March 2017 - Pharo-dev - lists.pharo.org

about smacc gt extensions

by Stephane Ducasse

hi andrei and doru I think that we should improve the gt extension of smacc. Right now when I have a token I do not see it when inspecting it in raw. The raw view should always show the object implementation. After we can have other view. Now with the token this is super strange to get a string (or the result of value). because when you are exploring a tree you want to know if this is a string or a token. As in the following attachment: the current situation is confusing. in the list I have a list of token and when I look at one, I get a string. Stef [image: Inline image 1]

March 26, 2017

Re: [Pharo-dev] about push up and pull up

by Stephane Ducasse

Tx John for the historical perspective.I see. I would like to have a uniform naming: "move method up :)" We will revisit everything with Calypso. And fix some problems: In Pharo we broke the precondition that pulling up a method should only work if the state it accessed is already in the destination. We want also to have a layer (refactoring not accessing UI element) and a layer on top to manage interaction with the user. Right now this is mixed up. We should be able to invoke refactorings without UI. Stef On Sun, Mar 26, 2017 at 12:08 AM, John Brant <brant(a)refactoryworkers.com> wrote: > > On Mar 25, 2017, at 2:27 PM, Stephane Ducasse <stepharo.self(a)gmail.com> > wrote: > > > > Hi guys > > > > I wonder why we have push up method and pull up instance variables. > > I remember that long time ago we got a discussion but I forgot. > > Do you remember? > > > > Stef (writing chapters for new books for total newbies: yes the future > pharoers) > > In the original RefactoringBrowser, if you wanted to move an instance > variable up in the hierarchy, you would select the superclass and perform > pull up instance variable. It would then open a dialog where you chose what > subclass instance variable you wanted to pull up. Since we were at the > superclass in the browser instead of the instance variable, I chose pull up > instance variable for the name. At that time, we hadnât implemented context > selection menu options for the class definition, so we didnât have a way to > select an instance variable in the browser without opening a dialog. > > Later, I implemented push up method. In this case, the browser already > supported selecting methods, so it didnât make sense to have the user move > to the superclass, perform pull up method, and then select the method you > want to pull up from a dialog. > > Essentially, the push vs. pull naming was based on where you were in the > browser. If the thing to move wasnât where you were in the browser, then > pull was used â you would reach over to the other place and pull it to you. > However, if the thing was currently where you were in the browser, then you > would push it to the new location. > > Of course, the current browser doesnât follow my push vs. pull naming > convention. When you pull up an instance variable, you need to have the > class that defines the variable selected. > > > John Brant >

