Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
Anita, Thanks for pointing out QEP#4, I wasn't aware of it. Tim has done an impressive work there. The above-mentioned QEP is a long term thing, what I was suggesting is a very short term (i.e. 2 cycles) proposal to try and satisfy the current needs for stability and devlopment momentum. I also am familiar with the discussion surrounding the 4 month cycle dates having been carefully chosen, hence why I was thinking that redistributing weeks within the context of two cycles wouldn't break that on the long term (i.e., by the end of the proposed two cycles, we're still 8 months from now, etc.) Math On Mon, Nov 10, 2014 at 2:11 PM, Anita Graser anitagra...@gmx.at wrote: Are you aware of QEP3? Please read Tim's suggestion. There are good reasons for this stable 4 month cycle at exactly the current release times of the year. Best wishes Anita On Nov 10, 2014 5:57 AM, Geo DrinX geodr...@gmail.com wrote: Yes yes yes. +1 but also +999 :) Roberto 2014-11-10 2:27 GMT+01:00 Mathieu Pellerin nirvn.a...@gmail.com: Guys, The recent thread Nyall kick-started with his “QGIS 3.0?” email got me to think about the eternal stability vs. development dilemma it (re-)exposed through the conversation. More specifically, it got me to brainstorm on the best way forward for QGIS at this juncture and whether there's a way to accommodate both the folks calling for a 2.8 LTS version, and others in need for space to further develop and expand QGIS' capability. And, I might just have found a way to do so. Here's the proposal, in a couple of points: - We make the 2.8 development cycle “fix and refinement”-only, and reduce the cycle's length to 6 to 8 weeks; - The reduced cycle will help everyone's focus on the above goal; - We append the freed 8-10 weeks to the subsequent development cycle, which would become QGIS 3.0; - The expanded cycle will help give space to develop some of the exciting features being cooked by developers (Nyall's Layouts, Marco's Geometry redesign, etc.) and bulletproof those. This, IMHO, caters to both groups demanding stability and space for development. It doesn't discourage or delay too much the grand scheme changes, and pushes out a 2.8 version focused on stability through a shorter cycle focusing on delivering a perfected tool. The above proposal does require a momentary lapse of the nice 4-month release cycle rhythm which the QGIS has successfully maintained for three releases now. But, it might actually be what's needed at this very time. Plus, the length of the two cycles stays the same, 8 months. Comments? I'm obviously particularly interested in what Jürgen has to say :) Cheers Math ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Python tests and C++ tests for the same functionality
Thank you for your comments I did not mean to exclude all C++ tests. If somebody wants or needs to write tests in C++ please continue to do so. I am referring only the ones that duplicate python tests. I don't think they have additional value and make maintainability harder. In general I think instead of rewriting the same test in another language it should be preferred to write a test for another functionality and increase coverage. Kind regards Matthias -- Help getting QGIS to the next level of quality before November 15! http://blog.vitu.ch/10102014-1046/crowdfunding-initiative-automated-testing ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
On Mon, Nov 10, 2014 at 5:57 AM, Geo DrinX geodr...@gmail.com wrote: Yes yes yes. +1 but also +999 :) And why not + ? ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
On Mon, Nov 10, 2014 at 3:23 PM, Luca Manganelli luc...@gmail.com wrote: On Mon, Nov 10, 2014 at 5:57 AM, Geo DrinX geodr...