gmane.comp.lang.smalltalk.pharo.user

Video: The Moose Roadmap

Stephan Eggermont — 2013-05-23T18:19:42

Another video from the MooseDay2013

Tudor Girba showing the Moose roadmap

http://www.youtube.com/watch?v=2JCDS3OVnGM

Stephan

Re: history graph widget

Martin Dias — 2013-05-22T17:30:50

Thanks Hernán, I'll take a look.

Martín

On Mon, May 20, 2013 at 8:19 PM, Hernán Morales Durand
<hernan.morales-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

Re: SmalltalkWorld chartFrameworks.

Tudor Girba — 2013-05-21T18:27:51

Hi,

My guess is that the problem with your old image is that you already loaded an old version of PetitParser in there and this loaded an old version of Glamour. When you loaded PetitParser with the script I gave you, it probably did not update the Glamour code.

In any case, I would suggest to never rely on an image more than one day. Images are for work, not for storage. This is why, Gofer and Metacello are great tools that liberate you from the pain of reconstructing an image. Please look into creating a script that loads everything you want in a new image.

Cheers,
Doru


On May 21, 2013, at 7:00 PM, Alain Busser <alain.busser-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:


--
www.tudorgirba.com

"Beauty is where we see it."

Re: SmalltalkWorld chartFrameworks.

Alain Busser — 2013-05-21T17:00:55

I didn't *have to* download file by file, I just wanted to try everything
that I could. Doing the Gofer trick still does not allow me to launch a
PPBrowser.

If I try with a fresh image, it works OK, that may prove that Petit works
well on Pharo 2.0, but on my own image it don't work, so there may well be
some competition between Athens and Petit (or MathsOntologie and Petit?)


Next step: I removed, by hand, any package which name begins with "petit",
then to crash Pharo I just have to do

Gofer new
         smalltalkhubUser: 'Moose' project: 'PetitParser';
         package: 'ConfigurationOfPetitParser';
         load.

and that's all!

With a fresh image, I will have to load Athens and MathsOntologie again,
won't I? Plus the changes that I already made to MathsOntologie...


On Tue, May 21, 2013 at 4:56 PM, stephane ducasse
<stephane.ducasse-GANU6spQydw< at >public.gmane.org>wrote:

Re: SmalltalkWorld chartFrameworks.

Tudor Girba — 2013-05-21T14:22:06

Exactly.

Take a fresh image. Do not download anything manually. Only execute the
script and it should work.

Doru


On Tue, May 21, 2013 at 2:56 PM, stephane ducasse
<stephane.ducasse-GANU6spQydw< at >public.gmane.org>wrote:

Re: SmalltalkWorld chartFrameworks.

stephane ducasse — 2013-05-21T12:56:42

On May 21, 2013, at 9:45 AM, Alain Busser <alain.busser-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:


why do you have to donwload file by file?

Gofer does it for you.

execute only these two expressions: 


Stef

Re: SmalltalkWorld chartFrameworks.

Alain Busser — 2013-05-21T07:45:20

Yes, I done it as you wrote with Gofer, then I used Monticello to download
petit (file by file), then I done the Gofer trick again, and I still have
the same problem: A nil as "bottom" which blocks a subtraction...

If I am the only one to encounter the problem, it probably comes from
somewhere else but where?


On Mon, May 20, 2013 at 10:37 PM, Tudor Girba <tudor-49T7vwAQlU/CX7Tz6oRo6w< at >public.gmane.org> wrote:

Re: SmalltalkWorld chartFrameworks.

Tudor Girba — 2013-05-20T18:37:31

Hi,

On May 20, 2013, at 6:05 PM, Alain Busser <alain.busser-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:


How are you downloading PetitParser? If you do not explicitly say what you do, it is hard to say what is wrong :)

I gave you detailed instructions for how to load PetitParser. Please follow them and let me know if you still have the problem:

- open a workspace in a Pharo 2.0 image
- execute (select and invoke do it):
Gofer new
         smalltalkhubUser: 'Moose' project: 'PetitParser';
         package: 'ConfigurationOfPetitParser';
         load.
(Smalltalk at: #ConfigurationOfPetitParser)
         loadDevelopment

Cheers,
Doru



--
www.tudorgirba.com

"Problem solving efficiency grows with the abstractness level of problem understanding."

Re: history graph widget

Hernán Morales Durand — 2013-05-20T18:19:04

For the grid part the SGrid supports embedding morphs, check the examples

http://stackoverflow.com/questions/4674116/are-there-any-open-source-spreadsheet-implementations-in-smalltalk-which-are-use

cheers,

Hernán

El 19/05/2013 13:24, Martin Dias escribió:

Re: history graph widget

Martin Dias — 2013-05-20T17:38:22

ah, yes, it's great! thanks


On Mon, May 20, 2013 at 6:38 PM, stephane ducasse
<stephane.ducasse-GANU6spQydw< at >public.gmane.org>wrote:

Re: history graph widget

stephane ducasse — 2013-05-20T16:38:35

in the image

timeprofiler

On May 20, 2013, at 5:59 PM, Martin Dias <tinchodias-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

Re: SmalltalkWorld chartFrameworks.

