No commit report on squeak-dev

> On Wed, May 13, 2020 at 12:47:22AM +0200, Nicolas Cellier wrote:
> > Hi all,
> > just to warn that my last commits to trunk did not generate a mail report
> > this evening.
> >
>
> Hmm... Most likely I am the guilty party.
>
> I restarted the source.squeak.org image yesterday after applying the
> following updates from the http://source.squeak.org/ss repository:
>
>   MonticelloConfigurations-dtl.161
>   TinyWiki-tpr.12
>   SqueakSource-tpr.1124
>   SqueakSource-tpr.1125
>   SqueakSource-tpr.1126
>   SqueakSource-tpr.1127
>   SqueakSource-dtl.1128
>
> Of these updates, two of them include changes that could affect mail
> delivery. The possibly relevant parts of the commit notices are:
>
> SqueakSource-tpr.1125
>  Use the admin email address when sending out error page reports;
>  'box-admins' is confusing.
>
> SqueakSource-tpr.1127
>  add support for using the email server user name and password; update
>  the settings page
>
> @tim - can you please check these and see if either update might be
> dangerous? I'll wait until tomorrow for guidance, but if in doubt I'll
> revert any of the updates that might be questionable.
>
> Dave
>
>

>
>
> > On 2020-05-12, at 5:45 PM, David T. Lewis <[hidden email]> wrote:
> >
> > Tim,
> >
> > Indeed, the SqueakSource-tpr.1127 update looks suspicious. It changes
> > #deliverMailFrom:to:text: and the changed method now calls an unimplemented
> > method:
> >
> >  SMTPClient>>deliverMailFrom:to:text:usingServer:userName:password:
>
> Hmm. According to the history I see before me this is in Network-tpr.238
> in trunk. It's in my running 19666 level image and MC swears it has not
>  been removed.
>

> On Tue, May 12, 2020 at 05:55:53PM -0700, tim Rowledge wrote:
> >
> >
> > > On 2020-05-12, at 5:45 PM, David T. Lewis <[hidden email]> wrote:
> > >
> > > Tim,
> > >
> > > Indeed, the SqueakSource-tpr.1127 update looks suspicious. It changes
> > > #deliverMailFrom:to:text: and the changed method now calls an unimplemented
> > > method:
> > >
> > >  SMTPClient>>deliverMailFrom:to:text:usingServer:userName:password:
> >
> > Hmm. According to the history I see before me this is in Network-tpr.238
> > in trunk. It's in my running 19666 level image and MC swears it has not
> >  been removed.
> >
>
> Ok, that makes sense. When I updated the source.squeak.org image,
> I updated the packages that I found in the ss repository. I did not
> update the other packages in the image (it is based on a Squeak 5.2
> release image) and I should not do that.
>
> I cannot follow up tonight, but it sounds like the best thing is for
> me to revert any non-essential changes, and restart the image with
> just the essential changes (Chris' patch.st to fix moving packages
> from inbox to treated or trunk, plus my update for commented MCMs).
>
> Sorry for the mail problem, I will get it straightened out tomorrow.
>
> Dave
>
>

>
> I restarted source.squeak.org and am not having difficulties getting it
> restarted. I'll report back as soon as possible.
>
> Dave
>
>
> On Tue, May 12, 2020 at 09:39:05PM -0400, David T. Lewis wrote:
> > On Tue, May 12, 2020 at 05:55:53PM -0700, tim Rowledge wrote:
> > >
> > >
> > > > On 2020-05-12, at 5:45 PM, David T. Lewis <[hidden email]> wrote:
> > > >
> > > > Tim,
> > > >
> > > > Indeed, the SqueakSource-tpr.1127 update looks suspicious. It changes
> > > > #deliverMailFrom:to:text: and the changed method now calls an unimplemented
> > > > method:
> > > >
> > > >  SMTPClient>>deliverMailFrom:to:text:usingServer:userName:password:
> > >
> > > Hmm. According to the history I see before me this is in Network-tpr.238
> > > in trunk. It's in my running 19666 level image and MC swears it has not
> > >  been removed.
> > >
> >
> > Ok, that makes sense. When I updated the source.squeak.org image,
> > I updated the packages that I found in the ss repository. I did not
> > update the other packages in the image (it is based on a Squeak 5.2
> > release image) and I should not do that.
> >
> > I cannot follow up tonight, but it sounds like the best thing is for
> > me to revert any non-essential changes, and restart the image with
> > just the essential changes (Chris' patch.st to fix moving packages
> > from inbox to treated or trunk, plus my update for commented MCMs).
> >
> > Sorry for the mail problem, I will get it straightened out tomorrow.
> >
> > Dave
> >
> >
>

> Thank you, Dave!
>
> On Wed, May 13, 2020 at 4:41 PM David T. Lewis <[hidden email]> wrote:
> >
> > I restarted source.squeak.org and am not having difficulties getting it
> > restarted. I'll report back as soon as possible.
> >
> > Dave
> >
> >
> > On Tue, May 12, 2020 at 09:39:05PM -0400, David T. Lewis wrote:
> > > On Tue, May 12, 2020 at 05:55:53PM -0700, tim Rowledge wrote:
> > > >
> > > >
> > > > > On 2020-05-12, at 5:45 PM, David T. Lewis <[hidden email]> wrote:
> > > > >
> > > > > Tim,
> > > > >
> > > > > Indeed, the SqueakSource-tpr.1127 update looks suspicious. It changes
> > > > > #deliverMailFrom:to:text: and the changed method now calls an unimplemented
> > > > > method:
> > > > >
> > > > >  SMTPClient>>deliverMailFrom:to:text:usingServer:userName:password:
> > > >
> > > > Hmm. According to the history I see before me this is in Network-tpr.238
> > > > in trunk. It's in my running 19666 level image and MC swears it has not
> > > >  been removed.
> > > >
> > >
> > > Ok, that makes sense. When I updated the source.squeak.org image,
> > > I updated the packages that I found in the ss repository. I did not
> > > update the other packages in the image (it is based on a Squeak 5.2
> > > release image) and I should not do that.
> > >
> > > I cannot follow up tonight, but it sounds like the best thing is for
> > > me to revert any non-essential changes, and restart the image with
> > > just the essential changes (Chris' patch.st to fix moving packages
> > > from inbox to treated or trunk, plus my update for commented MCMs).
> > >
> > > Sorry for the mail problem, I will get it straightened out tomorrow.
> > >
> > > Dave
> > >
> > >
> >
>

> I must be overlooking something.
>
> A process listing shows this:
> root       671   665  0  2017 ?        00:00:00 readproctitle service errors: ..............................................................................................................................................................................................................................................................................................................................................svscan: warning: unable to stat squeaksource: file does not exist
>
> And I can see the squeaksource image being endlessly restarted.
>
> What changed: I stopped the service (sudo -d /service/squeaksource), replaced
> the image, and started it again (sudo -u /service/squeaksource).
>
> I restored the previously working image, and still have the same problem (I
> put the right image back after that verification).
>
> I can't spot what I am overlooking here. I'll see if I can get it running
> manually outside of the supervise system.
>
> :-/
>
> Dave
>
>
> On Wed, May 13, 2020 at 04:42:18PM +0200, Fabio Niephaus wrote:
> > Thank you, Dave!
> >
> > On Wed, May 13, 2020 at 4:41 PM David T. Lewis <[hidden email]> wrote:
> > >
> > > I restarted source.squeak.org and am not having difficulties getting it
> > > restarted. I'll report back as soon as possible.
> > >
> > > Dave
> > >
> > >
> > > On Tue, May 12, 2020 at 09:39:05PM -0400, David T. Lewis wrote:
> > > > On Tue, May 12, 2020 at 05:55:53PM -0700, tim Rowledge wrote:
> > > > >
> > > > >
> > > > > > On 2020-05-12, at 5:45 PM, David T. Lewis <[hidden email]> wrote:
> > > > > >
> > > > > > Tim,
> > > > > >
> > > > > > Indeed, the SqueakSource-tpr.1127 update looks suspicious. It changes
> > > > > > #deliverMailFrom:to:text: and the changed method now calls an unimplemented
> > > > > > method:
> > > > > >
> > > > > >  SMTPClient>>deliverMailFrom:to:text:usingServer:userName:password:
> > > > >
> > > > > Hmm. According to the history I see before me this is in Network-tpr.238
> > > > > in trunk. It's in my running 19666 level image and MC swears it has not
> > > > >  been removed.
> > > > >
> > > >
> > > > Ok, that makes sense. When I updated the source.squeak.org image,
> > > > I updated the packages that I found in the ss repository. I did not
> > > > update the other packages in the image (it is based on a Squeak 5.2
> > > > release image) and I should not do that.
> > > >
> > > > I cannot follow up tonight, but it sounds like the best thing is for
> > > > me to revert any non-essential changes, and restart the image with
> > > > just the essential changes (Chris' patch.st to fix moving packages
> > > > from inbox to treated or trunk, plus my update for commented MCMs).
> > > >
> > > > Sorry for the mail problem, I will get it straightened out tomorrow.
> > > >
> > > > Dave
> > > >
> > > >
> > >
> >
>

> Hi Dave,
>
> On Wed, May 13, 2020 at 10:24 AM David T. Lewis <[hidden email]> wrote:
> >
> > OK, it should be back now, and hopefully the mail delivery will be working again.
> >
> > I don't know the details, but it looks like the startup script is looking for
> > a file called "patch.st" to load. Chris had previously told me to rename that
> > file to "patch.st.old", which I had done. But today I wanted to undo my
> > earlier changes, and I renamed patch.st.old to patch.st.
>
> "Undoing your changes" would also mean undoing the image, which I
> assume you had renamed to ".old"...?
>

> > On Wed, May 13, 2020 at 10:24 AM David T. Lewis <[hidden email]> wrote:
> > >
> > > OK, it should be back now, and hopefully the mail delivery will be working again.
> > >
> > > I don't know the details, but it looks like the startup script is looking for
> > > a file called "patch.st" to load. Chris had previously told me to rename that
> > > file to "patch.st.old", which I had done. But today I wanted to undo my
> > > earlier changes, and I renamed patch.st.old to patch.st.
> >
> > "Undoing your changes" would also mean undoing the image, which I
> > assume you had renamed to ".old"...?
> >
>
> No. I renamed the Sept 9, 2019 image per your direction, and I have not changed that.

> Dave,
>
> The normal response of any application with a boostrap failure is to
> exit.  And as long as you have supervise instructed to keep the server
> running, that's what it will do.  It's not a "death spiral".  :)
>
> > > On Wed, May 13, 2020 at 10:24 AM David T. Lewis <[hidden email]> wrote:
> > > >
> > > > OK, it should be back now, and hopefully the mail delivery will be working again.
> > > >
> > > > I don't know the details, but it looks like the startup script is looking for
> > > > a file called "patch.st" to load. Chris had previously told me to rename that
> > > > file to "patch.st.old", which I had done. But today I wanted to undo my
> > > > earlier changes, and I renamed patch.st.old to patch.st.
> > >
> > > "Undoing your changes" would also mean undoing the image, which I
> > > assume you had renamed to ".old"...?
> > >
> >
> > No. I renamed the Sept 9, 2019 image per your direction, and I have not changed that.
>
> I was referring to the proper way to "undo your changes".  If the
> system doesn't come right up and it isn't obvious, just execute your
> backout plan to backout 100% and regroup.  Putting the system into
> some "third" state on-the-fly that is neither what you tested, nor
> what the system was beforehand, is a recipe for problems.
>
> A new system with an old patch is an invalid state for Personal
> SqueakSource.  Removing that file is part of the normal deployment
> process.  I'm sorry if it was confusing, but the system worked exactly
> as designed by refusing to run an invalid configuration.  The message
> in the log was clear, and you knew instantly what to do once you saw
> it.
>
> What needs fixing at this point is that image -- we need to be using
> one built fresh with your MCConfigurations enhancement, so we know
> what code our server is running, can maintain and collaborate on it
> and be able to make fresh instantiations of it.
>

> > But looking that the configurations that you published for Personal
> > SqueakSource on SqueakMap, you are not referencing any of the copies
> > in the ss repository.
>
> The fixed versions used to run source.squeak.org do pull from /ss.
> "5.2.1" is the current.
>
> You're right that the head (development) version is pulling
> from squeaksource.com.  This is for instances besides
> source.squeak.org but the goal is for source.squeak.org to
> stay with it if it can.
>
> > If you're building on a Squeak 5.2 release
> > image, you would either need to add a reference to the new copy in
> > the ss repository, or you would have to backport it to the squeak52
> > repository (I don't know if that's a good thing to do).
>
> Yeah, I like your conservatism regarding backporting, that might
> actually be too aggressive for this feature.
>
> So, simply adding a line to load it from a copy of the 5.2 script should work.
>
> > Your patch (the one in the patch.st file) is a bit more complicated.
> > I added it to the head of the updates as SqueakSource-dtl.1128.
> > However, that version is incompatible with the Squeak 5.2 release
> > image due to some intermediate changes, so it cannot be used in the
> > source.squeak.org image.
>
> Yes, but aren't those "incompatibilities" rather trivial?  Just pass
> the email id and repository from the repository instead?  (sorry I
> only had time to glance at it, didn't analyze...)
>
> > I think this means that if you want a SqueakSource package that
> > contains your patch, and if you want to run it on a Squeak 5.2 image,
> > then you would need to branch the packages (i.e. make a package
> > with SqueakSource-cmm.1123 as the parent, and save it as
> > SqueakSource.squeak52-cmm.1124).
> >
> > Alternatively, you could just reinstate the patch.sh file on the
> > server and update its time stamp.
>
> I hope we could just fix whatever's wrong with the intermediate versions..
>
> > I'm happy to restore the patch.st file if you think that's the right
> > thing to do, but the rest of this stuff is a bit beyond my patience
> > threshold.
>
> Cleaning and scraping requires a lot of patience when all you want to
> do is get to the fun painting part, but it's necessary for the best
> results.  If your enhancement isn't eventually properly integrated, it
> could suddenly disappear one day if another fix or enhancement became
> necessary that resulted in a new 5.2-based SqueakSource image.
>
> Really, I think just adding the one line incantation to a new "5.2.2"
> Release is all that's needed, but the only way to know is for someone
> to try it, load it, upload it to andreas and restart the server.  This
> is what I thought your original plan was.
>
>  - Chris
>