- Pharo-users - lists.pharo.org

Re: [Pharo-users] LibC experiments in Pharo 7

by Alexandre Bergel

Excellent! How does it compare with OSSubProcess? LibC looks much lighter. Alexandre EnvoyÃ© de mon iPad > Le 16 mars 2019 Ã 08:50, Christopher Fuhrman <christopher.fuhrman(a)gmail.com> a Ã©crit : > > I wrote a blog entry about how I've used LibC in Pharo 7. > > https://fuhrmanator.github.io/LibC-Pharo-experiments/ > > Cheers! > > Christopher

March 16, 2019

Re: [Pharo-users] LibC experiments in Pharo 7

by Sven Van Caekenberghe

This is useful, thank you! > On 16 Mar 2019, at 12:50, Christopher Fuhrman <christopher.fuhrman(a)gmail.com> wrote: > > I wrote a blog entry about how I've used LibC in Pharo 7. > > https://fuhrmanator.github.io/LibC-Pharo-experiments/ > > Cheers! > > Christopher

March 16, 2019

Re: [Pharo-users] A strange iceberg/Monticello loading issue that has appeared in 7.0.2?

by Tim Mackinnon

Mulling this over - I have a few more experiments to try 1 - try loading from the playground and not the cmd line 2 - dig out a previous 7.0.1 zero conf image (that previously worked with an older master branch), clean out iceberg-local and see if itâs something in the new master that causes this If anyone can think of an obvious thing Iâm missing, I would appreciate the input Tim Sent from my iPhone > On 16 Mar 2019, at 09:20, Tim Mackinnon <tim(a)testit.works> wrote: > > Iâm really struggling a bit with new images in 7.0.2. Now it may be that itâs not .2âs fault - but as Iâve been doing lots of repeatable things in Pharo 6 and 7a and 7.0.1, you notice when things change. > > Iâd really appreciate some thoughts on an iceberg/Monticello issue i suddenly hit last night. > > I merged a large PR in my project and then tested it in a new zero conf 7.0.2 image and directory from master (via that github pr) with the following command line (which Iâve always used): > > ./pharo-ui Pharo.image eval " > Metacello new > baseline: 'Exercism'; > repository: 'github://exercism/pharo-smalltalk:master'; > load. > > The first time it seemed to work, but I found some bugs, and so I fixed them and merged 2 more tiny prâs. When I delete everything (rm -rf), and load it again, it doesn't load everything? Those commits are missing. When look in my image my fixes aren't there in the code (but they show do in github master) - and when I open up an iceberg window in that fresh image it shows 4 incoming pull requests? Why didnât it load everything? > > It never used to do this from a clean image using Metacello? > > Can anyone think of what might cause this? Iâm really confused. > > Tim > > Sent from my iPhone >

March 16, 2019

LibC experiments in Pharo 7

by Christopher Fuhrman

I wrote a blog entry about how I've used LibC in Pharo 7. https://fuhrmanator.github.io/LibC-Pharo-experiments/ Cheers! Christopher

March 16, 2019

Re: [Pharo-users] 7.0.2 and the font glitch - seems far worse than 7.0.1

by Tim Mackinnon

Hi Hilaire - I'm not changing any font settings at all - this is a clean zero conf image, load my project, open a system browser and start typing some code and it starts happening quite quickly. The only thing that might exasperate this is me loading in mirage after my project (I need to test that theory - as it might speed up the issue). Tim Sent from my iPhone > On 16 Mar 2019, at 10:03, Hilaire <hilaire(a)drgeo.eu> wrote: > > This problem also randomly shows up in Dr.Geo/Linux. > > I think I saw it recently in DrGeo build with P7.0.1 or P7.0.2. > > Tim, did you use an alternate true type font for your Pharo GUI? > > Until know I though the issue was related to the FreeType plugin and the > use of true type font in the UI. > > Hilaire > >> Le 16/03/2019 Ã 09:25, Sven Van Caekenberghe a Ã©crit : >> I am on the latest macOS, my 64-bit 7.0.2 image with lots of code (Bootstrap/Seaside/...) is running fine. I did not fiddle with the fonts, just switched to Dark look. >> >> I avoid loading any packages that use native code, unless it is very well proven and/or part of the standard image (IceBerg for example). >> >> I have none of the issues you describe. > > -- > Dr. Geo > http://drgeo.eu > > >

March 16, 2019

Re: [Pharo-users] 7.0.2 and the font glitch - seems far worse than 7.0.1

by Hilaire

This problem also randomly shows up in Dr.Geo/Linux. I think I saw it recently in DrGeo build with P7.0.1 or P7.0.2. Tim, did you use an alternate true type font for your Pharo GUI? Until know I though the issue was related to the FreeType plugin and the use of true type font in the UI. Hilaire Le 16/03/2019 Ã 09:25, Sven Van Caekenberghe a Ã©critÂ : > I am on the latest macOS, my 64-bit 7.0.2 image with lots of code (Bootstrap/Seaside/...) is running fine. I did not fiddle with the fonts, just switched to Dark look. > > I avoid loading any packages that use native code, unless it is very well proven and/or part of the standard image (IceBerg for example). > > I have none of the issues you describe. -- Dr. Geo http://drgeo.eu

