August 2013 - Pharo-users - lists.pharo.org

Re: [Pharo-users] Troubles using Pharo 2.0 one-click package on a secured WinXP

by StÃ©phane Ducasse

pay attention that also virus checker can think pharo is a virus because it modifies the change file. Stef On Aug 6, 2013, at 11:44 AM, Mun Mun <mun.sysdev(a)gmail.com> wrote: > Hi ! > > I'm quite new to Smalltalk and and so far, I've used Pharo on a regular XP system and everything went well. I'll add that my image occupy ~40 Mo in memory. > > I've recently tried to deploy my small app on the target system which is a very seccured package of WinXP. I can't help much on describing the specificities of the system as I'm just the user. But the system is cleared of unused services, there is a lot of account auditing, temporary files are explicitly written off the disk at logoff and so on. > > When launching the VM, I'm offered to choose an image to open and the following happen (I monitor the launching using the task manager) : > Â¤ The first ~40 Mo are quickly allocated in memory > Â¤ More memory is slowly allocated to the Pharo.exe process > Â¤ After ~5 min, 100Mo allocated and counting, I stop the process and never get to see the Pharo GUI > > > Is there a way for me to trace/dump what the VM is doing and maybe get help here to make my app work ? > > > Complementary information : > Â¤ I made a very crappy deployement: due to recent restrictions, I could not download a fresh and clean Pharo image. I picked the cleanest I had used, removed the various tutorial packages I had made (LOCell and stuff ;) and loaded the Monticello package of my little app. > Â¤ The target system run on a VM (with something like VSphere). My understanding is that we have a cheap virtualization setup which is not adapted at all for XP.

Aug. 6, 2013

Re: [Pharo-users] New version of PharoSound

by Xavier MESSNER

Hi, I have updated my blog to be used with pharo 2 summer edition <http://goo.gl/R8ZJJC> . I will write another to add vocal sound but it's a little more complicated. Hope it will help :) Xavier. -- View this message in context: http://forum.world.st/New-version-of-PharoSound-tp4702032p4702284.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Aug. 6, 2013

Re: [Pharo-users] Troubles using Pharo 2.0 one-click package on a secured WinXP

by Guillermo Polito

On Tue, Aug 6, 2013 at 2:49 PM, Mun Mun <mun.sysdev(a)gmail.com> wrote: > I found the source of the issue and it was because some files were > imported on the system in Read Only mode. Funny enough, there were allready > a PharoDebug.log from my dev environement weighting arround 20 Mo: removing > the file permited to launch the Pharo GUI but with a lot of file opening > failure and some code missing. It hinted me to the solution. > > Thanks for your time anyway ! > > Btw, it's the first time I use a mailling list and I wanted to answer to > my own question to save the hassle of other readers. I'm not sure how you > can do that without having received an answer. > Thanks for that, actually :). BTW, It would be nice to have a nice message and a clean quit instead of such ugly behavior... So, could you develop more? :) Guille > > > 2013/8/6 Mariano Martinez Peck <marianopeck(a)gmail.com> > >> >> >> >> On Tue, Aug 6, 2013 at 6:44 AM, Mun Mun <mun.sysdev(a)gmail.com> wrote: >> >>> Hi ! >>> >>> I'm quite new to Smalltalk and and so far, I've used Pharo on a regular >>> XP system and everything went well. I'll add that my image occupy ~40 Mo in >>> memory. >>> >>> I've recently tried to deploy my small app on the target system which is >>> a very seccured package of WinXP. I can't help much on describing the >>> specificities of the system as I'm just the user. But the system is cleared >>> of unused services, there is a lot of account auditing, temporary files are >>> explicitly written off the disk at logoff and so on. >>> >>> When launching the VM, I'm offered to choose an image to open and the >>> following happen (I monitor the launching using the task manager) : >>> Â¤ The first ~40 Mo are quickly allocated in memory >>> Â¤ More memory is slowly allocated to the Pharo.exe process >>> Â¤ After ~5 min, 100Mo allocated and counting, I stop the process and >>> never get to see the Pharo GUI >>> >>> >>> Is there a way for me to trace/dump what the VM is doing and maybe get >>> help here to make my app work ? >>> >>> >> Is there any PharoDebug.log next to the .image file? >> >> Are you sure the .sources is in the correct place and found when running? >> >> >>> Complementary information : >>> Â¤ I made a very crappy deployement: due to recent restrictions, I could >>> not download a fresh and clean Pharo image. I picked the cleanest I had >>> used, removed the various tutorial packages I had made (LOCell and stuff ;) >>> and loaded the Monticello package of my little app. >>> Â¤ The target system run on a VM (with something like VSphere). My >>> understanding is that we have a cheap virtualization setup which is not >>> adapted at all for XP. >>> >> >> >> >> -- >> Mariano >> http://marianopeck.wordpress.com >> > >

