- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Epicea user experience report

by Norbert Hartl

I don't like the extra menu. I think going three levels in a menu should only be done for thing used rarely. Feels awkward very time. Do we have three cases wirh epicea? - this session meaning all changed since start of the image? - last session/aborted session the last session that did not finished because of unexpected image stop. This should open automatically which does not work in current pharo6 - all other sessions In this situation when the aborted session opens automatically it means I deliberately open epicea through the menu for other reasons. In this case I'm with Doru that it should be a single menu entry which opens epicea and someone can choose there what to see. In case of list display this session will be the top most entry anyway so it can be discussed if it is important to separate this session and the others really. Norbert > Am 20.12.2016 um 17:17 schrieb Martin Dias <tinchodias(a)gmail.com>: > > Hi all, > > One point of Ben's feedback is how Epicea code changes tool is presented in the World Menu. Below you can see the current state + 3 options to discuss it. > > > 1. "Epicea" main entry > "Session Changes" + "All Changes" children [current state] > > > > > 2. Attach purpose to the name: "Epicea Code Changes" > > <Epicea Code Changes.png> > > > 3. Just "Code Changes" + rephrase children: > > â > â > > 4. Flatten > > <Flatten.png> > â > > > Reminder: In World Menu > Help > Help Browser > Epicea you can find a description of the tool and it's model. > > Cheers. > Martin > >> On Mon, Oct 31, 2016 at 1:02 AM, Martin Dias <tinchodias(a)gmail.com> wrote: >> Hello Ben, >> >> About discussion points in 2 (a, b and c): I agree with your arguments... somebody that just moved from Pharo 5 to 6 and crashed an image will look for a "Recover lost changes" in the menu and can have a problem to discover it the replacement in a World->Tools->Epicea->... entry. >> >> Then, as a first step we could flatten the 2 menu entries and then at least anybody will easily find an entry related to changes in World->Tools. >> >> Second, we could try to merge both Epicea GUIs into one (suggestions are welcome). >> >> I still have to read more in detail the remaining of your report to answer. Anyway, thanks a lot for it. >> >> Cheers, >> Martin >> >> >>> On Sat, Oct 29, 2016 at 5:22 AM, Ben Coman <btc(a)openinworld.com> wrote: >>> 1. Created fresh Pharo image (build 60269) >>> >>> >>> 2. Opened World > Tools > Epicea > All changes >>> >>> Points for discussion... >>> >>> a. How many submenu items are expected for Epicea? Can we push the >>> current ones up so the Tools menu remains a flat menu. >>> >>> b. Do we need the two current menu items? "Current session" is >>> encompassed by "All changes"? What expectations do people have of how >>> often they'll use the former rather than the latter? >>> >>> c. When people move from Pharo 5 to Pharo 6 and in a panic want to >>> "recover changes" for a crashed image, they'll be looking for that >>> familiar feature name, not a new app name. Could the app name be left >>> out or placed in brackets "Changes (Epicea)". >>> >>> btw, the interface looks really slick! nice work. >>> >>> >>> 3. Opened World > System Browser. >>> >>> 4. Added package AAA >>> All Changes window - no dynamic change. >>> On <refresh>, still no change, i.e. no sessions >>> #New All Changes window - not visible, no sessions. >>> >>> 5. Added class AA. >>> All Changes window - no dynamic change. >>> On <refresh>, shows new session with AAA & AA. >>> >>> 5. Added method... >>> AA>a >>> ^'something' >>> Prompted for author, entered 'BenComan' >>> All Changes window with session selected - dynamic update showing AA>>a. >>> >>> 6. Added package BBB. >>> All Changes window - no dynamic update. >>> On <refresh>, BBB still not visible in session. >>> >>> 7. Added class A to package AAA. >>> All Changes window - dynamic update showing A. >>> On <refresh>, BBB still not visible in session. >>> >>> 8. Added class BB to package BBB. >>> All Changes window - dynamic update showing BBB & BB. >>> >>> a. Package creation event seems not handled properly, being only >>> pushed through when a class is created in it. >>> >>> b. Since there is a dynamic update for class and method >>> modifications, could the session creation also dynamically update it >>> UI. >>> >>> ----------- >>> 9. Killed the vm from command line >>> $ ps -ef | grep pharo >>> $ kill 29349 >>> Restarted Pharo image >>> >>> 10. World > Tools > Epica > All changes. >>> Authorship is inconsistent: >>> * AAA and AA have blank author >>> * AA>>a, A, BBB, B have author 'BenComan'. >>> >>> a. I understand this follows on from Author not being requested until >>> the first method was defined. Did the old changes track the author of >>> packages and classes at all? >>> >>> b. Since Epicea can track package and class authors, can we trigger >>> the author prompt earlier for them? >>> >>> 11. Selected all previous changes AAA, AA, AA>>a, A, BBB, BB >>> and did <Apply Changes>. >>> Prompted for author. Entered 'DrWho' >>> Existing All Changes window - no change >>> New All Changes window - shows new session with all six changes. >>> Authorship is a little inconsistent: >>> * AAA and AA have author 'Unknown'. >>> * AA>>a, A, BBB, B have blank author. >>> >>> 12. Killed the vm from command line >>> $ ps -ef | grep pharo >>> $ kill 30696 >>> Restarted Pharo image >>> >>> 13. World > Tools > Epica > All changes. >>> Authorship is a little inconsistent: >>> First session >>> * AAA and AA have blank author >>> * AA>>a, A, BBB, B have author 'BenComan'. >>> Second session >>> * AAA and AA have blank author >>> * AA>>a, A, BBB, B have author 'DrWho'. >>> >>> a. Epicea changes are reapplied as the current author. This seems a >>> semantic change from Pharo 5 where changes were reapplied as their >>> original author. Is this accidental or by design? Can we have some >>> community discussion on this point (I don't remember seeing any)? >>> >>> cheers -ben >>> >>> P.S. I'll wait to see what arises out of discussion before creating any issues. >>> >> >

Dec. 21, 2016

Re: [Pharo-dev] [Moose-dev] Too frequent crashes :-(

by Nicolai Hess

Am 21.12.2016 8:10 vorm. schrieb "Alexandre Bergel" <alexandre.bergel(a)me.com >: I had plenty of elements in my Roassal visualization, and on some point it just crashed. I have the impression that having many elements and / or animations (as in the case of Johan Fabry) triggers the crash. One problem, the one peter (partly) fixed, is n issue with bitblt, Cairo (as a surface plugin) and garbage collection. Bitblt my call the surface plugin to request some information about the Form. NB-based ffi for Cairo just called gen. asm code. But with UFFI, we now callback to Smalltalk code, which may trigger a GC. So every time bitblt, makes a call to the surface plugin (Cairo) and this callback allocates a new ST-Object, the GC may run and moves(!)the objects, to which bitblt still holds a reference. Accessing this objects in bitblt, after the callback returns (may) result in a segfault. Depending on the concrete bltoperation, the object memory layout, the size(!) of the initial Cairo surface. With peters fix/hack, the code does not crash but it still may access and operates on moved objects. Pining the objects may help, but you need to know where in the (vm image) code a form is used for drawing, and if it is a real form or external data managed by the surface plugin. And of course, when to "unpin" the object. A saver way would be, to just disable any GC during the executing of the callback, if it was called by a plugin (here:bitblt) Alexandre > On Dec 20, 2016, at 9:12 AM, Stephane Ducasse <stepharo.self(a)gmail.com> wrote: > > So what was your crash problem? > > > On Tue, Dec 20, 2016 at 8:43 AM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > so did it work? > > Is there something that we should do? > > I had no problem, but as Henrik and Peter said, the problem is likely to be more complex > > Alexandre > > > > > > On Fri, Dec 16, 2016 at 5:06 PM, Alexandre Bergel < alexandre.bergel(a)me.com> wrote: > > Okay, tryingâ¦ > > > > Alexandre > > > > > > > On Dec 16, 2016, at 4:40 PM, Aliaksei Syrel <alex.syrel(a)gmail.com> wrote: > > > > > > According to crash.dmp TRMorph>drawOn: draws an image using FormCanvas>>#image:at:sourceRect:rule: at a coordinate represented by a Float point. > > > Try to round it to integer point. > > > > > > Cheers, > > > Alex > > > > > > On 16 December 2016 at 16:34, Alexandre Bergel < alexandre.bergel(a)me.com> wrote: > > > Hi! > > > > > > Vincent and I are experiencing many crashes from GrafPort>copyBits. It crashes every couple of minutes. > > > > > > I know that not much can be done so far. Stability is my biggest concern with Pharo :-( > > > I use the latest VM. > > > > > > Cheers, > > > Alexandre > > > > > > > > > > > > > > > > > > > > > -- > > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > > Alexandre Bergel http://www.bergel.eu > > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > > > > _______________________________________________ > > > Moose-dev mailing list > > > Moose-dev(a)list.inf.unibe.ch > > > https://www.list.inf.unibe.ch/listinfo/moose-dev > > > > > > > > > > -- > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > Alexandre Bergel http://www.bergel.eu > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