March 16, 2019

A strange iceberg/Monticello loading issue that has appeared in 7.0.2?

by Tim Mackinnon

Iâm really struggling a bit with new images in 7.0.2. Now it may be that itâs not .2âs fault - but as Iâve been doing lots of repeatable things in Pharo 6 and 7a and 7.0.1, you notice when things change. Iâd really appreciate some thoughts on an iceberg/Monticello issue i suddenly hit last night. I merged a large PR in my project and then tested it in a new zero conf 7.0.2 image and directory from master (via that github pr) with the following command line (which Iâve always used): ./pharo-ui Pharo.image eval " Metacello new baseline: 'Exercism'; repository: 'github://exercism/pharo-smalltalk:master'; load. The first time it seemed to work, but I found some bugs, and so I fixed them and merged 2 more tiny prâs. When I delete everything (rm -rf), and load it again, it doesn't load everything? Those commits are missing. When look in my image my fixes aren't there in the code (but they show do in github master) - and when I open up an iceberg window in that fresh image it shows 4 incoming pull requests? Why didnât it load everything? It never used to do this from a clean image using Metacello? Can anyone think of what might cause this? Iâm really confused. Tim Sent from my iPhone

March 16, 2019

Re: [Pharo-users] json parse problem

by Roelof Wobben

March 16, 2019

Re: [Pharo-users] Why doesn't collection have #excludes (the mirror of includes)?

by Tim Mackinnon

These are all good suggestions guys - but donât you find it odd that there isnât the mirror function #excludes: which would make all of them read more naturally? I know we canât have all combinations- but this one really struck me as odd by its absence (particularly when I was comparing the elegance of my little coding solution to other languages). Should I propose itâs inclusion in Collection? Tim Sent from my iPhone > On 16 Mar 2019, at 08:43, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > > >> On 15 Mar 2019, at 23:06, Tim Mackinnon <tim(a)testit.works> wrote: >> >> aString detect: [:c | ($0 to: $1) excludes: c] ifFound: aBlock. (Evaluate a block if the string isnât all 0 and 1âs) > > (aString allSatisfy: [ :each | '01' includes: each ]) not. > >

March 16, 2019

Re: [Pharo-users] 7.0.2 and the font glitch - seems far worse than 7.0.1

by Tim Mackinnon

Thanks for a confirmation - Iâm not sure if Mirage uses any native code, Iâll look - and will load up an image without it (possibly it is triggering something - as I know it has a background process that updates thumbnails to use for window switching). My other package is OSProcess (that I need to run an external program in a cross platform way. I think it does use native code - however Iâve been using it sine since Pharo 6, and itâs only in Pharo7 Iâve noticed this font glitching so readily). Other people have mentioned this, but it seems that this smallish combination of projects really triggers it quickly and reliably - so maybe there is a clue somewhere. It would be good to eliminate this problem as it makes Pharo look bad! Not ideal for exercism that is trying to attract new users... Tim Sent from my iPhone > On 16 Mar 2019, at 08:25, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > I am on the latest macOS, my 64-bit 7.0.2 image with lots of code (Bootstrap/Seaside/...) is running fine. I did not fiddle with the fonts, just switched to Dark look. > > I avoid loading any packages that use native code, unless it is very well proven and/or part of the standard image (IceBerg for example). > > I have none of the issues you describe. > >> On 16 Mar 2019, at 00:04, Tim Mackinnon <tim(a)testit.works> wrote: >> >> Moving this to a new thread - to avoid historic confusion as Iâm seeing a much clearer pattern now. >> >> In 7.0.1 (osx - high Sierra - launcher), I was seeing the odd font glitch (where characters suddenly start glyphing in code and menus - see photos) and it seemed to be fixable by saving your image the first time (e.g. after 30+ mins you would see this in a new install). Once the image was saved, this seemed to stick and never be a problem. >> >> In 7.0.2 this is now happening to me very 10-60 mins. The save fix still works - the moment you save the glyphs will all just disappear until it happens again. >> >> Is anyone else getting this so easily in 7.0.2 (are others using 7.0.2). Has something changed in 7.0.2 that would cause this (the change list was quite small - however its noted the welcome screen is broken in this build too , so maybe that is related)? >> >> My image is actually pretty small - just load Exercism , which relies on OSProcess, now Mocketry (with StateSpecs and Ghost) and in this occurrence I have also loaded Mirage (the lack of a decent keyboard window switcher was getting me down and mirage seems to nail that MRU style behaviour to switch between a test window and code window). >> >> It could be one of these things is the culprit (potentially mirage - as that is the newest edition, so I will create a new image and leave it out) - but wondering if others are seeing this? Or possibly Iâve found the right combination that reliably triggers this annoying issue. >> >> Tim >> >> <PastedGraphic-10.png> >> >> >> <C7359974-BA36-4D1D-A452-52EE8A91B9FF.png> >> >> >> <PastedGraphic-8.png> > >

March 16, 2019