OT: Call for politeness. Please people, keep the discussion constructively, I'm the first of apologies for/whe crossing the line. All opinions are relevant, every disagreement can put us a step closer to a consensus if we try to stay cool :)
On Oct 15, 2013, at 7:37 PM, Camille Teruel <[email protected]> wrote: > > On 15 oct. 2013, at 14:46, Camillo Bruni wrote: > >> processing.org uses monospaced font, these are the art guys that have more >> sense graphics >> than any one this mailinglist >> (BTW, how many of you have visited an art school?) > > Me, many times. > And surprisingly, most people there will tell you that (today's) art is not > concerned with aesthetics :) > >> >> Besides Smalltalk, I don't know any other language that would use proportial >> fonts. >> >> After that, anybody who really knows how to use Pharo can modify it. >> The newcomer is the only one you target... >> >> On 2013-10-15, at 13:57, Goubier Thierry <[email protected]> wrote: >>> Interesting discussion. I'll raise a few issues. >>> >>> Le 15/10/2013 13:29, Esteban Lorenzano a écrit : >>>> well... fonts and UX in general are two different (yet related) issues. >>>> >>>> UX is a huge an complicated task, and has to be taken very seriously if we >>>> want to succeed. To allow the appropriate/productive/happy flows in an >>>> environment requires a lot of effort and to put all the pieces together. >>>> Yes, I know, that sounds so general that is like not saying anything :) >>>> Here is the concrete: Put all the UX pieces together requires a lot of >>>> effort usually not taken into account. That's how the UX evolved more or >>>> less the same way as morphic: a patch over a patch without much thinking >>>> about the issue, just takign what is there and parching/extending as >>>> needed. As morphic, the current UX in pharo is broken: there is no >>>> coherence between tools and sometimes even inside the same tool (for >>>> example nautilus has different behavior inside the code panel than in the >>>> list panels on top). >>>> This is not the fault of any tool, just a consequence of how evolution was >>>> managed until now. >>> >>> Some of the thing most forgot is that when you do a GUI, what takes time is >>> not doing it, it's polishing it. Making sure all small things play together >>> nicely, and that you've spent days trying to get that drag and drop to work >>> in the perfect way, with the right feedback and all (and focus navigation, >>> and...). >>> >>>> So, we wanted a better UX for Pharo3 that included: a new Theme, new Icon >>>> set, and new tools that worked well together. But task demonstrated to be >>>> a hard to beat beast, and we just moved forward in small areas (there is >>>> for example a new centralized menu coming along with a new spotlight). >>>> And there is a prototype of a new theme and also some icons that where >>>> thought specially and that will fit nicely. But they will not be ready >>>> this year and after thinking a while (and getting feedback of people in >>>> community), we decided, for Pharo3: >>>> >>>> - adopt the glamour theme. This is a step forward our current one because >>>> glamour guys (specially Doru) continued working on it to have a really >>>> clean and simple theme. >>> >>> Is it the default theme coming with the latest 3.0, with that flat look? >>> Hate it because it breaks one HCI guideline visual cue: feedback when >>> pressing on a GUI element (scrollbars, buttons); there is none in that >>> theme. >>> >>> There it looks like a step backward, coming back to the squeak look (which >>> turned me away from squeak for many years: yes that's not rational but >>> can't get over it. Pharo was the first to give me back the feedback at the >>> GUI level) >>> >>>> - adopt the EclipsePack theme because is an iconset specially thought for >>>> programming that plays very well together. No matter if you do not like >>>> Eclipse (even if I think you are missing the relevance of Eclipse and a >>>> lot of good ideas that we could take from them), is about creating a >>>> unified vision. The old icon set (famfam) was not intended for programming >>>> environment and also there were a lot of different icons incorporated >>>> anarchically. >>> >>> Iconset are hard. But some of the Eclipse iconset are downright ugly >>> (packages). >>> >>>> - adopt a monospaced font for coding (right now Source Code Pro) and a >>>> non-monospaced for the rest (right now Open Sans). >>> >>> Hum. Once you're set on a non-monospaced font for coding, as most >>> smalltalkers have, going back to a monospaced font will hurt... I'm not >>> even using a monospaced font for coding in C :( >>> >>>> The objective is to offer a L&F that where visual elements plays well >>>> together. >>> >>> Work on that is the real key. Not sure the theme changes are part of it, >>> however. >>> >>> I value more a drive to get everything Spec-iffied: tends to create a lot >>> of common look and feel because applications tend to behave in the same way. >>> >>>> And there is another more important (IMHO) objective: to offer newcomers >>>> an environment easier to approach. Pharo (and all Smalltalk-inspired >>>> environments) is already very alien for newcomers. We get a lot of power >>>> in exchange of that alienish stuff, but very often the curve of learning >>>> or acceptance is too high and people that could step closer to us are >>>> pushed away. So, my idea is to keep been as alien as possible in the >>>> things that make us Pharo and be the less alien possible in the rest: A >>>> nice L&F that can be feel as "some kind" familiar, is part of it. >>> >>> It's a good objective, but... There is something there; Pharo is different >>> enough in it's approach that trying to match Eclipse won't work and may >>> even disrupt more, because you will make it alike where it is not. >>> >>>> Said so... well you still can switch back to the old and ugly (IMO) L&F >>>> executing some lines of code in your workspace. >>> >>> Or a setting somewhere :) >>> >>>> Same to fonts: monospaced fonts is the worldwide accepted way of present >>>> source code. Why should we stay different? >>> >>> I wouldn't be so sure of that. >>> >>>> In any case, please give it a chance before drop it (once I can actually >>>> see why the fonts are not really applied) and we'll see how it works. >>> >>> I will :) But, I'd be frank, here none of us is a HCI specialist, and it >>> shows. Sorry, but it does. No usability testing, no look into HCI >>> guidelines, but, at the same time, probably the most advanced GUI toolkit >>> available (Morphic), some of the best mind when it comes to architecturing >>> GUI code (and code on average), and the most productive environment around. >>> >>> If you want to make it familiar, look into Dolphin and VisualWorks and copy >>> that :) >>> >>> Thierry >>> -- >>> Thierry Goubier >>> CEA list >>> Laboratoire des Fondations des Systèmes Temps Réel Embarqués >>> 91191 Gif sur Yvette Cedex >>> France >>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >>> >> > >