Dec. 21, 2016

Re: [Pharo-dev] [Moose-dev] Too frequent crashes :-(

by Alexandre Bergel

Just to make clear, I am using the latest VM on Mac OS X, El capitan, latest Pharo. The crashes I experienced are not new. My original email was not a complain (even if it may be perceived as one), but more to trigger a discussion. Peter did an excellent report of the current state of this issue. Cheers, Alexandre > On Dec 16, 2016, at 4:34 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > Hi! > > Vincent and I are experiencing many crashes from GrafPort>copyBits. It crashes every couple of minutes. > > I know that not much can be done so far. Stability is my biggest concern with Pharo :-( > I use the latest VM. > > Cheers, > Alexandre > > > <crash.dmp> > > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > _______________________________________________ > Moose-dev mailing list > Moose-dev(a)list.inf.unibe.ch > https://www.list.inf.unibe.ch/listinfo/moose-dev -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

Dec. 21, 2016

Re: [Pharo-dev] [Moose-dev] Too frequent crashes :-(

by Alexandre Bergel

I had plenty of elements in my Roassal visualization, and on some point it just crashed. I have the impression that having many elements and / or animations (as in the case of Johan Fabry) triggers the crash. Alexandre > On Dec 20, 2016, at 9:12 AM, Stephane Ducasse <stepharo.self(a)gmail.com> wrote: > > So what was your crash problem? > > > On Tue, Dec 20, 2016 at 8:43 AM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > so did it work? > > Is there something that we should do? > > I had no problem, but as Henrik and Peter said, the problem is likely to be more complex > > Alexandre > > > > > > On Fri, Dec 16, 2016 at 5:06 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > Okay, tryingâ¦ > > > > Alexandre > > > > > > > On Dec 16, 2016, at 4:40 PM, Aliaksei Syrel <alex.syrel(a)gmail.com> wrote: > > > > > > According to crash.dmp TRMorph>drawOn: draws an image using FormCanvas>>#image:at:sourceRect:rule: at a coordinate represented by a Float point. > > > Try to round it to integer point. > > > > > > Cheers, > > > Alex > > > > > > On 16 December 2016 at 16:34, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > > Hi! > > > > > > Vincent and I are experiencing many crashes from GrafPort>copyBits. It crashes every couple of minutes. > > > > > > I know that not much can be done so far. Stability is my biggest concern with Pharo :-( > > > I use the latest VM. > > > > > > Cheers, > > > Alexandre > > > > > > > > > > > > > > > > > > > > > -- > > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > > Alexandre Bergel http://www.bergel.eu > > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > > > > _______________________________________________ > > > Moose-dev mailing list > > > Moose-dev(a)list.inf.unibe.ch > > > https://www.list.inf.unibe.ch/listinfo/moose-dev > > > > > > > > > > -- > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > Alexandre Bergel http://www.bergel.eu > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

