September 2019 - Pharo-dev - lists.pharo.org

[Pharo 8.0] Build #804: Deprecation should be optional and should be able to be disabled

by ci-pharo-ci-jenkins2＠inria.fr

There is a new Pharo build available! The status of the build #804 was: SUCCESS. The Pull Request #4775 was integrated: "Deprecation should be optional and should be able to be disabled" Pull request url: https://github.com/pharo-project/pharo/pull/4775 Issue Url: https://github.com/pharo-project/pharo/issues/deprecation Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

Sept. 30, 2019

Re: [Pharo-dev] Pumping FFI documentation

by Brainstorms

Guillermo Polito wrote >> El 29 sept 2019, a las 6:13, Brainstorms < > wild.ideas@ > > escribiÃ³: >> >> Hi Guillermo, >> >> I forked the uFFI booklet repo, branched your "version2", and revised & >> expanded the introduction section of the first chapter... >> >> I decided that before I got too far, I should submit a pull request for >> just >> that much and get some feedback in case I need trajectory tuning. >> >> Your prose is easy to edit. :^) > > haha, thanks! > I saw it and merged it already :) > >> >> And it looks like my submission promptly broke Travis... Oops. > > Iâll check it locally, probably there is a silly pillar syntax error > somewhere that generates broken latex... > >> I think I know what it is: ==ls -1 /lib/*/libc.so*== probably needs some escape characters. -t -- Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html

Sept. 30, 2019

[Pharo 8.0] Build #803: Add-comment-to-Compiledethod-isRequired

by ci-pharo-ci-jenkins2＠inria.fr

There is a new Pharo build available! The status of the build #803 was: SUCCESS. The Pull Request #4772 was integrated: "Add-comment-to-Compiledethod-isRequired" Pull request url: https://github.com/pharo-project/pharo/pull/4772 Issue Url: https://github.com/pharo-project/pharo/issues/Add Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

Sept. 30, 2019

[Pharo 8.0] Build #802: 4748-remove-slotsDictionary

by ci-pharo-ci-jenkins2＠inria.fr

There is a new Pharo build available! The status of the build #802 was: FAILURE. The Pull Request #4770 was integrated: "4748-remove-slotsDictionary" Pull request url: https://github.com/pharo-project/pharo/pull/4770 Issue Url: https://github.com/pharo-project/pharo/issues/4748 Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

Sept. 30, 2019

Re: [Pharo-dev] Pumping FFI documentation

by Guillermo Polito

> El 29 sept 2019, a las 6:13, Brainstorms <wild.ideas(a)gmail.com> escribiÃ³: > > Hi Guillermo, > > I forked the uFFI booklet repo, branched your "version2", and revised & > expanded the introduction section of the first chapter... > > I decided that before I got too far, I should submit a pull request for just > that much and get some feedback in case I need trajectory tuning. > > Your prose is easy to edit. :^) haha, thanks! I saw it and merged it already :) > > And it looks like my submission promptly broke Travis... Oops. Iâll check it locally, probably there is a silly pillar syntax error somewhere that generates broken latex... > > -Ted > > > > Guillermo Polito wrote >> Hi Ted, >> >> I split this in a separate thread to avoid noise :) >> >>> El 23 sept 2019, a las 23:14, Brainstorms < > >> wild.ideas@ > >> > escribiÃ³: >>> >>> Guillermo, >>> >>> I'm interested in helping, but at this point, I think I'd be most helpful >>> working at improving documentation (mainly editing) rather than working >>> on >>> Pharo code itself. (I'd like to work toward that, though.) >> >> Iâve been doing a pass on the structure, and I was thinking on a rough >> structure as follows: >> 1) Intro to FFI (callouts, function and library lookup, intro to value >> marshalling) >> 2) Marshalling (sending arguments, literal arguments, more on >> marshalling, basic C types: ints, floats, pointers and how they are >> transformed to pharo objects and vice-versaâ¦) >> 3) Complex types: strings, unions, arrays, opaque types >> 4) Derived types on the Pharo side: How to design nice classes with all >> this >> 5) Callbacks >> 6) Memory management >> >> I did already a pass on 1), and I got blocked in 2), though I want to >> release a version of it this week. >> >> If youâre up for it, there are several things we can do: >> - review the english :) >> - give feedback on what is missing, what is not understandable, what can >> be explained better >> - testing the examples? >> >>> >>> I'm still a newbie with Pharo, but I am a good writer/editor. And I >>> expect >>> that working with Pharo documentation would be another means of >>> increasing >>> my knowledge of the Pharo ecosystem -- so that's additional incentive for >>> me. >> >> Cool :) >> >>> I gather that the PDF books are written using Pillar, which I know >>> nothing >>> about. Are there resources & guides for this tool/format that would help >>> me >>> learn how to make & edit these kinds of documents? >> >> Pillar is a markup syntax (from Pierâs CMS, if you know it). >> https://github.com/pillar-markup/pillar >> <https://github.com/pillar-markup/pillar> >> >> Pillar comes with a document model, parser and generators to html, pdf >> (through latex), and othersâ¦ >> In Pillarâs readme there are the installation instructions + usage. >> >> If you check the travis file in the ffi booklet repository >> >> https://github.com/SquareBracketAssociates/Booklet-uFFI/blob/version2/.trav… >> <https://github.com/SquareBracketAssociates/Booklet-uFFI/blob/version2/.travis.yml> >> >> Youâll see it is built with pillar 7.4.1. In other words >> >> # install pillar >> $ git clone https://github.com/pillar-markup/pillar.git -b v7.4.1 >> $ cd pillar && ./scripts/build.sh && cd .. >> >> # go into the booklet repository and build the pdf >> $ ./pillar/build/pillar build pdf >> >> Although youâll need a mostly up-to-date latex version (latexmk required, >> plus several other packages, check Pillarâs readme) >> >>> Also, I've never contributed to an open source project; Pharo seems to be >>> a >>> good place to start doing so. I see that most of the documentation, web >>> pages, booklets, etc. are in English so there's the advantage that >>> English >>> is my first language (and I actually paid attention in school :^). I'm >>> also aware, from experience, that Documentation is rarely the first >>> choice >>> for developers to apply their time & enthusiasmâ¦ >> >> And itâs super important nevertheless ^^. >> >> Guille > > > > > > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html >

Sept. 30, 2019

Re: [Pharo-dev] Testing "Weekly Digest" for GitHub

by Marcus Denker

> On 23 Jul 2019, at 08:04, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > > >> On 17 Jul 2019, at 07:46, Marcus Denker <marcus.denker(a)inria.fr> wrote: >> >> Hi, >> >> This seems to have been caused by a faulty config (I really hate YAMLâ¦). >> >> I changed it *and* put the day to be saturday, as they seem to have run into rate limiting >> for the API on Sundays in the past. >> >> So hopefully this will work saturday again. >> > > I finally remove the config file and deinstalled/reinstalled the bot. > > https://github.com/pharo-project/pharo/issues/4024 > > "Last week, 64 pull requests were merged.â > Not very stable these botsâ¦ after some weeks, now we got a report: Weekly Digest (23 September, 2019 - 30 September, 2019) https://github.com/pharo-project/pharo/issues/4767 <https://github.com/pharo-project/pharo/issues/4767> I hope that it will work better the next weeks (it is configured to run modays). Marcus

Sept. 30, 2019

Re: [Pharo-dev] Cannot extract stable VM on Window

by Brainstorms

Good sleuthing, Ben! -- Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html

Sept. 30, 2019

Re: [Pharo-dev] Cannot extract stable VM on Window

by Ben Coman

On Mon, 30 Sep 2019 at 02:41, Cyril Ferlicot D. <cyril.ferlicot(a)gmail.com> wrote: > Le 29/09/2019 Ã 16:35, Brainstorms a Ã©crit : > > Hi Cyril, > > > > I downloaded it and tried it on Win7 Pro 64bit (running in Virtualbox), > and > > was able to open as expected. > > > > However, looking in the zip file itself, I noticed about two dozen > > "*_Zone.Identifier" files that I was not expecting to see. They likely > > should not be there; they have something to do with IT security > inspections > > on downloaded files, and I delete them as a matter of course whenever I > see > > them (as part of a download). I'm not sure why the Pharo build process > > would have these. > > > > I tried launching Pharo from this zip file before and after I removed > these > > files... It worked in both cases; no corruption reported. However, > since > > your error dialog was reporting one of these 'zone' files, I would trying > > removing them and see if that helps. > > > > Thanks! > > With your comment I succeeded to launch my image. What I needed to do > was to open the zip file without extracting it, delete all the > .Identifier files and extract it once done. > > I wonder how the vm zip files end up with those files in them. > Looking inside... http://files.pharo.org/vm/pharo-spur64/win/stable-20190916.zip I see filename "FT2Plugin.dll:Zone.Identifier" The colon in the filename is a clue that it was originally an Alternate Data Stream attached to the root file https://www.2brightsparks.com/resources/articles/NTFS-Alternate-Data-Stream… That ZoneIdentifier ADS wont show in a Opening "FT2Plugin.dll_Zone.Identifier" in Notepad shows it contains... [ZoneTransfer] ZoneId=3 ...which seems like its been tagged by a Windows system function into an untrusted "Internet Zone" http://woshub.com/how-windows-determines-that-the-file-has-been-downloaded-… So it seems not-malicious, since I'd guess an attacker would be putting it in a "Trusted Zone" Looking in directory... http://files.pharo.org/vm/pharo-spur64/win/ for the following at recent consecutive files, the ones marked "Y" have the ADS N 2019-09-13 22:18 8.1 MB pharo-win-x86_64-201909131927-218e97a.zip N 2019-09-15 00:16 8.1 MB pharo-win-x86_64-201909142122-f5de9a4.zip Y 2019-09-20 11:57 6.8 MB stable-20190916.zip Y 2019-09-20 11:57 6.8 MB pharo-win-x86_64-201909161029-19f5d00.zip N 2019-09-21 21:50 8.1 MB pharo-win-x86_64-201909211859-521c75a.zip N 2019-09-25 