- Pharo-users - lists.pharo.org

Re: [Pharo-users] Update of ConfigurationOfGlorp

by Herby VojÄÃk

stephan wrote: > On 11-10-17 20:33, Herby VojÄÃk wrote: >> 127 is general, is not sqlite-specific. Fixes any case where primary >> key is not primitive and has converter. > > Should 127 and 129 be merged first, and should current development be > promoted to release? And why is 128 missing? > > Stephan > > 129 is from different author, who skipped upload of its 127 and 128 (probably intermediates). 127 is mine, fixes IMO overoptimized case for DirectMapping which unlike its superclass (thus all the other mappings as there is no other specialization), stopped converting primary keys to db value in fk->pk relationships.

Oct. 11, 2017

Re: [Pharo-users] Update of ConfigurationOfGlorp

by stephan

On 11-10-17 20:33, Herby VojÄÃk wrote: > 127 is general, is not sqlite-specific. Fixes any case where primary key > is not primitive and has converter. Should 127 and 129 be merged first, and should current development be promoted to release? And why is 128 missing? Stephan

Oct. 11, 2017

Re: [Pharo-users] X11 options on Ubuntu VM / Athens rendering problems

by Hilaire

We tested it on P6.1 64bits, and this Cairo crash seems to be gone. It may need more testing though. Is it thanks to P6.1 or the 64bits image/VM? It looks strange because it looked like a lib Cairo crash, but in the other hand the faulty C function did not appear to be called from Pharo image. What need to be done porting to P6.1, is mainly image shrinking, fix of several bugs from P6.1 impacting drgeo, working setup for linux/windows/mac. Still some job. Hilaire Le 02/10/2017 Ã 15:19, Stephane Ducasse a Ã©critÂ : > How could we help? > > > On Sun, Oct 1, 2017 at 9:14 PM, Hilaire<hilaire(a)drgeo.eu> wrote: >> I had situation where the system libcairo:i386 does not work with Dr.Geo[1], >> Pharo crashes with what looks like a Cairo crash! Looks like not a Pharo >> fault, but I am definitely stuck. >> >> It is problematic, it is a system (LinuxMint based) used in Geneva primary >> schools and Dr. Geo was expected to be used there:( >> >> Hilaire >> >> [1]http://forum.world.st/Cairo-related-crash-with-Dr-Geo-tt4952023.html >> >> >> Le 01/10/2017 Ã 19:50, Igor Stasenko a Ã©crit : >>> All these difficulties are good argument to get the appropriate >>> libcairo/libpng shipped with the Linux VM, as done with Windows >>> and MacOSX VM >>> >>> Nah, in my case, there's nothing that can be lableled as "Pharo fault". It >>> is definitely a linux/ubuntu ecosystem fault. - by not providing parallel >>> updates for 32 & 64 bit variant of same library. >> -- >> Dr. Geo >> http://drgeo.eu >> >> >> -- Dr. Geo http://drgeo.eu

Oct. 11, 2017

Re: [Pharo-users] Update of ConfigurationOfGlorp

by Herby VojÄÃk

stephan wrote: > On 06-10-17 17:22, Herby VojÄÃk wrote: >> Any chance of incorporating fixes 127 / 129? > > Sure, as soon as someone tells me they are safe to > add. I am just testing Glorp with P3 and Postgres now, > and don't have the capacity to verify these changes > other than by just reading the delta. > > Stephan > > > 127 is general, is not sqlite-specific. Fixes any case where primary key is not primitive and has converter.