@gmail.com wrote: Yes yes yes. +1 but also +999 :) And why not + ? Seeing this I can't resist to quote a bit of PEP-10 [1] +1 I like it +0 I don't care, but go ahead -0 I don't care, so why bother? -1 I hate it You may occasionally see wild flashes of enthusiasm (either for or against) with vote scores like +2, +1000, or -1000. These aren't really valued much beyond the above scores, but it's nice to see people get excited about such geeky stuff. Cheers Martin [1] http://legacy.python.org/dev/peps/pep-0010/ ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 10/11/2014 09:31, Martin Dobias ha scritto: really valued much beyond the above scores, but it's nice to see people get excited about such geeky stuff. Hi all, I hate cooling down the enthusiasm, but I really see LTS as an empty word. To me, the whole issue boils down to having resources to do serious backporting of fixes. Without that, LTS will have no practical effect, as users will use the latest, more bugfixed version. This is exactly what has happened in the past. So from my point of view what we need is not +something, but funders supporting backporting. I'd be against spending our limited core funding for this. In short: power users, if you need stability, please set aside some funds to stably support a backporter, year round, and you'll have your much sought after long term stability. All the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgeowACgkQ/NedwLUzIr7k+gCeI1wmQ+nW7c7GepaNJewSDD6W jTcAnifJPwxN1WZ51CWaIB5gGzsssm26 =4OQL -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] qgis2threejs issue
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 10/11/2014 04:56, Minoru Akagi ha scritto: Please open a ticket if you like. done, thanks. all the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgfWUACgkQ/NedwLUzIr7NxgCgqRRB9mpwuMSC5KhUYQBXxork fjMAnR+nlT6KuhrTCuUudqhJKSvSVPmH =k5In -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
[Qgis-developer] Behaviour when create new features
I have an inexperienced QGIS user who needed to draw some simple line strings. She created a new shapefile layer (line type), enabled editing and used the Add Feature button to draw a line string. - She didn't right click to finished the feature when she was done! She saved the alterations (none because she didn't finish the feature creation process) and disabled editing. Upon opening the layer in another QGIS instance, the feature she created was missing - obvious since she didn't entirely created it. Why did she though that she had created it? Because even though she drew the line string (without saving it) and disabled the editing, the left over line string remains visible on the screen. You cannot click it, get feature info and so on, but it appears to be exactly where she created it. I don't know if this is created as an issue but is this the expected behavior when creating and aborting new features? This applies to QGIS 2.4 and 2.6. Kind regards, Casper ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
On Mon, Nov 10, 2014 at 9:42 AM, Paolo Cavallini cavall...@faunalia.it wrote: To me, the whole issue boils down to having resources to do serious backporting of fixes. Without that, LTS will have no practical effect, as users will use the latest, more bugfixed version. more bugfixed is not always true. We had issues with 1.8.0 and 2.2.0 and we refused to use them in our organization due to critical bugs that are fixed in newer version, but they have sometimes introduced other. So, I believe that in production environment the most stable (!= latest) version is used. For almost 2 years we used 1.7.4, for me the most stable QGIS version in earth (more than 2.4 and 2.6!). ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
On Mon, Nov 10, 2014 at 10:42:42AM +0100, Luca Manganelli wrote: So, I believe that in production environment the most stable (!= latest) version is used. For almost 2 years we used 1.7.4, for me the most stable QGIS version in earth (more than 2.4 and 2.6!). Yep, that little third number in the version is really special. Finally, it's being given the importance it deserves, thanks Tim for the effort in writing a plan for it: https://github.com/qgis/QGIS-Enhancement-Proposals/pull/6 --strk; Please help taking QGIS to the next level of quality. Before November 15 ! http://blog.vitu.ch/10102014-1046/crowdfunding-initiative-automated-testing ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
