- Pharo-users - lists.pharo.org

Re: [Pharo-users] Loading code from a local git repository

by FranÃ§ois Stephany

As Sven mentionned it, you can simply put your filetree files in a subdirectory. You can have a look at https://github.com/fstephany/hello-pharo It's a bit outdated but it will give you the idea. On Tue, Jul 28, 2015 at 12:42 PM, Tommaso Dal Sasso < tommaso.dalsasso(a)gmail.com> wrote: > Hello, > > I am trying to version a project on git using filetree. > Since I also have some external resources (like images), I would avoid > loading the whole project from the pharo image, because that would > download the external files in a cache directory, thus making them > difficult to reach and update. > > My idea is to update the git repository from the commandline, and then > run a script to build the image. > > The question then is: is there an easy way to programmatically load a > project from a local directory? > > Also, do you have any suggestions for a better workflow to handle this > situation? > > Thanks, > Tommaso > >

July 28, 2015

Re: [Pharo-users] Loading code from a local git repository

by Sven Van Caekenberghe

Just place the repository in a separate directory from the other stuff, like https://github.com/svenvc/zinc No ? > On 28 Jul 2015, at 12:42, Tommaso Dal Sasso <tommaso.dalsasso(a)gmail.com> wrote: > > Hello, > > I am trying to version a project on git using filetree. > Since I also have some external resources (like images), I would avoid > loading the whole project from the pharo image, because that would > download the external files in a cache directory, thus making them > difficult to reach and update. > > My idea is to update the git repository from the commandline, and then > run a script to build the image. > > The question then is: is there an easy way to programmatically load a > project from a local directory? > > Also, do you have any suggestions for a better workflow to handle this > situation? > > Thanks, > Tommaso >

July 28, 2015

Loading code from a local git repository

by Tommaso Dal Sasso

Hello, I am trying to version a project on git using filetree. Since I also have some external resources (like images), I would avoid loading the whole project from the pharo image, because that would download the external files in a cache directory, thus making them difficult to reach and update. My idea is to update the git repository from the commandline, and then run a script to build the image. The question then is: is there an easy way to programmatically load a project from a local directory? Also, do you have any suggestions for a better workflow to handle this situation? Thanks, Tommaso

July 28, 2015

Re: [Pharo-users] Announcments vs Events schism

by Matthieu Lacaton

It seems to me that sometimes the difference is a bit blurry. I used to consider announcements like some kinds of user-generated events, but I may be wrong here :/ I'll use SDL as an example because this is the technology chosen for Pharo. In SDL the concept of announcement does not exist. However, you can use SDL_PushEvent to push any event into the event queue. So if you manually push some mouse events, you'll have the events coming directly from the hardware and the ones you have pushed but both will be handled the same way. In this case, should user-generated events be considered as announcements and hardware-generated ones be considered as events ? Or are all these events just events ? In SDL, there is also another kind of event : SDL_UserEvent which is a custom event in which the user can put any data and then push it into the queue. I feel like this one is really close to be considered as an announcement but it is still handled like any other event. Basically in these cases, which ones are real events ? and which ones are announcements-like events ? An event is managed by the event loop and other. The idea is that other > objects do not register to an event. > it happens to me. > > An announcement is raised and unknown object can register and get > access to it. > Other register to react to it. > If an announcement is raised, you need to register to it to be able to react. Okay. But is it not the same thing for events ? If an event is raised you still need to somehow register to it to be able to react. In SDL, if you don't choose to handle the event in your switch then you won't be able to react. In OSWindow, if you don't choose to implement the "visitWhateverEvent:" method, then you won't be able to react. That's why for me, an announcement is like an event, except an event is raised by the hardware and you don't really have a choice about what it is made of, whereas an announcement is raised by the user and he can do whatever he wants with its definition. 2015-07-28 9:58 GMT+02:00 stepharo <stepharo(a)free.fr>: > I think that in Bloc alain did it on purpose and not just by accident. > Alain may correct me: > > An event is managed by the event loop and other. The idea is that > other objects do not register to an event. > it happens to me. > > An announcement is raised and unknown object can register and get > access to it. > Other register to react to it. > > So Event is not equal to Announcement > > Stef > > > > Le 27/7/15 23:57, Peter UhnÃ¡k a Ã©crit : > > Hi, >> >> is there a good reason why not just rename Announcement to Event since >> everyone is calling it that anyways? >> Everybody (almost irrespective of language) understands what an event, >> whilst Announcement is a new vocabulary. >> >> Most people when subclassing will anyways do >> >> Announcement subclass: MyEvent >> >> and so forth. >> >> Is there a benefit in naming it Announcement? Doesn't it just add more >> confusion? >> >> For example Bloc has >> Announcement subclass: #BlAbstractEvent >> and >> Announcement subclass: #BlMorphAnnouncement >> >> whilst Event is clearly favored. >> BlMorphAnnouncement allSubclasses size. "5" >> BlAbstractEvent allSubclasses size. "60" >> >> (Why does Bloc has this separation anyways?) >> >> Peter >> > > >