Dec. 21, 2016

Re: [Pharo-dev] Epicea user experience report

by Tudor Girba

Hi, I think I would prefer to not have to think about this choice at all when in the world menu level. The user interface from the Epicea window already allows me to switch between current session and all sessions, and usually the decision of what to look at is based on whether I find what I am looking for or not. So, rather then asking the user to think about it upfront, I would prefer to have one command that opens the Epicea window on the current session and allow the user to dive deeper. This would also imply that we would have only one command in the menu item. What do you think? Cheers, Doru > On Dec 20, 2016, at 10:16 PM, stepharong <stepharong(a)free.fr> wrote: > > I like the second because it associates the name with the function. > > Stef > > > Hi all, > > One point of Ben's feedback is how Epicea code changes tool is presented in the World Menu. Below you can see the current state + 3 options to discuss it. > > > 1. "Epicea" main entry > "Session Changes" + "All Changes" children [current state] > > <Mail Attachment.png> > > > 2. Attach purpose to the name: "Epicea Code Changes" > > <Mail Attachment.png> > > > 3. Just "Code Changes" + rephrase children: > > â<Mail Attachment.png> > â > > 4. Flatten > > <Mail Attachment.png> > â > > > Reminder: In World Menu > Help > Help Browser > Epicea you can find a description of the tool and it's model. > > Cheers. > Martin > > On Mon, Oct 31, 2016 at 1:02 AM, Martin Dias <tinchodias(a)gmail.com> wrote: > Hello Ben, > > About discussion points in 2 (a, b and c): I agree with your arguments... somebody that just moved from Pharo 5 to 6 and crashed an image will look for a "Recover lost changes" in the menu and can have a problem to discover it the replacement in a World->Tools->Epicea->... entry. > > Then, as a first step we could flatten the 2 menu entries and then at least anybody will easily find an entry related to changes in World->Tools. > > Second, we could try to merge both Epicea GUIs into one (suggestions are welcome). > > I still have to read more in detail the remaining of your report to answer. Anyway, thanks a lot for it. > > Cheers, > Martin > > > On Sat, Oct 29, 2016 at 5:22 AM, Ben Coman <btc(a)openinworld.com> wrote: > 1. Created fresh Pharo image (build 60269) > > > 2. Opened World > Tools > Epicea > All changes > > Points for discussion... > > a. How many submenu items are expected for Epicea? Can we push the > current ones up so the Tools menu remains a flat menu. > > b. Do we need the two current menu items? "Current session" is > encompassed by "All changes"? What expectations do people have of how > often they'll use the former rather than the latter? > > c. When people move from Pharo 5 to Pharo 6 and in a panic want to > "recover changes" for a crashed image, they'll be looking for that > familiar feature name, not a new app name. Could the app name be left > out or placed in brackets "Changes (Epicea)". > > btw, the interface looks really slick! nice work. > > > 3. Opened World > System Browser. > > 4. Added package AAA > All Changes window - no dynamic change. > On <refresh>, still no change, i.e. no sessions > #New All Changes window - not visible, no sessions. > > 5. Added class AA. > All Changes window - no dynamic change. > On <refresh>, shows new session with AAA & AA. > > 5. Added method... > AA>a > ^'something' > Prompted for author, entered 'BenComan' > All Changes window with session selected - dynamic update showing AA>>a. > > 6. Added package BBB. > All Changes window - no dynamic update. > On <refresh>, BBB still not visible in session. > > 7. Added class A to package AAA. > All Changes window - dynamic update showing A. > On <refresh>, BBB still not visible in session. > > 8. Added class BB to package BBB. > All Changes window - dynamic update showing BBB & BB. > > a. Package creation event seems not handled properly, being only > pushed through when a class is created in it. > > b. Since there is a dynamic update for class and method > modifications, could the session creation also dynamically update it > UI. > > ----------- > 9. Killed the vm from command line > $ ps -ef | grep pharo > $ kill 29349 > Restarted Pharo image > > 10. World > Tools > Epica > All changes. > Authorship is inconsistent: > * AAA and AA have blank author > * AA>>a, A, BBB, B have author 'BenComan'. > > a. I understand this follows on from Author not being requested until > the first method was defined. Did the old changes track the author of > packages and classes at all? > > b. Since Epicea can track package and class authors, can we trigger > the author prompt earlier for them? > > 11. Selected all previous changes AAA, AA, AA>>a, A, BBB, BB > and did <Apply Changes>. > Prompted for author. Entered 'DrWho' > Existing All Changes window - no change > New All Changes window - shows new session with all six changes. > Authorship is a little inconsistent: > * AAA and AA have author 'Unknown'. > * AA>>a, A, BBB, B have blank author. > > 12. Killed the vm from command line > $ ps -ef | grep pharo > $ kill 30696 > Restarted Pharo image > > 13. World > Tools > Epica > All changes. > Authorship is a little inconsistent: > First session > * AAA and AA have blank author > * AA>>a, A, BBB, B have author 'BenComan'. > Second session > * AAA and AA have blank author > * AA>>a, A, BBB, B have author 'DrWho'. > > a. Epicea changes are reapplied as the current author. This seems a > semantic change from Pharo 5 where changes were reapplied as their > original author. Is this accidental or by design? Can we have some > community discussion on this point (I don't remember seeing any)? > > cheers -ben > > P.S. I'll wait to see what arises out of discussion before creating any issues. > > > > > > > -- > Using Opera's mail client: http://www.opera.com/mail/ -- www.tudorgirba.com www.feenk.com "In a world where everything is moving ever faster, one might have better chances to win by moving slower."

