Re: [josm-dev] Localization of exception messages
My suggestion is to throw just english language exceptions. All exceptions should have some code and this code can be key in some wiki page with tutorials how to remove occurence of exceptions. Ideal is link in exception window. Googling for number could be sometimes problem. For developer could not be better exception than filename, line number and exception class. This is imho da best info which is developer able to get from random user. Trying to ask in user forum, if translated exception is better than english one is probably usefull idea. PS: who is developer and dont know english, rise your hand :D (joke) LuVar On Friday 08 January 2010 21:06:06 Matthias Julius wrote: I noticed that some exceptions get localized messages like this one: IllegalArgumentException(tr(Parameter ''{0}'' must not be null., tags)) Does this really help anyone? Those messages are intended to be for developers and users probably won't know what to do even with the best translation. On top of that most of the developers won't understand most of the localized messages (I understand only a very small subset of them) which is making it harder to understand bug reports. I would vote for removing those lokalizations. I don't know how many there are but this will also reduce the load on translators and the size of the language files a bit. Other opinions? Matthias ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] Java experts
Hi, I dont have info about that concrete situation, but as I know java gui api, there should be some function called revalidate() which should do revalidation also on parent components. So try to override this on core panel, than from overriden method call original revalidate (super.revalidate()) and than your updateMeta(). There are also AspectJ (ascpect programming paradigm) which could solve this, but imho this is not right solution. I hope revalidate, or paint or some update function overriding should help. GL and pls post also final solution to this mail thread. On Sunday 22 November 2009 13:35:00 Ævar Arnfjörð Bjarmason wrote: On Sun, Nov 22, 2009 at 10:09, Dirk Stöcker openstreet...@dstoecker.de wrote: Hello, a short question for the all Java experts here: In the file src/org/openstreetmap/josm/gui/preferences/ProjectionPreference.java the updateMeta() function should be called whenever one of the elements of the projSubPrefPanel is modified. Does anyone know how to do this? (I don' know anything about Java) Perhaps adding a component listener to projSubPrefPanel will do what you want? There are only listeners for hidden/moved/resized/shown. componentResized is Invoked when the component's size changes so that might do what you want: http://java.sun.com/docs/books/tutorial/uiswing/events/componentlistener.ht ml http://java.sun.com/j2se/1.4.2/docs/api/java/awt/event/ComponentListener.ht ml ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] Plugin Popularity
Vhat about to make it legal? As in eclipse or netbeans, show to user one page, where will be writen, that josm coul monitor his ussage off josm sw and report it to its autors to improve future versions. Only uploaded data will be a list of plugins. This upload will be done on any change in used plugins... What do you think about this? In future there could be an optional monitoring for monitoring ussages of all buttons in josm. PS: also some linux distributions (minimally kubuntu) ask for permission about to monitor popularity of installed packages. Default is this option off. On Tuesday 20 October 2009 19:14:34 Frederik Ramm wrote: Hi, the number of plugins becomes ever larger, which is good. I would like to amend the JOSM User-Agent header to contain not only version and language, but also a list of installed plugins - e.g. instead of today's JOSM/1.5 (2069 SVN en_GB) Java/1.6.0_0 I would transmit JOSM/1.5 (2069 SVN en_GB; wmsplugin, validator, slippymap) Java/1.6.0_0 This would enable us to make an estimate about which plugins are most popular, and we could rank them accordingly - assuming the most popular ones are also the most important ones. If you find this undesirable (it is a privacy issue but we have already sneaked in the version and language without asking anyone), then I would like to find some other way to indicate to new users which plugins are those commonly used. Bye Frederik -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] Remove internal help browser?
