Thank you Noury. Writing blog posts about Atlas was my plan to pave the way toward a presentation at ESUGâ24, but when the date of the conference was announced I was sad to conclude that I cannot attend this year. So I will write more blog posts ;-) I like the discussion too. As was already clear when I organized the Pharo Browser Survey <https://all-objects-all-the-time.st/#/blog/posts/3>, a lot of people have an opinion on the tool they use every day, which is good. I welcome all the feedback. Best regards, Koen
On 13 Mar 2024, at 17:43, Noury Bouraqadi <bouraqadi@gmail.com> wrote:
I read the blog post and I liked it. Thank you Koen!
I've also liked very much the discussion in this thread. I hope that you guys would expand your ideas and give talks at the ESUG conference next July.
FYI here is the page with the ESUG call for presentations https://esug.github.io/2024-Conference/call2024.html
Noury ;-) On Mar 11 2024, at 4:27 pm, Aik-Siong Koh <askoh@askoh.com> wrote: Atlas is indeed ambitious, and I am glad you want to push the boundaries.
Over the decades, GUIs have introduced a bewildering array of widgets and designs that I think we are in a mess again. I feel we should restart again with a blank window, allow select and rightclick for context menu, and proceed for there to introduce the base minimum for a functional GUI. I believe "select and rightclick context menu" can solve almost every GUI need simply. I also think the single column menu can be updated to a two-column menu which is more compact and balanced. The cursor can be centered between the two columns instead of being at the top left corner. The cursor has to travel less for any selection. https://askoh.com/index.html#TwoColumnMenu
All the best, Aik-Siong Koh
On Mon, 11 Mar 2024 09:58:40 +0000, "Tim Mackinnon" <tim@testit.works> wrote:
As browsers are a passion in the Smalltalk world it will be great to read your thoughts - as its certainly a hot potato, and we don't seem to have quite cracked it so far.
I recall the presentations on Calypso from Pharo days (might have been recorded for review if you haven't seen them). I recall being won over at the time (and I was hesitant) - there was lots of flexibility that had been thought about, and many useful and tricky browsing patterns were covered - but over time I think its proved tricky to work with. In Pharo 11, the browsers don't seem to work as well as they should (lots of funny focus issues and loss of context that I don't recall in previous version - which I think is more down to understanding how it was supposed to work than technical flaws).
It's definitely worth generating conversation and getting some consensus otherwise it will be a rise and fall scenario all over again. This said, continuing to find a good model that is both flexible and simple is useful.
Tim
p.s. On thing I recall from those Calypso presentations was that the model should have let us design browsers where we have different navigation models (e.g. you in theory you could design something where a class has a path to methods which are both instance and class so you don't have to have a mode to swap between them - something I find distracting when designing the interface of a class and trying to figure out how you instantiate/initialize it and you want to jump between the 2 view - I just want to see all all the methods in a list, differentiated in some way vs. hiding them).
On Sun, 10 Mar 2024, at 5:24 PM, Koen De Hondt wrote: Dear Pharo users and developers,
Some people already know that I am working on a browser for Pharo. With this announcement, I make it official ð. In my latest blog post <https://all-objects-all-the-time.st/#/blog/posts/6>, I introduce Atlas as an ambitious successor of Calypso. It is the first post in a series.
Happy reading!
Ciao, Koen