Dec. 21, 2016

Re: [Pharo-dev] a request to change the meaning of <example> pragma

by Sean P. DeNigris

Stephane Ducasse-3 wrote > In nautilus in Pharo 60 when you use > > <sampleInstance> > you can get an inspector on the object returned by the > method Cool! Great feature :) ----- Cheers, Sean -- View this message in context: http://forum.world.st/a-request-to-change-the-meaning-of-example-pragma-tp4… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Dec. 21, 2016

Re: [Pharo-dev] About Fraction and parentheses

by stepharong

On Tue, 20 Dec 2016 09:24:57 +0100, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Ah, OK, sorry about the misunderstanding. you are welcome :) Now I do not get why there would be ((1/3)) I have to check > >> On 20 Dec 2016, at 09:21, Stephane Ducasse <stepharo.self(a)gmail.com> >> wrote: >> >> I think that I know that :) >> And this is why I mentioned that if you print 1/3 it prints (1/3) so >> that we can continue typing + (2/3) >> >> On Mon, Dec 19, 2016 at 11:02 PM, Sven Van Caekenberghe <sven(a)stfx.eu> >> wrote: >> >> > On 19 Dec 2016, at 22:00, Stephane Ducasse <stepharo.self(a)gmail.com> >> wrote: >> > >> > Hi serge >> > >> > the printing should be normal because like that you are ready to type >> > 1/3 + 2/3 and get 1. >> >> No, not at all: >> >> 1/3 + 2/3 >> >> "(7/9)" >> >> These are all just binary messages evaluated left to right, like this: >> >> ((1/3) + 2)/3 >> >> "(7/9)" >> >> That is why you need the parenthesis in certain contexts. >> Now adding two of them is definitively wrong. >> >> (1/3) + (2/3) >> >> "1" >> >> (The first parenthesis are not needed, btw). >> >> > Now I do not get why you get ((1/2)) thanks for reporting it. >> > >> > Stef >> > >> > On Mon, Dec 19, 2016 at 9:08 AM, Serge Stinckwich >> <serge.stinckwich(a)gmail.com> wrote: >> > I have one question about Fraction. >> > >> > When I print a fraction like 1/2 in the Playground I obtain (1/2) and >> > when I inspect I obtain ((1/2)). Why do we need all these parentheses >> > ? >> > >> > I didn't try with a current Pharo 6.0 image, because my bandwidth is >> > quite limited right now (Vietnam). >> > >> > -- >> > Serge Stinckwich >> > UCBN & UMI UMMISCO 209 (IRD/UPMC) >> > Every DSL ends up being Smalltalk >> > http://www.doesnotunderstand.org/ >> > >> > >> >> >> > > -- Using Opera's mail client: http://www.opera.com/mail/

Dec. 20, 2016

Re: [Pharo-dev] [Vm-dev] Problem with OSSubprocess / signals / heartbeat ?

by stepharong

> I asked Santa an OSSubprocess that won't hang :) I really appreciate to see such not so small polar elfves helping santa :) Stef > Thanks, > Guille > > > -------- Original Message -------- >> >>> On 19 Dec 2016, at 14:41, Mariano Martinez Peck >>> <marianopeck(a)gmail.com> wrote: >>> >>> Hi guys, >>> >>> Guille Polito kept one of these images if someone can give us a hand. >>> He also proposed the great idea of using `strace` to see what was >>> going on. He (together with Pable Tesone) suspected that the >>> heartbeat could be interrupting the `clone()` function which is (I >>> think) called internally by the 'posix_spawn()' which is the one used >>> by OSSubprocess. >>> >>> When these images are "hung" they found at a infinitive loop like this: >> Okay, how many child processes do you have at that point? How many >> processes does the system have? >> >> >>> [pid 17477] --- SIGALRM {si_signo=SIGALRM, si_code=SI_KERNEL} --- >>> [pid 17477] gettimeofday({1482152630, 593498}, NULL) = 0 >>> [pid 17477] sigreturn() (mask []) = 120 >>> [pid 17477] clone(child_stack=0, >>> flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, >>> child_tidptr=0xf7578768) = ? ERESTARTNOINTR (To be restarted) >>> [pid 17477] --- SIGALRM {si_signo=SIGALRM, si_code=SI_KERNEL} --- >>> [pid 17477] gettimeofday({1482152630, 600126}, NULL) = 0 >>> [pid 17477] sigreturn() (mask []) = 120 >>> [pid 17477] clone(child_stack=0, >>> flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, >>> child_tidptr=0xf7578768) = ? ERESTARTNOINTR (To be restarted) >> so above.. 7ms between the two gettimeofday calls? Nothing else? Set a >> breakpoint on clone/fork in gdb and look at the c-stack at this point? >> Could you strace with timestamps to see how much time is spent? Is the >> process suspicious in other ways? >> >> So yes.. sounds like clone doesn't complete.. the question is why? Is >> it out of resources? Is something in the VM blocking longer than the >> heartbeat, is the heartbeat more frequent than expected? >> >> >>> As you can see, there is a SIGALARM involved. It also looks like the >>> `gettimeofday` is used by the heartbeat ? Could it be that someone >>> the heartbeat is interrupting the `clone()` ? >>> >>> Guille also showed me the `strace` output with a regular / working >>> image: >>> >>> >>> [pid 18647] --- SIGALRM {si_signo=SIGALRM, si_code=SI_KERNEL} --- >>> [pid 18647] gettimeofday({1482152829, 481014}, NULL) = 0 >>> [pid 18647] sigreturn() (mask []) = -1 EINTR (Interrupted system >>> call) >>> [pid 18647] getitimer(ITIMER_REAL, {it_interval={0, 2000}, >>> it_value={0, 1917}}) = 0 >>> [pid 18647] recvmsg(3, 0xff7b0734, 0) = -1 EAGAIN (Resource >>> temporarily unavailable) >>> [pid 18647] select(4, [3], [], [3], {0, 1000}) = 0 (Timeout) >>> [pid 18647] getitimer(ITIMER_REAL, {it_interval={0, 2000}, >>> it_value={0, 797}}) = 0 >>> [pid 18647] recvmsg(3, 0xff7b0734, 0) = -1 EAGAIN (Resource >>> temporarily unavailable) >>> [pid 18647] select(4, [3], [], [3], {0, 1000}) = ? ERESTARTNOHAND (To >>> be restarted if no handler) >>> [pid 18647] --- SIGALRM {si_signo=SIGALRM, si_code=SI_KERNEL} --- >>> >>> >>> Does anyone have any hint here? >> Get timestamps in there. How long does it take to fail/end in this >> situation? >> >> holger >> >> >> >> >> > > > -- Using Opera's mail client: http://www.opera.com/mail/

Dec. 20, 2016

Re: [Pharo-dev] Epicea user experience report

by stepharong

I like the second because it associates the name with the function. Stef > Hi all, > > One point of Ben's feedback is how Epicea code changes tool is presented > in the World Menu. Below you can see the current state + 3 options to > discuss it. > > > 1. "Epicea" main entry > "Session Changes" + "All Changes" children > [current state] > > > > > 2. Attach purpose to the name: "Epicea Code Changes" > > > > > 3. Just "Code Changes" + rephrase children: > > â > â > > 4. Flatten > > > â > > > Reminder: In World Menu > Help > Help Browser > Epicea you can find a > description of the tool and it's model. > > Cheers. > Martin > > On Mon, Oct 31, 2016 at 1:02 AM, Martin Dias <tinchodias(a)gmail.com> > wrote: >> Hello Ben, >> >> About discussion points in 2 (a, b and c): I agree with your >> arguments... somebody that just moved from Pharo 5 to 6 and crashed an >> image will look for a "Recover lost changes" in the menu >>and can have >> a problem to discover it the replacement in a World->Tools->Epicea->... >> entry. >> >> Then, as a first step we could flatten the 2 menu entries and then at >> least anybody will easily find an entry related to changes in >> World->Tools. >> Second, we could try to merge both Epicea GUIs into one (suggestions >> are welcome). >> >> I still have to read more in detail the remaining of your report to >> answer. Anyway, thanks a lot for it. >> >> Cheers, >> Martin >> >> >> On Sat, Oct 29, 2016 at 5:22 AM, Ben Coman <btc(a)openinworld.com> wrote: >>> 1. Created fresh Pharo image (build 60269) >>> >>> >>> 2. Opened World > Tools > Epicea > All changes >>> >>> Points for discussion... >>> >>> a. How many submenu items are expected for Epicea? Can we push the >>> current ones up so the Tools menu remains a flat menu. >>> >>> b. Do we need the two current menu items? "Current session" is >>> encompassed by "All changes"? What expectations do people have of how >>> often they'll use the former rather than the latter? >>> >>> c. When people move from Pharo 5 to Pharo 6 and in a panic want to >>> "recover changes" for a crashed image, they'll be looking for that >>> familiar feature name, not a new app name. Could the app name be left >>> out or placed in brackets "Changes (Epicea)". >>> >>> btw, the interface looks really slick! nice work. >>> >>> >>> 3. Opened World > System Browser. >>> >>> 4. Added package AAA >>> All Changes window - no dynamic change. >>> On <refresh>, still no change, i.e. no sessions >>> #New All Changes window - not visible, no sessions. >>> >>> 5. Added class AA. >>> All Changes window - no dynamic change. >>> On <refresh>, shows new session with AAA & AA. >>> >>> 5. Added method... >>> AA>a >>> ^'something' >>> Prompted for author, entered 'BenComan' >>> All Changes window with session selected - dynamic update showing >>> AA>>a. >>> >>> 6. Added package BBB. >>> All Changes window - no dynamic update. >>> On <refresh>, BBB still not visible in session. >>> >>> 7. Added class A to package AAA. >>> All Changes window - dynamic update showing A. >>> On <refresh>, BBB still not visible in session. >>> >>> 8. Added class BB to package BBB. >>> All Changes window - dynamic update showing BBB & BB. >>> >>> a. Package creation event seems not handled properly, being only >>> pushed through when a class is created in it. >>> >>> b. Since there is a dynamic update for class and method >>> modifications, could the session creation also dynamically update it >>> UI. >>> >>> ----------- >>> 9. Killed the vm from command line >>> $ ps -ef | grep pharo >>> $ kill 29349 >>> Restarted Pharo image >>> >>> 10. World > Tools > Epica > All changes. >>> Authorship is inconsistent: >>> * AAA and AA have blank author >>> * AA>>a, A, BBB, B have author 'BenComan'. >>> >>> a. I understand this follows on from Author not being requested until >>> the first method was defined. Did the old changes track the author of >>> packages and classes at all? >>> >>> b. Since Epicea can track package and class authors, can we trigger >>> the author prompt earlier for them? >>> >>> 11. Selected all previous changes AAA, AA, AA>>a, A, BBB, BB >>> and did <Apply Changes>. >>> Prompted for author. Entered 'DrWho' >>> Existing All Changes window - no change >>> New All Changes window - shows new session with all six changes. >>> Authorship is a little inconsistent: >>> * AAA and AA have author 'Unknown'. >>> * AA>>a, A, BBB, B have blank author. >>> >>> 12. Killed the vm from command line >>> $ ps -ef | grep pharo >>> $ kill 30696 >>> Restarted Pharo image >>> >>> 13. World > Tools > Epica > All changes. >>> Authorship is a little inconsistent: >>> First session >>> * AAA and AA have blank author >>> * AA>>a, A, BBB, B have author 'BenComan'. >>> Second session >>> * AAA and AA have blank author >>> * AA>>a, A, BBB, B have author 'DrWho'. >>> >>> a. Epicea changes are reapplied as the current author. This seems a >>> semantic change from Pharo 5 where changes were reapplied as their >>> original author. Is this accidental or by design? Can we have some >>> community discussion on this point (I don't remember seeing any)? >>> >>> cheers -ben >>> >>> P.S. I'll wait to see what arises out of discussion before creating >>> any issues. >>> >> > -- Using Opera's mail client: http://www.opera.com/mail/

Dec. 20, 2016

Re: [Pharo-dev] Epicea user experience report

by Sean P. DeNigris

Marcus Denker-4 wrote >> 3. Just "Code Changes" + rephrase children: > I like that most. It is the easiest to understand for newcomers. +1 ----- Cheers, Sean -- View this message in context: http://forum.world.st/Epicea-user-experience-report-tp4920653p4927604.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Dec. 20, 2016