June 2018 - Pharo-users - lists.pharo.org

Its too hard to contribute to the Iceberg wiki

by Tim Mackinnon

Ok - my love affair with Github ends at the wiki - what a load of $%$%^$%â¦ there is no sane and easy way to fork a wiki, make some changes and then submit a PR. (how the hell did they get into that state?). Can we agree to scrap the wiki Iceberg wiki and just have a longer readme.md (with sections of the current wiki?). Or if someone can explain easy instructions for both me (as a contributor) and the admins of iceberg then Iâm all earsâ¦. If not - Iâll submit a PR with my subtle changes to the readme. Tim

June 29, 2018

Re: [Pharo-users] Pharo Launcher multiplicated PhLDeploymentScript

by Peter UhnÃ¡k

PhLStartupManager class>>addToLauncherStartUpList: aClass self startUpClasses addIfNotPresent: aClass seems to be enough. Peter On Fri, Jun 29, 2018 at 9:58 AM, Peter UhnÃ¡k <i.uhnak(a)gmail.com> wrote: > Hi, > > I've noticed that when I save Pharo Launcher image (ctrl+shift+s), because > I was changing something, then PhLDeploymentScript is added to the the > startup classes list. > > PhLStartupManager startUpClasses "a SortedCollection( > PhLDirectoryBasedImageRepository PhLSettingBrowser PhLDeploymentScript > PhLDeploymentScript ...(atm 34x)... PhLDeploymentScript)" > > PhLStartupManager startUpClasses asSet "a Set( > PhLDirectoryBasedImageRepository PhLDeploymentScript PhLSettingBrowser)" > > I didn't notice any impact on functionality or speed (although for some > reason there is a significant delay between clicking on the PhL icon and > the image opening), but nevertheless it probably should be deduplicated. > > Thanks. > > Peter >

June 29, 2018

Re: [Pharo-users] image startup profiling

by Sven Van Caekenberghe

Hi Peter, I once did this: http://forum.world.st/WorkingSession-log-tt4957187.html HTH, Sven > On 29 Jun 2018, at 09:54, Peter UhnÃ¡k <i.uhnak(a)gmail.com> wrote: > > Hi, > > does anyone know any tips how to profile why opening of an image takes X amount of time and on what? > > Thanks, > Peter

June 29, 2018

Pharo Launcher multiplicated PhLDeploymentScript

by Peter UhnÃ¡k

Hi, I've noticed that when I save Pharo Launcher image (ctrl+shift+s), because I was changing something, then PhLDeploymentScript is added to the the startup classes list. PhLStartupManager startUpClasses "a SortedCollection(PhLDirectoryBasedImageRepository PhLSettingBrowser PhLDeploymentScript PhLDeploymentScript ...(atm 34x)... PhLDeploymentScript)" PhLStartupManager startUpClasses asSet "a Set(PhLDirectoryBasedImageRepository PhLDeploymentScript PhLSettingBrowser)" I didn't notice any impact on functionality or speed (although for some reason there is a significant delay between clicking on the PhL icon and the image opening), but nevertheless it probably should be deduplicated. Thanks. Peter

June 29, 2018

image startup profiling

by Peter UhnÃ¡k

Hi, does anyone know any tips how to profile why opening of an image takes X amount of time and on what? Thanks, Peter

June 29, 2018

PolyMath move to github !

by Serge Stinckwich

Thank to the terrific work of Cyril, all the PolyMath code is now available on github: https://github.com/PolyMathOrg/PolyMath With a lot of magic tricks, Cyril was able to preserve part of the history of the project: https://github.com/PolyMathOrg/PolyMath/graphs/contributors This is nice to all the commits that we have done sine more than 3 years, available as git commits now ! âThank you again Cyril ! >From now, we will use only github for the commits. You will have to use Pharo 6.1 or Pharo 7.0 to use Iceberg.â I will prepare a guide for contributing to PolyMath soon. ââ -- Serge Stinckwich UMI UMMISCO 209 (SU/IRD/UY1) "Programs must be written for people to read, and only incidentally for machines to execute."http://www.doesnotunderstand.org/

June 28, 2018

Re: [Pharo-users] Iceberg - finding deleted classes, reverting versions?

