- Pharo-users - lists.pharo.org

Re: [Pharo-users] Named edges and nodes in a Roassal graph

by Peter UhnÃ¡k

It seems there is an issue with the arrow head, I'll try to investigate it for now try running it without it =========================== v := RTView new. e1 := (RTEllipse new size: 20) elementOn: 'hello'. e2 := (RTEllipse new size: 20) elementOn: 'world'. e1 @ RTDraggable. e2 @ RTDraggable. l := RTEdge from: e1 to: e2. l + (RTLine new color: Color red). "l + (RTSimpleArrowHead new color: Color red)." e2 translateBy: 60 @ 80. v addAll: (Array with: e1 with: e2 with: l ). e1 @ RTLabelled. e2 @ RTLabelled. "Note that the RTLabelled interaction has to be set after having added the element in the view" l @ (RTLabelled new text: 'lining up!'). v open ========================= As far as documentation goes, it is still heavily work in progress, so right now Agile Visualization is probably the best source (and of course mailing list).

Nov. 23, 2014

Re: [Pharo-users] Named edges and nodes in a Roassal graph

by Offray Vladimir Luna CÃ¡rdenas

Hi, El 22/11/14 a las #4, Peter UhnÃ¡k escribiÃ³: > Check out RTRoassalExample>>exampleArrowedAndLabelledLine , it should > demonstrate the functionality for both elements and edges. You can also use > TRConstraint, but that is a tad bit more advanced. > > Peter > Thanks, I have seen it, but the demo is broken. When I move nodes the names and the text are moved, but the edge is not redrawn. People in the workshop were interested in Roasal/Pharo but there where some important lessons: - There is a lack of information on the basic issues (like this one). Agile Visualization is a good start as the Object Profile gallery, but we need more of that. This test case scenarios are useful in detecting the pending issues. - Would be nice to have tools for all Roassal visualizations like export as svg. Some demos have them, some other don't, so other casual issues has a bigger learning curve to adopt what they are seeing and use it in another context. I would like to post the working example on the meetup page, so if any can tell me why the example is broken or how to solve it, would be nice. Cheers, Offray

Nov. 23, 2014

Re: [Pharo-users] Named edges and nodes in a Roassal graph

by Offray Vladimir Luna CÃ¡rdenas

By the way, I have found exampleLabelledGraph but I don't understand where you add individual labels to nodes or edges. Thanks, Offray El 22/11/14 a las #4, Offray Vladimir Luna CÃ¡rdenas escribiÃ³: > Hi all. > > We're now in a small workshop about data narratives and I'm showing some > people Roassal. > > I would like to have named nodes and edges in a graph. I have created a > small understable graph from the object profile gallery here: > http://ws.stfx.eu/8QMELFJ4V370 > > Anyone knows how to add names to edges and nodes in this graph? > > If I could have a quick answer on this one would be awesome. By the way > I'm now 5:50 pm (GMT-5) in the pharo irc channel is someone pops up. > > Thanks, > > Offray > >

Nov. 22, 2014

Re: [Pharo-users] Named edges and nodes in a Roassal graph

by Peter UhnÃ¡k

Check out RTRoassalExample>>exampleArrowedAndLabelledLine , it should demonstrate the functionality for both elements and edges. You can also use TRConstraint, but that is a tad bit more advanced. Peter

Nov. 22, 2014

Named edges and nodes in a Roassal graph

by Offray Vladimir Luna CÃ¡rdenas

Hi all. We're now in a small workshop about data narratives and I'm showing some people Roassal. I would like to have named nodes and edges in a graph. I have created a small understable graph from the object profile gallery here: http://ws.stfx.eu/8QMELFJ4V370 Anyone knows how to add names to edges and nodes in this graph? If I could have a quick answer on this one would be awesome. By the way I'm now 5:50 pm (GMT-5) in the pharo irc channel is someone pops up. Thanks, Offray