March 26, 2017

Re: [Pharo-dev] Estebanâs list of blocking issues for release Pharo 6.0

by Ben Coman

On Sat, Mar 25, 2017 at 8:12 AM, Ben Coman <btc(a)openinworld.com> wrote: > On Sat, Mar 25, 2017 at 5:17 AM, Nicolas Cellier > <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> >> >> 2017-03-24 17:23 GMT+01:00 Ben Coman <btc(a)openinworld.com>: >>> >>> On Fri, Mar 24, 2017 at 9:30 PM, Esteban Lorenzano <estebanlm(a)gmail.com> >>> wrote: >>> > >>> > 2017-03-24 11:29 GMT+01:00 Esteban Lorenzano <estebanlm(a)gmail.com>: >>> >> >>> >> >>> >> VM 64bits >>> >> ========= >>> >> - Is incredibly hard to compile a 64bits version of libgit2 0.23 (the >>> >> one we are using) for macOS, which makes >>> >> impossible to use Iceberg. Not sure this is blocking, but well... I >>> >> needed to point it :) >>> > >>> > Hi Esteban, >>> > Doesn't the Travis build correctly succeed? >>> > https://bintray.com/opensmalltalk/vm/cog/201703222227#files >>> > >>> > yes, but the PharoVM for mac in 64bits is not including libgit2 (because >>> > I was unable to compile it) :P >>> > if you (or anyone) find the time and helps me doing that, I will be >>> > very, very, very grateful (and release will be earlier ;) >>> >>> Do you mean build libgit2 standalone? or as part of pharo-vm build? >>> >>> I dug out my old mac-mini to try. Its an older OSX Version 10.9.5. I >>> could upgrade to try if needed. >>> I don't know about the rpath warning. I ignored it. >>> >>> Standalone... >>> >>> 502 mkdir LIBGIT >>> 503 cd LIBGIT/ >>> 504 git clone https://github.com/libgit2/libgit2.git >>> 505 cd libgit2 >>> 521 git checkout v0.23.0 >>> 522 mkdir ../build && cd ../build >>> 525 cmake ../libgit2 >>> -- Configuring done >>> CMake Warning (dev): >>> Policy CMP0042 is not set: MACOSX_RPATH is enabled by default. Run >>> "cmake >>> --help-policy CMP0042" for policy details. Use the cmake_policy command >>> to >>> set the policy and suppress this warning. >>> MACOSX_RPATH is not specified for the following targets: >>> git2 >>> >>> 526 cmake --build . >>> 528 file libgit2.0.23.0.dylib >>> libgit2.0.23.4.dylib: Mach-O 64-bit dynamically linked shared library >>> x86_64 >>> >>> 529 cmake --version >>> cmake version 3.0.0 >>> >>> 530 uname -a >>> Darwin bens-mac-mini.home.gateway 13.4.0 Darwin Kernel Version 13.4.0: >>> Mon Jan 11 18:17:34 PST 2016; root:xnu-2422.115.15~1/RELEASE_X86_64 >>> x86_64 >>> >> >> Same for me here on OSX 10.11.6 El Capitan, >> libgit2 builds out of the box, same warning for RPATH >> Maybe we could reactivate on travis? > > Is this relevant regarding using ssh with using libgit ? > https://github.com/libgit2/libgit2/issues/2665 I'm having trouble with my mac-mini with extended pauses between shell commands so I've not been able to test whether this is relevant, but doing... $ cd ~/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/build.macos64x64/pharo.cog.spur $ ./mvm I see==> tar x -f ../../.thirdparty-cache/libgit2-v0.23.0.tar.gz -C builddbg/third-party cd builddbg/third-party/libgit2-0.23.0 \ && cmake \ -DCMAKE_INSTALL_PREFIX=/Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64 \ -DUSE_SSH=ON \ -DCURL=ON \ -DBUILD_CLAR=OFF \ -DOPENSSL_FOUND=ON \ -DLIBSSH2_FOUND=ON \ -DLIBSSH2_INCLUDE_DIRS=/Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/include \ -DLIBSSH2_LIBRARIES=/Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/lib/libssh2.1.dylib \ -DCMAKE_C_FLAGS="-L/usr/lib -L/Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/lib -lssl -lcrypto -liconv" \ -DCMAKE_OSX_ARCHITECTURES=x86_64 \ -DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath,@executable_path:@executable_path/Plugins \ && make \ && make install -- The C compiler identification is AppleClang 6.0.0.6000056 ... -- checking for module 'libssh2' -- package 'libssh2' not found -- Looking for libssh2_userauth_publickey_frommemory in /Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/lib/libssh2.1.dylib -- Looking for libssh2_userauth_publickey_frommemory in /Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/lib/libssh2.1.dylib - found ---------------- Trying again on its own in a clean dir... 533 cd ~/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/build.macos64x64/pharo.cog.spur/builddbg/third-party 534 rm -r libgit2-0.23.0 535 tar x -f ../../../../.thirdparty-cache/libgit2-v0.23.0.tar.gz -C . 536 cd libgit2-0.23.0/ 537 cmake ....same...params...as...above I get the same error, -- package 'libssh2' not found but if I up-arrow to run `cmake` again, that message doesn't appear. Maybe its caching the result from the next line from the -- Looking for libssh2_userauth_publickey_frommemory in /Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/lib/libssh2.1.dylib - found So does that mean SSH ended up found and cmake cached the result? $ make ==> Linking C shared library libgit2.dylib [100%] Built target git2 $ otool -L libgit2.dylib ==> /Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/build.macos64x64/pharo.cog.spur/builddbg/third-party/libgit2-0.23.0/libgit2.23.dylib (compatibility version 23.0.0, current version 0.23.0) /usr/lib/libssl.0.9.8.dylib (compatibility version 0.9.8, current version 0.9.8) /usr/lib/libcrypto.0.9.8.dylib (compatibility version 0.9.8, current version 0.9.8) /usr/lib/libiconv.2.dylib (compatibility version 7.0.0, current version 7.0.0) /usr/lib/libz.1.dylib (compatibility version 1.0.0, current version 1.2.5) /System/Library/Frameworks/Security.framework/Versions/A/Security (compatibility version 1.0.0, current version 55471.14.16) /System/Library/Frameworks/CoreFoundation.framework/Versions/A/CoreFoundation (compatibility version 150.0.0, current version 855.17.0) /Users/ben/Repos/PHAROVM/pharo-vm/opensmalltalk-vm/.thirdparty-cache/macOS/x86_64/lib/libssh2.1.dylib (compatibility version 2.0.0, current version 2.1.0) /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 1197.1.1) So does that look like libgit is set up to use ssh okay? ------ btw, I found a related topic... http://stackoverflow.com/questions/32744681/libgit2-and-libssh2-cmake-packa… and here is the quick reference diff... https://www.diffchecker.com/SOGHWKRC So I notice among other things that changes LIBSSH2_INCLUDE_DIRS to LIBSSH2_INCLUDE_DIR, and our build uses -DLIBSSH2_INCLUDE_DIRS thats all I got for now. cheers -ben

March 26, 2017

Re: [Pharo-dev] Good way to promote Smalltalk

by askoh

TechGig Code Gladiators is the world's largest code contest. Organized in India but invites world participation. Is in the news now. http://tech.economictimes.indiatimes.com/news/internet/techgig-code-gladiat… Aik-Siong Koh -- View this message in context: http://forum.world.st/Good-way-to-promote-Smalltalk-tp4939758p4939998.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

March 26, 2017

Re: [Pharo-dev] about push up and pull up

by John Brant

> On Mar 25, 2017, at 2:27 PM, Stephane Ducasse <stepharo.self(a)gmail.com> wrote: > > Hi guys > > I wonder why we have push up method and pull up instance variables. > I remember that long time ago we got a discussion but I forgot. > Do you remember? > > Stef (writing chapters for new books for total newbies: yes the future pharoers) In the original RefactoringBrowser, if you wanted to move an instance variable up in the hierarchy, you would select the superclass and perform pull up instance variable. It would then open a dialog where you chose what subclass instance variable you wanted to pull up. Since we were at the superclass in the browser instead of the instance variable, I chose pull up instance variable for the name. At that time, we hadnât implemented context selection menu options for the class definition, so we didnât have a way to select an instance variable in the browser without opening a dialog. Later, I implemented push up method. In this case, the browser already supported selecting methods, so it didnât make sense to have the user move to the superclass, perform pull up method, and then select the method you want to pull up from a dialog. Essentially, the push vs. pull naming was based on where you were in the browser. If the thing to move wasnât where you were in the browser, then pull was used â you would reach over to the other place and pull it to you. However, if the thing was currently where you were in the browser, then you would push it to the new location. Of course, the current browser doesnât follow my push vs. pull naming convention. When you pull up an instance variable, you need to have the class that defines the variable selected. John Brant

March 25, 2017

Crashes on snapshot with the new compactor

by Eliot Miranda

Hi All, a number of people are being affected by crashes on snapshotting the image, the worst possible time for a crash. There is a bug in the new compactor that unfortunately bites when saving. The compactor is invoked as part of a full garbage collect after the garbage collector has feed unreachable objects. Normally the new compactor makes only a single pass through the heap, which may not move all the objects that are possible to move. (The amount of objects that can be moved in a single pass is limited by available free space.) But on snapshot the compactor makes as may passes as are necessary to slide all movable objects down as far as possible. Unfortunately there is a bug in this second pass. Fixing this bug is now my priority. I have an example image from Esteban Lorenzano to test. I am asking anyone else that can provide an image that reliably crashes when trying to save it to make the image and changes available to me for testing if possible. In the mean time one may be able to work around the problem by doing a full garbage collect before snapshot. This should do a GC with a single compaction pass which should not fail, and then make it much more likely that the GC during snapshot will do a single compaction pass, since fewer objects should be mobile after the single pass compaction in the explicit GC. To do this in Pharo I would put a full gc here: SessionManager>>snapshot: save andQuit: quit | isImageStarting snapshotResult | ChangesLog default logSnapshot: save andQuit: quit. >> SmalltalkImage current primitiveGarbageCollect. self currentSession stop: quit. "Image not usable from here until the session is restarted!" ... In Squeak I would put a full GC here: snapshot: save andQuit: quit withExitCode: exitCode embedded: embeddedFlag "Mark the changes file and close all files as part of #processShutdownList. If save is true, save the current state of this Smalltalk in the image file. If quit is true, then exit to the outer OS shell. If exitCode is not nil, then use it as exit code. The latter part of this method runs when resuming a previously saved image. This resume logic checks for a document file to process when starting up." | resuming msg | Object flushDependents. Object flushEvents. ... Smalltalk processShutDownList: quit. >> SmalltalkImage current primitiveGarbageCollect. Cursor write show. save ifTrue: [resuming := embeddedFlag ifTrue: [self snapshotEmbeddedPrimitive] ifFalse: [self snapshotPrimitive]] "<-- PC frozen here on image file" ifFalse: [resuming := false]. I do apologise for the bug. I hope it will be fixed within a few days. _,,,^..^,,,_ best, Eliot

March 25, 2017

Re: [Pharo-dev] about push up and pull up

by Stephane Ducasse

Tx cyril. I'm remember that there was reason but it looks obscure now :) On Sat, Mar 25, 2017 at 8:32 PM, Cyril Ferlicot <cyril.ferlicot(a)gmail.com> wrote: > > On sam. 25 mars 2017 at 20:28, Stephane Ducasse <stepharo.self(a)gmail.com> > wrote: > >> Hi guys >> >> I wonder why we have push up method and pull up instance variables. >> I remember that long time ago we got a discussion but I forgot. >> Do you remember? >> >> Stef (writing chapters for new books for total newbies: yes the future >> pharoers) >> > I opened an issue some time ago > > https://pharo.fogbugz.com/f/cases/19607/Concistency- > question-pull-up-push-up > > > -- > Cheers > Cyril Ferlicot >

March 25, 2017

Re: [Pharo-dev] about push up and pull up

by Cyril Ferlicot

On sam. 25 mars 2017 at 20:28, Stephane Ducasse <stepharo.self(a)gmail.com> wrote: > Hi guys > > I wonder why we have push up method and pull up instance variables. > I remember that long time ago we got a discussion but I forgot. > Do you remember? > > Stef (writing chapters for new books for total newbies: yes the future > pharoers) > I opened an issue some time ago https://pharo.fogbugz.com/f/cases/19607/Concistency-question-pull-up-push-up -- Cheers Cyril Ferlicot

March 25, 2017

about push up and pull up

by Stephane Ducasse

Hi guys I wonder why we have push up method and pull up instance variables. I remember that long time ago we got a discussion but I forgot. Do you remember? Stef (writing chapters for new books for total newbies: yes the future pharoers)

March 25, 2017

Re: [Pharo-dev] ZnEasy problem?

by Ben Coman

On Sat, Mar 25, 2017 at 4:50 PM, phil(a)highoctane.be <phil(a)highoctane.be> wrote: > Very interesting findings. > > I am always delighted by your investigations Ben, that is inspiring me a lot > to be a better programmer Just a compulsion to pull on a loose thread and see where it leads. Don't ever lend me a knitted jumper with a loose thread ;) cheers -ben > > Security is a large beast and it becomes more and more pervasive in the > projects. I was busy with some more Kerberos yesterday and it would be very > useful for me to have support for it in Pharo natively. Kerberos looks > complicated until one realizes that there is nothing like it for doing what > it does in a simple way :-D And from Kerberos its only a hop, skip and jump to in-Image support of Microsoft's bastardised Kerberos for Windows authentication. (errr, maybe thats a wish too far .) cheers -ben > > HTTP/2 is also there now and we will have to support this too. > > Phil > > > > > > On Sat, Mar 25, 2017 at 12:37 AM, Ben Coman <btc(a)openinworld.com> wrote: >> >> On Sat, Mar 25, 2017 at 2:04 AM, Sven Van Caekenberghe <sven(a)stfx.eu> >> wrote: >> > >> > Some further notes. >> > >> > > On 24 Mar 2017, at 14:20, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: >> > > >> > > Hmm, we'll have to study this better, since the following still works >> > > for me (while Stef's original URL, http://pharo.org/web/files/pharo.png, >> > > also fails for me): >> > > >> > > ZnEasy getPng: 'http://pharo.org/files/pharo.png'. >> > > >> > > This is also the URL used in the unit tests. >> > >> > Of course that one works, because it does not do a redirect from HTTP to >> > HTTPS, duh. >> > Still it is good to known that at least for examples/demos/tests we >> > still have it. >> > >> > > Now, needless to say that the day Pharo cannot access its own server >> > > using TLS/SSL we are in trouble. I have warned about this before. We need to >> > > work on the SSL Plugin. Step one is the following: >> > > >> > > >> > > https://pharo.fogbugz.com/f/cases/19606/Switch-SSL-Plugin-Code-for-macOS-to… >> > > >> > > Hopefully we can do something similar for Windows, but that is out of >> > > my area of expertise. >> > > >> > > Just to be clear: this is related to TLS/SSL (Zodiac), not Zinc. The >> > > problem is probably not in the fundamental logic of encrypting/decrypting >> > > the stream, but it in the setup of the connection, most probably related to >> > > certificates. >> > >> > I took a file from one of my servers that it up to date and well >> > configured (nginx, let's encrypt certificate). That works fine. >> > >> > ZnEasy getJpeg: 'https://audio359.eu/img/audio359-main-800.jpg'. >> > >> > Testing our failing URL and this succeeding URL using curl on macOS, I >> > see the following: >> > >> > $ curl -v https://audio359.eu/img/audio359-main-800.jpg > /dev/null >> > * TCP_NODELAY set >> > * Connected to audio359.eu (146.185.177.20) port 443 (#0) >> > * TLS 1.2 connection using TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 >> > * Server certificate: audio359.eu >> > * Server certificate: Let's Encrypt Authority X3 >> > * Server certificate: DST Root CA X3 >> > > GET /img/audio359-main-800.jpg HTTP/1.1 >> > ... >> > >> > vs. >> > >> > $ curl -v https://pharo.org/web/files/pharo.png > /dev/null >> > * TCP_NODELAY set >> > * Connected to pharo.org (104.28.26.35) port 443 (#0) >> > * TLS 1.2 connection using TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 >> > * Server certificate: sni25155.cloudflaressl.com >> > * Server certificate: COMODO ECC Domain Validation Secure Server CA 2 >> > * Server certificate: COMODO ECC