Hi all! We are currently packaging RC1, which will be the first (and maybe only) release candidate of Squeak 5.3. Due to recent discussions on the list, here are some infos from yesterday's board meeting: - Release VM will be 201912311458 if there is no veto - No new sources file for 5.3; so slightly bigger changes file compared to 5.2 - No update of the localization files ("locale") due to time constraints - Same Linux directory structure as in 5.2 (compact bin/shared in Linux bundle; resources/bin/lib/squeak in all-in-one bundle) due to time constraints - Next Trunk version will be 6.0alpha ;-) So, we want to keep the CI infrastructure as is and push the release out as soon as possible. :-) *** If we overlooked important fixes from the Inbox, please speak up now! :-) Best, Marcel |
Thanks for the update, Marcel!
On Thu, Feb 20, 2020 at 3:52 PM Marcel Taeumel <[hidden email]> wrote: > > Hi all! > > We are currently packaging RC1, which will be the first (and maybe only) release candidate of Squeak 5.3. > > Due to recent discussions on the list, here are some infos from yesterday's board meeting: > > - Release VM will be 201912311458 if there is no veto > - No new sources file for 5.3; so slightly bigger changes file compared to 5.2 > - No update of the localization files ("locale") due to time constraints > - Same Linux directory structure as in 5.2 (compact bin/shared in Linux bundle; resources/bin/lib/squeak in all-in-one bundle) due to time constraints > - Next Trunk version will be 6.0alpha ;-) > > So, we want to keep the CI infrastructure as is and push the release out as soon as possible. :-) > > *** > > If we overlooked important fixes from the Inbox, please speak up now! :-) > > Best, > Marcel > > |
In reply to this post by marcel.taeumel
Hi Marcel, On Feb 20, 2020, at 6:52 AM, Marcel Taeumel <[hidden email]> wrote:
These are VMMaker.oscog-nice.2714/202002190233. The main reasons to use them are that a) that Nicolas recently got a fast primitive highBit working and b) they are otherwise almost identical to 201912311458. The highBit primitive would then be able to be enableded in 6.0 alpha. The Mac Pharo builds fail because of a technicality, but all the Squeak ones are good.
Eliot _,,,^..^,,,_ (phone) |
Hi Eliot. > The highBit primitive would then be able to be enableded in 6.0 alpha. Hmm... if there is no benefit for 5.3, we could also swap the VM directly after the 5.3 release to be used for all Trunk (6.0alpha) builds. Best, Marcel
|
Hi Marcel, this would require a second upgrade of VM just after the release... Most people will start the development with 5.3 released VM, which means that we cannot integrate image enhancements. Le jeu. 20 févr. 2020 à 16:45, Marcel Taeumel <[hidden email]> a écrit :
|
In reply to this post by marcel.taeumel
> If we overlooked important fixes from the Inbox, please speak up now! :-)
The fixes for properly working sound volume control in several subclasses of AbstractSound that we discussed a while ago were not in trunk last time I checked. Ref: http://forum.world.st/The-Inbox-Sound-dtl-67-mcz-td5107995.html#a5108064 Stef |
In reply to this post by marcel.taeumel
Hi Marcel, On Feb 20, 2020, at 7:45 AM, Marcel Taeumel <[hidden email]> wrote:
|
Hi Eliot. Agreed. It would be easier that way. :-) Best, Marcel
|
In reply to this post by Stéphane Rollandin
Thanks. Just merged. Best, Marcel
|
In reply to this post by marcel.taeumel
Then would someone please post a link here to the exact VM that will be the release VM? I would like to upgrade to it immediately and at least run my base tests on it. I don't see any download links at the link Eliot posted. Thanks. On Thu, Feb 20, 2020 at 12:01 PM Marcel Taeumel <[hidden email]> wrote:
|
In reply to this post by marcel.taeumel