Hi, I personally dont like external browsers much. Many times I dont have control, which browser is used. If there will be move to external browser, please try it make somehow configurable, that after click on link button it will show panel with more options about what browser use. I have firefox with many, many, opened tabs and I lunch it only when I know that I will need to browse more than one page :-) In other cases I use konqueror. It ends up in 10 opened konquerors with 2 or 3 tabs over all desktops, but it have smaler footprint in memory and also in cpu ussage compared to one firefox app with ~80 tabs opened. PS: probably Iam very unstandard user, so do not be very burdened by request in this mail. On Friday 09 October 2009 21:44:40 Sebastian Klein wrote: Karl Guggisberg wrote: Hi, I plan to remove JOSMs internal help browser and to always delegate to an external browser, for two reasons: Good idea. Having an internal browser for help seems quite strange and old-fashioned to me. Also, the current version is far from usable. Nowadays, if there is some problem, you open a browser and google it. If it isn't answered in some forum immediately, you will at least find the documentation and try to resolve it there. Regarding this F1 thing - no one will get it. Best you can do is put some big inviting help button on the dialog that might need explanation. (But only if there is some documentation available.) It's quite pointless to press F1 and get an empty page each time. Dirk Stöcker wrote: Yes. External browsers have too many drawbacks regarding visibility and launching. If you do not like the internal one, then make it configurable to disable it. So many applications do this browser launch thing, it can't be that awful. Cheers, Sebastian ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] shocking - unsecure password sending!
Issues with importing certificates to java can be solved by contacting me :-) It is simple and if it will be necessary, I could write some tutorial, or code part which will be importing/installing certificates. PS: (very simple explanation) Certificates are stored for example in: java-6-sun-1.6.0.16/jre/lib/security/cacerts Importing to it could be done by some java code found on web (InstallCert.java), or you can create by that miniprogram some mycertfile which can be instrued to be used by some properties parameter: java -Djavax.net.ssl.trustStore=/var/lib/someProgram/my_java_cert -jar someProgram.jar On Wednesday 07 October 2009 15:27:36 ste...@binaervarianz.de wrote: On Wed, 07 Oct 2009 14:26:37 +0200, Frederik Ramm frede...@remote.org wrote: Hi, I agree that it is possible to get proper certificate for less than $400 per year (more like $30 or so). The free and self-signed ones have a tendency to be frowned upon by Sun's Java stack but maybe certificate checking can be disabled somehow? I don't know about Java problems with free certificates. One has to manually import them into browsers or personal keychains to use them, but JOSM could handle that gracefully for the user. I think certificates (at least self signed) are mandatory for HTTPS, but SSL could be used on its own without any form of authentication. Transferring username/password via https and the rest without would require changes to the server which someone would have to code. Yes, that's a problem. But is OAuth already implemented? But you have a point: Better let someone with ambition code something which doesn't quite fit the requirements instead of finding the perfect solution and nobody to actually implement it. I better calm down on the matter. A username/password authentication for most people just implies a kind of security which is not garanteed by the implemantation. I'm easily able to make edits, delete tracks and write diary entries in the name of other people as long as I 'm able to catch a JOSM authentication packet. I don't think all useres are aware of or even asume that. Of course most users are NOT aware of this, much as they are NOT aware that anyone can sniff out their credit card number when they make a purchase, or read their e-mail when they use an unencrypted W-Lan, or ... it's a cruel world! I have amended the JOSM start page to say that username and password are transmitted unencrypted, and that people should not upload changes if they do not want that. As soon as someone comes along who is willing and able to make the changes to the API, get them rolled out, and modify JOSM accordingly, that note can be removed. That was wise. OSM could even get in legal trouble if it would somehow 'lose' some private data. Better tell the user that their data isn't private at all. Regards Stefan ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] shocking - unsecure password sending!
