March 2014 - Pharo-dev - lists.pharo.org

Bug in valueWithin:onTimeout: ?
by ClÃ©ment Bera March 11, 2014

March 11, 2014

Hello guys, Yesterday I had this problem that at startup a pharo image has several process running at priority 40. The problem is that my tests (run from command line) trigger #valueWithin:onTimeout: and the protected block (the block calling #valueWithin:onTimeout:) would always timed out for the first test as resuming the watchdog in #valueWithin:onTimeout: allows another runnable process at the same priority as the active process to be run. Example: Here we have two process at the activePriority, the second process takes 100ms to be run and as the transcript shows, the protected block always times out because this other process is started after the watchdog so during the protected block activation. | p endTime | p := Processor activeProcess priority. [ 'Process2Start' logCr. endTime := Time millisecondClockValue + 100. [ Time millisecondClockValue > endTime ] whileFalse: [ 1 + 2 + 3 ]. 'Process2End' logCr. ] forkAt: p. [ 'protected' logCr] valueWithin: 0.05 second onTimeout: [ 'timeout' logCr ]. Transcript displays: 'Process2Start' 'Process2End' 'timeout' I'm not very familiar with Pharo's process framework nor with processes in general so I would like to discuss the problem. *Question 1: Is it a bug ?* I fixed it by adding a Processor yield just before resuming the watchdog in #valueWithin:onTimeout: to allow other runnable process at the activePriority to be run before the protected block and therefore forbidding other process started outside of the protected block to trigger the time out. No process can be started in between the Processor yield and watchdog resume. *Question 2: Can someone tell me if this code looks ok ? I've just added Processor yield.* valueWithin: aDuration onTimeout: timeoutBlock "Evaluate the receiver. If the evaluation does not complete in less than aDuration evaluate the timeoutBlock instead" | theProcess delay watchdog tag | aDuration <= Duration zero ifTrue: [^ timeoutBlock value ]. "the block will be executed in the current process" theProcess := Processor activeProcess. delay := aDuration asDelay. tag := self. "make a watchdog process" watchdog := [ delay wait. "wait for timeout or completion" theProcess ifNotNil:[ theProcess signalException: (TimedOut new tag: tag)] ] newProcess. "Watchdog needs to run at high priority to do its job (but not at timing priority)" watchdog priority: Processor timingPriority-1. *Processor yield.* "catch the timeout signal" ^ [ watchdog resume. "start up the watchdog" self ensure:[ "evaluate the receiver" theProcess := nil. "it has completed, so ..." delay delaySemaphore signal. "arrange for the watchdog to exit" ]] on: TimedOut do: [ :e | e tag == tag ifTrue:[ timeoutBlock value ] ifFalse:[ e pass]].

2 1

About a Spec dedicated mailing-list
by Benjamin March 11, 2014

March 11, 2014

Hello guys :) I was wondering if it will make sense to have a spec dedicated mailing list :) As I have pros and cons, I am asking for your opinion :) Thanks in advance, Ben

5 4

May I have a write permission to MetaRepoForPharo30?
by Attila Magyar March 11, 2014

March 11, 2014

I'ld like to publis my app. My user is this one: http://smalltalkhub.com/#!/~zeroflag -- View this message in context: http://forum.world.st/May-I-have-a-write-permission-to-MetaRepoForPharo30-t… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

3 2

[FEEDBACK] Install Pharo on Windows
by Damien Cassou March 10, 2014

March 10, 2014

Hi, I've just made an executable file for installing Pharo on Windows (in program files, with a shortcut in the Start menu...). Please try it and report problems: https://ci.inria.fr/pharo-contribution/job/PharoLauncher-Win-Package/lastSu… Best, -- Damien Cassou http://damiencassou.seasidehosting.st "Success is the ability to go from one failure to another without losing enthusiasm." Winston Churchill