Sorry, but I lost track of the recent VM changes. Could both Nicolas and Eliot please confirm that <a dir="ltr" href="tel:2714/202002190233" style="word-spacing:1px;border-color:rgb(49,49,49);color:rgb(49,49,49)">202002190233 is the correct version that we should use in the upcoming Squeak release? Once confirmed, I will do another VM release, which we can then use for building our bundles. Fabio On Thu, 20 Feb 2020 at 7:01 pm, Marcel Taeumel <[hidden email]> wrote:
|
Hi Eliot, hi Nicolas. :-) Which one should it be? (https://bintray.com/opensmalltalk/vm/cog)
|
In reply to this post by marcel.taeumel
Hi Marcel, On Feb 21, 2020, at 6:33 AM, Marcel Taeumel <[hidden email]> wrote:
|
Or don't we want to integrate the BitBlt Convex Shape Fill fix? Eliot, any objection? Sorry in advance for the never ending litany of fixes ;) Le ven. 21 févr. 2020 à 17:24, Eliot Miranda <[hidden email]> a écrit :
|
On Feb 22, 2020, at 6:36 AM, Nicolas Cellier <[hidden email]> wrote:
|
In reply to this post by Nicolas Cellier
The answer should come from Eliot and Nicolas, but FWIW I am running
the 202002220842 64-bit VM on Linux and can confirm that it fixes the bug reported by Stef Rollandin. The BitBltTests are green. I see no problems with using this VM. Note to fellow Squeak oversight board members - I know I have been saying not to worry so much about which version we pick, the're all good. But this issue affects all Squeak users in a visible way, so I'm happy to contradict myself now :-) Dave On Sat, Feb 22, 2020 at 03:35:46PM +0100, Nicolas Cellier wrote: > Or don't we want to integrate the BitBlt Convex Shape Fill fix? > https://bintray.com/opensmalltalk/vm/cog/202002220842 > Eliot, any objection? > > Sorry in advance for the never ending litany of fixes ;) > > Le ven. 21 f??vr. 2020 ?? 17:24, Eliot Miranda <[hidden email]> a > ??crit : > > > ???Hi Marcel, > > > > > > On Feb 21, 2020, at 6:33 AM, Marcel Taeumel <[hidden email]> wrote: > > > > ??? > > Hi Eliot, hi Nicolas. :-) > > > > Which one should it be? (https://bintray.com/opensmalltalk/vm/cog) > > > > 202002190239 > > > > <http://twitter.com/share?text=Version+202002190239+of+cog+has+been+released%21+Grab+it+now+from+%23bintray+https%3A%2F%2Fbintray.com%2Fopensmalltalk%2Fvm%2Fcog%2F202002190239%21> > > 202002190233 > > 202002190019 > > 202002172228 > > 202002150156 > > > > > > 2714/202002190233 > > > > ... > > > > Best, > > Marcel > > > > > > Am 20.02.2020 23:28:51 schrieb Fabio Niephaus <[hidden email]>: > > Sorry, but I lost track of the recent VM changes. > > > > Could both Nicolas and Eliot please confirm that 202002190233 > > <2714/202002190233> is the correct version that we should use in the > > upcoming Squeak release? > > > > Once confirmed, I will do another VM release, which we can then use for > > building our bundles. > > > > Fabio > > > > On Thu, 20 Feb 2020 at 7:01 pm, Marcel Taeumel <[hidden email]> > > wrote: > > > >> Hi Eliot. > >> > >> Agreed. It would be easier that way. :-) > >> > >> Best, > >> Marcel > >> > >> Am 20.02.2020 17:25:32 schrieb Eliot Miranda <[hidden email]>: > >> Hi Marcel, > >> > >> On Feb 20, 2020, at 7:45 AM, Marcel Taeumel <[hidden email]> > >> wrote: > >> > >> ??? > >> Hi Eliot. > >> > >> > The highBit primitive would then be able to be enableded in 6.0 alpha. > >> > >> Hmm... if there is no benefit for 5.3, we could also swap the VM directly > >> after the 5.3 release to be used for all Trunk (6.0alpha) builds. > >> > >> > >> I would much rather base 6.0 alpha work on a release vm. The standard > >> work flow is going to be people installing 5.3, changing the update url to > >> trunk, and updating. It would be a shame if they also had to install a new > >> vm for this to work fully. > >> > >> > >> Best, > >> Marcel > >> > >> Am 20.02.2020 16:37:50 schrieb Eliot Miranda <[hidden email]>: > >> ???Hi Marcel, > >> > >> On Feb 20, 2020, at 6:52 AM, Marcel Taeumel <[hidden email]> > >> wrote: > >> > >> ??? > >> Hi all! > >> > >> We are currently