by Tim Mackinnon

Thanks Ben - I hadnât really thought of using the GitHub web ui itself (so many places to choose from ;) Effectively the ~1 syntax Guille mentioned is exactly what youâve explained (so that hangs together nicely). Now looking back at the tutorial I had followed, I didnât notice the ~1 in there. This one will be firmly embedded in my brain for sure. Still, its stuff like this that helps you learn the wider toolset in play. Tim p.s. Appreciate you guys following up with me - it was bugging me in the background. I kept returning to it, annoyed that it didnât work and should (now completely explained) > On 28 Jun 2018, at 12:25, Ben Coman <btc(a)openInWorld.com> wrote: > > > > On 28 June 2018 at 07:38, Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>> wrote: > Hey thanks Ben - this still doesnât work for meâ¦ good idea to try on the Pharo repo itself. > > I have a fork of it, which I catch up periodically - so I tried a similar suggestion to you - went into my directory (underneath the src folder) and typed: > > log --diff-filter=D --summary --graph --pretty='format:%cd | %h | %H | %cn%n-> %s%nâ > > This gave me: > > Tue Jun 19 19:28:37 2018 +0200 | 13e81343e | 13e81343e8a23a8d6ca43f1c77b804b8134fc986 | CyrilFerlicot > -> Remove now useless logs > > delete mode 100755 bootstrap/scripts/printFolderContent.sh > > Mon Jun 18 16:50:19 2018 +0200 | 0433d908d | 0433d908de2987fa0069a47dee1f0794ccbb52c2 | GitHub > -> Revert "22161-Add-more-log-to-bootstrap" > > delete mode 100755 bootstrap/scripts/printFolderContent.sh > > Thu Jun 14 11:41:16 2018 +0200 | b4a285a61 | b4a285a61d3eb0cc1e642b2fcc242dac7092617f | Marcus Denker > -> fix the last sender of #asIcon https://pharo.fogbugz.com/f/cases/22139/fix-the-last-sender-of-asIcon <https://pharo.fogbugz.com/f/cases/22139/fix-the-last-sender-of-asIcon> > > delete mode 100644 src/Polymorph-Widgets-Rules/IconShortcutRule.class.st <http://iconshortcutrule.class.st/> > > So then I tried to recover that last one (again at root underneath the src directory) > > git checkout b4a285a61 src/Polymorph-Widgets-Rules/IconShortcutRule.class.st <http://iconshortcutrule.class.st/> > error: pathspec 'src/Polymorph-Widgets-Rules/IconShortcutRule.class.st <http://iconshortcutrule.class.st/>' did not match any file(s) known to git. > > Concrete examples help. I get the same error > when I try checking that file out from b4a285a61 > > Now looking at https://github.com/pharo-project/pharo/tree/b4a285a61/src/Polymorph-Widgets… <https://github.com/pharo-project/pharo/tree/b4a285a61/src/Polymorph-Widgets…> > there is no file IconShortcutRule.class.st <http://iconshortcutrule.class.st/> > > So from try clicking the <History> button, > and currently at the top you'll see " b4a28" on Jun 14, > and the next one down is "d4b4464". > Click the "<>" button to the right of that, > then browse back to .../src/Polymorph-Widgets-Rules > and now you'll see IconShortcutRule.class.st <http://iconshortcutrule.class.st/> > > So the following works for me... > git checkout d4b4464 src/Polymorph-Widgets-Rules/IconShortcutRule.class.st <http://iconshortcutrule.class.st/> > > In summary ;), you can't check out a file from a commit that doesn't exist in that commit. > > HTH > cheers -ben

June 28, 2018

Re: [Pharo-users] Iceberg - finding deleted classes, reverting versions?