Aug. 6, 2013

Re: [Pharo-users] Troubles using Pharo 2.0 one-click package on a secured WinXP

by Mun Mun

I found the source of the issue and it was because some files were imported on the system in Read Only mode. Funny enough, there were allready a PharoDebug.log from my dev environement weighting arround 20 Mo: removing the file permited to launch the Pharo GUI but with a lot of file opening failure and some code missing. It hinted me to the solution. Thanks for your time anyway ! Btw, it's the first time I use a mailling list and I wanted to answer to my own question to save the hassle of other readers. I'm not sure how you can do that without having received an answer. 2013/8/6 Mariano Martinez Peck <marianopeck(a)gmail.com> > > > > On Tue, Aug 6, 2013 at 6:44 AM, Mun Mun <mun.sysdev(a)gmail.com> wrote: > >> Hi ! >> >> I'm quite new to Smalltalk and and so far, I've used Pharo on a regular >> XP system and everything went well. I'll add that my image occupy ~40 Mo in >> memory. >> >> I've recently tried to deploy my small app on the target system which is >> a very seccured package of WinXP. I can't help much on describing the >> specificities of the system as I'm just the user. But the system is cleared >> of unused services, there is a lot of account auditing, temporary files are >> explicitly written off the disk at logoff and so on. >> >> When launching the VM, I'm offered to choose an image to open and the >> following happen (I monitor the launching using the task manager) : >> Â¤ The first ~40 Mo are quickly allocated in memory >> Â¤ More memory is slowly allocated to the Pharo.exe process >> Â¤ After ~5 min, 100Mo allocated and counting, I stop the process and >> never get to see the Pharo GUI >> >> >> Is there a way for me to trace/dump what the VM is doing and maybe get >> help here to make my app work ? >> >> > Is there any PharoDebug.log next to the .image file? > > Are you sure the .sources is in the correct place and found when running? > > >> Complementary information : >> Â¤ I made a very crappy deployement: due to recent restrictions, I could >> not download a fresh and clean Pharo image. I picked the cleanest I had >> used, removed the various tutorial packages I had made (LOCell and stuff ;) >> and loaded the Monticello package of my little app. >> Â¤ The target system run on a VM (with something like VSphere). My >> understanding is that we have a cheap virtualization setup which is not >> adapted at all for XP. >> > > > > -- > Mariano > http://marianopeck.wordpress.com >

Aug. 6, 2013

Re: [Pharo-users] Troubles using Pharo 2.0 one-click package on a secured WinXP

by Mariano Martinez Peck

On Tue, Aug 6, 2013 at 6:44 AM, Mun Mun <mun.sysdev(a)gmail.com> wrote: > Hi ! > > I'm quite new to Smalltalk and and so far, I've used Pharo on a regular XP > system and everything went well. I'll add that my image occupy ~40 Mo in > memory. > > I've recently tried to deploy my small app on the target system which is a > very seccured package of WinXP. I can't help much on describing the > specificities of the system as I'm just the user. But the system is cleared > of unused services, there is a lot of account auditing, temporary files are > explicitly written off the disk at logoff and so on. > > When launching the VM, I'm offered to choose an image to open and the > following happen (I monitor the launching using the task manager) : > Â¤ The first ~40 Mo are quickly allocated in memory > Â¤ More memory is slowly allocated to the Pharo.exe process > Â¤ After ~5 min, 100Mo allocated and counting, I stop the process and never > get to see the Pharo GUI > > > Is there a way for me to trace/dump what the VM is doing and maybe get > help here to make my app work ? > > Is there any PharoDebug.log next to the .image file? Are you sure the .sources is in the correct place and found when running? > Complementary information : > Â¤ I made a very crappy deployement: due to recent restrictions, I could > not download a fresh and clean Pharo image. I picked the cleanest I had > used, removed the various tutorial packages I had made (LOCell and stuff ;) > and loaded the Monticello package of my little app. > Â¤ The target system run on a VM (with something like VSphere). My > understanding is that we have a cheap virtualization setup which is not > adapted at all for XP. > -- Mariano http://marianopeck.wordpress.com