Oct. 11, 2017

Re: [Pharo-users] Behold Pharo: The Modern Smalltalk

by Dimitris Chloupis

Well there is a move towards Pillar for class and method commands so who knows maybe we will have that soon enough ;) There is a solution, not a great solution but it can do this. You can kinda do this via metacello and filetree, because when you export to git , it coverts class and method comments to markdown that can be viewed like online documentation . They are saved as README.md which is the standard for github. That means you can use markdown inside the class and method comments , and then convert that to HTML . Tons of markdown to html coverters out there so it would not be a problem. Of course Pharo users may not be happy to read comments in markdown but then markdown has minimal syntax so it should not be such a major issue. The real problem is the class comments itself, if you lack comments talking about automated API reference documentation sounds the least of the problems. Cannot blame developers, documentation is as much work as writting the code and I can understand that some may find it much less pleasant than I do :) I tried my best to help on the documentation department but we need more people because its an effort not only to document but also to keep the documentation updated. One of the perk working in an enviroment improving very fast, but you wont get any complain from me about this ;) On Wed, Oct 11, 2017 at 6:02 PM Offray Vladimir Luna CÃ¡rdenas < offray.luna(a)mutabit.com> wrote: > Yes. I know them. I mean API docs as static files. I don't really sold on > them compared with a live system and I don't think static API docs are > critical for Pharo success. > > Cheers, > > Offray > > On 11/10/17 09:52, Dimitris Chloupis wrote: > > Ah and my static website was built with Pillar and Bootstrap, using > bootstrap templates was easy because Pillar supports mustache that makes > html manipulation much easier > > http://www.kilon-alios.com > > Pillar of course is not made for generating websites but itâs an awesome > Pharo library that allows for great degree of freedom so I thought , why > not ? > On Wed, 11 Oct 2017 at 17:48, Dimitris Chloupis <kilon.alios(a)gmail.com> > wrote: > >> Docs are available in static online html format , at least the book I was >> working on >> >> Pharo By Example >> >> You can find those links here >> >> https://github.com/SquareBracketAssociates/UpdatedPharoByExample >> >> Our documentation system , Pillar , outputs pdf , html and markdown >> files. >> >> If the book in question is built like PBE with CI of Inria where most >> Pharo related official projects are built then it should have at least pdf >> and html with online access so you can easily link to. >> >> Donât quote me on this but I think the html output of pillar generate >> links even for paragraphs you can do an even more process linking to the >> documentation. >> On Wed, 11 Oct 2017 at 17:40, Offray Vladimir Luna CÃ¡rdenas < >> offray.luna(a)mutabit.com> wrote: >> >>> The more I use Pharo, the less I use web documentation. For me seems >>> pretty suboptimal compared to live environment with code browser and >>> GT-Spotter. Regarding the comment on Medium, it also took me little to find >>> #raisedTo:, so the millage can vary. What I was missing was proper books >>> for particular domains, but Pharo books are covering that. I don't know if >>> a Q&A site could improve search-ability for newbies (certainly you can find >>> little stuff in Stack Overflow). >>> >>> My bet is about trying to create more "end user" tools (Grafoscopio is >>> kind of this), besides tools for developers. There is a broad community of >>> people who can be active contributors and members of the community, welcome >>> Pharo and live coding a lot and don't complain that much about stuff that >>> is not already pretty similar to what they already know (being that only >>> English MOOC or online static html docs). >>> >>> Cheers, >>> >>> Offray >>> >>> On 11/10/17 07:34, Dimitris Chloupis wrote: >>> >>> for me it is a yes and no situation, yes its very coold to have your >>> entire system in your fingertips but Pharo has serious issues with code >>> organisation and I find the lack of namespaces quite inconvenient. You have >>> to be careful how to name your classes which does not sound to me very OOP >>> friendly. >>> >>> Also the IDE does not handle spaggetification very well, sure you can >>> find implementors , senders etc but if the execution chain is complex , >>> welcome to spaggeti hell. But that is a problem with most other IDEs if not >>> all as well. Problem is in this case that we have the very good rule of >>> using sort methods which multiplies this problem and makes navigation even >>> harder. Code becomes much easier to read per method and messages but much >>> harder to understand in a bird eye view. >>> >>> Some of that pain has been aleviated with the introduction of GTSpotter >>> which I have praised quite a lot and I will continue to do so. But yeah >>> there are more needed to be done in the department to make Pharo code >>> navigation a more comfortable task. >>> >>> On Wed, Oct 11, 2017 at 2:57 PM Vitor Medina Cruz <vitormcruz(a)gmail.com> >>> wrote: >>> >>>> I dunno, maybe Iâm weird, but I find the System Browser a fantastic way >>>> to explore the class library. If you find a class or method that isnât well >>>> documented, write a comment and send a change request. Stef told me this >>>> ages ago. I might add, if you find a bug you should write a test that >>>> exercises the bug and submit it on fogbugz (the bug tracking system). >>>> >>>> >>>> I will reference of response of mine to a similar opinion made by >>>> Richard: >>>> https://medium.com/@vitormcruz/i-disagree-it-is-much-harder-to-find-anythin… >>>> >>>> My 2 cents. >>>> >>>> >>>> >>>> >>>> On Tue, Oct 10, 2017 at 11:59 PM, john pfersich <jpfersich(a)gmail.com> >>>> wrote: >>>> >>>>> >>>>> > On Oct 10, 2017, at 09:58, horrido <horrido.hobbies(a)gmail.com> >>>>> wrote: >>>>> > >>>>> > Interestingly, I'm getting a fair amount of pushback on this. >>>>> Personally, I >>>>> > think it would be very helpful to have a live (updatable, so as to >>>>> keep it >>>>> > current) reference page for the class library, something that >>>>> developers can >>>>> > easily look up what they need. After all, most of the power of Pharo >>>>> comes >>>>> > from the class library and we need to make it as accessible as >>>>> possible to >>>>> > less experienced Pharoers (i.e., beginners). >>>>> > >>>>> > Exploring the class library through the System Browser is very >>>>> inefficient. >>>>> > This is further exacerbated by the fact that many classes and >>>>> methods are >>>>> > simply not well-documented (containing a cursory remark which is >>>>> just barely >>>>> > useful). >>>>> > >>>>> I dunno, maybe Iâm weird, but I find the System Browser a fantastic >>>>> way to explore the class library. If you find a class or method that isnât >>>>> well documented, write a comment and send a change request. Stef told me >>>>> this ages ago. I might add, if you find a bug you should write a test that >>>>> exercises the bug and submit it on fogbugz (the bug tracking system). >>>>> >>>>> > I realize that creating a live reference page is not easy to do. In >>>>> fact, >>>>> > it's a lot of work. But the absence of such a page is a real >>>>> obstacle to >>>>> > Pharo acceptance. >>>>> > >>>>> > >>>>> > >>>>> > horrido wrote >>>>> >> Thanks. I gave your answer verbatim. I also added the following >>>>> paragraph: >>>>> >> >>>>> >> The problem I find with todayâs developers is that they are rather >>>>> >> closed-minded. They are rigid and inflexible, and not willing to >>>>> adapt to >>>>> >> new and different ways of doing things. In my generation (circa >>>>> >> 1980â1990), >>>>> >> people didnât have a problem with trying different technologies. >>>>> Thatâs >>>>> >> why >>>>> >> I had no issue with learning Smalltalk 10 years ago, after I had >>>>> retired >>>>> >> from a 20-year-long career in C systems programming and FORTRAN >>>>> scientific >>>>> >> programming. >>>>> >> >>>>> >> >>>>> >> >>>>> >> Sven Van Caekenberghe-2 wrote >>>>> >>>> On 6 Oct 2017, at 14:54, horrido < >>>>> >> >>>>> >>> horrido.hobbies@ >>>>> >> >>>>> >>> > wrote: >>>>> >>>> >>>>> >>>> I received this comment from someone who complained: >>>>> >>>> >>>>> >>>> *What about the lack of documentation? From time to time Iâve >>>>> checked >>>>> >>>> some >>>>> >>>> SmallTalk implementations like Squeak, GNU-Smalltalk and now >>>>> Pharo. Of >>>>> >>>> these, only GNU-SmallTalk appears to have a free, official >>>>> programming >>>>> >>>> guide >>>>> >>>> and core library reference that any serious programmer expects >>>>> from a >>>>> >>>> language. >>>>> >>>> >>>>> >>>> https://www.gnu.org/software/smalltalk/manual-base/html_node/* >>>>> >>>> >>>>> >>>> I pointed to Pharo's documentation but then he came back with: >>>>> >>>> >>>>> >>>> *Then show me a link of the free, maintained reference >>>>> documentation for >>>>> >>>> the >>>>> >>>> classes that form âthe core libraryâ, like this one for Python >>>>> >>>> (https://docs.python.org/3/library/index.html)* >>>>> >>>> >>>>> >>>> It's true, most Smalltalks do not have a core library reference, >>>>> not >>>>> >>>> even >>>>> >>>> VisualWorks! So what is the proper response to this complaint? >>>>> >>> >>>>> >>> The first answer is that Pharo/Smalltalk is unique in that a >>>>> running >>>>> >>> system/IDE contains _all_ source code, _all_ documentation (class, >>>>> >>> method, >>>>> >>> help, tutorial), _all_ unit tests and _all_ runnable examples in a >>>>> very >>>>> >>> easy, accessible way. It takes some getting used to, but this is >>>>> actually >>>>> >>> better and much more powerful than any alternative. >>>>> >>> >>>>> >>> The second answer is that there are lots of books and articles >>>>> that take >>>>> >>> the classic/structured book/paper approach. There is >>>>> >>> http://books.pharo.org, http://themoosebook.org, >>>>> >>> http://book.seaside.st/book, http://medium.com/concerning-pharo >>>>> and many >>>>> >>> more. >>>>> >>> >>>>> >>>> Thanks. >>>>> >>>> >>>>> >>>> >>>>> >>>> >>>>> >>>> -- >>>>> >>>> Sent from: >>>>> http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >>>>> >>>> >>>>> >> >>>>> >> >>>>> >> >>>>> >> >>>>> >> >>>>> >> -- >>>>> >> Sent from: >>>>> http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > -- >>>>> > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >>>>> > >>>>> >>>>> >>>> >>> >