by Ben Coman

On 28 June 2018 at 07:38, Tim Mackinnon <tim(a)testit.works> wrote: > Hey thanks Ben - this still doesnât work for meâ¦ good idea to try on the > Pharo repo itself. > > I have a fork of it, which I catch up periodically - so I tried a similar > suggestion to you - went into my directory (underneath the src folder) and > typed: > > log --diff-filter=D --summary --graph --pretty='format:%cd | %h | %H | > %cn%n-> %s%nâ > > This gave me: > > Tue Jun 19 19:28:37 2018 +0200 | 13e81343e | 13e81343e8a23a8d6ca43f1c77b804b8134fc986 > | CyrilFerlicot > -> Remove now useless logs > > delete mode 100755 bootstrap/scripts/printFolderContent.sh > > Mon Jun 18 16:50:19 2018 +0200 | 0433d908d | 0433d908de2987fa0069a47dee1f0794ccbb52c2 > | GitHub > -> Revert "22161-Add-more-log-to-bootstrap" > > delete mode 100755 bootstrap/scripts/printFolderContent.sh > > Thu Jun 14 11:41:16 2018 +0200 | b4a285a61 | b4a285a61d3eb0cc1e642b2fcc242dac7092617f > | Marcus Denker > -> fix the last sender of #asIcon https://pharo.fogbugz.com/f/ > cases/22139/fix-the-last-sender-of-asIcon > > delete mode 100644 src/Polymorph-Widgets-Rules/IconShortcutRule.class.st > > So then I tried to recover that last one (again at root underneath the src > directory) > > git checkout b4a285a61 src/Polymorph-Widgets-Rules/Ic > onShortcutRule.class.st > error: pathspec 'src/Polymorph-Widgets-Rules/IconShortcutRule.class.st' > did not match any file(s) known to git. > Concrete examples help. I get the same error when I try checking that file out from b4a285a61 Now looking at https://github.com/pharo-project/pharo/tree/b4a285a61/src/Polymorph-Widgets… there is no file IconShortcutRule.class.st <http://iconshortcutrule.class.st/> So from try clicking the <History> button, and currently at the top you'll see " b4a28" on Jun 14, and the next one down is "d4b4464". Click the "<>" button to the right of that, then browse back to .../src/Polymorph-Widgets-Rules and now you'll see IconShortcutRule.class.st <http://iconshortcutrule.class.st/> So the following works for me... git checkout d4b4464 src/Polymorph-Widgets-Rules/IconShortcutRule.class.st <http://iconshortcutrule.class.st/> In summary ;), you can't check out a file from a commit that doesn't exist in that commit. HTH cheers -ben

June 28, 2018

Re: [Pharo-users] Iceberg - finding deleted classes, reverting versions?