July 28, 2015

Re: [Pharo-users] SmaCC Warning on Pharo 5

by Dimitris Chloupis

Ah ok, thanks, no I dont use the GUI so far. On Mon, Jul 27, 2015 at 11:39 PM Thierry Goubier <thierry.goubier(a)gmail.com> wrote: > Hi Kilon, > > I haven't done the work to port the SmaCC Gui to Pharo 5. As long as you > don't use the GUI, you should be fine. > > Thierry > Le 27 juil. 2015 9:48 PM, "Dimitris Chloupis" <kilon.alios(a)gmail.com> a > Ã©crit : > >> I am installing my project Ephestos on Pharo 5 that depends on SmaCC and >> SmaCC gives me the following warning >> >> This package depends on the following classes: StringHolder You must >> resolve these dependencies before you will be able to load these >> definitions: SmaCCCodeHolder SmaCCCodeHolder>>#owner: >> SmaCCCodeHolder>>#owner:ownerAcceptSelector: >> SmaCCCodeHolder>>#contents:notifying: >> SmaCCCodeHolder>>#contentsMenu:shifted: SmaCCCodeHolder>>#contentsMorph >> SmaCCCodeHolder>>#hasUnacceptedEdits SmaCCCodeHolder>>#okToChange >> SmaCCCodeHolder>>#owner SmaCCCodeHolder>>#owner: >> SmaCCCodeHolder>>#owner:ownerAcceptSelector: >> SmaCCCodeHolder>>#ownerAcceptSelector >> SmaCCCodeHolder>>#ownerAcceptSelector: SmaCCCodeHolder>>#perform:orSendTo: >> SmaCCCodeHolder>>#readContentsFromFile SmaCCCodeHolder>>#saveContentsInFile >> Select Proceed to continue, or close this window to cancel the operation. >> >> Is this normal ? Should I be worried ? Is SmaCC ok on Pharo 5 ? >> >> Clicking proceed does not appear to create any side effect. >> >

July 28, 2015

Re: [Pharo-users] Announcments vs Events schism

by stepharo

I think that in Bloc alain did it on purpose and not just by accident. Alain may correct me: An event is managed by the event loop and other. The idea is that other objects do not register to an event. it happens to me. An announcement is raised and unknown object can register and get access to it. Other register to react to it. So Event is not equal to Announcement Stef Le 27/7/15 23:57, Peter UhnÃ¡k a Ã©crit : > Hi, > > is there a good reason why not just rename Announcement to Event since > everyone is calling it that anyways? > Everybody (almost irrespective of language) understands what an event, > whilst Announcement is a new vocabulary. > > Most people when subclassing will anyways do > > Announcement subclass: MyEvent > > and so forth. > > Is there a benefit in naming it Announcement? Doesn't it just add more > confusion? > > For example Bloc has > Announcement subclass: #BlAbstractEvent > and > Announcement subclass: #BlMorphAnnouncement > > whilst Event is clearly favored. > BlMorphAnnouncement allSubclasses size. "5" > BlAbstractEvent allSubclasses size. "60" > > (Why does Bloc has this separation anyways?) > > Peter

