headers
David Lichteblau | 2 Jul 2002 04:50
Gravatar

ASDF system for portable allegroserve on CMUCL

[cc'ed to the cclan people, please respond to the appropriate address]

I am too stupid to compile paserve using defsystem on CMUCL, so I have
made ASDF systems replacing the old CMUCL systems which essentially do
the same thing but happen to work.  (Sorry -- I'm too dumb and lazy to
figure out what the defsystem problems are about, but interested in
having a ASDF version anyway...)

(The attached patch also changes client.cl to use HTTP/1.0 instead of
1.1 as a default on CMUCL, since 1.1 requires chunked encoding, which is
not yet supported on CMUCL.)

Perhaps someone finds this useful,
David
--

-- 
FAILURE IS NOT AN OPTION - it comes bundled with the software.
Attachment (portableaserve-asdf-cmu.diff.bz2): application/octet-stream, 3961 bytes
Permalink | Reply | 4 comments |
headers
David Lichteblau | 2 Jul 2002 10:03
Gravatar

cclan-get: fetch and install vn-cclan packages

* http://www.lichteblau.com/code/cclan-get/
| 
| cclan-get fetches and installs vn-cclan packages from local or remote
| repositories.
| 
|   * (describe 'cclan-get)
|   CCLAN-GET is an external symbol in the CCLAN-GET package.
|   Function: #<Function CCLAN-GET {486D7231}>
|   Function arguments:
|     (name &key (operation 'load-op) (repositories *cclan-repositories*))
|   Function documentation:
|     Find ASDF system NAME.  If it is not present yet, find it in one of
|   the REPOSITORIES, download it to *CCLAN-TARBALL-DIRECTORY* (if needed),
|   extract it to *CCLAN-SOURCE-DIRECTORY*, and register the ASDF system in
|   *CCLAN-ASDF-REGISTRY*.  Recursively install all dependencies in the
|   same way.  Finally perform OPERATION.  Return the system.

Have fun,
David
--

-- 
FAILURE IS NOT AN OPTION - it comes bundled with the software.

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
Permalink | Reply | 1 comment |
headers
Daniel Barlow | 2 Jul 2002 12:57

Re: cclan-get: fetch and install vn-cclan packages

David Lichteblau <david <at> lichteblau.com> writes:

> | cclan-get fetches and installs vn-cclan packages from local or remote
> | repositories.

(using portableaserve, for anyone who was wondering and hadn't checked
the URL)

Very neat.  I'm not sure what you're doing with dependencies, though.
Personally I would do something like (handwavey pseudocode follows)

(defun get-system (name operation)
   (download-and-copy-system-somewhere name)
   (handler-bind
      ((asdf:missing-component (c)
           (when (null (asdf:missing-parent c))
             ;; nil parent indicates the component is a system
             (get-system (asdf:missing-requires c))
             (invoke-restart 'retry-component))))
      (asdf:operate operation name)))

i.e. let asdf flag up the missing dependencies as and when it finds them.

fwiw, I'm happy with your "conventions", because that's the format
that my cvs-tag and make-tar-file functions (in cclan.lisp, in the
same place as asdf) already use :-)

As to the more general issue: how much of a problem is it really to
mandate that the tarball name is the same as (or mechanically
derivable from) the asdf system name?  I've been saying all along that
(Continue reading)

Permalink | Reply | 0 comments |
headers
Rudi Schlatte | 3 Jul 2002 14:49
Picon
Gravatar

Re: [Portableaserve-discuss] ASDF system for portable allegroserve on CMUCL


David Lichteblau <david <at> lichteblau.com> writes:

> I am too stupid to compile paserve using defsystem on CMUCL, so I have
> made ASDF systems replacing the old CMUCL systems which essentially do
> the same thing but happen to work.  (Sorry -- I'm too dumb and lazy to
> figure out what the defsystem problems are about, but interested in
> having a ASDF version anyway...)

Thanks, so am I.  I'll check the asdf version in this weekend at the
latest.  It is a good idea to support all the different system
definition utilities that are utilized by Debian|(vn-)?cCLan -- at
least, until the Grand Unified Defsystem emerges from the fray...

Re defsystem: could you send a backtrace of what happens when you
evaluate (load "INSTALL.lisp")?  (If you're at all interested, that is.)

> (The attached patch also changes client.cl to use HTTP/1.0 instead of
> 1.1 as a default on CMUCL, since 1.1 requires chunked encoding, which is
> not yet supported on CMUCL.)

I'll have to think about that -- not that it's a bad idea to be
standards conformant, but now that Gray streams work on cmucl,
chunking is not so much work anymore, and I'd have to remember
removing the patch afterwards :)

As an aside: I noticed that vn-cclan has started using portableaserve
for package downloading.  From my POV, this is a Good Thing, of
course :)  Do you need multithreading for that?  Otherwise, I could
whip up a port of (single-threaded) paserve to SBCL on a rainy
(Continue reading)

Permalink | Reply | 3 comments |
headers
John DeSoi | 4 Jul 2002 03:41
Picon

Re: [Portableaserve-discuss] ASDF system for portable allegroserve on CMUCL


>
>>  I am too stupid to compile paserve using defsystem on CMUCL, so I have
>>  made ASDF systems replacing the old CMUCL systems which essentially do
>>  the same thing but happen to work.  (Sorry -- I'm too dumb and lazy to
>>  figure out what the defsystem problems are about, but interested in
>>  having a ASDF version anyway...)
>
>Thanks, so am I.  I'll check the asdf version in this weekend at the
>latest.  It is a good idea to support all the different system
>definition utilities that are utilized by Debian|(vn-)?cCLan -- at
>least, until the Grand Unified Defsystem emerges from the fray...
>

I plan on creating ASDF definitions for MCL and OpenMCL also.

