Re: [boost] boost cm practices

From: Beman Dawes (bdawes_at_[hidden])
Date: 2001-10-30 19:25:31

At 02:14 PM 10/30/2001, mda_at_[hidden] wrote:

 >all i know is that as i type this, http://www.boost.org/boost_all.zip
 >is 4.3mB as a zip file, expanding to over 12mB, and that the
 >top pigs are such things as:
 >libs/thread/test/bin/test_thread/msvc/debug/runtime-link-
 >dynamic/threading-multi:
 >-rwxr-xr-x    114815 Oct  1 14:14 test_thread.exe
 >-rw-r--r--    236804 Oct  1 14:14 test_thread.ilk
 >-rw-r--r--    211142 Oct  1 14:14 test_thread.obj
 >-rw-r--r--    476160 Oct  1 14:14 test_thread.pdb

That was a mistake on my part.  Forgot to run clean after running a thread 
test.

 >it would be nice if:
 >1. a release were identical to a snapshot of the CVS at some time,
 >sans the CVS cruft, without someone's personal build artifacts.

Yes, that would be useful.  If you would like to submit a script (Bash, 
Python, or Win2K batch file) that does that, I'll add it to the release 
procedure.

 >2. the release history were documented, with a date, scm label,
 >human-generated release notes, and (ideally) automatically generated
 >change summary

Can't you get that from CVS?  If you have a format that is better than the 
CVS output, consider submitting a script to generate it.

 >3. someone documented whether tar.gz and .zip releases are identical,
 >or line endings are different

That's already documented on the download page.

 >4. the hosted documentation files on boost.org showed a version id

That would require changing each page for each release, although it might 
be possible to do it in a single place with a style sheet or some other 
trick.  Why don't you put together a specific proposal, with sample HTML 
code?

 >5. the hosted documentation and source files matched the latest cvs.
 >(For example, http://www.boost.org/boost/array.hpp pointed to from
 >http://www.boost.org/libs/array/index.htm does not match cvs of
 >even several months ago.)

Consider writing a script to perform that error check.  It would be a nice 
addition to the release procedure.

There is also the similar problem of pages on the web site never being 
updated.

Note that some of the problems would go away if we used a different release 
mechanism. Perhaps more like the SourceForge model. OTOH, the current 
release procedure does regularly detect various other types of errors.

 >right now, afaik, none of those are true.
 >
 >-mda, releasing some pent-up annoyance (sorry)

Don't be annoyed - write some scripts to fix the problems!

--Beman