by Guillermo Polito

> You mentioned a Calypso plugin for versions - where is that? I loaded a recentish P7 image and I donât see those icons? Is this something I can easily load to try out? As soon as a class is inside a package that is managed by iceberg, you should see in Calypso a new button and menu item on methods (called History, with a little box icon). > Not sure how to get a repository but I cobbled together the following? > (IceLibgitRepository registry detect: [ :r | r name beginsWith: 'Prismic' ]) lookupCommit: '65363adâ. > This give me an IceGitCommit - but then what can I easily do with this to try and iterate over deleted classes to try and get the one Iâve identified back? Well there is no magic :). Either you walk the commits (#parent or #newCommitWalk (!! too low level)) and calculate for each a diff against its parent (#diffTo:). Other thing you can do to check how IceGitLog is implemented, using directly libgit. About the git error, I don't have time to check now but, I think it may be something easy that is going under your nose :) You got your commitish from the `git log --diff-filter=D --summary âpretty="format:%cd|%h|%cn%n%s%nâ`, didn't you? Then, this means the commitish you got has the file already deleted! So if you want to recover it, you have to check its parent! Can you try (notice the added the ~1) $ git checkout 65363ad~1 src/PrismicDemo/PrismicBlock.class.st ? On Thu, Jun 28, 2018 at 2:32 AM Tim Mackinnon <tim(a)testit.works> wrote: > For the record - I get the same error using those steps on Ubuntu as well. > So I must be doing something wrong? > > (Note in my message when I pasted the command used - I neglected the git > in front of: log --diff-filter=D --summary âpretty="format:%cd | %h | %H > | %cn%n-> %s%nâ) > > > > Tim > > On 28 Jun 2018, at 00:46, Tim Mackinnon <tim(a)testit.works> wrote: > > And just to add to this - I just did a git clone (git clone git(a)github.com > :pharo-project/pharo.git) in a fresh directory and can see that same > deleted file - and when I try to do a checkout, it gives the path spec > error. > > This is very weird - is it some osx thing? I will try it on ubuntu - but I > feel I must be missing something, like do I need a different prefix on path > names (but you didnât seem to need one)? > > Iâd be very worried if git is acting upâ¦ but I still suspect user error > here. > > Tim > > > On 28 Jun 2018, at 00:38, Tim Mackinnon <tim(a)testit.works> wrote: > > Hey thanks Ben - this still doesnât work for meâ¦ good idea to try on the > Pharo repo itself. > > I have a fork of it, which I catch up periodically - so I tried a similar > suggestion to you - went into my directory (underneath the src folder) and > typed: > > log --diff-filter=D --summary --graph --pretty='format:%cd | %h | %H | > %cn%n-> %s%nâ > > This gave me: > > Tue Jun 19 19:28:37 2018 +0200 | 13e81343e | > 13e81343e8a23a8d6ca43f1c77b804b8134fc986 | CyrilFerlicot > -> Remove now useless logs > > delete mode 100755 bootstrap/scripts/printFolderContent.sh > > Mon Jun 18 16:50:19 2018 +0200 | 0433d908d | > 0433d908de2987fa0069a47dee1f0794ccbb52c2 | GitHub > -> Revert "22161-Add-more-log-to-bootstrap" > > delete mode 100755 bootstrap/scripts/printFolderContent.sh > > Thu Jun 14 11:41:16 2018 +0200 | b4a285a61 | > b4a285a61d3eb0cc1e642b2fcc242dac7092617f | Marcus Denker > -> fix the last sender of #asIcon > https://pharo.fogbugz.com/f/cases/22139/fix-the-last-sender-of-asIcon > > delete mode 100644 src/Polymorph-Widgets-Rules/IconShortcutRule.class.st > <http://iconshortcutrule.class.st/> > > So then I tried to recover that last one (again at root underneath the src > directory) > > git checkout b4a285a61 src/Polymorph-Widgets-Rules/ > IconShortcutRule.class.st <http://iconshortcutrule.class.st/> > error: pathspec 'src/Polymorph-Widgets-Rules/IconShortcutRule.class.st > <http://iconshortcutrule.class.st/>' did not match any file(s) known to > git. > > And I still get that dreaded pathspec error. This is on OSX (not windows) > - so Iâm struggling to understand what I could be doing wrong. It must be > something obvious - but damned if I can see how Iâm doing anything > different to you. And it wouldnât make sense that git is broken on OSX? > > Tim > > On 25 Jun 2018, at 14:21, Ben Coman <btc(a)openInWorld.com> wrote: > > On 25 June 2018 at 19:41, Tim Mackinnon <tim(a)testit.works> wrote: > >> Iâd be really interested if someone with lower level GIT knowledge might >> try a: >> >> git checkout <gitid> src/<proj>/<class>.class.st >> <http://prismicblock.class.st/> >> >> For their project - as I donât understand what Iâm doing wrong - and Iâd >> like the comfort of knowing that our source is in a place/state where we >> can rely on normal git tools in a case of emergency. At the moment, Iâm a >> bit nervous that we are corrupting something . >> >> Tim >> > > I'm not sure if this is what you wanted, but I found a test case for for > Pharo 7, a deleted class "MultiByteFileStreamTest.class.st > <http://multibytefilestreamtest.class.st/>" > > https://github.com/pharo-project/pharo/pull/1031/files#diff-750a25fb99d29cd… > > From Windows 10 cmd.exe I tried the following (I can't remember which > tool installed `git`)... > > > mkdir C:\temp\test > > cd C:\temp\test > > git clone git@github.com:pharo-project/pharo.git > > cd pharo\src\Deprecated70 > > dir Multi* > no result > > git checkout e74308e67d9f84 MultiByteFileStreamTest.class.st > <http://multibytefilestreamtest.class.st/> > > dir Multi* > MultiByteFileStreamTest.class.st > <http://multibytefilestreamtest.class.st/> > > Then I compared the file I checked out to the raw file on github and they > were identical... > https://www.diffchecker.com/gVCpJzFe > > > btw, I did get a momentary error "error: pathspec 'src/Deprecated70/ > MultiByteFileStreamTest.class.st > <http://multibytefilestreamtest.class.st/>' did not match any file(s) > known to git." > when I incorrectly did... > > cd pharo\src\Deprecated70 > > git git checkout e74308e67d9f84 src/Deprecated70/ > MultiByteFileStreamTest.class.st > <http://multibytefilestreamtest.class.st/> > > > This worked with the longer path... > > cd pharo > > git git checkout e74308e67d9f84 src/Deprecated70/ > MultiByteFileStreamTest.class.st > <http://multibytefilestreamtest.class.st/> > > cheers -ben > > > > > -- Guille Polito Research Engineer Centre de Recherche en Informatique, Signal et Automatique de Lille CRIStAL - UMR 9189 French National Center for Scientific Research - *http://www.cnrs.fr <http://www.cnrs.fr>* *Web:* *http://guillep.github.io* <http://guillep.github.io> *Phone: *+33 06 52 70 66 13

June 28, 2018

Re: [Pharo-users] pharo bash script with startup

by Otto Behrens

Thanks for the effort guys. I tried to download the image, sources and vm separately (basically extracted what https://get.pharo.org/64/61+vm does), but ran into fresh trouble. Firstly, is "wget -O - https://get.pharo.org/64/61+vm | bash" not risky in terms of security? It should be quite possible to inject a lovely trojan horse with this, or not? Secondly, the pharo bash script that it generates is different to the one in the zip. The directory structure (where the image & vm goes) is also different. Why is that? I started the image (http://files.pharo.org/get-files/61/pharo64.zip) with the vm (http://files.pharo.org/get-files/61/pharo-linux-stable.zip) and got the following, which I tried to do, but that did not take the message away: pthread_setschedparam failed: Operation not permitted This VM uses a separate heartbeat thread to update its internal clock and handle events. For best operation, this thread should run at a higher priority, however the VM was unable to change the priority. The effect is that heavily loaded systems may experience some latency issues. If this occurs, please create the appropriate configuration file in /etc/security/limits.d/ as shown below: cat <<END | sudo tee /etc/security/limits.d/pharo.conf * hard rtprio 2 * soft rtprio 2 END and report to the pharo mailing list whether this improves behaviour. You will need to log out and log back in for the limits to take effect. (I did do this). I reverted to the deb package for now, because this have take some time and I can't focus on it now. I'll give it another shot soon, I hope On Wed, Jun 27, 2018 at 8:58 PM, Tim Mackinnon <tim(a)testit.works> wrote: > When we get confirmation of success - we need to make sure this gets > applied to both zeroconf and official downloads. > > Anything we can do to simplify and make it robust is gratefully > appreciated as there is nothing worse than falling at the lunch hurdle. > > Itâs cool to see so many clever minds on this. > > Tim > > Sent from my iPhone > > > On 27 Jun 2018, at 19:52, HernÃ¡n Morales Durand < > hernan.morales(a)gmail.com> wrote: > > > > 2018-06-27 10:50 GMT-03:00 K K Subbu <kksubbu.ml(a)gmail.com>: > >>> On Wednesday 27 June 2018 06:39 PM, K K Subbu wrote: > >>> > >>> > >>> The double quotes are required here to skip splitting arguments with > >>> embedded spaces into different words. > >>> > >>> I suspect the error is earlier in the image=$@ assignment. This > requires > >>> double quotes. Double quotes are also required while calling zenity to > avoid > >>> word splitting. > >> > >> > >> My earlier fix is also in error, Sorry! > >> > >> Essentially, we are mixing up single value variable (image) with > argument > >> array ($@). I found a cleaner fix : > >> > >> ```` > >> if [ $# -eq 0 ]; then > >> image_count=`ls "$RESOURCES"/*.image 2>/dev/null |wc -l` > >> if [ "$image_count" -eq 1 ]; then > >> set -- "$RESOURCES"/*.image > >> elif which zenity &>/dev/null; then > >> set -- "$(zenity --title 'Select an image' > --file-selection > >> --filename "$RESOURCES/" --file-filter '*.image' --file- > >> filter '*')" > >> else > >> set -- "$RESOURCES/Pharo6.1-64.image" > >> fi > >> fi > >> > > > > Use $() instead of backticks for command substitution: > > http://mywiki.wooledge.org/BashFAQ/082 > > > > Cheers, > > > > HernÃ¡n > > > >> # execute > >> exec "$LINUX/pharo" \ > >> --plugins "$LINUX" \ > >> --encoding utf8 \ > >> -vm-display-X11 \ > >> "$@" > >> ```` > >> > >> HTH .. Subbu > >> > > > > >

June 28, 2018