packaging RC1, which will be the first (and maybe only) > >> release candidate of Squeak 5.3. > >> > >> Due to recent discussions on the list, here are some infos from > >> yesterday's board meeting: > >> > >> - Release VM will be 201912311458 if there is no veto > >> > >> > >> Not a veto but a request that we use the latest VMs here: > >> > >> https://travis-ci.org/OpenSmalltalk/opensmalltalk-vm/jobs/652284937 > >> > >> These are VMMaker.oscog-nice.2714/202002190233. The main reasons to use > >> them are that a) that Nicolas recently got a fast primitive highBit working > >> and b) they are otherwise almost identical to 201912311458. The highBit > >> primitive would then be able to be enableded in 6.0 alpha. > >> > >> The Mac Pharo builds fail because of a technicality, but all the Squeak > >> ones are good. > >> > >> - No new sources file for 5.3; so slightly bigger changes file compared > >> to 5.2 > >> - No update of the localization files ("locale") due to time constraints > >> - Same Linux directory structure as in 5.2 (compact bin/shared in Linux > >> bundle; resources/bin/lib/squeak in all-in-one bundle) due to time > >> constraints > >> - Next Trunk version will be 6.0alpha ;-) > >> > >> So, we want to keep the CI infrastructure as is and push the release out > >> as soon as possible. :-) > >> > >> *** > >> > >> If we overlooked important fixes from the Inbox, please speak up now! :-) > >> > >> Best, > >> Marcel > >> > >> > >> Eliot > >> _,,,^..^,,,_ (phone) > >> > >> > >> > >> > > > > > |
In reply to this post by marcel.taeumel
Hi, What about the endless debugger issues? Have any of the proposed fixes (mostly by Christoph) been reviewed and accepted? If not, will it happen before the release? Stepping over the yield: in this still ruins the image: Generator on: [:g | g yield: #foo]. Kind regards, Jakob Am Do., 20. Feb. 2020 um 19:07 Uhr schrieb Marcel Taeumel <[hidden email]>:
|
On Sun, Feb 23, 2020 at 11:11:02PM +0100, Jakob Reschke wrote:
> Hi, > > What about the endless debugger issues? Have any of the proposed fixes > (mostly by Christoph) been reviewed and accepted? I don't know. > If not, will it happen before the release? Hopefully the answer is _ _ ___ | \ | |/ _ \ | \| | | | | | |\ | |_| | _ |_| \_|\___/ (_) There is a lot of good stuff waiting in the inbox, and the sooner we get the 5.3 release published, the sooner we can get to work on merging the inbox. So let's get it done and move forward :-) Dave |
> On Feb 23, 2020, at 6:20 PM, David T. Lewis <[hidden email]> wrote: > > On Sun, Feb 23, 2020 at 11:11:02PM +0100, Jakob Reschke wrote: >> Hi, >> >> What about the endless debugger issues? Have any of the proposed fixes >> (mostly by Christoph) been reviewed and accepted? > > > I don't know. > > >> If not, will it happen before the release? > > > Hopefully the answer is > > _ _ ___ > | \ | |/ _ \ > | \| | | | | > | |\ | |_| | _ > |_| \_|\___/ (_) > > > There is a lot of good stuff waiting in the inbox, and the sooner we get > the 5.3 release published, the sooner we can get to work on merging the > inbox. So let's get it done and move forward :-) +1. Don’t let the perfect be the enemy of the good. > > Dave > > |
Hi Jakob, hi all! > What about the endless debugger issues? Have any of the proposed fixes (mostly by Christoph) been reviewed and accepted? If not, will it happen before the release? The proposed fix will not be included in the 5.3 release because we are short on the side of automated tests for the debugger. While "step over #yield: in generator" is not a new bug (see Squeak 4.3 for example), I agree that it got worse because [cmd]+[dot] is not working anymore. Btw: "step over an object-as-method call" has never been working either ;-) I will mention this issue in the release notes. Sorry for any inconvenience cause by this decision. Best, Marcel
|
Free forum by Nabble | Edit this page |