July 28, 2015

Re: [Pharo-users] Personal wiki / information manager

by Offray Vladimir Luna CÃ¡rdenas

Thanks Evan for the detailed info. This is a busy week for me, but next weekend I will be giving a more detailed look at your software and answering your mail. Cheers, Offray Ps: for some reason your post don't become part of the thread in my mailing client. They start like new post, but keeping the title. :-/ On 27/07/15 18:44, Evan Donahue wrote: > Ah, yes, by 'offline' I did mean outside of the mailing list, but that > was purely out of consideration for the list. If the list wants to > hear it, far be it from me to keep it from them. > > For my part, I am a graduate student working on my dissertation, and I > have been trying to build tools to facilitate that as I go, from > taking and sharing notes, to visualizing primary material, to > (eventually) organizing and writing the dissertation. I have been > following Offray's updates with interest, as I think my goals are very > much allied, but I am starting in a slightly different place, if I > have understood correctly (research and organizing, and not yet > writing, sharing, or publishing). My efforts thus far have been > concentrated in two repositories: > > Chancery (http://smalltalkhub.com/#!/~EvanDonahue/Chancery > <http://smalltalkhub.com/#%21/%7EEvanDonahue/Chancery>) is an early > prototype for a note taking system. I am coming from taking notes in > emacs org-mode, and so Chancery right now is the most straightforward > collapsable tree outliner/note taker that could give me functionality > basically equivalent to what I had in org-mode, and serve as a > platform to co-evolve my research methods and their supporting > technology. The near-term goals for this involve adding metadata > (authors, dates, etc) to the sources I am annotating and using that > for various kinds of historical visualization (and further, attendant > annotation) eg who wrote what when, referencing who, influenced by > what, etc), and can I edit all that directly in its graphical form. > The status of this project is that it is usable by myself and those I > have worked with on it (and I use it as my primary note taking > system), but I'm not sure what it would look like to someone loading > it at the moment (no real intro/docs, as it was just a quick prototype > to provoke further design discussion). > > Trantor (http://smalltalkhub.com/#!/~EvanDonahue/Trantor > <http://smalltalkhub.com/#%21/%7EEvanDonahue/Trantor>) is a > disttributed p2p system for incrementally keeping data structures in > sync between multiple, asynchronously connected pharo instances across > the internet without requiring human attention to resolve merge > conflicts (think laptos where data is edited offline and then > automatically synced when they connect to one another); it is my > answer to sharing and collaboration for things created in Chancery > (although it is a more general purpose project). Functionally, Trantor > provides a set of common datatypes (set, dict, text blob, etc) and a > means to run nodes and peer them for synchronization of said > datatypes. The goal with respect to Chancery/personal information > management is being able to easily grab some structured set of notes > or writing out of the note taking system and share it with > collaborators or colleagues with a button click, and without having to > go through the whole export/send/manage versions dance. This project > is about 85% complete, the proof of concept works, it just needs a big > refactor to make it usable and quash some bugs. > > I guess my observations on using Pharo thus far are first, that it is > absolutely ideal for evolving prototypes, since you get a sort of poor > man's gui and persistance (object inspector + image serialization) > basically for free, which is 90% of the work of such a simple > information management app prototype. I have a fully functional tree > outliner at a point when in another language I would still be trying > to hack a low functionality command line interface. The other thing > I've found refreshing is that my application can live as one among > many within the image, sharing data with other programs by passing > objects around in a way that is impossible to do with a standalone > application running on a modern operating system. the dream of the web > is, I think, heading more in this direction, but working with the > browser (as I was before I came to pharo) requires coordinating a > separate server process, compressing everything into http/html/json, > and a host of other problems that vanish when you can just serialize > objects between pharo instances. In short, I'm getting a lot done in > this language and I expect I'll be around for a while. > > Cheers, > Evan

