Monthly Archives: October 2019

[ann] Material Design Lite for seaside 2.2.0

Hi everyone!

I am happy to announce the version 2.2.0 of Material Design Lite for seaside.

This version brings 3 new features:
– Expansion panel in the form of brushes or widget
– A table component that allows one to describe the columns content
instead of providing a collection of rows
– Predefined alerts as an extension of MDL

Thanks to Julien Delplanque who helped a lot to release this version.

You can find a demo of the MDL version at:

Have a nice day

Cyril Ferlicot

2019 Lands Platform presentation

The Lands Platform: Lan.guages and D.omain S.yntax,
@sleconf 2019, co-located with @splashcon

Showcasing live magic tricks powered by the @pharoproject

[Ann] alpha … Pillar rendering


Rendering the Pillar markup inside the Pharo image

This is work in (early) progress, but I think it is far enough for a “technology preview”.

Goal of the project

The goal of the project is to be able to do better in-image documentation. To do this, I am working on several fronts:

  1. Rendering more pillar, expecially I would like to get a solution for tables.
  2. Rendering github markdown in addition to pillar markdown. This will allow me to pull markdown from git and display it in-image.
  3. Giving the “help-system” some love.
  4. In-image tools for writing markdown – a markdown editor
  5. A design for keeping the mark-down texts seamlessly outside of the image

Reg. 1., I can only figure out how to transform a morphic table into an image, and then add that image to the text (not included atm as I find it so ugly a solution). Any other ideas are greatly appreciated.

Reg. 2. I am nearly done on this one, at least for the easy aspects (headers, lists, emphasis, code-blocks).

Reg 3 and 4. Only loose ideas atm.

Trying it out

Load it using Metacello:

Metacello new
   baseline: 'PillarRichTextRender';
   repository: 'github://kasperosterbye/PillarRichTextRender';

At the moment it will load the whole of Pillar, but I hope to fix that soon.

Once loaded, the class PRRichTextComposer has some examples on the class side.

[Ann] bootstrapping tools

I released a first version of the bootstrapping tools in

Follow the instructions in the readme for generating the environment for bootstrapping small kernels.
Once you build your development environment, two windows are shown in Pharo by default, each for bootstrapping a different small image: one is Pharo Candle, a tiny image that weights around 160kb and which logs some small benchmarks in a log.txt file. The second one is our minimal image of Pharo, which weights 1.7MB and is able to load precompiled code  using a tool called Hermes. (The generated images will initially have a size of 30mb, and will be resized after executing them for the firs time with the VM, this is a bug that I will fix soon)
The images are generated taking as input the source code of all the classes that will belong to the system (in tonel format) and they are located in the folders pharo/src for the sources of Pharo minimal and pharo/bootstrap/scripts/pharo/PharoCandleSrc for Pharo Candle.
You can check the content of these source files using the button show sources and you can export them in tonel format using the link “export to Tonel” shown in the Calypso view. (If you do that, you need to close the tool and open it again .. I will add an auto refresh soon :p )
When you click the button write , generated images are saved in the folder pharo/bootstrap/scripts/pharo/bootstrapped . You can execute these images using the pharo headless vm 32 bits (you can use the one in pharo/bootstrap/scripts/pharo/.pharo
The current tool also allows you to execute and debug code in the generated image after generating it and before writing it to disk (it has some limitations whose cause and extent I need to explore in the future), and it also allows you to inspect an already generated image (inspect button), showing you it’s special objects array for now.
This is just an early version, but it already allows to play with the bootstrapping in a more friendly way

[Ann] Beta of the Commander20 booklet


Just a little post to notify you that we will be releasing the documentation of the new Command v2.0 framework in the coming weeks.

[ANN] HeySql v1.1

Hi, I have just released HeySql, version 1.1 – a database orm for Postgresql.

– Support for migrations
– Generation of migration templates based on classes
– Models can now be subclassed
– Updated license to MIT
– Updated docs
– Some minor stuff


[Ann] ActiveX Data Objects (ADO) on Microsoft Windows

Dear all,

A working prototype of PharoADO is now available at I’d be glad if you can find the time to
test it and report your observations to this list or on the GitHub. PRs are
welcome and much needed 🙂

Firstly, I would like to focus on various database data types. When using
ADOClient and ADORecordset, the availability of data types heavily depends
from PharoWin32/PharoCOM packages, so we have to update both repositories

The PharoADO package is not huge after all, since ActiveX Data Objects
itself is an abstraction layer and provides much of the needed
functionalities by itself.

Best wishes,

[Ann] Embeddable Image in VM

Dear Consortium members
At ESUG 2019 LabWare people (Mark is on CC) told us that a key requirement for their business 
was to be able to embed an image into the executable.
It was the first time people ask us this feature (It requires that the image does not write on any external resources
and a modification of the VM). Having a silent image is something that we are working since a couple of years
from time to time with the help of V. Blondeau from LAMRC and now Lifeware. 
So we decided to show that we can do it and that we are listening to people.  
Here is the email that Pablo sent us internally and I wanted to share it with you. 

    I have made a VM with an embeddable image inside, I have done it
for Windows, but the same can be done for other operating systems.

If you want to try it (of course for windows) I have put a zip here:

It should not require any account or access.

I have modified some methods and added a way of detecting if the image
is embedded.

Do you consider is a good thing to have in the CI?


Pablo Tesone.