So, I believe that in production environment the most stable (!= latest) version is used. For almost 2 years we used 1.7.4, for me the most stable QGIS version in earth (more than 2.4 and 2.6!). Oh man. I couldn't even use 1.7.4 anymore it's so old ;) Anyway the point is a valid one. Running the latest != most stable. IMO we don't need resources to do bug fixing. The dev that does the bug fix in master can do it in the 2.x branch for that stable release if it is relevant.. This obviously has to be done smart but using the recent crash and project corruption as an example that Martin fixed right away, to me this warrants a new release off that branch, LTS or not, as project corruption is a really really bad look. - Nathan On Mon Nov 10 2014 at 7:43:27 PM Luca Manganelli luc...@gmail.com wrote: On Mon, Nov 10, 2014 at 9:42 AM, Paolo Cavallini cavall...@faunalia.it wrote: To me, the whole issue boils down to having resources to do serious backporting of fixes. Without that, LTS will have no practical effect, as users will use the latest, more bugfixed version. more bugfixed is not always true. We had issues with 1.8.0 and 2.2.0 and we refused to use them in our organization due to critical bugs that are fixed in newer version, but they have sometimes introduced other. So, I believe that in production environment the most stable (!= latest) version is used. For almost 2 years we used 1.7.4, for me the most stable QGIS version in earth (more than 2.4 and 2.6!). ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 10/11/2014 10:42, Luca Manganelli ha scritto: So, I believe that in production environment the most stable (!= latest) version is used. For almost 2 years we used 1.7.4, for me the most stable QGIS version in earth (more than 2.4 and 2.6!). There is no such a thing as the most stable version: what is blocking for an user is not relevant for another. I have customers and friends that cannot upgrade to various versions for very specific bugs. I'm sorry to insist: backporting fixes is the only relevant thing in this issue, IMO. All the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgi80ACgkQ/NedwLUzIr7pLwCePSD5Z+1SXav/K3L3dVa5wQDS fT4An1zPtLE2u6L94t4El8siPZvEDOat =1S/2 -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 10/11/2014 10:56, Nathan Woodrow ha scritto: IMO we don't need resources to do bug fixing. The dev that does the bug fix in master can do it in the 2.x branch for that stable release if Sorry I do not agree here: we had many cases of fixes breaking other stuff, so backporting should be done with great care, and lots of extra work; that's why I believe that without significant resources we are not going to solve the problem satisfactorily. All the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgjH0ACgkQ/NedwLUzIr72yQCdHYe2bCc19H3eyfsammiMur1P KtEAnik3lf96t2D1daJ4ZaiST8Rw5Qtd =0gqs -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
Hi Paolo and all developers, Il 10/11/2014 09.42, Paolo Cavallini ha scritto: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 10/11/2014 09:31, Martin Dobias ha scritto: really valued much beyond the above scores, but it's nice to see people get excited about such geeky stuff. Hi all, I hate cooling down the enthusiasm, but I really see LTS as an empty word. To me, the whole issue boils down to having resources to do serious backporting of fixes. Without that, LTS will have no practical effect, as users will use the latest, more bugfixed version. This is exactly what has happened in the past. So from my point of view what we need is not +something, but funders supporting backporting. I'd be against spending our limited core funding for this. Reading this post i get a bit confused about the future of QGIS and its target, because in the QGIS Release schedule for 2015 (QGIS web site) there is a 2.8 LTR release, maybe this LTR release is to be bugfixed, maybe just a release to stuck with if you don't like to switch but with no planned bug fixing, this is no clear to me ... In short: power users, if you need stability, please set aside some funds to stably support a backporter, year round, and you'll have your much sought after long term stability. It is also nice to have someone who keeps us with our feet on the ground :-) thanks and regards, Roy. ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
On Mon, Nov 10, 2014 at 09:56:22AM +, Nathan Woodrow wrote: relevant.. This obviously has to be done smart but using the recent crash and project corruption as an example that Martin fixed right away, to me this warrants a new release off that branch, LTS or not, as project corruption is a really really bad look. Well, while I'd agree on that if there was no concept of LTS, in presence of LTS whatever corruption happens would have to be expected by users in all but LTS releases, meaning there'd be no rush to ever cut a new release unless the week of silence following fix in LTS branch event happens. How far away would next (first) LTS be ? --strk; Please help taking QGIS to the next level of quality. Before November 15 ! http://blog.vitu.ch/10102014-1046/crowdfunding-initiative-automated-testing ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
