Is there some more information about this? I opened that issue and I asks quesions about the NewList implementation on the mailing(With not so much feedback). And making NewListRender working with athens caused quite some headache to me. And, as we have already quite some mess in our widget libraries, because we thanks nicolai---------- Forwarded message ---------- From: Pharo Issue Tracker <[hidden email]> Date: 2015-05-22 11:23 GMT+02:00 Subject: FogBugz (Case [Issue]13315) Morphic - NewList without icons To: "[hidden email]" <[hidden email]>
|
Hi nicolai
Esteban is rewriting the list/table widgets and it is much much faster. He should announce a beta any soon. With it we will be able to remove - NewList (because its implementation is not good) - pluggableListMorph - and potentially the paginated and morph tree morph. Stef Le 23/5/15 00:20, Nicolai Hess a
écrit :
I agree. I would like to remove them all and get only one
Why the migration to NewList was stopped is that we realised that it was not good. Alain told us that 8 months ago.
|
you can already play with it: Gofer it smalltalkhubUser: 'estebanlm' project: 'FastTable'; configuration; loadStable. not in my place this weekend so I cannot finish it as I wanted before pushing it. My idea is to ask community help finish it. Implementation uses datasources to feed the list, so in theory you can browse infinite rows without loosing speed. according to my tests (in my machine): [ Object browse ] timeToRun asMilliSeconds. -> ~3000ms (without cache) [ Object browse ] timeToRun asMilliSeconds. -> ~350ms (with cache active) [ Object browse ] timeToRun asMilliSeconds. -> ~150ms with FastTable, always. There is an adaptor FTPluggableIconListMorphAdaptor that has the same format as a PluggableIconListMorph so changing already existing calls will be easy (even if there should not be done like that in the long term :) So, so far, list works fine, but - still missing drag&drop support - there are some issues remaining: -- cells escaping his width -- last elements in long lists sometimes are not visible some details has to be around too... but this is a good design, taken from Cocoa... is years better than the naive implementation of old PluggableListMorph and better than the strategy used by NewList (spawning threads all around). Please, help me finish! Repository is open so anybody can contribute. FTExamples has a set of wxamples with all things implemented :) Esteban ps: I already has the adaptor working with Nautilus, but since it depends on some changes on it I will wait until is integrated to push it. But basically if you want to try you just need to go to: CategoryWidget>>#buildCategoriesList, ClassWidget>>#buildClassesList and MethodWidget>>#buildMethodList and change PluggableIconListMorph with FTPluggableIconListMorphAdaptor On Sun, May 24, 2015 at 7:41 AM, stepharo <[hidden email]> wrote:
|
Thank your reply
2015-05-24 10:02 GMT+02:00 Esteban Lorenzano <[hidden email]>:
Some info about this? Is this in some mac developer guide or a paper written about the design behind cocoa?
+1
Thank you for not only having examples but class comments too!
|
In reply to this post by EstebanLM
2015-05-24 10:02 GMT+02:00 Esteban Lorenzano <[hidden email]>:
Morphic-Widgets-FastTable-NicolaiHess.7 calculate column width (needed if all are undefined / see FTExamples exampleTable1) |
2015-05-25 13:23 GMT+02:00 Nicolai Hess <[hidden email]>:
Name: Morphic-Widgets-FastTable-NicolaiHess.8 better calculation of visible number of rows for scrolling, not perfect. |
Thanks
|
In reply to this post by EstebanLM
2015-05-24 10:02 GMT+02:00 Esteban Lorenzano <[hidden email]>:
Any special reasons why you omitted doubleClick and drag&drop ? Didn't it fit into the actual design or is it ok to implement it the same way like it is in PluggableListMorph? |
In reply to this post by Nicolai Hess
thanks! this is exactly what I was hoping for (community help) :)
Esteban
|
In reply to this post by Nicolai Hess
Hi,
I didn’t omit, just still did not implemented part of them: - double click is implemented, it announces a FTStrongSelectionChanged (and the examples show it). - double click is still not implemented for the PluggableEtcAdaptor… but this was lack of time, not by design :) - drag&drop is not implemented at all but again, because lack of time. I was targeting to work on it this week, but I’m also full of other stuff, so you will made me happy if you took it :) cheers, Esteban |
In reply to this post by Nicolai Hess
Hi,
there is some that I use as reference frequently: this one: is a “retired” document, but is retired because is old (old version of objective-c and old version of cocoa), but fundamentals remains ok. instead that document, apple refers now to here: last, I use as a source of inspiration this project: http://www.cocotron.org which basically is a rewrite of cocoa but open source (and portable). Project is at least slow (if not stagnated), but I use it to consult algorithms and solutions they found to UI problems I find time to time. cheers, Esteban
|
Hi, We do some kind of a List for Bloc: Main features: - List doesn't care what kind of morphs to contain. - Flexible height of elements. Any will do. - Items can resize. - Elements can be partially visible. - Different kind of caches. - Selection strategies. Short demo video: If you want to play load: http://ws.stfx.eu/F4DRC0CVSDII Examples included. Cheers, Alex |
It looks nice I want to play with it2015-05-29 21:24 GMT+02:00 Aliaksei Syrel <[hidden email]>:
|
In reply to this post by Aliaksei Syrel
EXCELLENT
Aliaksei I can tell you that we presented bloc to thales people and they loved it. Bloc passed the acid test! Alain also shows me the logic of the PopUp with the event listener and this event listeners are great. Now I imagine that we can do inplace edit of the list items (for example for changing the protocols we will be able to double click on them :). Stef PS: Alain proposed to ask an Summer Code for you and indeed please do it.
|
In reply to this post by Aliaksei Syrel
Aliaksei
do you have a configuration of Brick? I can do one and do you have a jenkins job? Because we can do it. Stef Le 29/5/15 21:24, Aliaksei Syrel a
écrit :
|
In reply to this post by Aliaksei Syrel
https://pharoweekly.wordpress.com/2015/05/30/future-new-widget-sets-is-coming-to-you/
Le 29/5/15 21:24, Aliaksei Syrel a écrit : > If you want to play load: http://ws.stfx.eu/F4DRC0CVSDII > Examples included. > > Cheers, > Alex |
In reply to this post by stepharo
No, I don't have both of them. It would be great to have a Jenkins job. With increasing amount of tests it already becomes complicated to check stability. Cheers, On May 30, 2015 7:41 AM, "stepharo" <[hidden email]> wrote:
|
Ok I will build one when I have time.
Do you have an account on ci.inria.fr? If not please request one for the Pharo project. Le 30/5/15 11:29, Aliaksei Syrel a
écrit :
|
EXCELLENT Thanks:) Alain also shows me the logic of the PopUp with the event listener and this event listeners are great. Exactly! Event listeners are extremely powerful. And popup is very nice example in sense that from the beginning it wasn't obvious that it can be done using only event listener. And I'm sure there will be more use cases. BTW, key binding logic in the list is added as local event listener, so there is absolutely no need to subclass list morph - just add custom listener and be done) Yes, I would like to do the summer project ;) Do you have an account on ci.inria.fr? yes, login is my gmail address Ok I will build one when I have time. Perfect :) Cheers, Alex On Sat, May 30, 2015 at 11:36 AM, stepharo <[hidden email]> wrote:
|
Free forum by Nabble | Edit this page |