Re: scsh.net broken link(s)
Nunez Steve <[email protected]> Sun, 8 Mar 2009 21:02:05 -0700 (PDT)
| Newsgroups | gmane.lisp.scheme.scsh |
|---|---|
| Message-ID | <[email protected]> |
--0-1734661576-1236571325=:51758 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Being new here, I don't have but in reference to low-level internals:=0A=0A= > 2. Extend scheme48 to include everything scheme-with-scsh needs to functi= on?=0A=0ACertainly seems like the best option. Sort of killing two birds wi= th one stone, and you may get help from the Scheme people. (?)=0A=0ARegards= ,=0A - SteveN=0A=0A=0A=0A________________________________=0AFrom: "sand= @blarg.net" <[email protected]>=0ATo: Michael Sperber <[email protected]= >=0ACc: [email protected]; Scsh Users <[email protected]>=0ASent: Monday, Ma= rch 9, 2009 7:43:55 AM=0ASubject: Re: scsh.net broken link(s)=0A=0AMichael = Sperber writes:=0A> [email protected] writes:=0A> =0A> > Or, more specifically= , are there any plans to port SCSH from its=0A> > current Scheme 48 impleme= ntation to something more recent? Version=0A> > 1.8 released a month ago a= dvertises 64-bit support.=0A> =0A> Plans yes, but we're short some manpower= . Help would be quite welcome!=0A> =0A=0AWhere should the line of division= between the two be? From chapter 9=0Aof the manual:=0A=0A> Scheme 48=E2= =80=99s POSIX interface differs from Scsh=E2=80=99s [11, 12] in several=0A>= ways. The interface here lacks Scsh=E2=80=99s high-level constructs and=0A= > utilities, such as the process notation, awk procedure, and parsing=0A> u= tilities. Scheme 48 uses distinct types for some values that Scsh=0A> leave= s as symbols or unboxed integers; these include file types, file=0A> modes,= and user and group ids. Many of the names and other interface=0A> details = are different, as well.=0A=0AThe high-level stuff should be straightforward= to port. For the=0Alow-level external interface, do we:=0A=0A1. Keep the = scheme-with-scsh interface identical to what we have now?=0A=0A2. Keep the = scheme-with-scsh interface identical to what we have now,=0Ato support back= wards-compatibility, but implement it on top of=0Aanother "native" interfac= e (e.g., package scheme48-with-scsh)?=0A=0A3. Rework the scheme-with-scsh i= nterface from 0.6.7 to 1.8, to use=0Athe 1.8 types and functions as much as= possible (so only=0Ascheme48-with-scsh exists)?=0A=0AFor the low-level int= ernals, do we:=0A=0A1. Put into scheme-with-scsh everything not already imp= lemented in=0Ascheme48?=0A=0A2. Extend scheme48 to include everything schem= e-with-scsh needs to=0Afunction?=0A=0A> Scheme 48=E2=80=99s POSIX interface= will likely change significantly in the=0A> future. The implementation is = new and may have significant bugs.=0A=0AIs this still true?=0A=0ADerek=0A= =0A-- =0ADerek [email protected]=0A=0A=0A --0-1734661576-1236571325=:51758 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable <html><head><style type=3D"text/css"><!-- DIV {margin:0px;} --></style></he= ad><body><div style=3D"font-family:arial, helvetica, sans-serif;font-size:1= 0pt"><div>Being new here, I don't have but in reference to low-level intern= als:</div><div><br></div><div>>=C2=A02. Extend scheme48 to include everythi= ng scheme-with-scsh needs to=C2=A0function?</div><div><br></div><div>Certai= nly seems like the best option. Sort of killing two birds with one stone, a= nd you may get help from the Scheme people. (?)</div><div><br></div><div>Re= gards,</div><div>=C2=A0=C2=A0 =C2=A0- =C2=A0SteveN<div style=3D"font-family= :arial, helvetica, sans-serif;font-size:10pt"><br><div style=3D"font-family= :arial, helvetica, sans-serif;font-size:13px"><font size=3D"2" face=3D"Taho= ma"><hr size=3D"1"><b><span style=3D"font-weight: bold;">From:</span></b> "= [email protected]" <[email protected]><br><b><span style=3D"font-weight: bold;= ">To:</span></b> Michael Sperber <[email protected]><br><b><span s= tyle=3D"font-weight: bold;">Cc:</span></b> [email protected]; Scsh Users <[email protected]><= br><b><span style=3D"font-weight: bold;">Sent:</span></b> Monday, March 9, = 2009 7:43:55 AM<br><b><span style=3D"font-weight: bold;">Subject:</span></b= > Re: scsh.net broken link(s)<br></font><br>=0AMichael Sperber writes:<br>>= <a ymailto=3D"mailto:[email protected]" href=3D"mailto:[email protected]">sand@b= larg.net</a> writes:<br>> <br>> > Or, more specifically, are there any plan= s to port SCSH from its<br>> > current Scheme 48 implementation to somethin= g more recent?=C2=A0 Version<br>> > 1.8 released a month ago advertises 64-= bit support.<br>> <br>> Plans yes, but we're short some manpower.=C2=A0 Hel= p would be quite welcome!<br>> <br><br>Where should the line of division be= tween the two be?=C2=A0 From chapter 9<br>of the manual:<br><br>> Scheme 48= =E2=80=99s POSIX interface differs from Scsh=E2=80=99s [11, 12] in several<= br>> ways. The interface here lacks Scsh=E2=80=99s high-level constructs an= d<br>> utilities, such as the process notation, awk procedure, and parsing<= br>> utilities. Scheme 48 uses distinct types for some values that Scsh<br>= > leaves as symbols or unboxed integers; these include file types, file<br>= > modes, and user and group ids. Many of the names and other interface<br>> details are different, as well.<br><br>The high-level stuff= should be straightforward to port.=C2=A0 For the<br>low-level external int= erface, do we:<br><br> 1. Keep the scheme-with-scsh interface identical to = what we have now?<br><br> 2. Keep the scheme-with-scsh interface identical = to what we have now,<br> to support backwards-compatibility, but implement = it on top of<br> another "native" interface (e.g., package scheme48-with-sc= sh)?<br><br> 3. Rework the scheme-with-scsh interface from 0.6.7 to 1.8, to= use<br> the 1.8 types and functions as much as possible (so only<br> schem= e48-with-scsh exists)?<br><br>For the low-level internals, do we:<br><br> 1= . Put into scheme-with-scsh everything not already implemented in<br> schem= e48?<br><br> 2. Extend scheme48 to include everything scheme-with-scsh need= s to<br> function?<br><br>> Scheme 48=E2=80=99s POSIX interface will likely= change significantly in the<br>> future. The implementation is new and may have significant bugs.<br><br>Is this still true?<br><br>Derek<br><br>-- <= br>Derek Upham<br><a ymailto=3D"mailto:[email protected]" href=3D"mailto:sand@= blarg.net">[email protected]</a><br><br></div></div><div style=3D"position:fix= ed"></div></div></div><br>=0A=0A </body></html> --0-1734661576-1236571325=:51758--