Aug. 6, 2013

Troubles using Pharo 2.0 one-click package on a secured WinXP

by Mun Mun

Hi ! I'm quite new to Smalltalk and and so far, I've used Pharo on a regular XP system and everything went well. I'll add that my image occupy ~40 Mo in memory. I've recently tried to deploy my small app on the target system which is a very seccured package of WinXP. I can't help much on describing the specificities of the system as I'm just the user. But the system is cleared of unused services, there is a lot of account auditing, temporary files are explicitly written off the disk at logoff and so on. When launching the VM, I'm offered to choose an image to open and the following happen (I monitor the launching using the task manager) : Â¤ The first ~40 Mo are quickly allocated in memory Â¤ More memory is slowly allocated to the Pharo.exe process Â¤ After ~5 min, 100Mo allocated and counting, I stop the process and never get to see the Pharo GUI Is there a way for me to trace/dump what the VM is doing and maybe get help here to make my app work ? Complementary information : Â¤ I made a very crappy deployement: due to recent restrictions, I could not download a fresh and clean Pharo image. I picked the cleanest I had used, removed the various tutorial packages I had made (LOCell and stuff ;) and loaded the Monticello package of my little app. Â¤ The target system run on a VM (with something like VSphere). My understanding is that we have a cheap virtualization setup which is not adapted at all for XP.

Aug. 6, 2013

Re: [Pharo-users] Pharo 2.0 on Fedora 18

by StÃ©phane Ducasse

hi edwin we are sad not to be able to help more. But we are really busy (even more than that). Stef On Aug 5, 2013, at 10:19 PM, edwin ancaer <eancaer(a)gmail.com> wrote: > Esteban, Mariano, > > thanks for helping out. It's probably a lack of knowledge on my side, but until now, everything I tried failed. > > There seems to ben no package openssl098e.i686in the sources of Fedora 18, so I downmoaded it, and opened it with yumex. The answer was that there was nothing to do for libssl. > > I tried to create a symbolic link to /opt/Adobe/Reader9/Reader/intellinux/lib/libssl.so.0.9.8, after installing a package that contained the library in question, but to no avail. With ldd I found: > > [edwin@ottopedi bin]$ ldd libSqueakSSL.so > ./libSqueakSSL.so: /lib/libssl.so.0.9.8: no version information available (required by ./libSqueakSSL.so) > linux-gate.so.1 => (0x00795000) > libssl.so.0.9.8 => /lib/libssl.so.0.9.8 (0x001e3000) > libc.so.6 => /lib/libc.so.6 (0x0085c000) > libcrypto.so.0.9.8 => /lib/libcrypto.so.0.9.8 (0x0035d000) > libdl.so.2 => /lib/libdl.so.2 (0x00b13000) > /lib/ld-linux.so.2 (0x4ed07000) > libz.so.1 => /lib/libz.so.1 (0x00c9c000) > > I even tried to install from source, but I only managed to create libssl.a, a static library. > For now, I'll postpone the tests for the sending of emails with Pharo to a later date. > > Thanks a lot, > > E. Ancaer > > > 2013/8/5 Mariano Martinez Peck <marianopeck(a)gmail.com> > Hi Edwin. > > I wrote some details on how to install Pharo in CentOS for a client. I paste them here in case it may help you: > > --------------------------- > > > Installing Pharo in CentOS is not difficult. However, it does require some packages and dependencies. Pharo will only work with 32 bits libraries, therefore, we need to download them. > [edit]Get Pharo > > You can get the Pharo from http://www.pharo-project.org/pharo-download. To install it you just need to uncompress it. > [edit]Installing 32 bits needed libraries > > To install the minimal needed libraries, we can execute: > sudo yum install libX11.i686 libX11-devel.i686 mesa-libGL.i686 mesa-libGL-devel.i686 > To test if the VM works, just execute the following with a demo.image: > ./pharo xxx.image > And there shouldn't be any error in the console. > [edit]Installing OpenSSL > > If we plan to use Zodiac or SSL we have to install openSSL 32 bits libraries. To do that: > sudo yum install openssl098e.i686 openssl.i686 > sudo cp /usr/lib/libssl.so.0.9.8e /usr/lib/libssl.so.0.9.8 > Notice the second line is because otherwise Pharo fails to link to /usr/lib/libssl.so.0.9.8. To test if the link is correct, do: > ldd /opt/pharo/pharoVM/libSqueakSSL.so > And you should see something like: > libssl.so.0.9.8 => /usr/lib/libssl.so.0.9.8 (0x00291000) > Otherwise, there is a problem in the link. > Now, all Zodiac and Zinc-Zodiac tests should be green. > > > On Mon, Aug 5, 2013 at 7:25 AM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > Hi, > > That's a hard reference, sadly. > You need to install 0.9.8 (32bits) in your distribution. I don't know how to do it in fedora, but on ubiuntu is usually plain simple... probably in fedora is also available with the apt-get equivalent tool :) > > cheers, > Esteban > > > On Aug 5, 2013, at 6:43 AM, edwin ancaer <eancaer(a)gmail.com> wrote: > >> Hello, >> >> I'm running Pharo 2.0 on Fedora 18. I seem to have an issue with libssl.so which on my system is 1.0.1e, while libSquekSSL.so looks for version 0.9.8, as described here:Zodiac mail sending in Pharo smalltalk . >> >> Does anyone know how to solve this? Creating a symbolic link from 0.9.8 to 1.0.1e does not solve the problem >> >> [edwin@ottopedi bin]# ldd libSqueakSSL.so >> ./libSqueakSSL.so: /lib/libssl.so.0.9.8: version `OPENSSL_0.9.8' not found (required by ./libSqueakSSL.so) >> linux-gate.so.1 => (0x001d2000) >> libssl.so.0.9.8 => /lib/libssl.so.0.9.8 (0x00161000) >> libc.so.6 => /lib/libc.so.6 (0x003a5000) >> libgssapi_krb5.so.2 => /lib/libgssapi_krb5.so.2 (0x00342000) >> libkrb5.so.3 => /lib/libkrb5.so.3 (0x001d3000) >> libcom_err.so.2 => /lib/libcom_err.so.2 (0x009ec000) >> libk5crypto.so.3 => /lib/libk5crypto.so.3 (0x00b76000) >> libresolv.so.2 => /lib/libresolv.so.2 (0x002b3000) >> libcrypto.so.10 => /lib/libcrypto.so.10 (0x0068a000) >> libdl.so.2 => /lib/libdl.so.2 (0x0087a000) >> libz.so.1 => /lib/libz.so.1 (0x00300000) >> /lib/ld-linux.so.2 (0x4ed07000) >> libkrb5support.so.0 => /lib/libkrb5support.so.0 (0x00bcd000) >> libkeyutils.so.1 => /lib/libkeyutils.so.1 (0x00c36000) >> libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x0091d000) >> libpthread.so.0 => /lib/libpthread.so.0 (0x002cc000) >> libselinux.so.1 => /lib/libselinux.so.1 (0x0097f000) >> libpcre.so.1 => /lib/libpcre.so.1 (0x0055b000) >> : >> >> Is there anyone who managed to install the version 0.9.8 and 1.0.1e of the openssl rpm on the same machine? Or is nobidy on Fedora 18 using Pharo? >> >> Thanks in advance, >> >> Edwin Ancaer > > > > > -- > Mariano > http://marianopeck.wordpress.com >

Aug. 5, 2013

Re: [Pharo-users] Pharo 2.0 on Fedora 18

by edwin ancaer

Esteban, Mariano, thanks for helping out. It's probably a lack of knowledge on my side, but until now, everything I tried failed. There seems to ben no package openssl098e.i686in the sources of Fedora 18, so I downmoaded it, and opened it with yumex. The answer was that there was nothing to do for libssl. I tried to create a symbolic link to /opt/Adobe/Reader9/Reader/intellinux/lib/libssl.so.0.9.8, after installing a package that contained the library in question, but to no avail. With ldd I found: [edwin@ottopedi bin]$ ldd libSqueakSSL.so *./libSqueakSSL.so: /lib/libssl.so.0.9.8: no version information available (required by ./libSqueakSSL.so)* linux-gate.so.1 => (0x00795000) libssl.so.0.9.8 => /lib/libssl.so.0.9.8 (0x001e3000) libc.so.6 => /lib/libc.so.6 (0x0085c000) libcrypto.so.0.9.8 => /lib/libcrypto.so.0.9.8 (0x0035d000) libdl.so.2 => /lib/libdl.so.2 (0x00b13000) /lib/ld-linux.so.2 (0x4ed07000) libz.so.1 => /lib/libz.so.1 (0x00c9c000) I even tried to install from source, but I only managed to create libssl.a, a static library. For now, I'll postpone the tests for the sending of emails with Pharo to a later date. Thanks a lot, E. Ancaer 2013/8/5 Mariano Martinez Peck <marianopeck(a)gmail.com> > Hi Edwin. > > I wrote some details on how to install Pharo in CentOS for a client. I > paste them here in case it may help you: > > --------------------------- > > > Installing Pharo in CentOS is not difficult. However, it does require some > packages and dependencies. Pharo will only work with 32 bits libraries, > therefore, we need to download them. > [edit<http://debrispublishing.com/wiki/index.php?title=Infrastructure:Installing_…> > ]Get Pharo > > You can get the Pharo from http://www.pharo-project.org/pharo-download. > To install it you just need to uncompress it. > [edit<http://debrispublishing.com/wiki/index.php?title=Infrastructure:Installing_…> > ]Installing 32 bits needed libraries > > To install the minimal needed libraries, we can execute: > sudo yum install libX11.i686 libX11-devel.i686 mesa-libGL.i686 > mesa-libGL-devel.i686 > > To test if the VM works, just execute the following with a demo.image: > ./pharo xxx.image > > And there shouldn't be any error in the console. > [edit<http://debrispublishing.com/wiki/index.php?title=Infrastructure:Installing_…> > ]Installing OpenSSL > > If we plan to use Zodiac or SSL we have to install openSSL 32 bits > libraries. To do that: > sudo yum install openssl098e.i686 openssl.i686 > sudo cp /usr/lib/libssl.so.0.9.8e /usr/lib/libssl.so.0.9.8 > > Notice the second line is because otherwise Pharo fails to link to > /usr/lib/libssl.so.0.9.8. To test if the link is correct, do: > ldd /opt/pharo/pharoVM/libSqueakSSL.so > > And you should see something like: > libssl.so.0.9.8 => /usr/lib/libssl.so.0.9.8 (0x00291000) > > Otherwise, there is a problem in the link. > > Now, all Zodiac and Zinc-Zodiac tests should be green. > > > On Mon, Aug 5, 2013 at 7:25 AM, Esteban Lorenzano <estebanlm(a)gmail.com>wrote: > >> Hi, >> >> That's a hard reference, sadly. >> You need to install 0.9.8 (32bits) in your distribution. I don't know how >> to do it in fedora, but on ubiuntu is usually plain simple... probably in >> fedora is also available with the apt-get equivalent tool :) >> >> cheers, >> Esteban >> >> >> On Aug 5, 2013, at 6:43 AM, edwin ancaer <eancaer(a)gmail.com> wrote: >> >> Hello, >> >> I'm running Pharo 2.0 on Fedora 18. I seem to have an issue with >> libssl.so which on my system is 1.0.1e, while libSquekSSL.so looks for >> version 0.9.8, as described here:Zodiac mail sending in Pharo smalltalk<http://stackoverflow.com/questions/12293670/zodiac-mail-sending-in-pharo-sm…>. >> >> Does anyone know how to solve this? Creating a symbolic link from 0.9.8 >> to 1.0.1e does not solve the problem >> >> [edwin@ottopedi bin]# ldd libSqueakSSL.so >> *./libSqueakSSL.so: /lib/libssl.so.0.9.8: version `OPENSSL_0.9.8' not >> found (required by ./libSqueakSSL.so)* >> linux-gate.so.1 => (0x001d2000) >> libssl.so.0.9.8 => /lib/libssl.so.0.9.8 (0x00161000) >> libc.so.6 => /lib/libc.so.6 (0x003a5000) >> libgssapi_krb5.so.2 => /lib/libgssapi_krb5.so.2 (0x00342000) >> libkrb5.so.3 => /lib/libkrb5.so.3 (0x001d3000) >> libcom_err.so.2 => /lib/libcom_err.so.2 (0x009ec000) >> libk5crypto.so.3 => /lib/libk5crypto.so.3 (0x00b76000) >> libresolv.so.2 => /lib/libresolv.so.2 (0x002b3000) >> libcrypto.so.10 => /lib/libcrypto.so.10 (0x0068a000) >> libdl.so.2 => /lib/libdl.so.2 (0x0087a000) >> libz.so.1 => /lib/libz.so.1 (0x00300000) >> /lib/ld-linux.so.2 (0x4ed07000) >> libkrb5support.so.0 => /lib/libkrb5support.so.0 (0x00bcd000) >> libkeyutils.so.1 => /lib/libkeyutils.so.1 (0x00c36000) >> libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x0091d000) >> libpthread.so.0 => /lib/libpthread.so.0 (0x002cc000) >> libselinux.so.1 => /lib/libselinux.so.1 (0x0097f000) >> libpcre.so.1 => /lib/libpcre.so.1 (0x0055b000) >> : >> >> Is there anyone who managed to install the version 0.9.8 and 1.0.1e of >> the openssl rpm on the same machine? Or is nobidy on Fedora 18 using Pharo? >> >> Thanks in advance, >> >> Edwin Ancaer >> >> >> > > > -- > Mariano > http://marianopeck.wordpress.com >

Aug. 5, 2013

Re: [Pharo-users] OSProcess execution from specific directory

by Juraj Kubelka

Cool, thank you. Jura El 05-08-2013, a las 12:46, "David T. Lewis" <lewis(a)mail.msen.com> escribiÃ³: > On Mon, Aug 05, 2013 at 08:42:12AM -0400, Juraj Kubelka wrote: >> Hi! >> >> On execution of >> >> (PipeableOSProcess command: 'pwd') output >> >> , it always says root directory ("/"). Is it possible to execute command from specific directory? Or the only correct solution is like this?: >> >> (PipeableOSProcess command: 'cd /Users/jura/ ; pwd') output >> >> Thanks a lot. >> Cheers, >> Jura > > By default, PipeableOSProcess class>command: uses whatever current working > directory is being used by the VM (which may vary depending on how your > VM and image are being run). You can specify a different working directory > like this: > > (PipeableOSProcess command: 'pwd' > environment: nil > workingDir: '/usr/bin' > input: nil > output: nil > error: nil > errorPipelineStream: nil) output > > HTH, > Dave > >

Aug. 5, 2013

Re: [Pharo-users] OSProcess execution from specific directory

by David T. Lewis

On Mon, Aug 05, 2013 at 08:42:12AM -0400, Juraj Kubelka wrote: > Hi! > > On execution of > > (PipeableOSProcess command: 'pwd') output > > , it always says root directory ("/"). Is it possible to execute command from specific directory? Or the only correct solution is like this?: > > (PipeableOSProcess command: 'cd /Users/jura/ ; pwd') output > > Thanks a lot. > Cheers, > Jura By default, PipeableOSProcess class>command: uses whatever current working directory is being used by the VM (which may vary depending on how your VM and image are being run). You can specify a different working directory like this: (PipeableOSProcess command: 'pwd' environment: nil workingDir: '/usr/bin' input: nil output: nil error: nil errorPipelineStream: nil) output HTH, Dave

Aug. 5, 2013