Re: A less invasive rcs
David Levner <[email protected]> Wed, 24 Feb 2021 14:48:00 +0000 (UTC)
| Newsgroups | gmane.comp.gnu.rcs |
|---|---|
| Message-ID | <[email protected]> |
------=_Part_2044709_975689792.1614178080532
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Another way to implement John's .rcs sub-tree suggestion is bywriting wrapp=
er=C2=A0functions that invoke the rcs executables. I havealready written wr=
apper=C2=A0functions that provide some functionalitythat RCS lacks, such as=
=C2=A0invoking a text editor to edit check-incomments. (My wrapper=C2=A0fun=
ctions run under a custom shell writtenin Perl. The functions=C2=A0are Perl=
subroutines that can be invoked ascommands from the shell's=C2=A0command l=
ine.)
Building on the wrapper functions that I currently have, I couldimplement J=
ohn's .rcs idea in a couple of days (one day if I gotlucky).
My concern about the .rcs subtree is that I sometimes make changesto my sou=
rce directory structure. For example, I might create a filelib/Something.pm=
and later decide to create a lib/Some directoryand rename the file to lib/=
Some/Thing.pm. With the RCS directoriesthat I currently use or a .rcs subtr=
ee, manual changes are needed.This could=C2=A0be addressed by making wrappe=
rs for mv and mkdir that areRCS-aware. I have not done this--perhaps it is =
not worth the extracomplexity. Perhaps there is a better solution.
Another thought is to make .rcs into a small configuration file. If itconta=
ined the name of the root of the RCS sub-tree, then theRCS sub-tree could g=
o anywhere.
John Yates and Dario Niedermann wrote:
>I propose a third location, a mirror of that working directory
>located by searching upward, through parent directories, for
>a "dominating" .rcs directory.
Personally, I like to have the RCS directory in foo/ to be a symlink=20
to /usr/local/share/RCS/foo; and the RCS directory in foo/bar/ as a=20
symlink to ../RCS/bar .
Doesn't this look like what you're trying to achieve? =20
------=_Part_2044709_975689792.1614178080532
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
<html><head></head><body><div class=3D"ydp8f4ccafbyahoo-style-wrap" style=
=3D"font-family: Helvetica Neue, Helvetica, Arial, sans-serif; font-size: 1=
6px;"><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2">An=
other way to implement John's .rcs sub-tree suggestion is by</font></div><d=
iv dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2">writing wr=
apper </font><span style=3D"font-size: small;">functions that invoke t=
he rcs executables. I have</span></div><div dir=3D"ltr" data-setdir=3D"fals=
e" style=3D""><span style=3D"font-size: small;">already written wrapper&nbs=
p;</span><span style=3D"font-size: small;">functions that provide some func=
tionality</span></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><sp=
an style=3D"font-size: small;">that RCS lacks, such as </span><span st=
yle=3D"font-size: small;">invoking a text editor to edit check-in</span></d=
iv><div dir=3D"ltr" data-setdir=3D"false" style=3D""><span style=3D"font-si=
ze: small;">comments. (My wrapper </span><span style=3D"font-size: sma=
ll;">functions run under a custom shell written</span></div><div dir=3D"ltr=
" data-setdir=3D"false" style=3D""><span style=3D"font-size: small;">in Per=
l. The functions </span><span style=3D"font-size: small;">are Perl sub=
routines that can be invoked as</span></div><div dir=3D"ltr" data-setdir=3D=
"false" style=3D""><span style=3D"font-size: small;">commands from the shel=
l's </span><span style=3D"font-size: small;">command line.)</span></di=
v><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2"><br></=
font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"=
2">Building on the wrapper functions that I currently have, I could</font><=
/div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2">imp=
lement John's .rcs idea in a couple of days (one day if I got</font></div><=
div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2">lucky).</=
font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"=
2"><br></font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font=
size=3D"2">My concern about the .rcs subtree is that I sometimes make chan=
ges</font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font siz=
e=3D"2">to my source directory structure. For example, I might create a fil=
e</font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=
=3D"2">lib/Something.pm and later decide to create a lib/Some directory</fo=
nt></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2"=
>and rename the file to lib/Some/Thing.pm. With the RCS directories</font><=
/div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2">tha=
t I currently use or a .rcs subtree, manual changes are needed.</font></div=
><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font size=3D"2">This co=
uld </font><span style=3D"font-size: small;">be addressed by making wr=
appers for mv and mkdir that are</span></div><div dir=3D"ltr" data-setdir=
=3D"false" style=3D""><font size=3D"2">RCS-aware. I have not done this--per=
haps it is not worth the extra</font></div><div dir=3D"ltr" data-setdir=3D"=
false" style=3D""><font size=3D"2">complexity. Perhaps there is a better so=
lution.</font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><font=
size=3D"2"><br></font></div><div dir=3D"ltr" data-setdir=3D"false" style=
=3D""><font size=3D"2">Another thought is to make .rcs into a small configu=
ration file. If it</font></div><div dir=3D"ltr" data-setdir=3D"false" style=
=3D""><font size=3D"2">contained the name of the root of the RCS sub-tree, =
then the</font></div><div dir=3D"ltr" data-setdir=3D"false" style=3D""><fon=
t size=3D"2">RCS sub-tree could go anywhere.</font></div><div dir=3D"ltr" d=
ata-setdir=3D"false" style=3D""><font size=3D"2"><br></font></div></div><di=
v id=3D"ydpb31a8928yahoo_quoted_4833625685" class=3D"ydpb31a8928yahoo_quote=
d"><div style=3D"font-family:'Helvetica Neue', Helvetica, Arial, sans-serif=
;font-size:13px;color:#26282a;"><div><div dir=3D"ltr" data-setdir=3D"false"=
>John Yates and Dario Niedermann wrote:</div><div dir=3D"ltr" data-setdir=
=3D"false"><br></div><div dir=3D"ltr">>I propose a third location, a mir=
ror of that working directory<br></div><div dir=3D"ltr">>located by sear=
ching upward, through parent directories, for<br></div><div dir=3D"ltr">>=
;a "dominating" .rcs directory.<br></div><div dir=3D"ltr"><br></div><div di=
r=3D"ltr">Personally, I like to have the RCS directory in foo/ to be a syml=
ink <br></div><div dir=3D"ltr">to /usr/local/share/RCS/foo; and the RCS dir=
ectory in foo/bar/ as a <br></div><div dir=3D"ltr">symlink to ../RCS/bar .<=
br></div><div dir=3D"ltr"><br></div><div dir=3D"ltr">Doesn't this look like=
what you're trying to achieve?</div></div>
</div>
</div></body></html>
------=_Part_2044709_975689792.1614178080532--