I think that secure mechanism are today here and are a little bit standard (ssl, https). Why dont use them? If osm.org dont want to play with certificates, or to have some cpu power burned for ssl, I think that this problem isnt't josm problem. Imho osm.org should introduce some https page, where josm software could get pass token for some time period. This change of username+passwd for token, will be crypted (one https page). Other request will be http (no cpu burned for ssl) and will use token. This is second version how to get secured. First is to introduce https over all requests and imho should be implemented. This thread is from my point of wiev just waste of time and some Oauth, etc, would be also waste of time. This is my opinion, Iam a little bit paranoid in IT world, but I dont have any want (in meaning of have to be) for secured osm. Does anyone know what opinion does hava osm core group? On Tuesday 06 October 2009 19:11:04 Karl Guggisberg wrote: I think that people would be disappointed if one explained them how OAuth would work from JOSM. My understanding is, that it would work along the following steps: 1. User starts JOSM and clicks on Sign In 2. JOSM displays an internal, modal window saying We now launch a Web Browser. Please follow the instructions you are given there. At the end a so called request token will be generated for you. Please copy/paste it in the text field below and click 'Authorize' (did I mention that the window includes a text field and a button Authorize?) 2. An external (or internal) Web Browser is launched. It shows the normal www.openstreetmap.org login sreen. The user has to login with his user id/passwort. Since OSM still doesn't support HTTPS, neither for the login page nor for any other page, and since it only supports the Basic Auth schem, not digest authentication, the user id and the password are transferred in cleartext over the net, in exactly the same way JOSM transfers it today. 4. The user follows the steps required by OAuth, gets a request token, copies it, and pastes it to the field it JOSM. Then he clicks 'Authorize'. 5. JOSM requsts an access token from OSM and uses it in subsequent calls. The request token can be saved in the JOSM-profile (agreed, that this avoids having userid/password unencrypted in the profile) and it will be used to get another access token the next time JOSM is started, but using OAuth doesn't protect us from sending uid/password in cleartext over the net. Not much of a improvement, IMHO. Or do you I miss something? Regards Karl -Ursprüngliche Nachricht- Von: josm-dev-boun...@openstreetmap.org [mailto:josm-dev-boun...@openstreetmap.org] Im Auftrag von Valent Turkovic Gesendet: Dienstag, 6. Oktober 2009 09:56 An: josm-...@openstreetmap.org Betreff: Re: [josm-dev] shocking - unsecure password sending! On Sat, 26 Sep 2009 13:49:00 +, Ævar Arnfjörð Bjarmason wrote: On OSM.org you can give out tokens that allow the holder to *only* edit the map data. As opposed to also getting access to your private GPX tracks, making diary entries / comments etc. So transfering plaintext OAuth tokens would be more secure as in the event of a breach the access the attacker would gain to OSM.org in your name would at least be compartmentalized. Not to mention that the OAuth token would *only* work on OSM.org whereas users are likely to supply the same email/password pair for multiple websites that they're using. This definitely sounds like a step forward in the right direction. This seams like a nice feature to secure users account, and you are right, this would be much better than nothing. -- pratite me na twitteru - www.twitter.com/valentt http://kernelreloaded.blog385.com/ linux, blog, anime, spirituality, windsurf, wireless registered as user #367004 with the Linux Counter, http://counter.li.org. ICQ: 2125241, Skype: valent.turkovic ___ josm-dev mailing list josm-...@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev ___ josm-dev mailing list josm-...@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-...@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] shocking - unsecure password sending!
Thanks for that improvement. Now we are a little bit safer :-) Have a nice day. On Friday 02 October 2009 21:20:15 Dirk Stöcker wrote: On Thu, 24 Sep 2009, Valent Turkovic wrote: I was amazed when my OSM username and password appeared on Wall of Sheep during the conference at which I was presenting OpenStreetMap project! I was using JOSM only to download some data, and I wasn't aware that JOSM sends login data even when it is only downloading data and not sending. In rev. the capabilities request no longer sends username/password. This means again only uploads require authentication. Ciao -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] [PATCH 1/2] QuadBuckets (using quad tiling) for node storage in JOSM
AFAIK this is possible right now. Just use PostGis database of world, connect GeoServer like renderer / transformer to WMS / WFS / WSC and add WMS layer to JOSM. So in JOSM you could view whole world from PostGis (postgres) database. One problem is imho sld style for rendering which I doesnt found and it probably have to be writen manually. PS: searching thru WFS could be probably done by cql query. But Iam not sure about this. On Saturday 19 September 2009 21:36:04 Ævar Arnfjörð Bjarmason wrote: On Sat, Sep 19, 2009 at 6:15 PM, Dave Hansen d...@sr71.net wrote: On Sat, 2009-09-19 at 08:36 -0400, Greg Troxel wrote: I have wondered about hooking up postgis to josm as a working data storage format, so you can use spatial queries and indexes. That's got a lot of downsides, but I wonder about it being an optional setup for those working with big datasets. It's an interesting idea, but I don't know enough about postgis to give you an intelligent answer. PostGIS can be used as a backend for other GIS programs such as ArcGIS. So I think any potential troubles in this regard would have more to do with what sort of requirements JOSM makes of its data model than PostGIS limitations. Such an integration could bring some very neat features. Like being able to view the whole planet in JOSM with PostGIS on the backend with JOSM only requesting the data that it needed to display at any one time. With other operations such as search being (optionally) global. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
[josm-dev] svn not working?
Hi. Is svn down? Or do I have miss some migration? lu...@blackhorse:~/eclipseworkspace/josm$ svn up svn: OPTIONS of 'http://josm.openstreetmap.de/svn/trunk': could not connect to server (http://josm.openstreetmap.de) Thanks for response. (I want to publish last binary of slippymap plugin to solve some problems) -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] Cloudmade's Editor
Hi, You probably mean Specialized Junction Editing chapter of that url. If yes, I think that should not be problem to program something like that. I think that cloud made had a wery fresh look to mapping and this sort of editing wizard or guide is cutting edge. PS: does anyone know of some resource, where are writen some info about how to tag junctions? If it is too informations to keep in mind, why not to make some help for mappers? On Friday 24 July 2009 19:16:23 Frederik Ramm wrote: Hi, Cloudmade has the turn restriction editor we always wanted: http://blog.cloudmade.com/2009/07/22/mapzen-an-easy-to-use-editor-for-opens treetmap/ Let's get something like that for JOSM. Hope they haven't patented the thing ;-) Bye Frederik -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] [PATCH 01/14] make image scaling smarter
Hi. I have sucessfully merged patches from Dave Hansen. It probably works well :) Please try it from here http://dawn.ynet.sk/~luvar/openStreetMap/slippymap.jar or build your own from svn, where are all patches merged. After some approving, or after monday (Ill get my mouse :) Ill publish it to autoupdate to all josm users. Thanks to Dave code is now nicer, but I want to make it better because of lack of documentation and a veeery long class SlippyMapLayer (~800 lines). Also many new features for comfortable use was added. Thanks a lot. PS: Does anybody know, how to do logging in josm plugins? Should I use system.out, or logger from log4j, or something like Main.debug(bla bla)? On Saturday 13 June 2009 23:04:04 Dave Hansen wrote: I've been playing with the SlippyMap code again over the last month or two. The following set of patches (I think) clean up the code substantially. They also add some cool things like being able to draw images that are not from the current zoom level. -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] Czech house numbers
btw. I think Slovak numbering is same. I use just number which is connected to street name, but we also have that second number... PS: if the only problem is to implement it, I could have a look at it. But I think problem is to stabilize tags in house numbers... On Monday 11 May 2009 12:33:35 Radomir Cernoch wrote: Hi, the system is following: We have the standard 'housenumber', which defines the order of houses on a street and therefore this number makes sense only in combination with the street name. This system is used by most people as it is easy to remember and hence this number is rendered on OSM maps. In Czech this number is called číslo orientační. The second system, which we currently store in 'alternatenumber' (číslo popisné) defines a unique number of a house within a village/town or suburb. Houses are ordered historically, the oldest has number 1, the newest might have 4029. Therefore houses 4028 and 4029 can be on opposite sides of the town. This system is used for administration and in small villages it serves as the primary system, since it is not crucial to adopt the 'housenumber' system if eg. the municipality consists only of one street. Lastly 'alternatenumber' is used on Czech land register maps and so we need it for finding position of individual houses. This is why we would appreciate if by looking at the map in JOSM, we could see both the 'alternatenumber' of a node together with the background register map (provided by WMS). I think we should currently resolve the following question: Is it better to use 'addr:alternatenumber' (or any other similar key) as the preferred way of storing secondary housing numbers for _most_ countries or is it better to adopt 'addr:some_czech_label' for Czech Republic and 'addr:some_german_label' for Germany, etc.? Personally I would set up 'addr:alternatenumber' wherever possible and switch to 'addr:some_local_label' if local system was too different. However this would probably require to push 'alternatenumber' into official list of attributes... Yours, Radek Cernoch ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] R: R: R: R: R: FW: [OSM-newbies] JOSM problems - is it just me?