Hi all, QEP #4 allows to do backports for every release. Not only LTR. 2.6.1 will be very welcome. LTR releases will be available for 1 year and will receive bugfixes during that time. That's not going to happen magically. That requires power users and organizations to help the development. Thank you Paolo for raising this important point. QEP #4 IMO outlines a frame in which organizations have more security in what they invest. Documentation will be valid for a longer timeframe. Things are less likely to break. There is the possibility of introducing a new bug with another fix. But that's not a reason at all to not fix something. Organizations which are using LTR will probably now already be testing new releases heavily before they deploy. That will help to discover such bugs fast. Backporting features is not something I would like to see. That really introduces a bigger risk of breaking things. If that's done we could just stick to the current release schedule. (For every feature you will find somebody who asks for a backport to LTR). To me, QEP #4 is a wonderful plan. It needs the support of organizations! Maybe some organizations can build a pool with funds and Q/A that is determined to maintain that version? Regards, Matthias -- Help getting QGIS to the next level of quality before November 15! http://blog.vitu.ch/10102014-1046/crowdfunding-initiative-automated-testing ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
[Qgis-developer] import gdalconst
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi all. I noticed that import gdalconst fails on current win version (2.6, gdal 1.11), whereas it is ok on Debian (2.6, gdal 1.10.1): has something changed, or is this a (packaging?) issue? All the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgkaQACgkQ/NedwLUzIr4ozwCgk8kywDa08Y6sdT5ykmxotWzr HykAn1rlTZ0cT6JOAlt+DpowL3c9a7Lr =Ax4H -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
[Qgis-developer] QGIS-Server Coredump printing with lizmap webclient
Dear all, i'm testing qgis-server 2.6.0 with lizmap web client on ubuntu 12.04 server and qgis server crashes trying to print a vector layer with advanced symbols ( line pattern fill and marker line); apache log following: # apache2 error.log #mod_fcgid: error reading data, FastCGI server closed connection #[Mon Nov 10 11:28:47 2014] [error] [client 127.0.0.1] Premature end of script headers: qgis_mapserv.fcgi #[Mon Nov 10 11:28:51 2014] [error] mod_fcgid: process /usr/lib/cgi-bin/qgis_mapserv.fcgi(1998) exit(communication error), get signal #11, possible coredump generated # apache2 error.log END using more simple (simple fill) symbology this does not appens is this possibly a bug or are we not supposed to use fancy symbols printig with qgis-server ... thanks, Roy. P.S. is this the right list to ask questions about lizmap and qgis-server? ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] QGIS-Server Coredump printing with lizmap webclient
2014-11-10 11:40 GMT+01:00 Roy royr...@outlook.com: Dear all, i'm testing qgis-server 2.6.0 with lizmap web client on ubuntu 12.04 server and qgis server crashes trying to print a vector layer with advanced symbols ( line pattern fill and marker line); apache log following: # apache2 error.log #mod_fcgid: error reading data, FastCGI server closed connection #[Mon Nov 10 11:28:47 2014] [error] [client 127.0.0.1] Premature end of script headers: qgis_mapserv.fcgi #[Mon Nov 10 11:28:51 2014] [error] mod_fcgid: process /usr/lib/cgi-bin/qgis_mapserv.fcgi(1998) exit(communication error), get signal #11, possible coredump generated # apache2 error.log END using more simple (simple fill) symbology this does not appens is this possibly a bug or are we not supposed to use fancy symbols printig with qgis-server ... thanks, Roy. P.S. is this the right list to ask questions about lizmap and qgis-server? See if this is your case: http://hub.qgis.org/issues/9763 More info here: http://www.itopen.it/2014/03/13/qgis-server-setup-notes/ In short: make sure you have an (fake) X sever running. -- Alessandro Pasotti w3: www.itopen.it ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] QGIS-Server Coredump printing with lizmap webclient
