Lex Spoon wrote:

>Bert Freudenberg <bert at impara.de> wrote:
>  
>
>>This is the developer's site where you get the latest and greatest. 
>>Recently moved to http://squeak.hpl.hp.com/unix/
>>    
>>
>
>In practice, that site is not actually updated so much, for
>whatever reason.  I actually point people to the download page on the
>Swiki, which does get updated and has more info.  The first link goes to
>this page, so in a sense it is strictly a superset of the information. 
>As a particular example, Debian users definitely want to go to the swiki
>site.  The site at the above URL has had obsolete packages for over a
>year.  The Swiki page is at:
>
>	"Download for Unix"
>	http://minnow.cc.gatech.edu/squeak/298
>
>Whatever else you say, the hp.com site is certainly not a "master" site
>in the sense of giving you full information.  Just look at the swiki
>page, and then look at the hp.com site.  The Swiki page has a great deal of
>information that is not on the hp.com site, and the hp.com site doesn't
>even link to the swiki at all.  Thus, the hp.com site is not a page you should start
>at, if you want to learn about Squeak on Unix in general.  It's a place
>to find one person's work.  At the very least, if this one person is to
>be officially blessed, then let's be honest about what we are doing and
>stop calling it a "master" site.
>
>Aside from that, this site has a long history of not reflecting
>community contributions.  Updated Debian packages have been around for
>well over a year now.  Further, there was a multi-year period when I
>collected people's bug fixes on a separate site because people kept
>reporting these bugs on the mailing list.  Some of these were serious
>bugs -- things like "I delete a morph and my image hangs", or "my Swiki
>won't start at all", or "Squeak doesn't start at all".  These only got
>incorporated when a second distribution, called SHA, was on the verge of
>becoming the de facto standard.  At the time, there was a suggestion
>that things would change, and the site would be more responsive, and
>thus everyone dumped SHA in favor of going with the above distribution. 
>Now that time has passed, things seem about the same as they were before
>(albeit, with those important bug fixes now included).
>
>Nowadays, as before, when I email the address listed on the page, I have
>about a 50% chance of getting any response at all.
>
>Thus for multiple reasons, this does not seem like it should be the
>official Unix VM distribution site.  It does not have complete
>information, it will leave many users without what they need, and its
>maintainer has a long history of not coordinating the sub-community that
>runs Squeak on Unix.  While this site is notable, and people should know
>about it, people should also know what else is going on with Squeak on
>Unix.  We owe it to our new users not to give them broken code and
>incomplete information if we can avoid it.
>
>
>To be clear, I would also be happy to see some alternative resolution
>where that site gets updated more often.  For example, if more
>maintainers were added, so that the site could get updated more
>responsively, that would be great.  I can nominate people, but just go
>look who has contributed to the old SourceForge CVS repository to get
>some ideas.  I am not hopeful about this idea, though.  I suggest that a
>well-meaning promise of change should not be enough to keep this as the
>master site, given that we have already had a claim that things will
>change.  We should insist on some sort of structural change that could
>plausibly make the site become a good gathering ground for the
>Squeak-on-Unix Squeakers.
>
>But really, changing that site's organization seems not only unlikely,
>but overkill.  We already have a way to do community web pages: the
>swiki.  It seems by far the simplest thing, to post Unix download
>information on the swiki.  It's what it is there for.  Further, the
>swiki page provides a great forum for Unix Squeakers to work out these
>issues among ourselves.  It's a simple and effective way to give
>this part of the community a voice.
>
>
>I certainly agree that we should give recognition where it is due.  Ian
>Piumarta's original Squeak port to Unix has served many of us very well,
>and should rightfully be mentioned in any early history of Squeak.  It's
>really cool that Squeak ports started appearing just weeks after the
>Mac-only first version of Squeak was posted on the web.  Further, Ian's
>VM is the one that most Unix Squeakers use by far.  No matter how the
>Squeak web site ends up being organized, people should be made aware of
>this excellent resource.
>
>
>Nevertheless, we should think about today, not just history.  Surely
>www.squeak.org should point users in the direction that is most likely
>to be helpful for them.  Also, surely our organization should seek ways
>to leverage people's contributions -- and should carefully react
>whenever contributions are being blocked.  Finally, porting Squeak is
>*by design* not a major effort--part of the coolness of the above story,
>is that it did *not* require a huge amount of time to port it.  How much
>should we defer to someone who does an initial port to a platform?
>
>
>You want to grab the bull by the horns, guys?  Here is a chance to take
>on a difficult leadership issue.
>
>-Lex
>  
>
Thanks for the information, Lex. It is helpful to see the history and I 
agree with you that there should be one place to dnl the stable 
release(s).  If that initial location takes you to another site, I think 
that is less than desirable, but sometimes it's preferable because of 
complexity (for instance, *nix distributions are many so it *might not* 
be best to have a long list of platforms on the initial web page) 
However, the page you mentioned:

http://minnow.cc.gatech.edu/squeak/298

is, in my opinion, in need of cleanup. But, that's a minor point.

But, more importantly, there is a clear misunderstanding among the community as to where exactly is the best place to dnl *nix releases! Perhaps the squeaders can place this issue on their ToDo list. (squeaders or castaways? Hmmm... maybe squealers? <-- I say that in the "laughing with you" way)

Oh... and I would like to respond to Bruce:


>It is up to the VM maintainers to release VMs.  I don't think it is a good
>idea to hold up everyone just because someone is busy.

why? If the community wants to push forward solid releases, why should Linux be left behind? 
There are two issues here:  
- What are the major platforms
 and 
- Should releases (VMs and images) be brought to "GM/stable" on all major platforms simultaneously (maybe final releases are called something else, I'll look later -- you get my drift, though.)

I contend that the major platforms are: Mac, Windows and Linux.
No question on Mac and Windows -- anyone? Linux is becoming more and more popular by the end user and especially by developers. Why should it not be included?

Secondly, I think it would be an encouraging process change to add a rule that mandates (tough word to use in a open source community, I know) all major platform releases be released simultaneously. It makes sure that no user/developer on major platforms are left out to dry waiting.

I bet that I'm the only one here that believes that this is important -- so go ahead, I'll take my lumps now....

brad