- Esug-list - lists.esug.org

Re: [Esug-list] Few thoughts about Google Summer of Code
by mkobetic＠cincom.com Oct. 29, 2010

Oct. 29, 2010

"Paolo Bonzini"<bonzini(a)gnu.org> wrote: > On 10/29/2010 04:15 PM, mkobetic(a)cincom.com wrote: > > Another significant hurdle was that each dialect has it's own fileout > > format. CVST allowed using SIF as well for that reason, but that's > > foreign to every dialect. Would you be happy to commit your sources > > to git in SIF? > > I don't think inter-dialect communication is a problem. Any dialect > should be able to use whatever it prefers, like XML for VW or chunks for > Squeak or its own syntax for GST. Did CVST actually rely on SIF or did > it just "allow" using it? No, it had configurable "source codec" for that reason, so you could use native chunks or SIF (when I started, VW 5i wasn't out yet, so no XML). But SIF was critical when we had our our cross-dialect workshop with Steve Wart (Dolphin) and Brian Hogan (VA) at Camp Smalltalk in Colorado Springs. And we did get it going between the three dialects there. Anyway, yes, for dialect specific project, SIF wouldn't be necessary, but I feel we'd be missing an opportunity here. If you commit your project into something like github in a portable format, the chances of the project taking off in another dialect as well are greatly increased IMO. It may actually help our community to stop diverging so much and promote more collaboration. Look how adopting monticello in Gemstone helped the portability of Seaside between the two dialects. Why raise arbitrary barriers ? Another point I'd like to emphasize, if we want to attract people from other languages claiming integration with these VCS, it's not enough to merely use them as just a back-end store for our source. We have to step outside of our comfort zone and adopt the development model that comes with it (and it might actually do us good, this new crop of VCS has some neat stuff). We need to learn the tool and use the tool for version-control, not just hide it away and only push it from time to time with a ten foot pole. Otherwise people will see right away that we're just faking the integration. So in VW I wouldn't try to swap git in as a just a different "database back end" behind Store. I think it needs to be full on file based VC integration (ala CVST or whatever) that can run side by side with Store if you wanted.

3 2

Re: [Esug-list] Few thoughts about Google Summer of Code
by mkobetic＠cincom.com Oct. 29, 2010

Oct. 29, 2010

"Paolo Bonzini"<bonzini(a)gnu.org> wrote: > For the VCS, I've always been surprised there's no git or svn backend > for Monticello. It wouldn't seem _too_ hard to have a directory per > package and replace each .mcz file with a commit in that directory. > > Alternatively, there's actually very little GST-specific code in > http://github.com/timfel/gitocello (a Squeak+Monticello <-> GST+git > bridge), so if someone wants to play with it... I played with this back in 2000, when VW didn't have it's own VCS yet. Some might remember a project called CVST (http://cvstproj.sourceforge.net/) Ultimately I stopped working on it, because VW got Store, and Squeak got Monticello shortly after, and each of the dialect specific VCSs had some interesting features that CVS (back then) couldn't provide (not easily at least). The communities clearly valued the additional capabilities that their solutions provided. I tried to push CVST as a cross-dialect solution, but there seemed to be little interest in that. Another significant hurdle was that each dialect has it's own fileout format. CVST allowed using SIF as well for that reason, but that's foreign to every dialect. Would you be happy to commit your sources to git in SIF? Anyway, one aspect that I liked about CVST was that the image could be treated as a sort of "fancy editor" for the working copy of your code in the working directory, which might appeal to newcomers from other languages too. It's not tied to CVS in any way, you could use svn or git with it just as well (Here are slides that talk a bit more about how it worked http://www.slideshare.net/mkobetic/01-stscvst) It would require some update to plug into current packaging structures (e.g. packages in VW, it was integrated with parcels back then), but that wouldn't be difficult. Some level of dialect specific tool integration would be needed (I wrote some for VW, but those would mostly be replaced by rather simple RB extensions), overall I would expect it to be much work. What I wonder about though, if we did bring it up to date (or came up with something completely new), would people actually use it ? Because if most would still stick to their Store, Monticello X, Envy, etc, presumably because of some cool feature that they'd miss, the "yet another VCS" would be just a waste of time. But maybe times have changed and we, Smalltalkers, are more likely to accept something like svn or git even if we have to give up some of the niceties of our home-brew solutions.

