[Moose-dev] Re: Moose/Glamour is cool (and I have questions!)

Hi, On 20 Apr 2011, at 13:09, Mariano Martinez Peck wrote: wrote: marianopeck(a)gmail.com> wrote: marianopeck(a)gmail.com> wrote: learning VM and related stuff. I never did anything because I was lazy. Last week, I took Glamour and without knowing almost anything about it, in less than 3 hours (that includes everything, downloading the image, browsing examples, searching for something similar to what I need, and code the browsers I needed) , I come up with what I want to share with you. Basically, I only have the browsers for the moment. My knowledge is very limited in the Slang to C translator, or machine code simulation. However, the "real VM hackers" will try to provide me such functionalities. in slang. When you select a method, you have 2 panels: smalltalk code and C translated code. for every method, you have 3 panels: smalltalk code, bytecodes and machine code (using the simple mapping) packages and classes of the system and for every method, you have 5 panels: smalltalk code and one panel per strategy to map to machine code (simple cogit, sista, etc). the same time: smalltalk code, bytecodes, machine code, etc. The idea is to easily learn. as well? I would love to link to it from the Moose website. waiting for some Eliot changes to be able to get the simulated machine code, and hopefully the same from Dave about C code. subclasses from GLMGlobalBrowserTemplate. example, look this screenshot: http://img4.imageshack.us/f/vmbrowsermappings.png/ love that somehow (which a little X in the tab title, or a checkbox somewhere with the available panes), one or more of those panes can be automatically removed (and if added again even better). And when I remove one of them, the others auto-adjust to ocupy the full container. It is really difficult to explain by email.. glamorousCancel entitled: 'Close' with: [:text :aClass | self halt. ]. entitled: 'Close' with: [:text :aClass | self halt. ]. super container. ) perspective. the browsing model with rendering. present the same object in 5 different ways, and only look at some of them as you wish. problem is that imageine this: (I don't need the machine code). Tomorrow I may need only the bytecodes and the machine code. Etc...so, creating a new browser for each possible combination is not a good idea. Even more in this case: http://img4.imageshack.us/f/vmbrowsermappings.png/ because this is exactly for learning how to map from smalltalk to bytecodes, or from btyecodes to machine code, or from....so tabs doesn't work in this case. It looks to me like you basically want to parameterize the structure of the browser and you are mainly worried for code reuse.

If this is the case, the bottom part should probably become one single pane and in there you could put a separate browser. Like that you could implement multiple browsers for the lower part and compose them programatically. You could also use a DynamicPresentation that decides what presentations should go inside at runtime, but it is less likely that you will need that.

presentations (like you did for Class list / Hierarchy). Hierarchy all of the image. could put them in tabs. through scripting. If you need a user interface interaction that does not match your desire, you should create your own browser or presentation type. Just do not mix the UI with the conceptual browser. look to customize that? Should I create a subclass from GLMTabulator and you that from my browsers instead ? Sorry for the newbie question..I am just getting started ;) No, the GLMTabulator is not what you want. Please, read the Glamour chapter from the Moose Book first, and then we can resume the discussion on what can be done.

Cheers, Doru -- www.tudorgirba.com "We can create beautiful models in a vacuum. But, to get them effective we have to deal with the inconvenience of reality." _______________________________________________ Moose-dev mailing list Moose-dev(a)iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev