Skip Menu |

This queue is for tickets about the Map-Tube-Prague CPAN distribution.

Report information
The Basics
Id: 103078
Status: open
Priority: 0/
Queue: Map-Tube-Prague

People
Owner: SKIM [...] cpan.org
Requestors: GWS [...] cpan.org
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: 0.10
Fixed in: (no value)



Subject: Artefacts for stations where lines cross (Mustek, Muzeum, Florenc)
After the recent introduction of the other_line feature into Map::Tube, having the above-named stations twice in the maps produces artefacts, e.g., in M::T::GraphViz (see attached example), in that these stations occur twice (with identical names, but different colouring) in the maps. For the naive viewer of the map (i.e., me) this is slightly confusing. This may not be the use that was intended for the other_line feature, as it is currently implemented. E.g., M::T::London does not have double entries for each station where lines cross; rather, these are reserved for cases where two stations (on different lines) are so close to each other that changes by foot are easily doable.
Subject: Prague.png
Download Prague.png
image/png 253k
Prague.png
Dne Út 24.bře.2015 17:51:46, GWS napsal(a): Hi, Show quoted text
> After the recent introduction of the other_line feature into > Map::Tube, having the above-named stations twice in the maps produces
[SNIP] Show quoted text
> reserved for cases where two stations (on different lines) are so > close to each other that changes by foot are easily doable.
I know. I am refactoring code for it. M.
Subject: AW: [rt.cpan.org #103078] Artefacts for stations where lines cross (Mustek, Muzeum, Florenc)
Date: Wed, 25 Mar 2015 11:37:02 +0000
To: "bug-Map-Tube-Prague [...] rt.cpan.org" <bug-Map-Tube-Prague [...] rt.cpan.org>
From: "Selke, Gisbert W." <gisbert.selke [...] wido.bv.aok.de>
Ahoj Míšo -- Wouldn't it be easiest just to revert to the old contents of the Prague XML file? That seemed to be just perfect to me. \Gisbík Show quoted text
> -----Ursprüngliche Nachricht----- > Von: Michal Špaček via RT [mailto:bug-Map-Tube-Prague@rt.cpan.org] > Gesendet: Mittwoch, 25. März 2015 07:30 > An: GWS@cpan.org > Betreff: [rt.cpan.org #103078] Artefacts for stations where lines cross > (Mustek, Muzeum, Florenc) > > <URL: https://rt.cpan.org/Ticket/Display.html?id=103078 > > > Dne Út 24.bře.2015 17:51:46, GWS napsal(a): > Hi, >
> > After the recent introduction of the other_line feature into > > Map::Tube, having the above-named stations twice in the maps produces
> [SNIP]
> > reserved for cases where two stations (on different lines) are so > > close to each other that changes by foot are easily doable.
> > I know. I am refactoring code for it. > > M. >
Dne St 25.bře.2015 07:37:19, gisbert.selke@wido.bv.aok.de napsal(a): Ahoj Gisbíku, Show quoted text
> Wouldn't it be easiest just to revert to the old contents of the > Prague XML file? That seemed to be just perfect to me.
The old version of Prague doesn't use resolution of link. I added it and this cause problems with Map::Tube::GraphViz. I am thinking, that is right way to resolve this issue. btw: This new version is better maintainable than old. Prague is simple, but Singapore or KualaLumpur has problems. Měj se hezky, M.
Am 2015-03-25 07:45:33, SKIM schrieb: Show quoted text
> The old version of Prague doesn't use resolution of link. I added it > and > this cause problems with Map::Tube::GraphViz. > I am thinking, that is right way to resolve this issue.
The problem is that other modules/methods are also affected by this, e.g., M::T::get_shortest_route(). Show quoted text
> > btw: This new version is better maintainable than old. Prague is > simple, but Singapore or KualaLumpur has problems.
Point taken! Show quoted text
> Měj se hezky,
Ty taky! \Gisbert
Dne St 25.bře.2015 08:13:58, GWS napsal(a): Show quoted text
> Am 2015-03-25 07:45:33, SKIM schrieb:
> > The old version of Prague doesn't use resolution of link. I added it > > and > > this cause problems with Map::Tube::GraphViz. > > I am thinking, that is right way to resolve this issue.
Show quoted text
> The problem is that other modules/methods are also affected by this, > e.g., M::T::get_shortest_route().
oh. There are two issues: * You say, that get_shortest_route() has problems. * Map::Tube::GraphViz has problems with multiple station with same id. This cause images as Map::Tube::Prague has on man page. M.
Am 2015-03-25 08:19:53, SKIM schrieb: Show quoted text
> oh. > There are two issues: > * You say, that get_shortest_route() has problems. > * Map::Tube::GraphViz has problems with multiple station with same id.
Well, yes, the id should be unique. Any station may occur more than once, of course, in different lines, but one station is one station and probably should not be linked to itself. Here's the output from Florenc to Flora as generated by get_shortest_route()->preferred(): Florenc (Linka C), Hlavni nadrazi (Linka C), Muzeum (Linka C), Muzeum (Tunnel), Namesti Miru (Linka A), Jiriho z Podebrad (Linka A), Flora (Linka A) And here's the output from M::T::Text::Shortest: From Florenc to Flora ===================== -- Route 1 (cost ?) --------- [ C ] Florenc [ * ? ] Florenc [ C ] Hlavni nadrazi [ C ] Muzeum [ * ? ] Muzeum [ A ] Muzeum [ * ? ] Muzeum [ A ] Namesti Miru [ A ] Jiriho z Podebrad [ A ] Flora C Linka C A Linka A B Linka B *: Transfer to other line +: Transfer to other station
Dne St 25.bře.2015 09:26:49, GWS napsal(a): Show quoted text
> Am 2015-03-25 08:19:53, SKIM schrieb: >
> > oh. > > There are two issues: > > * You say, that get_shortest_route() has problems. > > * Map::Tube::GraphViz has problems with multiple station with same > > id.
Sorry. That was bad sentence. ... with same name. e.g. KualaLumpur or Singapur has situations, when two stations has same name and on each line has different unique ids. Show quoted text
> Well, yes, the id should be unique. Any station may occur more than > once, of course, in different lines, but one station is one station > and probably should not be linked to itself.
Yes. Show quoted text
> Here's the output from Florenc to Flora as generated by > get_shortest_route()->preferred():
I don't test new preferred() method. Show quoted text
> Florenc (Linka C), Hlavni nadrazi (Linka C), Muzeum (Linka C), Muzeum > (Tunnel), Namesti Miru (Linka A), Jiriho z Podebrad (Linka A), Flora > (Linka A)
I am expecting something like: Florenc (Linka C), Hlavni nadrazi (Linka C), Muzeum (Linka C), Muzeum (Linka A), Namesti Miru (Linka A), Jiriho z Podebrad (Linka A), Flora (Linka A) And will be good see some method to get link type between stations (in Map::Tube::Route). Links are: (Florenc C) Tube (Hlavni nadrazi C), (Hlavni nadrazi C) Tube (Muzeum C), (Muzeum C) Tunnel (Muzeum A), (Muzeum A) Tube (Namesti Miru A), (Namesti Miru A) Tube (Jiriho z Podebrad A), (Jiriho z Podebrad A) Tube (Flora (A) Show quoted text
> And here's the output from M::T::Text::Shortest: > > From Florenc to Flora > ===================== > > -- Route 1 (cost ?) --------- > [ C ] Florenc > [ * ? ] Florenc > [ C ] Hlavni nadrazi > [ C ] Muzeum > [ * ? ] Muzeum > [ A ] Muzeum > [ * ? ] Muzeum > [ A ] Namesti Miru > [ A ] Jiriho z Podebrad > [ A ] Flora > > C Linka C > A Linka A > B Linka B > > *: Transfer to other line > +: Transfer to other station
Yes, this is bad. M.
Subject: AW: [rt.cpan.org #103078] Artefacts for stations where lines cross (Mustek, Muzeum, Florenc)
Date: Wed, 25 Mar 2015 23:30:40 +0000
To: "bug-Map-Tube-Prague [...] rt.cpan.org" <bug-Map-Tube-Prague [...] rt.cpan.org>
From: "Selke, Gisbert W." <gisbert.selke [...] wido.bv.aok.de>
Show quoted text
> -----Ursprüngliche Nachricht----- > Von: Michal Špaček via RT [mailto:bug-Map-Tube-Prague@rt.cpan.org] > Gesendet: Mittwoch, 25. März 2015 15:02 > I don't test new preferred() method.
Ya, ok -- get_shortest_route() gives similar results, only slightly longer and less easy to read; I was lazy in including only this abbreviated output. Show quoted text
> And will be good see some method to get link type between stations (in Map::Tube::Route).
Agree! I'd even prefer something along this line: Florenc <Tube C> Hlavni nadrazi <Tube C> Muzeum <Tube A> Namesti Miru <Tube A> Jiriho z Podebrad <Tube A> Flora Because this makes it clear that there are nodes (stations) and links (tube lines). Even apart from graph theory, this is the kind of information I need as a user of public transport. Of course, other, more compact styles could also be envisaged, like Florenc <Tube C, 2 stations> Muzeum <Tube A, 3 stations> Flora But these are all just variations on output styles. The underlying data model (i.e., the graph) would be the same. \Gisbert