Since ASDF is a single small file, I would also like to propose 
including it the distribution (it can just be ignored if a local 
version is loaded). I would really like to see Portable AllegroServe 
load and run without having to worry about finding and loading other 
software.

Best,

John DeSoi, Ph.D.

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
No, I will not fix your computer.
http://thinkgeek.com/sf
(Continue reading)

Permalink | Reply | 2 comments |
headers
Daniel Barlow | 4 Jul 2002 04:35

Re: Re: [Portableaserve-discuss] ASDF system for portable allegroserve on CMUCL


What an exciting collection of subject taglines we're going to collect
here ...

John DeSoi <desoi <at> mac.com> writes:
> Since ASDF is a single small file, I would also like to propose
> including it the distribution (it can just be ignored if a local

Wearing my asdf maintainer hat I'd be happy to see this happen, but
I am reluctant to end up in the position that mk-defsystem has been in
at times, with several out-of-sync versions of it floating around
different parts of the net.  So, all I'll say is: please feed back 
bug reports instead of hacking locally, and if you must hack locally
(e.g. because I am being a lazy slacker and not responding to email)
do something to the file to make it obvious it's changed.  

Of course, if you want to hack locally and send me nice diffs that I
can merge, I'll be delighted.  I'm just concerned about distributed
divergent versions.

-dan

--

-- 

  http://ww.telent.net/cliki/ - Link farm for free CL-on-Unix resources 

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
No, I will not fix your computer.
http://thinkgeek.com/sf
(Continue reading)

Permalink | Reply | 1 comment |
headers
Jochen Schmidt | 4 Jul 2002 05:27
Picon

Re: Re: [Portableaserve-discuss] ASDF system for portable allegroserve on CMUCL

Am Don, 2002-07-04 um 04.35 schrieb Daniel Barlow:
> 
> What an exciting collection of subject taglines we're going to collect
> here ...
> 
> John DeSoi <desoi <at> mac.com> writes:
> > Since ASDF is a single small file, I would also like to propose
> > including it the distribution (it can just be ignored if a local
> 
> Wearing my asdf maintainer hat I'd be happy to see this happen, but
> I am reluctant to end up in the position that mk-defsystem has been in
> at times, with several out-of-sync versions of it floating around
> different parts of the net.  So, all I'll say is: please feed back 
> bug reports instead of hacking locally, and if you must hack locally
> (e.g. because I am being a lazy slacker and not responding to email)
> do something to the file to make it obvious it's changed.  
> 
> Of course, if you want to hack locally and send me nice diffs that I
> can merge, I'll be delighted.  I'm just concerned about distributed
> divergent versions.

Ok - I think we will then include ASDF into the PAServe CVS.

If we do that we could bite the bullet and switch to only ASDF as
Defsystem facility (leaving out MK:Defsystem and LW-Defsystem). 

I personally do not see any alternative to ASDF. MK:Defsystem 4 looks
much to complicated to me (I prefer a single-file Defsystem facility).

To the question where to add it: I would propose to create a "contrib"
(Continue reading)

Permalink | Reply | 0 comments |
headers
Daniel Barlow | 19 Jul 2002 17:35

[Daniel Barlow <dan <at> telent.net>] [Daniel Barlow <dan <at> telent.net>] Re: Re: [Portableaserve-discuss] ASDF system for portable allegroserve on CMUCL


[ week-old mail that got bounced by the smtp filter at LSM.  Sorry ]

I replied on this subject (at more length than I'm about to write
here) yesterday morning, but my laptop crashed before I could get
internet access to send it :-(

David Lichteblau <david <at> lichteblau.com> writes:

> Are compilers obliged to signal an error if there are errors in the
> sources?  

No.  According to 3.2.5, COMPILE-FILE may or may not catch errors
signalled during the compilation process.  And for your real-life
example, CMUCL does, SBCL doesn't.

There are a couple of points worth making in connection with asdf.
Neither solves your problem, but either may at least ameliorate it
slightly.

1) you can set ``:on-failure :warn'' on individual components, instead
   of having to use the same *compile-file-failure-behaviour* across
   the whole system.  This will work with CMUCL because it stops on
   compile errors anyway.  It doesn't help with SBCL

2) for interactive development, the (badly named, I now realise)
   SKIP-COMPONENT restart can be used to carry on after an error that
   the user knows was not important.

For a proper fix there are three things I'd like to look at
(Continue reading)

Permalink | Reply | 0 comments |
headers
Stig E. Sandoe | 19 Jul 2002 18:09
Picon
Picon

cCLan-test, please ignore


testing news-gateway, please bear with me. 

--

-- 
------------------------------------------------------------------
Stig E Sandoe       stig <at> ii.uib.no     http://www.ii.uib.no/~stig/

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
Permalink | Reply | 0 comments |

Gmane