Oct. 11, 2017

Re: [Pharo-users] Behold Pharo: The Modern Smalltalk

by Dimitris Chloupis

major mistake i meant to say "providing a full live enviroment it would NOT be easy because I would have to find a mechanism to accomodate for tons of Python code that does not follow live coding standards " On Wed, Oct 11, 2017 at 8:08 PM Dimitris Chloupis <kilon.alios(a)gmail.com> wrote: > Yes sorry for not making this clear. I was wrong in this case. A module > reload will only replace class objects not instance objects. So if you > create a new instance after the reload it will have the updated version but > your old instance would not. > > I try to test something before I make a claim but in this case I failed > because I forgot the obvious. The tiny fact I was reinitialising my objects > in the python project I am currently using live coding, so of course my > instances were updating instantenously because I was in a loop that was > recreating them every few milliseconds. > > When webwarrior mentioned this it draw my supisions and I tested it > outside my code in the interpreter and it became obvious I was wrong. This > why my reply I dont claim he is wrong and I am talking about injecting > methods to old objects. > > I have no problem admitting when I am wrong but in this case maybe I was > not clear enough considering I am responsible for exploding this thread , I > hope in a good way. > > So yes its possible to do in Python but no it is not coming already > available to you, so you can restrain yourself from abandoning Pharo for > Python ;) > > Python is not there yet. I still keep the opinion that is easy to do , > method injection is just regular assignment, you cannot get any easier than > assignment but this means you have to keep track of the instances of a > class and repace their methods. The good news is that if you replace just > the methods and not the entire object , you can keep the references to the > old instance because keeping track of the references as well would be > trickier. So this way you would not brake your references and having to > rebuild them to point to a new instance because you continue to use the old > instance but with updated methods. You can do this also with adding / > removing / editing variables and/or methods. So it does have a wide > application. Python can give you back the references to and from an object > (via gc module) but the problem is that when an object is created its > already referenced by many internal stuff so it can get messy soon. Plus > Python is not exactly a high performance language , unless you use third > party libraries focusing on performance (basically C libraries wrapped for > Python). > > In my case its easy to do because I care about live coding my own project > and do not care about code outside of it, because I wont be changing it > anyway. But if I had to do this the Pharo way , providing a full live > enviroment it would be easy because I would have to find a mechanism to > accomodate for tons of Python code that does not follow live coding > standards by a long margin. So my goal is not to come up with full blown > live coding enviroment like Pharo does, that is simply not possible because > there is like a ton of python code out there and I am sure there so many > scenarios that live coding in python can go wrong that I am not even aware > of. But if live code is only for my code and my code alone , I dont think I > will have any major issues. Afterall even when I use Pharo I use live > coding strictly for my code and rarely touch the enviroment with the > exception of once messing with UI theme classes. > > But then again I am so new to this that at any point I may come up for a > solution about this too, who knows :) > > On Wed, Oct 11, 2017 at 7:05 PM Pierce Ng <pierce(a)samadhiweb.com> wrote: > >> On Sat, Oct 07, 2017 at 01:41:17AM +0000, Dimitris Chloupis wrote: >> > execution. Hence live coding, the Python VM replaces objects lively. >> Python >> > can also compile any size of code including individual methods. >> > That happens with one line of code importlib.reload(mymodule) >> >> AFAIK in Python when you modify and reload source, existing instances are >> not >> "reshaped" to the modified code. Not live enough. Of course with Python >> objects >> it is easy to add per-instance inst-vars, but still this needs to be >> scripted >> and doesn't come free with the programming environment. >> >> Pierce >> >>

Oct. 11, 2017

Re: [Pharo-users] Behold Pharo: The Modern Smalltalk

by Dimitris Chloupis

Yes sorry for not making this clear. I was wrong in this case. A module reload will only replace class objects not instance objects. So if you create a new instance after the reload it will have the updated version but your old instance would not. I try to test something before I make a claim but in this case I failed because I forgot the obvious. The tiny fact I was reinitialising my objects in the python project I am currently using live coding, so of course my instances were updating instantenously because I was in a loop that was recreating them every few milliseconds. When webwarrior mentioned this it draw my supisions and I tested it outside my code in the interpreter and it became obvious I was wrong. This why my reply I dont claim he is wrong and I am talking about injecting methods to old objects. I have no problem admitting when I am wrong but in this case maybe I was not clear enough considering I am responsible for exploding this thread , I hope in a good way. So yes its possible to do in Python but no it is not coming already available to you, so you can restrain yourself from abandoning Pharo for Python ;) Python is not there yet. I still keep the opinion that is easy to do , method injection is just regular assignment, you cannot get any easier than assignment but this means you have to keep track of the instances of a class and repace their methods. The good news is that if you replace just the methods and not the entire object , you can keep the references to the old instance because keeping track of the references as well would be trickier. So this way you would not brake your references and having to rebuild them to point to a new instance because you continue to use the old instance but with updated methods. You can do this also with adding / removing / editing variables and/or methods. So it does have a wide application. Python can give you back the references to and from an object (via gc module) but the problem is that when an object is created its already referenced by many internal stuff so it can get messy soon. Plus Python is not exactly a high performance language , unless you use third party libraries focusing on performance (basically C libraries wrapped for Python). In my case its easy to do because I care about live coding my own project and do not care about code outside of it, because I wont be changing it anyway. But if I had to do this the Pharo way , providing a full live enviroment it would be easy because I would have to find a mechanism to accomodate for tons of Python code that does not follow live coding standards by a long margin. So my goal is not to come up with full blown live coding enviroment like Pharo does, that is simply not possible because there is like a ton of python code out there and I am sure there so many scenarios that live coding in python can go wrong that I am not even aware of. But if live code is only for my code and my code alone , I dont think I will have any major issues. Afterall even when I use Pharo I use live coding strictly for my code and rarely touch the enviroment with the exception of once messing with UI theme classes. But then again I am so new to this that at any point I may come up for a solution about this too, who knows :) On Wed, Oct 11, 2017 at 7:05 PM Pierce Ng <pierce(a)samadhiweb.com> wrote: > On Sat, Oct 07, 2017 at 01:41:17AM +0000, Dimitris Chloupis wrote: > > execution. Hence live coding, the Python VM replaces objects lively. > Python > > can also compile any size of code including individual methods. > > That happens with one line of code importlib.reload(mymodule) > > AFAIK in Python when you modify and reload source, existing instances are > not > "reshaped" to the modified code. Not live enough. Of course with Python > objects > it is easy to add per-instance inst-vars, but still this needs to be > scripted > and doesn't come free with the programming environment. > > Pierce > >

Oct. 11, 2017

Re: [Pharo-users] Behold Pharo: The Modern Smalltalk

by Pierce Ng

On Sat, Oct 07, 2017 at 01:41:17AM +0000, Dimitris Chloupis wrote: > execution. Hence live coding, the Python VM replaces objects lively. Python > can also compile any size of code including individual methods. > That happens with one line of code importlib.reload(mymodule) AFAIK in Python when you modify and reload source, existing instances are not "reshaped" to the modified code. Not live enough. Of course with Python objects it is easy to add per-instance inst-vars, but still this needs to be scripted and doesn't come free with the programming environment. Pierce

Oct. 11, 2017

Re: [Pharo-users] Behold Pharo: The Modern Smalltalk

by Offray Vladimir Luna CÃ¡rdenas

Seaside for html live docs browsing seems a good idea. How do you create something like [1]?: [1] http://files.pharo.org/doc/4.0/#packageList=package.html&classList=package/… Cheers, Offray On 11/10/17 10:31, Marcus Denker wrote: > > >> On 11 Oct 2017, at 17:18, Thierry Goubier <thierry.goubier(a)gmail.com >> <mailto:thierry.goubier@gmail.com>> wrote: >> >> What about a seaside-based system browser oriented towards code >> documentation? That would fit both paradigms: html-like reference + >> live browsing. >> >> You could even make it buzzword compliant: have a REST interface to >> documentation and code. >> >> Note how the panes in a system browser could make for a nice URI: >> just look for >> >> pharo://Kernel/Number/mathematical%20functions/raisedTo: >> >> And we could even include version numbers, etc.... >> >> pharo://6.1/Kernel/Number/mathematical%20functions/raisedTo: >> >> Not very different from the github url for the pharo project relevant >> file... >> >> https://github.com/pharo-project/pharo/blob/development/src/Kernel.package/… >> > > There isÂ http://files.pharo.org/doc/, too.Â > > Marcus >

Oct. 11, 2017

Re: [Pharo-users] Behold Pharo: The Modern Smalltalk

by Marcus Denker

> On 11 Oct 2017, at 17:18, Thierry Goubier <thierry.goubier(a)gmail.com> wrote: > > What about a seaside-based system browser oriented towards code documentation? That would fit both paradigms: html-like reference + live browsing. > > You could even make it buzzword compliant: have a REST interface to documentation and code. > > Note how the panes in a system browser could make for a nice URI: just look for > > pharo://Kernel/Number/mathematical%20functions/raisedTo: > > And we could even include version numbers, etc.... > > pharo://6.1/Kernel/Number/mathematical%20functions/raisedTo: > > Not very different from the github url for the pharo project relevant file... > > https://github.com/pharo-project/pharo/blob/development/src/Kernel.package/… <https://github.com/pharo-project/pharo/blob/development/src/Kernel.package/…> > There is http://files.pharo.org/doc/ <http://files.pharo.org/doc/>, too. Marcus

Oct. 11, 2017