I have one question. Consider Iam random josm user, without any interest in bugreporting, finding why is something broken etc. (normal, flegmatic user :D) How will be presented this information to me? Will it after crash write me an link to bug and to solution? Is it possible when problem is in jvm? (I think it isnt possible to handle it in java code) Could be something done in running scripts which will detect this and make suggestions to users? Just asking. Probably there is some known bugs list, which everyone will see when downloading josm. Have a nice day. On Sunday 10 May 2009 09:40:27 Fabrizio Carrai wrote: Yes! After some days of tests and the Michael's confirmation, the problem can be consideres solved. I was trying to correlate the HW characteristics of two failing HWs, but I couldn't. So it remains a warning to Nvidia users and other with low video ram. Ciao! Fabrizio -Messaggio originale- Da: josm-dev-boun...@openstreetmap.org [mailto:josm-dev-boun...@openstreetmap.org]per conto di Michael Bemmerl Inviato: domenica 10 maggio 2009 2.18 A: josm-dev@openstreetmap.org Oggetto: Re: [josm-dev] R: R: R: R: FW: [OSM-newbies] JOSM problems - is it just me? Thanks Fabrizio, that did the trick! Fabrizio Carrai schrieb: It looks like I found the solution!!! It seems to happen with machines with low video ram. I've an NVidia Geoforce 6150 (shared memory 512MB): no idea if it falls in such category, but now it works! I have a GeForce 7300 GT with 256 MB on board. As in [1] the solution is to start JOSM including the Java option sun.java2d.d3d=false: C:\WINDOWS\system32\javaw.exe -Dsun.java2d.d3d=false -Xmx800M -jar C:\Programmi\Josm\josm-latest.jar I'm now working since several hours with JOSM doing several changes and uploads: no crashes till now! Before it was happening after very short editing sessions. Me too. Continuing testing! Ciao! Fabrizio [1] http://stackoverflow.com/questions/379574/possible-java-jna-issue Bye, Michi Nessun virus nel messaggio in arrivo. Controllato da AVG - www.avg.com Versione: 8.5.325 / Database dei virus: 270.12.23/2106 - Data di rilascio: 05/09/09 06:54:00 ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
Re: [josm-dev] New GPX implementation
It will run everywhere. Only problem with portability is that, that on java 1.5 systems is some aditional lib (packable into josm.jar) needed. afaik. But I see it like handicap, because if Iam not wrong, that lib has about 7 MB. That is realy big think to bundle with josm. I hope Iam wrong... PS: Iam not using primary java because of portability. It is useles on windows mobile based devices (no I/O posibility, no GPS, no access to BT...) I like java because of realy big framework which is bundled with its executive environment. On Wednesday 06 May 2009 02:10:31 Russ Nelson wrote: Ľubomír Varga writes: I like system not reinventing wheel. If josm will more rely on third party code, it will gain better maintability, stability and less code. Only drawback is in relying on third party code. There's nothing *wrong* with third party code. The problem is having a package that runs everywhere. If portability is unimportant to you, why write in Java? -- Odborník na všetko je zlý odborník. Ja sa snažím byť výnimkou potvrdzujúcou pravidlo. ___ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev