[Catalog-sig] Package comments

P.J. Eby pje at telecommunity.com
Wed Nov 4 01:39:45 CET 2009

At 10:49 PM 11/3/2009 +0100, Martin v. Löwis wrote:
>Telling them to use the tracker is, IMO, depriving them of
>their right to explain their evaluation of the package; a bug tracker is
>not an adequate means for that.

No, but their personal blog(s) are a wonderfully available and 
appropriate place for them to do so...  assuming that one believes 
that this "right" exists in the first place.

After all, if the general public has a "right" to make such comments, 
then the download page of Python.org should allow similarly 
unmoderated comments, and disallow any Python developer from deleting 
those comments.   (That would be "censorship", after 
all.)  Microsoft's Windows Update webpage should likewise be open to 
comment, as should Apple's iTunes page.  That would really help other 
users!  (Not.)

Yes, let's all have a right to comment on every web page, just 
because the creator of that web page has chosen to offer some 
software for us to use.

The idea of a "right" to comment on a software package's download 
page is a *really* bad idea, and it deserves much worse ridicule than 
it's currently receiving in this discussion.

Cleaning off graffiti is not censorship.

If you want to provide a comment process, why not include a link to 
search Google for backlinks to the package page, and let people write 
blog posts about the package that way?  Then the authors and users 
alike can search to see what's being said about the package.

Then, if you really feel that PyPI must provide free hosting for 
trolls, put the comments on a web page somewhere with a backlink, and 
let them compete for relevance with those who care enough about what 
they have to say to write something more thoughtful and relevant than 
"this sucks! boo!" or "it's awesome! yay!".

IMO, it should be easier to find thoughtful commentary and experience 
reports about a package, than to read a list of low-quality, 1-or-2 
line graffiti scrawls(which is what the current system demonstrably 

More information about the Catalog-SIG mailing list