Nov. 22, 2014

Re: [Pharo-users] question on syntax "negate numbers"

by Nicolai Hess

2014-11-22 13:23 GMT+01:00 Werner Kassens <wkassens(a)libello.com>: > >Thanks for all the answers, > >so, for better portability, maybe we should dissallow opals current > behavior > >and only accept "-5" and not "- 5" as negative numbers? > > Hi Nicolai, > perhaps. if you mean portability pharo->otherLanguage certainly. if you > mean otherLanguage->pharo, well you are the specialist, i only know pharo > and squeak as smalltalk dialects. of course you know that there exist > non-OO-languages that return the same result as opal if you enter "1 + - 2 > -->-1". and of course i know that if i want to translate something from > those other languages, deleting a <space> is the most simple of my problems. > portability pharo -> other smalltalk languages. Consider you have a library that could be used on other smalltalk systems. And it just don't work because there is a space between the number and the sign. > > changing the pov slightly, when do you have to enter a negative number in > a program by hand? essentially only if you use that number as a constant > (apart from tests of course). numbers are entered often automatically from > outside files. ok, then you have those parsers that read in a string in a > more flexible way. but wouldnt it make sense if the compiler reacts > somewhat similar to those parsers? i for example do have a program, where > the user, admittedly not a usual user but essentially me, enters simple > inequalities (and here negative numbers are very common) as strings and the > compiler eats those strings more or less directly without any additional > parser put in between. > > i realize that with your reply you'd prefer a fact based argumentation and > i readily admit that as a simple user, i see it simple stupid emotionally. > if i understand pharo's history correctly, it came into existence because > some language developers wanted more freedom. of course syntax controls > thinking. i dont have any real problems if you disallow - 5, i use -5 > anyway, if it makes sense do it, you are the specialist. but what comes > next? will everything you enter into nautilus automatically be > pretty-printed? i understand that pharo has grown up now, and it makes > complete sense to me that it wants to play with the big boys. perhaps you > need a clear-cut simple structured syntax to get accepted by the business > community, but not every businessman is a complete idiot and for example > mathematica, which understands "1 + - 2", _is_ occasionally used to make > some real money. i'd think about how far i'd wanna go with this thought > control thing. so much <friendly grin> for my personal pov. > werner > > thank you werner, for your point of view. If someone came up with a good reason why opals behavior is actually wrong, I would change it. But I think now, it is not that important. We have much more things with higher priority. nicolai

Nov. 22, 2014

Re: [Pharo-users] question on syntax "negate numbers"

by Werner Kassens

>Thanks for all the answers, >so, for better portability, maybe we should dissallow opals current behavior >and only accept "-5" and not "- 5" as negative numbers? Hi Nicolai, perhaps. if you mean portability pharo->otherLanguage certainly. if you mean otherLanguage->pharo, well you are the specialist, i only know pharo and squeak as smalltalk dialects. of course you know that there exist non-OO-languages that return the same result as opal if you enter "1 + - 2 -->-1". and of course i know that if i want to translate something from those other languages, deleting a <space> is the most simple of my problems. changing the pov slightly, when do you have to enter a negative number in a program by hand? essentially only if you use that number as a constant (apart from tests of course). numbers are entered often automatically from outside files. ok, then you have those parsers that read in a string in a more flexible way. but wouldnt it make sense if the compiler reacts somewhat similar to those parsers? i for example do have a program, where the user, admittedly not a usual user but essentially me, enters simple inequalities (and here negative numbers are very common) as strings and the compiler eats those strings more or less directly without any additional parser put in between. i realize that with your reply you'd prefer a fact based argumentation and i readily admit that as a simple user, i see it simple stupid emotionally. if i understand pharo's history correctly, it came into existence because some language developers wanted more freedom. of course syntax controls thinking. i dont have any real problems if you disallow - 5, i use -5 anyway, if it makes sense do it, you are the specialist. but what comes next? will everything you enter into nautilus automatically be pretty-printed? i understand that pharo has grown up now, and it makes complete sense to me that it wants to play with the big boys. perhaps you need a clear-cut simple structured syntax to get accepted by the business community, but not every businessman is a complete idiot and for example mathematica, which understands "1 + - 2", _is_ occasionally used to make some real money. i'd think about how far i'd wanna go with this thought control thing. so much <friendly grin> for my personal pov. werner

Nov. 22, 2014

Re: [Pharo-users] Maps from Wikipedia loaded on Roassal's RTMetricMap

by Alexandre Bergel

Yesterday I've tried Cirela, impressive work! I have never thought Roassal could be pushed so far. Alexandre > Le 22-11-2014 Ã 6:04, Thierry Goubier <thierry.goubier(a)gmail.com> a Ã©crit : > > Hi Vladimir, > > 2014-11-18 14:30 GMT+01:00 Offray Vladimir Luna CÃ¡rdenas <offray(a)riseup.net>: >> >> Notice that this tries to be a leaner approach to the one I had already asked about importing maps from open street map. > > Another option which you could look to is: > > - use Cirela-OSM (http://forum.world.st/ANN-NetGen-OSM-Roassal-Demo-tp4791158.html) which will give you the OpenStreetMap background for any place / any zoom level > > - and a combination of FreeBase, GeoJSON in a Roassal view > > (See the geometry properties of freebase at: http://www.freebase.com/location/location/geometry?schema=) > > In Cirela-OSM, you will find examples of queries to FreeBase to retrieve country / cities Lat/Lon coordinates and use them in Roassal. You also have an example of querying raw OpenStreetMap data as well. > > Good luck for your workshop. > > Thierry > >> >> Thanks, >> >> Offray >

Nov. 22, 2014

Re: [Pharo-users] Maps from Wikipedia loaded on Roassal's RTMetricMap

by Thierry Goubier

Hi Vladimir, 2014-11-18 14:30 GMT+01:00 Offray Vladimir Luna CÃ¡rdenas <offray(a)riseup.net> : > > Notice that this tries to be a leaner approach to the one I had already > asked about importing maps from open street map. > Another option which you could look to is: - use Cirela-OSM ( http://forum.world.st/ANN-NetGen-OSM-Roassal-Demo-tp4791158.html) which will give you the OpenStreetMap background for any place / any zoom level - and a combination of FreeBase, GeoJSON in a Roassal view (See the geometry properties of freebase at: http://www.freebase.com/location/location/geometry?schema=) In Cirela-OSM, you will find examples of queries to FreeBase to retrieve country / cities Lat/Lon coordinates and use them in Roassal. You also have an example of querying raw OpenStreetMap data as well. Good luck for your workshop. Thierry > > Thanks, > > Offray > >