Il 10/11/2014 11.44, Alessandro Pasotti ha scritto: 2014-11-10 11:40 GMT+01:00 Roy royr...@outlook.com: Dear all, i'm testing qgis-server 2.6.0 with lizmap web client on ubuntu 12.04 server and qgis server crashes trying to print a vector layer with advanced symbols ( line pattern fill and marker line); apache log following: # apache2 error.log #mod_fcgid: error reading data, FastCGI server closed connection #[Mon Nov 10 11:28:47 2014] [error] [client 127.0.0.1] Premature end of script headers: qgis_mapserv.fcgi #[Mon Nov 10 11:28:51 2014] [error] mod_fcgid: process /usr/lib/cgi-bin/qgis_mapserv.fcgi(1998) exit(communication error), get signal #11, possible coredump generated # apache2 error.log END using more simple (simple fill) symbology this does not appens is this possibly a bug or are we not supposed to use fancy symbols printig with qgis-server ... thanks, Roy. P.S. is this the right list to ask questions about lizmap and qgis-server? See if this is your case: http://hub.qgis.org/issues/9763 More info here: http://www.itopen.it/2014/03/13/qgis-server-setup-notes/ In short: make sure you have an (fake) X sever running. Thank you Alessandro that's working ! regards, Roy. ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
A couple of thoughts from a non-dev looking inwards: Sorry I do not agree here: we had many cases of fixes breaking other stuff Would not something like Unit Tests help ameliorate that? That's what they're designed for isn't it? I realise the state of QGIS' unit test infrastructure isn't optimal currently, but I thought I saw a project to fix get funding recently. : what is blocking for an user is not relevant for another. I have customers and friends that cannot upgrade to various versions for very specific bugs. Then why not fix the bugs and require them to be backported? I know that seems flippant, but is there a reason that backporting by the submitter/committer can't be required for any bugfix submitted? If a bugfix breaks other stuff, then either the bugfix should be regressed or the breakage fixed with another fix. Neither of these suggestions would require any outlay from the QGIS core fund, though they may increase the cost of any individual feature/bugfix. I believe GeoServer does both of these and has a healthy 30-day release schedule consisting of up to 3 branches despite having considerably fewer resources than QGIS. Just my 2p. Cheers, Jonathan -Original Message- From: qgis-developer-boun...@lists.osgeo.org [mailto:qgis-developer-boun...@lists.osgeo.org] On Behalf Of Paolo Cavallini Sent: Monday, November 10, 2014 9:59 AM To: qgis-developer@lists.osgeo.org Subject: Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 10/11/2014 10:56, Nathan Woodrow ha scritto: IMO we don't need resources to do bug fixing. The dev that does the bug fix in master can do it in the 2.x branch for that stable release if Sorry I do not agree here: we had many cases of fixes breaking other stuff, so backporting should be done with great care, and lots of extra work; that's why I believe that without significant resources we are not going to solve the problem satisfactorily. All the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgjH0ACgkQ/NedwLUzIr72yQCdHYe2bCc19H3eyfsammiMur1P KtEAnik3lf96t2D1daJ4ZaiST8Rw5Qtd =0gqs -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer This message has been scanned for viruses by MailControl - www.mailcontrol.com Click https://www.mailcontrol.com/sr/IIyLovGDSdTGX2PQPOmvUizKrmxxhcEGFn0HB2PmJETu!FF59CPCcZfM1iyAeuy4hu5YqLyvUFQhei1KLnaPDA== to report this email as spam. HR Wallingford and its subsidiaries uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Behaviour when create new features
