On Mon, 02 Aug 2004 08:18:00 +0200, Boris wrote in message 

> Arnt Karlsen wrote:
> > On Sun, 01 Aug 2004 20:31:49 +0200, Boris wrote in message 
> > 
> > 
> >>one other thing: I mentioned already that I didn't have the original
> >>pre-release available to create that patch, meanwhile
> >>Stewart & Steven have released a patch that's based on the
> >_original_>pre2-release, which *differs* from mine, you can get that
> >file
> >at:>http://www.geocities.com/sandreas41/data/base-0.9.5p2-0.9.5.tar.
> >gz
> > 
> > 
> > ..ah, that means at least one of you guys has been a naughty boy 
> > and worked from something _other_ than an official FG release.  ;-)
> Arnt, how about starting to actually *read* my postings - at least
> those that you reply to ? :-)

..heh, good catch, could looong length be an issue?  ;-)

> I did mention _all the time_ that I didn't have the original
> (pre)releases available and hence decided to use a CVS checkout
> as reference basis for the patch.
> Having meanwhile had the chance to test it, it does seem
> work anyway, just more files being put into the folder - but
> I didn't really check it thorougly.
> Contary to that, the 0.9.4 final => 0.9.5 final patch is based
> on official releases, which I also mentioned ;-)
> >>As I said already: I would not mind creating such a patch chain,
> >>but first we would need to know whether things are working as
> >>expected and THEN I would still need access to the original
> >>pre-releases in order to create the necessary patch archives.
> > 
> > 
> > ..another idea to test these; provide test scripts.  I have 
> > bandwith and disk space and vacant cpus, but no time.
> that would then be very specific to FlightGear, 

..yup, thats precisely the idea.

> and I think Steven & Stewart are right in trying to keep things as
> general as possible, e.g. that way they can use that script
> for _many_ purposes, so it does have its justification outside
> the FG world.

..it (tardiff) does, and it looks good, so build on it.

> IF such an extension is considered a good idea by several
> users here, one could think about providing externals
> means for it.

..in this meritocraty, _only_ those ideas that are _acted_ upon,
prevails.  ;-)

> > ..basically, something like "for FG in FlightGear SimGear plib ;
> >     ;for V in 0.9.5 0.9.4 # etc for SimGear plib too
> >             ;do wget -c $FG.org/downloads/FG-$V.tar.bz2 
> >             ;tar jxvf $FG$V.tar.bz2 
> >     ;done 
> >     # etc
> > ;done
> so you are talking of an automated updater ?

..define "automated".  The idea is the user should 
find an update script over at fg.org, and be able to
update to the latest official release, and at least 
"say Yes".  ;-)

> regarding that one really has to be careful, not
> everybody  has a full GNU toolchain available,
> even though there are things like Cygwin they
> do significantly complicate things for novice
> users - or at least for those who are not really
> familiar with Unix.
> (I know that stuff like wget is also available as
> a standard Win32 compiled version, but it's not
> per default available on windows ...)

..so test for it and haul it home where needed. ;-)

> > diff -ruN $FG$V $FG$($V-1) >diff-from-$FG$($V-1)-to-$FG$V
> >     md5sum diff-from-$FG$($V-1)-to-$FG$V
> >     bzip2 diff-from-$FG$($V-1)-to-$FG$V
> >     md5sum diff-from-$FG$($V-1)-to-$FG$V.bz2  
> > # etc".
> > ..the md5sums are neat to verify that we wind up with the same 
> > source tarballs, without having to build them.
> not sure about how much sense something like that would
> make, we will have to wait for other opinions, 

..what suddenly stopped you from forming and voicing 
your own opinions here?  ;-)

> but it's  gonna certainly be beyond the scope of "tardiff".

..it _is_.  ;-)
> > ..expanding on this idea, it is possible to have newbies use "this
> > upgrade script" to update their old FG to the current,
> I really doubt, how feasible something like that would be for
> for "newbies", I know a lot of windows users who would certainly
> not manage to make use of something like that - and as soon as
> you are a user of a unix-based OS the debate becomes pointless
> as you are likely to be somewhat more familiar with your system
> anyway and certainly would not care doing the required steps
> manually.
> > first chking 
> > for their old FG, then fetch Boris' tardiffs
> tardiff itself comes from Steven & Stewart Andreason - so
> it certainly was _not_ mine idea  - just to clarify things
> and give credit where it's due.

..true.  And when you use their tardiff script to diff tarballs,
those tardiffs or better, diffballs are "yours."

> > and patch up their FG 
> > install to the latest official FG, SimGear and plib.
> I think we'll really have to wait for other opinions, I really
> doubt that it would pay off - simply because the work that needs
> to be done would probably take relatively long compared to that
> group of users who might really make use of something like that,
> but that's my personal view ...

..see above.  ;-)

> > ..at some stage, the official tarballs (or a cvs co to the latest
> > cvs release tag) becomes more comvenient, so don't over-engineer it.
> > ;-)
> I agree, I've talked to Steven & Stewart about that and they also
> think that the current version is going to be the final version for
> the near future, maybe there'll be one or two small fixes but not
> many enhancements anymore. It might still become useful to add one
> or two small features when changes in the fgfs base archive require
> more sophisticated tracking mechanisms.
> The only thing that I can currently think of would be an addition
> to support simultaneous creation of ZIP archives, simply as there
> are a lot more common for winows and more familiar to its users
> so it might really make things simpler for those users...

..I'd rather see them suggest useing tgz, if the idea is get Winzip

..med vennlig hilsen = with Kind Regards from Arnt... ;-)
...with a number of polar bear hunters in his ancestry...
  Scenarios always come in sets of three: 
  best case, worst case, and just in case.

Flightgear-devel mailing list

Reply via email to