps, direct addressing is safe >> + only when storing the numbers of the completed steps in files, whereas reverse >> + relative addressing it is safe in any case and can often be more convenient. > На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется > одинаковым. Хотя бы current-1 OK. Давай переделаю так: step<N> = pipe<N> -- прямая адресация, так первоначально и было. curr-<N> -- обратная относительная адресация. Так будет лучше? > >> - Allows you to work with shorter and more familiar paths to special files >> devices thanks to the use of `DEVNAME` along with `dev`. >> >> diff --git a/features/bootchain-core/data/bin/bootchain-sh-functions b/features/bootchain-core/data/bin/bootchain-sh-functions >> index e73fd35..0eebbaf 100644 >> --- a/features/bootchain-core/data/bin/bootchain-sh-functions >> +++ b/features/bootchain-core/data/bin/bootchain-sh-functions >> @@ -27,19 +27,46 @@ get_parameter() >> resolve_target() >> { >> local target="$1" >> + local where="${target%%/*}" >> + local n="${#where}" >> >> - case "${target%%/*}" in >> + target="${target:$n}" >> + >> + case "$where" in >> '') >> ;; >> pipe[0-9]|pipe[0-9][0-9]|pipe[0-9][0-9][0-9]) >> - target="$mntdir/dst/step${target:4}" >> + [ -d "$mntdir/dst/step${where:4}" ] || >> + return 0 >> + target="$(readlink-e "$mntdir/dst/step${where:4}")${target}" >> + ;; >> + step-[1-9]|step-[1-9][0-9]|step-[1-9][0-9][0-9]) >> + case "$destdir" in >> + "$mntdir"/dst/step[0-9]*) >> + n="${destdir##*/}" >> + n="$(( ${n:4} - ${where:5} ))" >> + ;; >> + *) >> + return 0 >> + ;; >> + esac >> + [ "$n" -ge 0 ] && [ -d "$mntdir/dst/step$n" ] || >> + return 0 >> + target="$(readlink-e "$mntdir/dst/step$n")${target}" >> ;; >> *) >> if [ -z "${prevdir-}" ]; then >> - message "no previous stop to use" >> - return >> + message "no previous step results to use with $name" >> + return 0 >> + fi >> + if [ -z "$target" ] && >> + [ "$where" = DEVNAME ] && >> + [ -s "$prevdir"/DEVNAME ] >> + then >> + read -r target <"$prevdir"/DEVNAME ||: >> + else >> + target="$prevdir/${where#/}${target}" >> fi >> - target="$prevdir/${target#/}" >> ;; >> esac >> >> -- >> 2.24.1 >> >> _______________________________________________ >> Make-initrd mailing list >> Make-initrd@lists.altlinux.org >> https://lists.altlinux.org/mailman/listinfo/make-initrd >> -- Best regards, Leonid Krivoshein.
On Tue, Oct 26, 2021 at 02:28:53PM +0300, Leonid Krivoshein wrote: > > > 26.10.2021 14:12, Alexey Gladkov пишет: > > On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote: > > > Adds support for reverse addressing relative to the current step, > > > as well as calculating the path to the special device node obtained > > > in the previous step. > > > > > > Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com> > > > --- > > > features/bootchain-core/README.md | 6 +++ > > > .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++--- > > > 2 files changed, 38 insertions(+), 5 deletions(-) > > > > > > diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md > > > index 8d341e0..ba66259 100644 > > > --- a/features/bootchain-core/README.md > > > +++ b/features/bootchain-core/README.md > > > @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need. > > > - Modularity: loading methods are initially separated from the common > > > code and daemon. > > > +- Via resolve_target() supports not only forward, but also reverse addressing, > > > + relative to the current step. For example, a record like `step-3/dir1/dev` > > > + will process the result of `dir1/dev`, made in the third step from the current > > > + one. Together with the overload of the chain of steps, direct addressing is safe > > > + only when storing the numbers of the completed steps in files, whereas reverse > > > + relative addressing it is safe in any case and can often be more convenient. > > На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется > > одинаковым. Хотя бы current-1 > > OK. Давай переделаю так: > > step<N> = pipe<N> -- прямая адресация, так первоначально и было. > curr-<N> -- обратная относительная адресация. > > Так будет лучше? Мне будет понятнее. Но я хотел бы спросить помощи зала. Глеб, Антон, как вы думаете, как будет понятнее ? > > > > > > - Allows you to work with shorter and more familiar paths to special files > > > devices thanks to the use of `DEVNAME` along with `dev`. > > > diff --git a/features/bootchain-core/data/bin/bootchain-sh-functions b/features/bootchain-core/data/bin/bootchain-sh-functions > > > index e73fd35..0eebbaf 100644 > > > --- a/features/bootchain-core/data/bin/bootchain-sh-functions > > > +++ b/features/bootchain-core/data/bin/bootchain-sh-functions > > > @@ -27,19 +27,46 @@ get_parameter() > > > resolve_target() > > > { > > > local target="$1" > > > + local where="${target%%/*}" > > > + local n="${#where}" > > > - case "${target%%/*}" in > > > + target="${target:$n}" > > > + > > > + case "$where" in > > > '') > > > ;; > > > pipe[0-9]|pipe[0-9][0-9]|pipe[0-9][0-9][0-9]) > > > - target="$mntdir/dst/step${target:4}" > > > + [ -d "$mntdir/dst/step${where:4}" ] || > > > + return 0 > > > + target="$(readlink-e "$mntdir/dst/step${where:4}")${target}" > > > + ;; > > > + step-[1-9]|step-[1-9][0-9]|step-[1-9][0-9][0-9]) > > > + case "$destdir" in > > > + "$mntdir"/dst/step[0-9]*) > > > + n="${destdir##*/}" > > > + n="$(( ${n:4} - ${where:5} ))" > > > + ;; > > > + *) > > > + return 0 > > > + ;; > > > + esac > > > + [ "$n" -ge 0 ] && [ -d "$mntdir/dst/step$n" ] || > > > + return 0 > > > + target="$(readlink-e "$mntdir/dst/step$n")${target}" > > > ;; > > > *) > > > if [ -z "${prevdir-}" ]; then > > > - message "no previous stop to use" > > > - return > > > + message "no previous step results to use with $name" > > > + return 0 > > > + fi > > > + if [ -z "$target" ] && > > > + [ "$where" = DEVNAME ] && > > > + [ -s "$prevdir"/DEVNAME ] > > > + then > > > + read -r target <"$prevdir"/DEVNAME ||: > > > + else > > > + target="$prevdir/${where#/}${target}" > > > fi > > > - target="$prevdir/${target#/}" > > > ;; > > > esac > > > -- > > > 2.24.1 > > > > > > _______________________________________________ > > > Make-initrd mailing list > > > Make-initrd@lists.altlinux.org > > > https://lists.altlinux.org/mailman/listinfo/make-initrd > > > > > -- > Best regards, > Leonid Krivoshein. > > _______________________________________________ > Make-initrd mailing list > Make-initrd@lists.altlinux.org > https://lists.altlinux.org/mailman/listinfo/make-initrd -- Rgrds, legion
[-- Attachment #1: Type: text/plain, Size: 2686 bytes --] On Tue, Oct 26, 2021 at 03:58:34PM +0200, Alexey Gladkov wrote: > On Tue, Oct 26, 2021 at 02:28:53PM +0300, Leonid Krivoshein wrote: > > > > > > 26.10.2021 14:12, Alexey Gladkov пишет: > > > On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote: > > > > Adds support for reverse addressing relative to the current step, > > > > as well as calculating the path to the special device node obtained > > > > in the previous step. > > > > > > > > Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com> > > > > --- > > > > features/bootchain-core/README.md | 6 +++ > > > > .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++--- > > > > 2 files changed, 38 insertions(+), 5 deletions(-) > > > > > > > > diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md > > > > index 8d341e0..ba66259 100644 > > > > --- a/features/bootchain-core/README.md > > > > +++ b/features/bootchain-core/README.md > > > > @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need. > > > > - Modularity: loading methods are initially separated from the common > > > > code and daemon. > > > > +- Via resolve_target() supports not only forward, but also reverse addressing, > > > > + relative to the current step. For example, a record like `step-3/dir1/dev` > > > > + will process the result of `dir1/dev`, made in the third step from the current > > > > + one. Together with the overload of the chain of steps, direct addressing is safe > > > > + only when storing the numbers of the completed steps in files, whereas reverse > > > > + relative addressing it is safe in any case and can often be more convenient. > > > На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется > > > одинаковым. Хотя бы current-1 > > > > OK. Давай переделаю так: > > > > step<N> = pipe<N> -- прямая адресация, так первоначально и было. > > curr-<N> -- обратная относительная адресация. > > > > Так будет лучше? > > Мне будет понятнее. Но я хотел бы спросить помощи зала. > > Глеб, Антон, как вы думаете, как будет понятнее ? Если из предложенного выбирать, то curr-<N> однозначно лучше. Но мне кажется, что тут - не читается как минус. Я бы скорее предложил что-то вроде prev<N> или back<N>. -- glebfm [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 801 bytes --]
30.10.2021 2:50, Gleb Fotengauer-Malinovskiy пишет:
> On Tue, Oct 26, 2021 at 03:58:34PM +0200, Alexey Gladkov wrote:
>> On Tue, Oct 26, 2021 at 02:28:53PM +0300, Leonid Krivoshein wrote:
>>> 26.10.2021 14:12, Alexey Gladkov пишет:
>>>> On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote:
>>>>> Adds support for reverse addressing relative to the current step,
>>>>> as well as calculating the path to the special device node obtained
>>>>> in the previous step.
>>>>>
>>>>> Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com>
>>>>> ---
>>>>> features/bootchain-core/README.md | 6 +++
>>>>> .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++---
>>>>> 2 files changed, 38 insertions(+), 5 deletions(-)
>>>>>
>>>>> diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md
>>>>> index 8d341e0..ba66259 100644
>>>>> --- a/features/bootchain-core/README.md
>>>>> +++ b/features/bootchain-core/README.md
>>>>> @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need.
>>>>> - Modularity: loading methods are initially separated from the common
>>>>> code and daemon.
>>>>> +- Via resolve_target() supports not only forward, but also reverse addressing,
>>>>> + relative to the current step. For example, a record like `step-3/dir1/dev`
>>>>> + will process the result of `dir1/dev`, made in the third step from the current
>>>>> + one. Together with the overload of the chain of steps, direct addressing is safe
>>>>> + only when storing the numbers of the completed steps in files, whereas reverse
>>>>> + relative addressing it is safe in any case and can often be more convenient.
>>>> На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется
>>>> одинаковым. Хотя бы current-1
>>> OK. Давай переделаю так:
>>>
>>> step<N> = pipe<N> -- прямая адресация, так первоначально и было.
>>> curr-<N> -- обратная относительная адресация.
>>>
>>> Так будет лучше?
>> Мне будет понятнее. Но я хотел бы спросить помощи зала.
>>
>> Глеб, Антон, как вы думаете, как будет понятнее ?
> Если из предложенного выбирать, то curr-<N> однозначно лучше.
>
> Но мне кажется, что тут - не читается как минус. Я бы скорее предложил
> что-то вроде prev<N> или back<N>.
Отлично, тогда переделаю на back<N> !
Как я понял, по остальным моментам Алексей пока взял паузу...
--
Best regards,
Leonid Krivoshein.
On Sat, Oct 30, 2021 at 05:13:49AM +0300, Leonid Krivoshein wrote:
>
> 30.10.2021 2:50, Gleb Fotengauer-Malinovskiy пишет:
> > On Tue, Oct 26, 2021 at 03:58:34PM +0200, Alexey Gladkov wrote:
> > > On Tue, Oct 26, 2021 at 02:28:53PM +0300, Leonid Krivoshein wrote:
> > > > 26.10.2021 14:12, Alexey Gladkov пишет:
> > > > > On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote:
> > > > > > Adds support for reverse addressing relative to the current step,
> > > > > > as well as calculating the path to the special device node obtained
> > > > > > in the previous step.
> > > > > >
> > > > > > Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com>
> > > > > > ---
> > > > > > features/bootchain-core/README.md | 6 +++
> > > > > > .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++---
> > > > > > 2 files changed, 38 insertions(+), 5 deletions(-)
> > > > > >
> > > > > > diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md
> > > > > > index 8d341e0..ba66259 100644
> > > > > > --- a/features/bootchain-core/README.md
> > > > > > +++ b/features/bootchain-core/README.md
> > > > > > @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need.
> > > > > > - Modularity: loading methods are initially separated from the common
> > > > > > code and daemon.
> > > > > > +- Via resolve_target() supports not only forward, but also reverse addressing,
> > > > > > + relative to the current step. For example, a record like `step-3/dir1/dev`
> > > > > > + will process the result of `dir1/dev`, made in the third step from the current
> > > > > > + one. Together with the overload of the chain of steps, direct addressing is safe
> > > > > > + only when storing the numbers of the completed steps in files, whereas reverse
> > > > > > + relative addressing it is safe in any case and can often be more convenient.
> > > > > На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется
> > > > > одинаковым. Хотя бы current-1
> > > > OK. Давай переделаю так:
> > > >
> > > > step<N> = pipe<N> -- прямая адресация, так первоначально и было.
> > > > curr-<N> -- обратная относительная адресация.
> > > >
> > > > Так будет лучше?
> > > Мне будет понятнее. Но я хотел бы спросить помощи зала.
> > >
> > > Глеб, Антон, как вы думаете, как будет понятнее ?
> > Если из предложенного выбирать, то curr-<N> однозначно лучше.
> >
> > Но мне кажется, что тут - не читается как минус. Я бы скорее предложил
> > что-то вроде prev<N> или back<N>.
>
> Отлично, тогда переделаю на back<N> !
>
> Как я понял, по остальным моментам Алексей пока взял паузу...
В прошлый раз я вроде всё прокомментировал, что нашёл.
--
Rgrds, legion
06.11.2021 15:47, Alexey Gladkov пишет:
> On Sat, Oct 30, 2021 at 05:13:49AM +0300, Leonid Krivoshein wrote:
>> 30.10.2021 2:50, Gleb Fotengauer-Malinovskiy пишет:
>>> On Tue, Oct 26, 2021 at 03:58:34PM +0200, Alexey Gladkov wrote:
>>>> On Tue, Oct 26, 2021 at 02:28:53PM +0300, Leonid Krivoshein wrote:
>>>>> 26.10.2021 14:12, Alexey Gladkov пишет:
>>>>>> On Sun, Oct 24, 2021 at 08:21:28PM +0300, Leonid Krivoshein wrote:
>>>>>>> Adds support for reverse addressing relative to the current step,
>>>>>>> as well as calculating the path to the special device node obtained
>>>>>>> in the previous step.
>>>>>>>
>>>>>>> Signed-off-by: Leonid Krivoshein <klark.devel@gmail.com>
>>>>>>> ---
>>>>>>> features/bootchain-core/README.md | 6 +++
>>>>>>> .../data/bin/bootchain-sh-functions | 37 ++++++++++++++++---
>>>>>>> 2 files changed, 38 insertions(+), 5 deletions(-)
>>>>>>>
>>>>>>> diff --git a/features/bootchain-core/README.md b/features/bootchain-core/README.md
>>>>>>> index 8d341e0..ba66259 100644
>>>>>>> --- a/features/bootchain-core/README.md
>>>>>>> +++ b/features/bootchain-core/README.md
>>>>>>> @@ -49,6 +49,12 @@ us to optimize fill in `initramfs` only which we are need.
>>>>>>> - Modularity: loading methods are initially separated from the common
>>>>>>> code and daemon.
>>>>>>> +- Via resolve_target() supports not only forward, but also reverse addressing,
>>>>>>> + relative to the current step. For example, a record like `step-3/dir1/dev`
>>>>>>> + will process the result of `dir1/dev`, made in the third step from the current
>>>>>>> + one. Together with the overload of the chain of steps, direct addressing is safe
>>>>>>> + only when storing the numbers of the completed steps in files, whereas reverse
>>>>>>> + relative addressing it is safe in any case and can often be more convenient.
>>>>>> На мой взгляд очень непонятный синтаксис. step1 и step-1 кажется
>>>>>> одинаковым. Хотя бы current-1
>>>>> OK. Давай переделаю так:
>>>>>
>>>>> step<N> = pipe<N> -- прямая адресация, так первоначально и было.
>>>>> curr-<N> -- обратная относительная адресация.
>>>>>
>>>>> Так будет лучше?
>>>> Мне будет понятнее. Но я хотел бы спросить помощи зала.
>>>>
>>>> Глеб, Антон, как вы думаете, как будет понятнее ?
>>> Если из предложенного выбирать, то curr-<N> однозначно лучше.
>>>
>>> Но мне кажется, что тут - не читается как минус. Я бы скорее предложил
>>> что-то вроде prev<N> или back<N>.
>> Отлично, тогда переделаю на back<N> !
>>
>> Как я понял, по остальным моментам Алексей пока взял паузу...
> В прошлый раз я вроде всё прокомментировал, что нашёл.
Тогда перехожу к финальной итерации? Просто, мне показалось, что
остались какие-то спорные моменты.
--
Best regards,
Leonid Krivoshein.