Personally I’d think there are two bugs here, at least from a user-interface perspective, even if they’re not from a technical one: a) Edits showing when editing was disabled (I’ve seen this one myself). b)No warning that the edited features hadn’t been “finished” when saving and/or when disabling editing. (2.6 Testing indicates that disabling and then re-enabling editing while editing but not “finishing” a feature shows that you continue editing the same feature; I’m not sure that’s very intuitive, hence item (b).) Cheers, Jonathan From: qgis-developer-boun...@lists.osgeo.org [mailto:qgis-developer-boun...@lists.osgeo.org] On Behalf Of Casper Børgesen (CABO) Sent: Monday, November 10, 2014 9:09 AM To: qgis-developer Subject: [Qgis-developer] Behaviour when create new features I have an inexperienced QGIS user who needed to draw some simple line strings. She created a new shapefile layer (line type), enabled editing and used the “Add Feature” button to draw a line string. - She didn’t right click to finished the feature when she was done! She saved the alterations (none because she didn’t finish the feature creation process) and disabled editing. Upon opening the layer in another QGIS instance, the feature she created was missing – obvious since she didn’t entirely created it. Why did she though that she had created it? Because even though she drew the line string (without saving it) and disabled the editing, the left over line string remains visible on the screen. You cannot click it, get feature info and so on, but it appears to be exactly where she created it. I don’t know if this is created as an issue but is this the expected behavior when creating and aborting new features? This applies to QGIS 2.4 and 2.6. Kind regards, Casper This message has been scanned for viruses by MailControlhttp://www.mailcontrol.com/, a service from BlackSpider Technology Click herehttps://www.mailcontrol.com/sr/6TqvtsvWU1!GX2PQPOmvUqy!8K4P2+30kzTs6GKpZSmWH+w0+RKE!5fM1iyAeuy4hu5YqLyvUFQhei1KLnaPDA== to report this email as spam. HR Wallingford and its subsidiaries uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Behaviour when create new features
b)No warning that the edited features hadn’t been “finished” when saving and/or when disabling editing. (2.6 Testing indicates that disabling and then re-enabling editing while editing but not “finishing” a feature shows that you continue editing the same feature; I’m not sure that’s very intuitive, hence item (b).) These are both bugs. Just need to open a ticket for them. - Nathan On Mon Nov 10 2014 at 10:22:16 PM Jonathan Moules j.mou...@hrwallingford.com wrote: Personally I’d think there are two bugs here, at least from a user-interface perspective, even if they’re not from a technical one: a) Edits showing when editing was disabled (I’ve seen this one myself). b)No warning that the edited features hadn’t been “finished” when saving and/or when disabling editing. (2.6 Testing indicates that disabling and then re-enabling editing while editing but not “finishing” a feature shows that you continue editing the same feature; I’m not sure that’s very intuitive, hence item (b).) Cheers, Jonathan *From:* qgis-developer-boun...@lists.osgeo.org [mailto: qgis-developer-boun...@lists.osgeo.org] *On Behalf Of *Casper Børgesen (CABO) *Sent:* Monday, November 10, 2014 9:09 AM *To:* qgis-developer *Subject:* [Qgis-developer] Behaviour when create new features I have an inexperienced QGIS user who needed to draw some simple line strings. She created a new shapefile layer (line type), enabled editing and used the “Add Feature” button to draw a line string. - She didn’t right click to finished the feature when she was done! She saved the alterations (none because she didn’t finish the feature creation process) and disabled editing. Upon opening the layer in another QGIS instance, the feature she created was missing – obvious since she didn’t entirely created it. Why did she though that she had created it? Because even though she drew the line string (without saving it) and disabled the editing, the left over line string remains visible on the screen. You cannot click it, get feature info and so on, but it appears to be exactly where she created it. I don’t know if this is created as an issue but is this the expected behavior when creating and aborting new features? This applies to QGIS 2.4 and 2.6. Kind regards, Casper This message has been scanned for viruses by MailControl http://www.mailcontrol.com/, a service from BlackSpider Technology Click here https://www.mailcontrol.com/sr/6TqvtsvWU1!GX2PQPOmvUqy!8K4P2+30kzTs6GKpZSmWH+w0+RKE!5fM1iyAeuy4hu5YqLyvUFQhei1KLnaPDA== to report this email as spam. -- *HR Wallingford and its subsidiaries* uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 -- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Stability (2.8 LTS) vs development (3.0), a proposed way forward
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi all. Il 10/11/2014 13:13, Jonathan Moules ha scritto: Then why not fix the bugs and require them to be backported? I know that seems flippant, but is there a reason that backporting by the submitter/committer can't be required for any bugfix submitted? If a bugfix breaks other stuff, then either the bugfix should be regressed or the breakage fixed with another fix. Agreed. Please consider, however, that this generally not something that comes for free, and it will increase the cost of the bugfix, so the customers must be clearly aware of this. All the best. - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRgwFkACgkQ/NedwLUzIr67jACdHvj10iJOnyFZAQL3iZedquqF ingAnRqGBy5hs2mKjCbWvlppuwo5+h71 =y+Y+ -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Behaviour when create new features
b)No warning that the edited features hadn’t been “finished” when saving and/or when disabling editing. I have created a ticket for this: http://hub.qgis.org/issues/11607 (2.6 Testing indicates that disabling and then re-enabling editing while editing but not “finishing” a feature shows that you continue editing the same feature; I’m not sure that’s very intuitive, hence item (b).) Jonathan, could you create a ticket for the other one, I guess You are more into this part? Kind regards, Casper From: Nathan Woodrow [mailto:madman...@gmail.com] Sent: 10. november 2014 13:26 To: Jonathan Moules; Casper Børgesen (CABO); qgis-developer Subject: Re: [Qgis-developer] Behaviour when create new features b)No warning that the edited features hadn’t been “finished” when saving and/or when disabling editing. (2.6 Testing indicates that disabling and then re-enabling editing while editing but not “finishing” a feature shows that you continue editing the same feature; I’m not sure that’s very intuitive, hence item (b).) These are both bugs. Just need to open a ticket for them. - Nathan On Mon Nov 10 2014 at 10:22:16 PM Jonathan Moules j.mou...@hrwallingford.commailto:j.mou...@hrwallingford.com wrote: Personally I’d think there are two bugs here, at least from a user-interface perspective, even if they’re not from a technical one: a) Edits showing when editing was disabled (I’ve seen this one myself). b)No warning that the edited features hadn’t been “finished” when saving and/or when disabling editing. (2.6 Testing indicates that disabling and then re-enabling editing while editing but not “finishing” a feature shows that you continue editing the same feature; I’m not sure that’s very intuitive, hence item (b).) Cheers, Jonathan From: qgis-developer-boun...@lists.osgeo.orgmailto:qgis-developer-boun...@lists.osgeo.org [mailto:qgis-developer-boun...@lists.osgeo.orgmailto:qgis-developer-boun...@lists.osgeo.org] On Behalf Of Casper Børgesen (CABO) Sent: Monday, November 10, 2014 9:09 AM To: qgis-developer Subject: [Qgis-developer] Behaviour when create new features I have an inexperienced QGIS user who needed to draw some simple line strings. She created a new shapefile layer (line type), enabled editing and used the “Add Feature” button to draw a line string. - She didn’t right click to finished the feature when she was done! She saved the alterations (none because she didn’t finish the feature creation process) and disabled editing. Upon opening the layer in another QGIS instance, the feature she created was missing – obvious since she didn’t entirely created it. Why did she though that she had created it? Because even though she drew the line string (without saving it) and disabled the editing, the left over line string remains visible on the screen. You cannot click it, get feature info and so on, but it appears to be exactly where she created it. I don’t know if this is created as an issue but is this the expected behavior when creating and aborting new features? This applies to QGIS 2.4 and 2.6. Kind regards, Casper This message has been scanned for viruses by MailControlhttp://www.mailcontrol.com/, a service from BlackSpider Technology Click herehttps://www.mailcontrol.com/sr/6TqvtsvWU1!GX2PQPOmvUqy!8K4P2+30kzTs6GKpZSmWH+w0+RKE!5fM1iyAeuy4hu5YqLyvUFQhei1KLnaPDA== to report this email as spam. HR Wallingford and its subsidiaries uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.orgmailto:Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Save selection as
