Squeak 3.7, SM2, TFNR etc (was Re: [UPDATES] 34 for Squeak3.7alpha)

goran.krampe at bluefish.se goran.krampe at bluefish.se
Sat Jan 17 11:40:37 UTC 2004


Hi all!

dway at mailcan.com wrote:
> Latest approved items.
> 
> By the way, 3.7 is tentatively scheduled to move to beta on February 7. 
> (Going by a 3.7 final date of April 3.)  So, we're getting toward the
> end of the 3.7 alpha stage.
> 
> - Doug

Ok, just wanted to write down a bit of stuff so that you all know what I
am doing etc.


SM2
====

Yesterday I sat down and wrote the proper loadscript for SM2 that seems to work in 3.5 (with or without SM2), 3.6 and 3.7alpha (as of yesterday).

It uses the bootstrap that I put in there a while back so that whenever someone tries to load updates using SM1 it will discover that the server is running SM2 and ask the user if he/she wants to upgrade - and then it upgrades automatically. This will also install MCInstaller and VersionNumber if those packages aren't installed.

Also - if someone using 3.5 tries to install SM it will install SM2 - because the bootstrap asks the SM server for the URL to the package "SqueakMap" and I intend to rename "SqueakMap2" to "SqueakMap" and rename the current "SqueakMap" to "SqueakMapOld" or something. In fact, with a bit of luck this might even work for 3.4 using the old script posted on the Swiki - need to look into that.

Today I am sitting down to start the actual deployment. We will see, but with a bit of luck we might be going over to SM2 today or tomorrow! I will be on #squeak on IRC while I work btw.


TFNR
======

TFNR has as you should have noticed gone into a slumber. I blame christmas. :) But as soon as I have SM2 up and running I will move my attention over to get TFNR going again. It would be very nice if we could wrap it up and squeeze those things it needs into 3.7 before alpha ends. That means february 7 is the deadline where TFNR should be finished IMHO.

I will report in more detail on TFNR in a day or two.


Squeak 3.7
============

I just looked up the ToDo list on the Swiki:

http://minnow.cc.gatech.edu/squeak/3491

I think we all should look it over and try to dig in one more time before alpha ends. There are LOTs of things on that page to look into.


Gift from heaven
====================

I have gotten an offer of a LOT of Squeak base documentation that probably (but no promises) will be sent to me before end of january!

Without modification it relies on Scamper since it is in Scamper-friendly HTML. I have thus promised to make sure Scamper works in 3.7, and Ken Causey (thanks) informed me on IRC that there are FIXes posted to get Scamper to work again.

The idea is to package this hypertextual documentation on SM2 and make it very easy to install it into a 3.7. Thus we probably want to add some form of link/button/text in the welcome screen so that people don't miss it. Since the docs are over 1200 HTML files totalling more than 2Mb we may not want to have it included in Basic or Full - not sure though, we will have to see how it looks.

If someone else could step up and make sure the FIXes for Scamper get into the image (or Scamper) I would be most grateful. Otherwise I will of course need to do it myself - but at the moment it would help me a lot if someone else did it.


License hell
==============

As most of you are aware there has been a long thread about Squeak-L again. The annual plague that we just have to put up with I guess. ;-) The final result of this year's thread is still not fully clear to me.

The most disturbing thing I have learned this time around is that Andrew, who is an expert in the field and even though sparks fly when we discuss ;-) is someone to listen carefully to, thinks that code developed *in* Squeak can not be released under any other license than Squeak-L - at least not unless the author wants to take a risk or undertakes a detailed legal analysis.

Of course - I may have gotten it wrong this time too :), but I don't think so. This has lead me to the following conclusions/discoveries:

- The code I have released which is meant for base Squeak (especially SM) will be changed into Squeak-L only. If someone today is relying on that code being under MIT - then I would like to hear about it! 

- We *can* get Squeak into Debian (somewhat) using a so called "installer package". But it would require someone to maintain such a package of course. For more info, use google and the name of a similar package "xanime-modules".

- Personally I would like to have a SqueakSub-L which I can use as a template when I release code for use in base Squeak. I have a hard time understanding how Squeak-L works for *me* since it only blabbers on about Apple. :)

- And last but not least, if Andrew thinks that we actually have a chance to get Squeak licensed under something better than Squeak-L (which I assume includes talking to Disney) then I am of course very interested in that. But in order for that to ever happen *for real* some people need to be involved, I would guess *at least* Andrew Greenberg and Alan Kay. So it bounces right back - are *you two guys* prepared to dig into this? Otherwise it practically ends right there.

Well, that was it for now I think.

regards, Göran

PS. Not even sure Andrew nor Alan reads this posting of course.



More information about the Squeak-dev mailing list