July 28, 2015

Re: [Pharo-users] Personal wiki / information manager

by Evan Donahue

Ah, yes, by 'offline' I did mean outside of the mailing list, but that was purely out of consideration for the list. If the list wants to hear it, far be it from me to keep it from them. For my part, I am a graduate student working on my dissertation, and I have been trying to build tools to facilitate that as I go, from taking and sharing notes, to visualizing primary material, to (eventually) organizing and writing the dissertation. I have been following Offray's updates with interest, as I think my goals are very much allied, but I am starting in a slightly different place, if I have understood correctly (research and organizing, and not yet writing, sharing, or publishing). My efforts thus far have been concentrated in two repositories: Chancery (http://smalltalkhub.com/#!/~EvanDonahue/Chancery) is an early prototype for a note taking system. I am coming from taking notes in emacs org-mode, and so Chancery right now is the most straightforward collapsable tree outliner/note taker that could give me functionality basically equivalent to what I had in org-mode, and serve as a platform to co-evolve my research methods and their supporting technology. The near-term goals for this involve adding metadata (authors, dates, etc) to the sources I am annotating and using that for various kinds of historical visualization (and further, attendant annotation) eg who wrote what when, referencing who, influenced by what, etc), and can I edit all that directly in its graphical form. The status of this project is that it is usable by myself and those I have worked with on it (and I use it as my primary note taking system), but I'm not sure what it would look like to someone loading it at the moment (no real intro/docs, as it was just a quick prototype to provoke further design discussion). Trantor (http://smalltalkhub.com/#!/~EvanDonahue/Trantor) is a disttributed p2p system for incrementally keeping data structures in sync between multiple, asynchronously connected pharo instances across the internet without requiring human attention to resolve merge conflicts (think laptos where data is edited offline and then automatically synced when they connect to one another); it is my answer to sharing and collaboration for things created in Chancery (although it is a more general purpose project). Functionally, Trantor provides a set of common datatypes (set, dict, text blob, etc) and a means to run nodes and peer them for synchronization of said datatypes. The goal with respect to Chancery/personal information management is being able to easily grab some structured set of notes or writing out of the note taking system and share it with collaborators or colleagues with a button click, and without having to go through the whole export/send/manage versions dance. This project is about 85% complete, the proof of concept works, it just needs a big refactor to make it usable and quash some bugs. I guess my observations on using Pharo thus far are first, that it is absolutely ideal for evolving prototypes, since you get a sort of poor man's gui and persistance (object inspector + image serialization) basically for free, which is 90% of the work of such a simple information management app prototype. I have a fully functional tree outliner at a point when in another language I would still be trying to hack a low functionality command line interface. The other thing I've found refreshing is that my application can live as one among many within the image, sharing data with other programs by passing objects around in a way that is impossible to do with a standalone application running on a modern operating system. the dream of the web is, I think, heading more in this direction, but working with the browser (as I was before I came to pharo) requires coordinating a separate server process, compressing everything into http/html/json, and a host of other problems that vanish when you can just serialize objects between pharo instances. In short, I'm getting a lot done in this language and I expect I'll be around for a while. Cheers, Evan