Certification Authority >> > > GET /web/files/pharo.png HTTP/1.1 >> > ... >> > >> > I see 2 differences: >> > >> > - a different cipher suite (combination of algorithms used) >> > TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 vs. >> > TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 with the key differences being RSA >> > vs ECDSA >> > - at least an indication of SNI (server name identification) being used >> > >> > I guess we have no control over how our website is served, since the CMS >> > used (and cloud flare CDN being involved). >> >> >> For Cloudflare customers at the free level of service >> Sites using UniversalSSL are issued SHA2+ECDSA certificates, which >> require clients that support TLSv1.2 and SNI. >> >> For customers at a paid level of service >> Cloudflare's server configuration for TLS cipher suites is set in >> nginx (which we use extensively) with the following configuration >> command: >> ssl_protocols TLSv1 TLSv1.1 TLSv1.2; >> ssl_ciphers >> EECDH+CHACHA20:EECDH+AES128:RSA+AES128:EECDH+AES256:RSA+AES256:EECDH+3DES:RSA+3DES:!MD5; >> ssl_prefer_server_ciphers on; >> >> >> https://support.cloudflare.com/hc/en-us/articles/200933580-What-cipher-suit… >> >> The Real Cost of a CloudFlare âFreeâ SSL Certificate >> https://info.ssl.com/the-real-cost-of-a-cloudflare-free-ssl-certificate/ >> >> SSL on my free plan isn't working on an old OS/Browser >> https://support.cloudflare.com/hc/en-us/articles/204144518-SSL-FAQ >> >> cheers -ben >> >> > >> > >> > > Sven >> > > >> > >> On 24 Mar 2017, at 09:08, Esteban Lorenzano <estebanlm(a)gmail.com> >> > >> wrote: >> > >> >> > >> I can confirm it too on mac. >> > >> A problem in the SSL plugin? (we know is not in the best shapeâ¦ but >> > >> Iâd like to know why this was working before and not it stopped). >> > >> >> > >> Esteban >> > >> >> > >>> On 24 Mar 2017, at 08:48, Ben Coman <btc(a)openInWorld.com> wrote: >> > >>> >> > >>> >> > >>> >> > >>> On Fri, Mar 24, 2017 at 3:02 PM, Stephane Ducasse >> > >>> <stepharo.self(a)gmail.com> wrote: >> > >>> Hi guys >> > >>> >> > >>> in the mooc (first example) we have the following expression >> > >>> >> > >>> (ZnEasy getPng: 'http://pharo.org/web/files/pharo.png') >> > >>> >> > >>> and it does not work anymore. I'm pair programming with Amal in >> > >>> Rennes >> > >>> from the university in italy and from her univ in Rennes >> > >>> we both get on 60 and 50 >> > >>> >> > >>> and time out https >> > >>> >> > >>> sslException: text code: code >> > >>> self error: >> > >>> (String streamContents: [ :stream | >> > >>> stream << 'SSL Exception: ' << text << ' [code:'. >> > >>> stream print: code. >> > >>> stream << ']' ]) >> > >>> >> > >>> Any idea. >> > >>> >> > >>> Tx in advance >> > >>> >> > >>> What platform are you on. >> > >>> >> > >>> On Windows 7, Pharo 60447, Win32 VM built on Mar 9 2017 16:38:45 >> > >>> CUT Compiler: 4.9.2 VMMaker versionString VM: 201703091627 >> > >>> I can confirm the same error with this... >> > >>> (ZnEasy getPng: 'http://pharo.org/web/files/pharo.png'). >> > >>> >> > >>> but picking some random SSL PNG, this works fine... >> > >>> (ZnEasy getPng: >> > >>> 'https://blog.keycdn.com/blog/wp-content/uploads/2015/10/http1-vs-http2.png'). >> > >>> >> > >>> It is strange that you are getting an SSL error using http rather >> > >>> than https. >> > >>> It seems the server is redirecting to http to https and maybe this >> > >>> is not handled? >> > >>> https://geekflare.com/http-to-https-redirection/ >> > >>> >> > >>> When was the last update to rewrite rules on the server? >> > >>> >> > >>> cheers -ben >> > >>> >> > >>> >> > >>> >> > >> >> > > >> > >> > >> >

March 25, 2017