2 1

Re: [Esug-list] Few thoughts about Google Summer of Code
by Damien Pollet Oct. 29, 2010

Oct. 29, 2010

On Oct 29, 2010, at 12:19, Serge Stinckwich wrote: > But comparing to Ruby, there is no standard UI in Ruby and this not > really annoying for the people, because they deploy their application > on the web. Yes, but Ruby has really nice bindings to at least GTK and Cocoa. Applications developed using that look just like their C / ObjC counterparts, they do not pop an image window, etc -- Damien Pollet type less, do more [ | ] http://people.untyped.org/damien.pollet

2 1

Re: [Esug-list] Few thoughts about Google Summer of Code
by Damien Pollet Oct. 29, 2010

Oct. 29, 2010

On Oct 29, 2010, at 14:29, Davorin Rusevljan wrote: > On Fri, Oct 29, 2010 at 2:23 PM, Paolo Bonzini <bonzini(a)gnu.org> wrote: >> On 10/29/2010 02:12 PM, Geert Claes wrote: >>> Sounds very interesting, what would the main advantages be of using a GIT >>> backend? >> >> It is what you use for other assets. So the Monticello git-backed repo >> could be simply a subtree of the project-wide repo. > > There would be also one trivial, but possibly important social aspect. > Some Smalltalk projects might be hosted on GitHub, and become visible > to larger audience. ...and benefit from github's issue tracker, from a VCS that handles branches in a way that actually makes sense, from any continuous integration tool that you can trigger on post-commit hooks, and so on. The hard point is how to fit in the file-based crowd, without imposing too many idiosyncrasies, but without loosing the power of smalltalk-specific tools. -- Damien Pollet type less, do more [ | ] http://people.untyped.org/damien.pollet

1 0

Re: [Esug-list] Few thoughts about Google Summer of Code
by Steven Kelly Oct. 29, 2010

Oct. 29, 2010

Davorin Rusevljan wrote: > Steven Kelly <stevek(a)metacase.com> wrote: > > People choose a programming environment based on (most important > > first): > > > > 1) what everybody else is using > > 2) what seems familiar to them > > 3) what is cool > > 4) what is good > > IMHO there is one criteria that comes in front of those. First > environment needs to be able to produce cool results. If environment > can do some others can not, people will be using it even if it > requires use of line editor. So being able to produce cool aps is our > priority zero. To some extent that intersects with 3), and I'd include it in 4) too: what you can produce, and how productive you are in doing it. I stand by my claim that 1) and 2) are more important in achieving mass success. 4) just helps guarantee a long life. Mind you, 1) should be rephrased to "what everybody else seems to be using", and a big score on 3) can outweigh a smaller difference on 1) and 2) over time: your message will get media coverage and some evaluators, helping to address 2) and 1) bit by bit. Steve

1 0

Re: [Esug-list] Few thoughts about Google Summer of Code
by Steven Kelly Oct. 29, 2010

Oct. 29, 2010

> On 10/29/2010 12:48 PM, Geert Claes wrote: > >> It sure cares! If the VCS doesn't allow easy integration of code > >> with other kinds of assets (pictures, CSS, schemas expressed as > >> DDL, etc.) and with other people on the team who work on those > >> assets, developers _are_ going to complain. VCS and source code organization is something that has now become standard across all mainstream languages (i.e. Java and C# :->). Your source code should be saved in class files, in a directory structure that mirrors the package/namespace structure. Other assets are stored and versioned in a similar way. You can use any VCSs with any programming language; indeed, you may be more attached to a VCS than a language. You will use one VCS for a project, but you may use multiple languages. Similar things can be said for IDEs and VMs: you probably pick one and stick to it, although you use multiple languages. People choose a programming environment based on (most important first): 1) what everybody else is using 2) what seems familiar to them 3) what is cool 4) what is good Based on that, it's obvious Smalltalk is not going to succeed in a big way in the foreseeable future. What we need to look at are what areas we lose out on in 2), and could be changed to fit what everybody else is doing, without losing much in 4). To my mind, source code in class files and normal VCS would be doable in Smalltalk, without losing anything major. In contrast, image-based development is a major factor in our competitive advantage, allowing debugging, experimentation etc. It's hard to imagine image-based development with Eclipse or VisualStudio, and implementing Smalltalk on another existing VM is currently too hard, so I'll skip those and focus on the low-hanging fruit. Which Smalltalks (+/- add-on projects) have code in class files, interface with standard VCS, and yet still have an image-based, GUI IDE? And for those that don't, how hard would it be to get there? Steve

4 3

Re: [Esug-list] Few thoughts about Google Summer of Code
by 'Markus Rips' Oct. 29, 2010

Oct. 29, 2010

>Squeak uses ToolBuilder to describe UI elements - lists, buttons, code >panes, menus, and the like. Has anyone considered writing a >SeasideToolBuilder to generate HTML/CSS? > >frank Hi Frank seaBreeze is such a tool for seaside. If someone is interested, have a look at: http://seabreeze.heeg.de/ Markus ________________________________ Markus Rips * Senior Consultant / certified Scrum Master * markus.rips(a)heeg.de phone: +49 231 9 75 99 24 * fax: +49 231 9 75 99 20 Georg Heeg eK Dortmund Handelsregister: Amtsgericht Dortmund A 12812

