Hi,
We are almost there for the release, but we still need a little effort more. Would be terrific if everyone of you took the latest 2.0 image and test it with your usual workflows (we cannot imagine all, they are infinite :) And please report bugs, and JUST bugs :) all the "nice to have" can be reported, but in 3.0 milestone (or not milestone at all). cheers, Esteban |
Crashes during update to the latest version (all updates used to work so far)
MNU: SmallInteger>>ifFail: Phil 2013/2/19 Esteban Lorenzano <[hidden email]>: > Hi, > > We are almost there for the release, but we still need a little effort more. > Would be terrific if everyone of you took the latest 2.0 image and test it with your usual workflows (we cannot imagine all, they are infinite :) > > And please report bugs, and JUST bugs :) > > all the "nice to have" can be reported, but in 3.0 milestone (or not milestone at all). > > cheers, > Esteban PharoScreenshot.2.png (233K) Download Attachment |
In reply to this post by EstebanLM
Even weirder: played with the debugger and sayed Over Over Over then
Proceed and it worked. Strange.... Phil 2013/2/19 Esteban Lorenzano <[hidden email]>: > Hi, > > We are almost there for the release, but we still need a little effort more. > Would be terrific if everyone of you took the latest 2.0 image and test it with your usual workflows (we cannot imagine all, they are infinite :) > > And please report bugs, and JUST bugs :) > > all the "nice to have" can be reported, but in 3.0 milestone (or not milestone at all). > > cheers, > Esteban |
Even weirder: played with the debugger and sayed Over Over Over then What do you mean exactly ? You were not able to Proceed after clicking on Over several times on Pharo 1.4 ? I've just made a fix in the debugger that was integrated today.Give me more details to know if I have to look at it.
Clement
|
Pharo 2.0 not 1.4
The World Menu > System >Software Update started working and then threw me in teh debugger where I had the error. I inspected several variables and then did some Restart,, Over, Over, ..Proceed and things went fine. There was a ZnClient in there but I can't figure out why there was a mention of SmallInteger. Phil 2013/2/19 Clément Bera <[hidden email]>: >> Even weirder: played with the debugger and sayed Over Over Over then >> Proceed and it worked. Strange.... >> > What do you mean exactly ? You were not able to Proceed after clicking on > Over several times on Pharo 1.4 ? > I've just made a fix in the debugger that was integrated today.Give me more > details to know if I have to look at it. > > Clement > >> Phil >> >> 2013/2/19 Esteban Lorenzano <[hidden email]>: >> > Hi, >> > >> > We are almost there for the release, but we still need a little effort >> > more. >> > Would be terrific if everyone of you took the latest 2.0 image and test >> > it with your usual workflows (we cannot imagine all, they are infinite :) >> > >> > And please report bugs, and JUST bugs :) >> > >> > all the "nice to have" can be reported, but in 3.0 milestone (or not >> > milestone at all). >> > >> > cheers, >> > Esteban >> > |
In reply to this post by EstebanLM
On 2/19/13, Esteban Lorenzano <[hidden email]> wrote:
...> And please report bugs, and JUST bugs :) .... > > cheers, > Esteban > Maybe not a bug, Esteban, but somehow strange. Selecting a *.mcz file in the File Browser shows two 'load' buttons with identical tool tip. Are they the same or are they different. So far I have not seen a difference. Regards --Hannes |
In reply to this post by philippeback
I start to believe that it might be your VM?
Which version are you using? Smalltalk vm version does the trick. On 2013-02-19, at 21:45, "[hidden email]" <[hidden email]> wrote: > Pharo 2.0 not 1.4 > > The World Menu > System >Software Update started working and then > threw me in teh debugger where I had the error. I inspected several > variables and then did some Restart,, Over, Over, ..Proceed and things > went fine. > > There was a ZnClient in there but I can't figure out why there was a > mention of SmallInteger. > > Phil > > 2013/2/19 Clément Bera <[hidden email]>: >>> Even weirder: played with the debugger and sayed Over Over Over then >>> Proceed and it worked. Strange.... >>> >> What do you mean exactly ? You were not able to Proceed after clicking on >> Over several times on Pharo 1.4 ? >> I've just made a fix in the debugger that was integrated today.Give me more >> details to know if I have to look at it. >> >> Clement >> >>> Phil >>> >>> 2013/2/19 Esteban Lorenzano <[hidden email]>: >>>> Hi, >>>> >>>> We are almost there for the release, but we still need a little effort >>>> more. >>>> Would be terrific if everyone of you took the latest 2.0 image and test >>>> it with your usual workflows (we cannot imagine all, they are infinite :) >>>> >>>> And please report bugs, and JUST bugs :) >>>> >>>> all the "nice to have" can be reported, but in 3.0 milestone (or not >>>> milestone at all). >>>> >>>> cheers, >>>> Esteban >>> >> > |
Latest PharoVM from the CI
'NBCoInterpreter NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: a53445f9-c0c0-4015-97a3-be7db8d9ed6b Feb 18 2013 NBCogit NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: a53445f9-c0c0-4015-97a3-be7db8d9ed6b Feb 18 2013 git://gitorious.org/cogvm/blessed.git Commit: 05ed728452dc92732f9f23ce24a3c32bfdc0f4b8 Date: 2013-02-15 17:04:09 +0100 By: Igor Stasenko <[hidden email]> Jenkins build #14519 ' 2013/2/19 Camillo Bruni <[hidden email]>: > I start to believe that it might be your VM? > Which version are you using? > > Smalltalk vm version > > does the trick. > > On 2013-02-19, at 21:45, "[hidden email]" <[hidden email]> wrote: >> Pharo 2.0 not 1.4 >> >> The World Menu > System >Software Update started working and then >> threw me in teh debugger where I had the error. I inspected several >> variables and then did some Restart,, Over, Over, ..Proceed and things >> went fine. >> >> There was a ZnClient in there but I can't figure out why there was a >> mention of SmallInteger. >> >> Phil >> >> 2013/2/19 Clément Bera <[hidden email]>: >>>> Even weirder: played with the debugger and sayed Over Over Over then >>>> Proceed and it worked. Strange.... >>>> >>> What do you mean exactly ? You were not able to Proceed after clicking on >>> Over several times on Pharo 1.4 ? >>> I've just made a fix in the debugger that was integrated today.Give me more >>> details to know if I have to look at it. >>> >>> Clement >>> >>>> Phil >>>> >>>> 2013/2/19 Esteban Lorenzano <[hidden email]>: >>>>> Hi, >>>>> >>>>> We are almost there for the release, but we still need a little effort >>>>> more. >>>>> Would be terrific if everyone of you took the latest 2.0 image and test >>>>> it with your usual workflows (we cannot imagine all, they are infinite :) >>>>> >>>>> And please report bugs, and JUST bugs :) >>>>> >>>>> all the "nice to have" can be reported, but in 3.0 milestone (or not >>>>> milestone at all). >>>>> >>>>> cheers, >>>>> Esteban >>>> >>> >> > > |
thanks, I will try it with that one and see what happens.
On 2013-02-19, at 23:28, "[hidden email]" <[hidden email]> wrote: > Latest PharoVM from the CI > > 'NBCoInterpreter NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: > a53445f9-c0c0-4015-97a3-be7db8d9ed6b Feb 18 2013 > NBCogit NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: > a53445f9-c0c0-4015-97a3-be7db8d9ed6b Feb 18 2013 > git://gitorious.org/cogvm/blessed.git Commit: > 05ed728452dc92732f9f23ce24a3c32bfdc0f4b8 Date: 2013-02-15 17:04:09 > +0100 By: Igor Stasenko <[hidden email]> Jenkins build #14519 > ' > > 2013/2/19 Camillo Bruni <[hidden email]>: >> I start to believe that it might be your VM? >> Which version are you using? >> >> Smalltalk vm version >> >> does the trick. >> >> On 2013-02-19, at 21:45, "[hidden email]" <[hidden email]> wrote: >>> Pharo 2.0 not 1.4 >>> >>> The World Menu > System >Software Update started working and then >>> threw me in teh debugger where I had the error. I inspected several >>> variables and then did some Restart,, Over, Over, ..Proceed and things >>> went fine. >>> >>> There was a ZnClient in there but I can't figure out why there was a >>> mention of SmallInteger. >>> >>> Phil >>> >>> 2013/2/19 Clément Bera <[hidden email]>: >>>>> Even weirder: played with the debugger and sayed Over Over Over then >>>>> Proceed and it worked. Strange.... >>>>> >>>> What do you mean exactly ? You were not able to Proceed after clicking on >>>> Over several times on Pharo 1.4 ? >>>> I've just made a fix in the debugger that was integrated today.Give me more >>>> details to know if I have to look at it. >>>> >>>> Clement >>>> >>>>> Phil >>>>> >>>>> 2013/2/19 Esteban Lorenzano <[hidden email]>: >>>>>> Hi, >>>>>> >>>>>> We are almost there for the release, but we still need a little effort >>>>>> more. >>>>>> Would be terrific if everyone of you took the latest 2.0 image and test >>>>>> it with your usual workflows (we cannot imagine all, they are infinite :) >>>>>> >>>>>> And please report bugs, and JUST bugs :) >>>>>> >>>>>> all the "nice to have" can be reported, but in 3.0 milestone (or not >>>>>> milestone at all). >>>>>> >>>>>> cheers, >>>>>> Esteban >>>>> >>>> >>> >> >> > |
All right. 'nuff Pharo 2.0 for the day for me.
Pharo 2.0 looks great BTW. Much cleaner and all. Keep up the good work guys! Phil 2013/2/19 Camillo Bruni <[hidden email]>: > thanks, I will try it with that one and see what happens. > > On 2013-02-19, at 23:28, "[hidden email]" <[hidden email]> wrote: > >> Latest PharoVM from the CI >> >> 'NBCoInterpreter NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: >> a53445f9-c0c0-4015-97a3-be7db8d9ed6b Feb 18 2013 >> NBCogit NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: >> a53445f9-c0c0-4015-97a3-be7db8d9ed6b Feb 18 2013 >> git://gitorious.org/cogvm/blessed.git Commit: >> 05ed728452dc92732f9f23ce24a3c32bfdc0f4b8 Date: 2013-02-15 17:04:09 >> +0100 By: Igor Stasenko <[hidden email]> Jenkins build #14519 >> ' >> >> 2013/2/19 Camillo Bruni <[hidden email]>: >>> I start to believe that it might be your VM? >>> Which version are you using? >>> >>> Smalltalk vm version >>> >>> does the trick. >>> >>> On 2013-02-19, at 21:45, "[hidden email]" <[hidden email]> wrote: >>>> Pharo 2.0 not 1.4 >>>> >>>> The World Menu > System >Software Update started working and then >>>> threw me in teh debugger where I had the error. I inspected several >>>> variables and then did some Restart,, Over, Over, ..Proceed and things >>>> went fine. >>>> >>>> There was a ZnClient in there but I can't figure out why there was a >>>> mention of SmallInteger. >>>> >>>> Phil >>>> >>>> 2013/2/19 Clément Bera <[hidden email]>: >>>>>> Even weirder: played with the debugger and sayed Over Over Over then >>>>>> Proceed and it worked. Strange.... >>>>>> >>>>> What do you mean exactly ? You were not able to Proceed after clicking on >>>>> Over several times on Pharo 1.4 ? >>>>> I've just made a fix in the debugger that was integrated today.Give me more >>>>> details to know if I have to look at it. >>>>> >>>>> Clement >>>>> >>>>>> Phil >>>>>> >>>>>> 2013/2/19 Esteban Lorenzano <[hidden email]>: >>>>>>> Hi, >>>>>>> >>>>>>> We are almost there for the release, but we still need a little effort >>>>>>> more. >>>>>>> Would be terrific if everyone of you took the latest 2.0 image and test >>>>>>> it with your usual workflows (we cannot imagine all, they are infinite :) >>>>>>> >>>>>>> And please report bugs, and JUST bugs :) >>>>>>> >>>>>>> all the "nice to have" can be reported, but in 3.0 milestone (or not >>>>>>> milestone at all). >>>>>>> >>>>>>> cheers, >>>>>>> Esteban >>>>>> >>>>> >>>> >>> >>> >> > > |
In reply to this post by Camillo Bruni-3
On 2013-02-19, at 23:30, Camillo Bruni <[hidden email]> wrote:
> thanks, I will try it with that one and see what happens. I just took the latest Mac VM and updated from Pharo 20550 with! a crash! => problem with the latest VM!! I can easily update the image with an older VM, maybe I was just lucky it didn't crash with the old build: CoInterpreter VMMaker-oscog-IgorStasenko.162 uuid: e4554f9a-cc90-4826-a807-ac282b782fe4 Jul 26 2012 StackToRegisterMappingCogit VMMaker-oscog-IgorStasenko.162 uuid: e4554f9a-cc90-4826-a807-ac282b782fe4 Jul 26 2012 https://git.gitorious.org/cogvm/blessed.git Commit: aeb0705cde4b8fc57cb262dc051c5ee6dfa72a14 Date: 2012-07-26 16:38:48 +0200 By: Igor Stasenko <[hidden email]> |
26/07/2012 ? Huh?
2013/2/19 Camillo Bruni <[hidden email]>: > On 2013-02-19, at 23:30, Camillo Bruni <[hidden email]> wrote: >> thanks, I will try it with that one and see what happens. > > I just took the latest Mac VM and updated from Pharo 20550 with! a crash! > > => problem with the latest VM!! > > I can easily update the image with an older VM, maybe I was just lucky it didn't > crash with the old build: > > CoInterpreter VMMaker-oscog-IgorStasenko.162 uuid: e4554f9a-cc90-4826-a807-ac282b782fe4 Jul 26 2012 > StackToRegisterMappingCogit VMMaker-oscog-IgorStasenko.162 uuid: e4554f9a-cc90-4826-a807-ac282b782fe4 Jul 26 2012 > https://git.gitorious.org/cogvm/blessed.git Commit: aeb0705cde4b8fc57cb262dc051c5ee6dfa72a14 Date: 2012-07-26 16:38:48 +0200 By: Igor Stasenko <[hidden email]> > > |
I guess I should start sleeping, tired beyond return (aka, my emails start to just be gibberish). ;)
On 2013-02-19, at 23:50, "[hidden email]" <[hidden email]> wrote: > 26/07/2012 ? Huh? yes the update runs fine with a "rather" old VM which we built. However it fails with one of the new series built this week. I will forward that to our VM masters ;) > 2013/2/19 Camillo Bruni <[hidden email]>: >> On 2013-02-19, at 23:30, Camillo Bruni <[hidden email]> wrote: >>> thanks, I will try it with that one and see what happens. >> >> I just took the latest Mac VM and updated from Pharo 20550 with! a crash! >> >> => problem with the latest VM!! >> >> I can easily update the image with an older VM, maybe I was just lucky it didn't >> crash with the old build: >> >> CoInterpreter VMMaker-oscog-IgorStasenko.162 uuid: e4554f9a-cc90-4826-a807-ac282b782fe4 Jul 26 2012 >> StackToRegisterMappingCogit VMMaker-oscog-IgorStasenko.162 uuid: e4554f9a-cc90-4826-a807-ac282b782fe4 Jul 26 2012 >> https://git.gitorious.org/cogvm/blessed.git Commit: aeb0705cde4b8fc57cb262dc051c5ee6dfa72a14 Date: 2012-07-26 16:38:48 +0200 By: Igor Stasenko <[hidden email]> |
On Tue, Feb 19, 2013 at 8:12 PM, Camillo Bruni <[hidden email]> wrote: I guess I should start sleeping, tired beyond return (aka, my emails start to just be gibberish). ;) |
In reply to this post by EstebanLM
On 19/02/13 12:22 PM, Esteban Lorenzano wrote: > > We are almost there for the release, but we still need a little > effort more. Would be terrific if everyone of you took the latest 2.0 > image and test it with your usual workflows (we cannot imagine all, > they are infinite :) I usually keep a build tracking the upcoming Pharo release, but when I tried to do it during early Pharo-2.0 development, the code was not stable. Then I never got back to it. Here's what happened, when I tried to build with the newest version. First try: "Notice: Your VM is too old for this image. Please download the latest VM." Get latest coglinux, from http://www.mirandabanda.org/files/Cog/VM/ Second try: "Can't infer base LD_LIBRARY_PATH. Aborting. Try adding a line for /lib/i686/cmov/libc.so.6 ..." Patch squeak shell script with a case for "/lib/i686/cmov/libc*" Third try: No complaint about LD_LIBRARY_PATH, but still says "VM is too old ..." Switch to a CogVM (early version of PharoVM) that was lying around. More tries: Can't load Seaside. Update to newer seaside31.st script from https://github.com/renggli/builder/tree/master/scripts Magritte3/Pier3 has deprecations and undefined class. Revert to trying Seaside30. Get: *** Warning: Warning: This package depends on the following classes: BlockContext You must resolve these dependencies before you will be able to load these definitions: renderOn: Go back to trying Magritte3/Pier3 on Seaside31 again. Ignore the deprecations, find: FileSystem-Legacy, at: http://smalltalkhub.com/mc/dh83/fisleg/main Can now load Pier3 into Pharo-20558. Figure out where to get the latest Zinc-Seaside. (Found at http://www.squeaksource.com/ZincHTTPComponents). Start Seaside, go to Pier link, get deprecation warnings about #includesSubString:. Try to disable warnings in System Settings. Still get warnings in webpage. Change to #includesSubString: to not call #deprecated:on:. Now, can see default Pier page. Next step, try to load my custom code... |
Hi Yanni,
On 20 Feb 2013, at 22:23, Yanni Chiu <[hidden email]> wrote: > On 19/02/13 12:22 PM, Esteban Lorenzano wrote: >> >> We are almost there for the release, but we still need a little >> effort more. Would be terrific if everyone of you took the latest 2.0 >> image and test it with your usual workflows (we cannot imagine all, >> they are infinite :) > > I usually keep a build tracking the upcoming Pharo release, but when I tried to do it during early Pharo-2.0 development, the code was not stable. Then I never got back to it. Here's what happened, when I tried to build with the newest version. Thanks for testing, thanks for the feedback. > First try: "Notice: Your VM is too old for this image. Please download the latest VM." > > Get latest coglinux, from http://www.mirandabanda.org/files/Cog/VM/ > > Second try: "Can't infer base LD_LIBRARY_PATH. Aborting. Try adding a line for /lib/i686/cmov/libc.so.6 ..." > > Patch squeak shell script with a case for "/lib/i686/cmov/libc*" > > Third try: No complaint about LD_LIBRARY_PATH, but still says "VM is too old ..." > > Switch to a CogVM (early version of PharoVM) that was lying around. This is a bit to be expected when you use older VMs, the one-click or designated Pharo (NBCog) VMs would have prevented these issues. > More tries: > > Can't load Seaside. Update to newer seaside31.st script from https://github.com/renggli/builder/tree/master/scripts > > Magritte3/Pier3 has deprecations and undefined class. > > Revert to trying Seaside30. Get: > *** Warning: Warning: This package depends on the following classes: > BlockContext > You must resolve these dependencies before you will be able to load these definitions: > renderOn: > > Go back to trying Magritte3/Pier3 on Seaside31 again. Ignore the deprecations, find: FileSystem-Legacy, at: http://smalltalkhub.com/mc/dh83/fisleg/main > Can now load Pier3 into Pharo-20558. > > Figure out where to get the latest Zinc-Seaside. (Found at http://www.squeaksource.com/ZincHTTPComponents). Start Seaside, go to Pier link, get deprecation warnings about #includesSubString:. Try to disable warnings in System Settings. Still get warnings in webpage. Change to #includesSubString: to not call #deprecated:on:. Now, can see default Pier page. It will be a while before all third party packages are properly ported to Pharo 2.0. Many are, but that depends mostly on their authors. Now, Seaside 3.1 and friends will load and work proper on Pharo 2.0, it is just not an out-of-the-box it-just-works experience. The #includesSub[S|s]tring: issue is easy to solve, it just requires a decision on the part of the Seaside developers. See http://forum.world.st/Broken-build-because-of-use-of-includesSubstring-tt4662939.html There also seem to be some technical reasons why a Metacello Configuration is not yet there. I guess that user pressure, once 2.0 is out, is the only thing that will make a difference. These things sometimes go slowly. This is all open source. > Next step, try to load my custom code… Let us know how it goes ! Regards, Sven -- Sven Van Caekenberghe http://stfx.eu Smalltalk is the Red Pill |
On 21/02/13 3:12 AM, Sven Van Caekenberghe wrote: > > On 20 Feb 2013, at 22:23, Yanni Chiu <[hidden email]> wrote: >> >> Switch to a CogVM (early version of PharoVM) that was lying around. > > This is a bit to be expected when you use older VMs, the one-click or designated Pharo (NBCog) VMs would have prevented these issues. Actually, the older VM worked fine - it's just a version of PharoVM, before it's name was changed from CogVM. The problem is that even the latest cog vm from http://mirandabanda.org will not work with Pharo-2.0. >> Next step, try to load my custom code… > > Let us know how it goes ! I just had to change my build script to stop loading an older version of Fuel on top of the version already present in Pharo-2.0. After some quick manual testing, things look good. But, the newer Fuel version means I have to migrate Fuel versions again, which is not something I expected to have to do. Hmmm, will have to back away from moving to 2.0, for now. |
Hi Yanni,
> I just had to change my build script to stop loading an older version of > Fuel on top of the version already present in Pharo-2.0. After some quick > manual testing, things look good. What version was you loading? > But, the newer Fuel version means I have to migrate Fuel versions again, > which is not something I expected to have to do. Hmmm, will have to back > away from moving to 2.0, for now. The code of Fuel that was integrated today is the same as installing with: Gofer it squeaksource3: 'Fuel'; package: 'ConfigurationOfFuel'; load. (ConfigurationOfFuel project version: '1.9') load. (actually, the code that has been integrated should be the definitive one for Fuel 1.9, but the configuration still needs a revision for old Pharos) Tell me if I can help with the migration. Martin |
On 25/02/13 4:25 PM, Martin Dias wrote:
> > What version was you loading? Fuel-MartinDias.685 which corresponds to version 1.8 > (actually, the code that has been integrated should be the definitive > one for Fuel 1.9, but the configuration still needs a revision for old > Pharos) Would it be easy to unload Fuel 1.9 from Pharo-2.0, and load Fuel 1.8 instead? > Tell me if I can help with the migration. I previously migrated from Fuel-MartinDias.259.mcz (which was something between version 1.4 and 1.5 of Fuel). It just takes time to set up the scripts and test the results. |
On Mon, Feb 25, 2013 at 10:55 PM, Yanni Chiu <[hidden email]> wrote:
> On 25/02/13 4:25 PM, Martin Dias wrote: >> >> >> What version was you loading? > > > Fuel-MartinDias.685 which corresponds to version 1.8 > > >> (actually, the code that has been integrated should be the definitive >> one for Fuel 1.9, but the configuration still needs a revision for old >> Pharos) > > > Would it be easy to unload Fuel 1.9 from Pharo-2.0, and load Fuel 1.8 > instead? I tried to load Fuel 1.8 in Pharo 2.0 but it seems that there is an incompatibility with Date and DateAndTime... |
Free forum by Nabble | Edit this page |