July 28, 2015

Re: [Pharo-users] Using geographic data

by Arturo Zambrano

Hi HernÃ¡n On Sat, Jul 25, 2015 at 4:54 AM, HernÃ¡n Morales Durand < hernan.morales(a)gmail.com> wrote: > Hi Arturo, > > Do you want to implement a Geolocator like NAC? > http://www.nactag.info/map.asp > Nope, just to count points for each different province. > > > The GADM world tree is built from a CSV which I generated from the GADM > 2.0 dBase files for the whole world. I checked the dBase format and doesn't > include the polygon information you find in the .kmz files. I saw they > released 2.5 this month (July 2015) but only in ESRI geodatabase format > (which I didn't investigated yet, but not the same as ESRI shapefile > format) and will release 2.7 in August 2015. > > Another way could be just to parse the KML (is just an XML file) using > XPath or XMLPullParser. > > Of course,thanks. I wanted to know if there were a library supporting KMZ, and doing some queries on the objects contained there. >From your reply and Sven's , I see there is no such library, but it seems very easy to implement that functionality. My only concern now is performance, I will try it and let you know. > About locating the point in a polygon the problem of traditional 2D > inclusion tests (ray crossing methods) is that require large storage, given > polygons of enough size, complexity and amount of edges. This is a common > situation with level 1 and 2 polygons. New methods for example classiffy > polygon edges into layers, results in faster inclusion test. > > Thanks again. best regards art > HernÃ¡n > > > > > 2015-07-24 20:19 GMT-03:00 Arturo Zambrano <arturo.zambrano(a)gmail.com>: > >> Hi All, >> I would like to get some advice regarding which path I should take. >> >> Problem: >> Given a coordinate I want to know to which geographical object it >> belongs to. >> >> Path 1 >> Using Glorp to make a query to a Postgis instance containing the >> boundaries of the geobjects >> >> >> >> Path 2 >> Use the data form GADM data (according to this thread >> <http://forum.world.st/PostGIS-in-Pharo-td4802233.html>). >> >> >> I would avoid path 1 in order to avoid installation of postgi. From the >> thread mentioned above, I thing that it requires some development. >> >> >> Regarding Path 2, I don't see which framework or library should I use to >> import and query the data contained in the KMZ from GADM. >> >> >> Any advice is welcome. >> >> Best regards. >> Art >> >> >> > >

July 28, 2015

Re: [Pharo-users] Using geographic data

by Arturo Zambrano

Thanks Sven, I will test the performance... some my objects have around 10000 points. Maybe I will reduce the amount of point and sacrifice some precision. I thought something like that was already available in SH. Best regards On Sat, Jul 25, 2015 at 4:31 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Arturo, > > In Pharo itself, you could just do it yourself, like this: > > T3GeoTools class>>#is: position inside: polygon > "T3GeoTools is: 5.33732(a)50.92601 inside: { 5.48230(a)50.82249. > 5.49523(a)50.81288. 5.50138(a)50.82008. 5.50228(a)50.82595. 5.49265(a)50.82560. > 5.48230(a)50.82249 }" > > "T3GeoTools is: 5.49007(a)50.82205 inside: { 5.48230(a)50.82249. > 5.49523(a)50.81288. 5.50138(a)50.82008. 5.50228(a)50.82595. 5.49265(a)50.82560. > 5.48230(a)50.82249 }" > > "Point in polygon - ray casting algorithm - > http://www.ecse.rpi.edu/Homepages/wrf/Research/Short_Notes/pnpoly.html" > > | inside otherIndex | > inside := false. > otherIndex := polygon size. > polygon doWithIndex: [ :each :index | | other | > other := polygon at: otherIndex. > (((each y > position y) ~= (other y > position y)) > and: [ position x < (((other x - each x) * > (position y - each y) / (other y - each y)) + each x) ]) > ifTrue: [ inside := inside not ]. > otherIndex := index ]. > ^ inside > > Inside circle, rectangle etc are trivial. > > Sven > > > On 25 Jul 2015, at 01:19, Arturo Zambrano <arturo.zambrano(a)gmail.com> > wrote: > > > > Hi All, > > I would like to get some advice regarding which path I should take. > > > > Problem: > > Given a coordinate I want to know to which geographical object it > belongs to. > > > > Path 1 > > Using Glorp to make a query to a Postgis instance containing the > boundaries of the geobjects > > > > > > > > Path 2 > > Use the data form GADM data (according to this thread). > > > > > > I would avoid path 1 in order to avoid installation of postgi. From the > > thread mentioned above, I thing that it requires some development. > > > > > > Regarding Path 2, I don't see which framework or library should I use to > > import and query the data contained in the KMZ from GADM. > > > > > > Any advice is welcome. > > > > Best regards. > > Art > > > > > > >

July 28, 2015