1 0

Re: [Esug-list] Few thoughts about Google Summer of Code
by Steven Kelly Oct. 29, 2010

Oct. 29, 2010

Jan Vrany wrote: > > On 10/29/2010 01:36 PM, Steven Kelly wrote: > > > Which Smalltalks (+/- add-on projects) have code in class files, > > > interface with standard VCS > > Smalltalk/X supports CVS and SubVersion. > > > > and yet still have an image-based, GUI IDE? > > Smalltalk/X for instance :-) Thanks! But to return to my criteria: 1) what everybody else is using 2) what seems familiar to them 3) what is cool 4) what is good Even in the Smalltalk community, Smalltalk/X isn't 1). And looking at the web pages, with not even a single picture of the IDE or Smalltalk/X applications on the Products > Smalltalk/X pages, it may find itself in pretty big trouble on 3). That's a real shame, in an era when even one-person companies and research projects have smart videos showing off their products. Particularly since from what I hear, Smalltalk/X does pretty well on 4). Sorry if that sounds harsh - I do understand the other side of the coin, i.e. that there's hardly any money to be made by selling software development environments these days. Good luck to eXept with expecco, which I guess is what they focus on now. Steve

1 0

Re: [Esug-list] Few thoughts about Google Summer of Code
by Steven Kelly Oct. 29, 2010

Oct. 29, 2010

James Robertson wrote: > -- the not quite "standard" widgets in VisualWorks make him think > -- the "all in one windows" thing in Pharo and Squeak make him think I don't think these are the major issues these days. VW's widgets are close to standard widgets these days, and most other apps' widgets diverge from the standards anyway. Squeak will suffer far more, as its look is further from the standard (last time I looked). Most IDEs are very much "all in one window", so I think VW suffers more there. The single window of Pharo and Squeak feels odd to a VW user, not to a normal person :-). Of course Pharo and Squeak could do more to have a single big IDE window with fixed panes, rather than the multiple floating windows-within-a-window that they had last time I looked. Then again, the CodeBubbles demo has got people excited, and that's pretty close to a Smalltalk "multiple browser windows" model. http://www.cs.brown.edu/people/acb/codebubbles_site.htm Steve

1 0

Few thoughts about Google Summer of Code
by Dennis Schetinin Oct. 29, 2010

Oct. 29, 2010

… and observations …and questions … mostly inspired by GSoC Mentor Summit 1. Mentor Summit was a great event! I didn't expect it to be THIS :) Very pleasant …and very hard for me: so many people, language and culture shock :) Great experience for me personally. Thank ESUG for choosing and sending me there. 2. Bad news: Smalltalk is not popular. Well, it's not that new actually, we know it. But I didn't manage to fix it :) Seriously: actually, it looks like everybody knows Smalltalk (they know it existed I mean; few knows it still exists), but it is treated like a black and white movie. People say "wow! it's cool" and go to see Avatar. :( So developers don't take Smalltalk seriously. We (Smalltalkers) know it's a mistake. But we have to show that for others. And I found myself unable to do that there. I found language and, more likely, culture barrier is too big to overcome. I didn't manage to set up a two-way communication with others. So I listened mostly. 3. And I've heard many interesting ideas… apparently, organizational ideas mostly. I need some time to process them. So, for now I'll just outline some directions… 4. Did we summarize our GSoC results? I thought I've missed them, but apparently we didn't do it. At least I found only this page http://code.google.com/p/google-summer-of-code-2010-esug/downloads/list. I didn't care much before, but now I think it is very important to look closer at the projects we did within GSoC, understand where we succeeded; where we failed; did we benefit and could we get more; make some plans for the future... etc. 5. I think we should pay more attention to GSoC. It is very important for our small society. Of course, we can't be sure smalltalkers will be invited again next year. (One of the things I tried to understand at summit but still I don't: how does Google select mentor organizations?) But this work will be very helpful, useful, advantageous… what ever. 6. We should be more serious about monitoring and controlling our projects. Since money are involved here, we have a right to. That's money from Google, but still. I think it should be a bit different from the way free projects are evolved within Smalltalk society. Since we have such opportunity we have to benefit as much as we can. 7. Smalltalk gives me competitive advantages. For that matter I don't want to :) but now I understand even better: we have to promote Smalltalk. I know everybody knows that, but I still want to state it once again. And promoting is not only about advertising and praising. We should be more open. We should find a way to start some cooperative projects with other societies/languages. I don't know how to do it. And it's extremely hard for me as I believe in Smalltalk superiority :) But still we have if we don't want Smalltalk to die. In general, it looks like making Smalltalk more open should be our priority, and this is one of the few ways to enhance prestige of Smalltalk among developers. -- Dennis Schetinin

2 1