19:12 8.1 MB pharo-win-x86_64-201909251551-3023fbc.zip So it seems to be a blip related around the moment that the "stable" zip was created. I note that the 2019-09-20 zips include the following additional files the surrounding zips don't have. SurfacePlugin.dll SqueakSSL.dll libfontconfig-1.dll libexpat-1.dll libbz2 FTPlugin.dll Highly speculative, but one scenario could be that "pharo-win-x86_64-201909161029-19f5d00.zip" was downloaded to someone's machine to add those files and it got tagged by Windows. They then got included when " pharo-win-x86_64-201909161029-19f5d00.zip " was rezipped and also copied to "stable-20190916.zip" ?? It would be good to understand what happened here. A solution might be to exclude ADS when "stable-20190916.zip" is created. btw, The `dir` command is not aware of ADS. This can be used to investigate... https://docs.microsoft.com/en-us/sysinternals/downloads/streams cheers -ben

Sept. 30, 2019

Re: [Pharo-dev] Cannot extract stable VM on Window

by Brainstorms

I'm glad it did the trick. And that's a VERY good question... Those files shouldn't be there, and it suggests that the files were downloaded/transferred over a network, then packaged into the zip. Why? (I don't know any other function that creates these, as I've only ever seen them when downloading on a Windows machine. They're not valid filenames (to Windows), which is an additional annoyance.) What I don't like about the look of it is that it has that kind of feel of a "man in the middle" type of attack. Not that this occurred, but it makes me a little uncomfortable. (Enough that I snapshotted my Vbox VM to test it on, so that I can destroy it afterward.) The maintainers might fill us in... -t CyrilFerlicot wrote > Le 29/09/2019 Ã 16:35, Brainstorms a Ã©critÂ : >> Hi Cyril, >> >> I downloaded it and tried it on Win7 Pro 64bit (running in Virtualbox), >> and >> was able to open as expected. >> >> However, looking in the zip file itself, I noticed about two dozen >> "*_Zone.Identifier" files that I was not expecting to see. They likely >> should not be there; they have something to do with IT security >> inspections >> on downloaded files, and I delete them as a matter of course whenever I >> see >> them (as part of a download). I'm not sure why the Pharo build process >> would have these. >> >> I tried launching Pharo from this zip file before and after I removed >> these >> files... It worked in both cases; no corruption reported. However, >> since >> your error dialog was reporting one of these 'zone' files, I would trying >> removing them and see if that helps. >> > > Thanks! > > With your comment I succeeded to launch my image. What I needed to do > was to open the zip file without extracting it, delete all the > .Identifier files and extract it once done. > > I wonder how the vm zip files end up with those files in them. > >> -Ted >> >> >> >> >> -- >> Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html >> > > > -- > Cyril Ferlicot > https://ferlicot.fr > > > > signature.asc (836 bytes) > <http://forum.world.st/attachment/5104673/0/signature.asc> -- Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html

Sept. 29, 2019

[Pharo 8.0] Build #801: Making the ReadWrite buffered stream to be used only by SourceFile

by ci-pharo-ci-jenkins2＠inria.fr

There is a new Pharo build available! The status of the build #801 was: SUCCESS. The Pull Request #4752 was integrated: "Making the ReadWrite buffered stream to be used only by SourceFile" Pull request url: https://github.com/pharo-project/pharo/pull/4752 Issue Url: https://github.com/pharo-project/pharo/issues/moving Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

Sept. 29, 2019