Nov. 22, 2014

Re: [Pharo-users] Maps from Wikipedia loaded on Roassal's RTMetricMap

by HernÃ¡n Morales Durand

Ok, download the wikimedia SVG and open it with a text editor. If you look for "Ulster, NY" you will see it enclosed in a path node. The contents of the "d" attribute is what you need to parse/extract to pass as parameter to RTSVGPath>>path: <path style="font-size:12px;fill:#d0d0d0;fill-opacity:1;fill-rule:nonzero;stroke:#000000;stroke-opacity:1;stroke-width:0.1;stroke-miterlimit:4;stroke-dasharray:none;stroke-linecap:butt;marker-start:none;stroke-linejoin:bevel" d="M 501.96598,97.8 L 502.15198,100.009 L 502.68298,101.739 L 502.88598,102.153 L 503.20498,104.028 L 502.29498,104.366 L 501.61098,104.502 L 501.02498,104.483 L 500.32998,104.231 L 499.33398,104.921 L 498.91498,104.299 L 497.86898,103.844 L 496.94598,103.415 L 497.13498,102.933 L 497.50398,102.086 L 497.69798,101.505 L 495.56698,100.824 L 494.24598,100.423 L 496.85998,97.791 L 498.36098,98.174 L 499.08798,98.084 L 500.56998,97.894 L 500.65098,96.89 L 501.06998,96.754 L 501.99798,97.146 L 501.96598,97.8" id="36111" inkscape:label="Ulster, NY" /> I have tried using the Roassal2 from the Configuration Browser in Pharo 3.0. Cheers, HernÃ¡n 2014-11-21 22:53 GMT-03:00 Offray Vladimir Luna CÃ¡rdenas <offray(a)riseup.net> : > Hi Hernan, > > Thanks for your answer. I got an error which says "Error: Reading a number > failed: a digit between 0 and 9 expected" and "element" of your example is > selected. I think that I'm not understanding well how to connect the > downloaded file (in fact I have not downloaded anything) to the part where > you render the svg and I don't see anything in your example which links > them. Whats the thing I'm missing? > > Thanks for your help. Hopefully, tomorrow we can show maps for new > territories in Roassal. > > Cheers, > > Offray > > El 21/11/14 a las #4, HernÃ¡n Morales Durand escribiÃ³: > >> >> >> 2014-11-18 10:30 GMT-03:00 Offray Vladimir Luna CÃ¡rdenas < >> offray(a)riseup.net >> <mailto:offray@riseup.net>>: >> >> Hi, >> >> This Thursday and Saturday we will have a small workshop on data >> storytelling and I would like to present Pharo/Roassal to the people >> there >> as a tool for creating and sharing visualizations. >> >> My idea is to use Playgrounds to make agile visualization and then >> share >> them using cloud sharing at stfx and etherpads. >> The maps I would like to make are similar to the ones that have >> colored >> countries [1][2], but using internal country geopolitical divisions. >> For the >> case of Colombia I would like to use the ones of [3], specifically >> something >> like [4] and use the same syntax of RTMetricMap inside the country. >> >> [1] http://bit.ly/1yiedrE >> [2] http://bit.ly/1xTAQV0 >> [3] https://commons.wikimedia.org/__wiki/User:Shadowxfox/Maps#__ >> Colombia >> <https://commons.wikimedia.org/wiki/User:Shadowxfox/Maps#Colombia> >> [4] >> https://commons.wikimedia.org/__wiki/File:Colombia___ >> departamentos_otros.svg >> >> <https://commons.wikimedia.org/wiki/File:Colombia_ >> departamentos_otros.svg> >> >> So my question is: >> >> - There is any way to upload this SGV maps and treat them as the maps >> already in RTMetricMap, telling the names of the internal >> geopolitical >> divisions inside a Country and how to color them according to a >> value and >> make them zoomable? >> >> - Where can I start to look to support this feature? >> >> >> SVG is basically XML, so you can parse it like this: >> >> | xmlTree | >> xmlTree := (XMLDOMParser parseFileNamed: 'Argentina.svg') firstNode. >> xmlTree >> >> The coordinates of a map are in the SVG path description, and you can add >> them >> separately. For example if you download this file >> http://commons.wikimedia.org/wiki/File:USA_Counties_with_ >> FIPS_and_names.svg and >> take the node with "Ulster, NY" you can render it this way: >> >> | view svg | >> view := RTView new. >> svg := (RTSVGPath new >> path: 'M 501.96598,97.8 L 502.15198,100.009 L 502.68298,101.739 L >> 502.88598,102.153 L 503.20498,104.028 L 502.29498,104.366 L >> 501.61098,104.502 L >> 501.02498,104.483 L 500.32998,104.231 L 499.33398,104.921 L >> 498.91498,104.299 L >> 497.86898,103.844 L 496.94598,103.415 L 497.13498,102.933 L >> 497.50398,102.086 L >> 497.69798,101.505 L 495.56698,100.824 L 494.24598,100.423 L >> 496.85998,97.791 L >> 498.36098,98.174 L 499.08798,98.084 L 500.56998,97.894 L 500.65098,96.89 L >> 501.06998,96.754 L 501.99798,97.146 L 501.96598,97.8'; >> fillColor: (Color r: 0.612 g: 0.488 b: 0.3 alpha: 1.0); >> scale: 5) element.http://commons.wikimedia.org/wiki/File:USA_ >> Counties_with_FIPS_and_names.svg >> view add: svg. >> view open >> >> There should be an easier way like using an SVG importer for Roassal, but >> that's >> all I have for now. >> >> Cheers, >> >> HernÃ¡n >> >> >> > >

Nov. 22, 2014