9 16

Crashing the vm in a reproducible way
by Pharo4Stef March 10, 2014

March 10, 2014

Hi We already experienced this with Pavel in December. Once I loaded UnReloader from SystemConfigurations from STHUB/Pharo and I run unload I crash systematically the latest vm obtain with wget -O - http://get.pharo.org/30+vmLatest | bash the method crashing the system is checkState "self new checkState" | difference | difference := SystemNavigation default allUnimplementedCalls copyWithoutAll: unimplementedCalls. difference ifNotEmpty: [ self log nextPutAll: 'new unimplemented calls: '; cr; nextPutAll: difference asString; cr.]. difference := (SystemNavigation default obsoleteClasses collect: #name) asSet copyWithoutAll: obsoleteClasses. difference ifNotEmpty: [ self log nextPutAll: 'new obsoleteClasses:'; cr; nextPutAll: difference asString; cr]. difference := (Undeclared collect: #name) asSet copyWithoutAll: undeclared. difference ifNotEmpty: [ self log nextPutAll: 'new undeclared: '; cr; nextPutAll: difference asString; cr]. unimplementedCalls := SystemNavigation default allUnimplementedCalls. obsoleteClasses := (SystemNavigation default obsoleteClasses collect: #name) asSet. undeclared := (Undeclared collect: #name) asSet. Now executing the expressions one by one once the tools have been removed does not crash. Stef

1 1

base64.b64encode(data) equivalent
by Pharo4Stef March 10, 2014

March 10, 2014

Iâm sure that we have that. I remember svenâs code using encoder but I do not know where. # Method for encoding ints with base64 encoding def encode(n): data = struct.pack("i", n) s = base64.b64encode(data) return s

3 5

Is there some shared memory space?
by roberto.minelli＠usi.ch March 10, 2014

March 10, 2014

Hi guys, I am experiencing a very strange problem that makes me think there is some shared memory problem with Pharo. Here are the steps that lead to the problem: - I have an image, I do something (i.e., materializing objects with Fuel) - I encounter a problem and a debugger pops up with a âself error: something" - I close the debugger - I close the image - I reopen the image - I fix the bug (i.e., commenting the self error) - I re-run the same thing as in the first step - An OutOfMemory error appears. Any clue? Thanks, Roberto

4 10

Re: [Pharo-dev] Glorp version in Pharo
by Mariano Martinez Peck March 10, 2014

March 10, 2014

Seems it was the one that came with VW 7.7.1..... On Mon, Mar 10, 2014 at 9:38 AM, Mariano Martinez Peck < marianopeck(a)gmail.com> wrote: > > > > On Wed, Mar 5, 2014 at 4:36 PM, Esteban A. Maringolo <emaringolo(a)gmail.com > > wrote: > >> [I'm cross posting Pharo-dev mailing list too] >> >> Hi Pharo developers, >> >> I'm having an issue while performing a not so common db query (but not >> esoteric either). >> And it seems the version I'm using in Pharo is outdated compared with >> the latest version they're using in VW. >> >> Is there an official maintaner and/or updater of the Glorp port for Pharo? >> >> > No. The first time, we did ir (me, Esteban Lorenzano, Guillermo Polito, > etc). We took certain version of the VW version. I will try to search which > one. After that, no one that I am aware of did a another port again. > > >> I'm using the GlorpDriverPostgreSQL, and the Glorp package from >> SmalltalkHub (Glorp-MarianoMartinezPeck.88). >> I don't have any DBXTalk stuff in my image. >> >> > So you are using Glorp + native postgres driver. > > >> Do we know how old is the Pharo branch from VisualWorks? >> >> > Quite a bit. let me see if I find which one... > > >> Regards! >> >> Esteban A. Maringolo >> >> >> 2014-03-05 15:50 GMT-03:00 Alan Knight <knight(a)acm.org>: >> > I think that whatever version it is it's an old one, because that looks >> like >> > a bug that was fixed some while ago. It's not putting in the subselect >> table >> > in the from clause. I would hope that the Pharo port would list >> somewhere >> > which Glorp version it was built from. There's code to generate a >> > Squeak/Pharo port from the VisualWorks version, and the code tries to be >> > entirely portable in the core, but I never got fixes back from the last >> > porting effort to incorporate, so there are probably some divergences. >> > >> > >> > On 5 March 2014 10:31, Esteban A. Maringolo <emaringolo(a)gmail.com> >> wrote: >> >> >> >> Hi Tom, >> >> >> >> It might have to do with the Pharo version of Glorp. I get the exact >> >> same error as with Alan's suggestion. >> >> ... >> >> I there any way I can know which version of Glorp I'm running in Pharo? >> >> >> 2014-03-03 18:21 GMT-03:00 Tom Robinson <zxrobinson(a)gmail.com>: >> >> > Hi Esteban, >> >> > >> >> > I was able to execute the following query, equivalent to what you're >> >> > trying >> >> > to do on the current dev build of VisualWorks connected to a Store >> >> > repository running in PostgreSQL. >> >> > >> >> > | session query | >> >> > session := StoreLoginFactory currentStoreSession. >> >> > query read: StorePackage >> >> > where: [:each | | subQuery | >> >> > subQuery := Query read: StorePackage >> >> > where: [:eachPkg| eachPkg name = each name]. >> >> > subQuery retrieveMax: [:eachPkg | eachPkg timestamp ]. >> >> > each timestamp = subQuery]. >> >> > session execute: query. >> >> > >> >> > This appears to be what you're trying to do. >> >> -- >> You received this message because you are subscribed to the Google Groups >> "glorp-group" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to glorp-group+unsubscribe(a)googlegroups.com. >> To post to this group, send email to glorp-group(a)googlegroups.com. >> Visit this group at http://groups.google.com/group/glorp-group. >> For more options, visit https://groups.google.com/groups/opt_out. >> > > > > -- > Mariano > http://marianopeck.wordpress.com > -- Mariano http://marianopeck.wordpress.com