Marcus Denker — 2013-05-20T16:13:03

On May 20, 2013, at 6:09 PM, Alain Busser <alain.busser-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

There are no daily updates in 2.0.

Marcus

Re: SmalltalkWorld chartFrameworks.

Alain Busser — 2013-05-20T16:05:15

I dropped the image on Pharo and I still have this illegal subtraction
message.

I don't build anything, I just downloaded Pharo 2.0 "one click" and added
MathsOntologie, DrGeo and PetitParser. Plus the (almost) daily updates.

The debugger says that GLMSystemWindow's "addMorph: fullFrame" method has a
subtraction where WindowBorderWidth (which equals 4) is subtracted to
bottom; and bottom is nil, so it don't understand the "-" message (I doubt
a nil object understands anything).

So I guess the question is "why is bottom nil?" I see that at the beginning
there is bottom := frame bottomOffset. Whereas frame is a LayoutFrame ( (0< at >0)
+ (4< at >4)px corner: (1< at >1) + (-4< at >nil)px ) so I imagine the nil comes from this
point, no?


On Sun, May 19, 2013 at 4:01 PM, Tudor Girba <tudor-49T7vwAQlU/CX7Tz6oRo6w< at >public.gmane.org> wrote:

Re: history graph widget

Martin Dias — 2013-05-20T15:59:18

No, I think. Where can I find it?


On Mon, May 20, 2013 at 4:42 PM, stephane ducasse
<stephane.ducasse-GANU6spQydw< at >public.gmane.org>wrote:

Re: history graph widget

stephane ducasse — 2013-05-20T14:42:56

did you check whtat alain did for the profiler?


On May 19, 2013, at 6:24 PM, Martin Dias <tinchodias-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

Re: SmalltalkWorld chartFrameworks.

Juan Ignacio Vaccarezza — 2013-05-20T12:51:16

Doru,
  I loaded the moose stuf from smalltalk hub (following your instructions)
and worked perferct thanks

Regars,
Juan


On Sun, May 19, 2013 at 9:01 AM, Tudor Girba <tudor-49T7vwAQlU/CX7Tz6oRo6w< at >public.gmane.org> wrote:

Re: Fwd: widget to show progress

Igor Stasenko — 2013-05-19T22:20:41

Completely agree with that. Even without reading that post i can say why:
- you changing the global state, and if your UI system complex enough
(and it tends to grow
more and more complex in time), eventually you may end up with having
two pieces of functionality,
both trying to set cursor (unique global resource) to something they
prefer right now,
without being aware that other party also wants to change it..
The result of it, you can see in many today's desktop applications
(not pharo) - cursor flickering changing from one shape into another,
while application processing something .. which is very annoying.

However pharo image right now having like 90 senders of this message
and getting rid all of them
is going to be tedious.
I just recommended something which is available now.. if people can't
wait for better solution,
which obviously should replace existing one.
:)

Re: Fwd: widget to show progress

Denis Kudriashov — 2013-05-19T20:09:49

Hi

2013/5/19 Igor Stasenko <siguctua-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org>


There is nice post about why it is bad
http://www.cincomsmalltalk.com/userblogs/travis/blogView?showComments=true&printTitle=Cursor_consider_showWhile:_%20[Harmful]&entry=3432339015

Re: Fwd: widget to show progress

Igor Stasenko — 2013-05-19T19:46:26

Here is a simple "widget" for you:

Cursor wait showWhile: [
 ... do something which may end, but you cannot predict when...
]

:)

On 19 May 2013 13:12, Benjamin <benjamin.vanryseghem.pharo-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

history graph widget

Martin Dias — 2013-05-19T16:24:02

Hi

I would like to have a Morphic widget like this. Does anyone know if there
is something already done that I could reuse?

[image: Inline image 1]

Maybe do it with Roassal in this way:

Use the list morph, where each row knows an image morph with its
corresponding slice of the graph.

Each time the graph changes (for example, if a node is appended), render
using Roassal the part of the graph has to be updated, and cut the rendered
surface in the new slices for each updated row.

Any opinion?

Maybe I should think more in vectors (i.e. Athens) instead of pixels/images?

Cheers,
Martin

Re: SmalltalkWorld chartFrameworks.

Tudor Girba — 2013-05-19T12:01:43

Hi Alain,

As I said, something does not add up.

What do you do exactly when you are building your image?

Also, could you try following the instructions from the previous mail and let me know if it still fails?

Cheers,
Doru


On May 19, 2013, at 1:24 PM, Alain Busser <alain.busser-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:


--
www.tudorgirba.com

"If you can't say why something is relevant, 
it probably isn't."