1 0

Glorp version in Pharo
by Esteban A. Maringolo March 10, 2014

March 10, 2014

[I'm cross posting Pharo-dev mailing list too] Hi Pharo developers, I'm having an issue while performing a not so common db query (but not esoteric either). And it seems the version I'm using in Pharo is outdated compared with the latest version they're using in VW. Is there an official maintaner and/or updater of the Glorp port for Pharo? I'm using the GlorpDriverPostgreSQL, and the Glorp package from SmalltalkHub (Glorp-MarianoMartinezPeck.88). I don't have any DBXTalk stuff in my image. Do we know how old is the Pharo branch from VisualWorks? Regards! Esteban A. Maringolo 2014-03-05 15:50 GMT-03:00 Alan Knight <knight(a)acm.org>: > I think that whatever version it is it's an old one, because that looks like > a bug that was fixed some while ago. It's not putting in the subselect table > in the from clause. I would hope that the Pharo port would list somewhere > which Glorp version it was built from. There's code to generate a > Squeak/Pharo port from the VisualWorks version, and the code tries to be > entirely portable in the core, but I never got fixes back from the last > porting effort to incorporate, so there are probably some divergences. > > > On 5 March 2014 10:31, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: >> >> Hi Tom, >> >> It might have to do with the Pharo version of Glorp. I get the exact >> same error as with Alan's suggestion. >> ... >> I there any way I can know which version of Glorp I'm running in Pharo? >> 2014-03-03 18:21 GMT-03:00 Tom Robinson <zxrobinson(a)gmail.com>: >> > Hi Esteban, >> > >> > I was able to execute the following query, equivalent to what you're >> > trying >> > to do on the current dev build of VisualWorks connected to a Store >> > repository running in PostgreSQL. >> > >> > | session query | >> > session := StoreLoginFactory currentStoreSession. >> > query read: StorePackage >> > where: [:each | | subQuery | >> > subQuery := Query read: StorePackage >> > where: [:eachPkg| eachPkg name = each name]. >> > subQuery retrieveMax: [:eachPkg | eachPkg timestamp ]. >> > each timestamp = subQuery]. >> > session execute: query. >> > >> > This appears to be what you're trying to do.

1 1

Pharo3.0 and Halos: very annoying to work with
by phil＠highoctane.be March 10, 2014

March 10, 2014

I am working on moving TWM to 3.0 and the way morphs are controlled now is really annoying. Now, we get the new menu when doing the special clicking (e.g. Shift-Alt-Click in Windows, Cmd-Shift Click in OSX). How is one now supposed to be accessing the internal morphs by clicking multiple times and having the halo follow suit? The menu prevents that as it gets in the way. Thanks. Phil

4 8