Hi, Irrespective of which is the way to do it (I lean toward checking it when there's a selection), I'd suggest that the behaviour should be consistent. So either this default be changed, or the field calculator one be, otherwise you end up with confused users. Cheers, Jonathan -Original Message- From: qgis-developer-boun...@lists.osgeo.org [mailto:qgis-developer-boun...@lists.osgeo.org] On Behalf Of Jürgen E. Fischer Sent: Saturday, November 08, 2014 11:41 AM To: qgis-developer@lists.osgeo.org; qgis...@lists.osgeo.org Subject: Re: [Qgis-developer] Save selection as Hi Anita, On Sat, 08. Nov 2014 at 12:32:52 +0100, Anita Graser wrote: On Sat, Nov 8, 2014 at 11:53 AM, Paolo Cavallini cavall...@faunalia.it wrote: Il 08/11/2014 10:18, Matthias Kuhn ha scritto: I think this could result in unexpected outcome for the users. All the best. Seems like a UX question ;-) I'm with Jürgen, it's very likely to confuse users because they don't realize that the had a selection and then they find that features are missing in the export. But I don't have a opinion on it - there's probably no save default for it. In that way having the two menu entries might have been better. Jürgen -- Jürgen E. Fischer norBIT GmbH Tel. +49-4931-918175-31 Dipl.-Inf. (FH) Rheinstraße 13 Fax. +49-4931-918175-50 Software Engineer D-26506 Norden http://www.norbit.de QGIS release manager (PSC) GermanyIRC: jef on FreeNode HR Wallingford and its subsidiaries uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] issues in QGIS 2.6 when saving/working with projects saved in previous QGIS releases
Hi, Is there an ETA on this? We're thinking of moving up to 2.6, but I can see this hitting us if we do. Thanks, Jonathan -Original Message- From: qgis-developer-boun...@lists.osgeo.org [mailto:qgis-developer-boun...@lists.osgeo.org] On Behalf Of Paolo Cavallini Sent: Saturday, November 08, 2014 7:37 AM To: qgis-developer@lists.osgeo.org Subject: Re: [Qgis-developer] issues in QGIS 2.6 when saving/working with projects saved in previous QGIS releases -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Il 08/11/2014 01:32, Mathieu Pellerin ha scritto: IMO, this is serious enough to package a 2.6.1 release speedily. There are enough ppl out there that don't backup their documents on a regular basis for this bug to be hugely problematic. seems reasonable to me all the best - -- Paolo Cavallini - www.faunalia.eu QGIS PostGIS courses: http://www.faunalia.eu/training.html -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlRdyDcACgkQ/NedwLUzIr7TbQCgg8MNiVXmwrXkFXvPnL1bh7kM OnAAni654JHAkWQqdNBCcAw1HXw7yo8H =YjGY -END PGP SIGNATURE- ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer This message has been scanned for viruses by MailControl - www.mailcontrol.com Click https://www.mailcontrol.com/sr/E6m3fMEKmGjGX2PQPOmvUkHxlm0SZLPCJKgYTJXEtXtEcF1xMGmboSw3pDhQYZTTb6gPET3dChj1VsBVsOva5A== to report this email as spam. HR Wallingford and its subsidiaries uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Save selection as
Hi, I think that functionality that works on selections and not on the whole layer should always be optional, and not the default. And yes - consistency would be nice ;-) Thanks, Andreas On 10.11.2014 15:48, Jonathan Moules wrote: Hi, Irrespective of which is the way to do it (I lean toward checking it when there's a selection), I'd suggest that the behaviour should be consistent. So either this default be changed, or the field calculator one be, otherwise you end up with confused users. Cheers, Jonathan -Original Message- From: qgis-developer-boun...@lists.osgeo.org [mailto:qgis-developer-boun...@lists.osgeo.org] On Behalf Of Jürgen E. Fischer Sent: Saturday, November 08, 2014 11:41 AM To: qgis-developer@lists.osgeo.org; qgis...@lists.osgeo.org Subject: Re: [Qgis-developer] Save selection as Hi Anita, On Sat, 08. Nov 2014 at 12:32:52 +0100, Anita Graser wrote: On Sat, Nov 8, 2014 at 11:53 AM, Paolo Cavallini cavall...@faunalia.it wrote: Il 08/11/2014 10:18, Matthias Kuhn ha scritto: I think this could result in unexpected outcome for the users. All the best. Seems like a UX question ;-) I'm with Jürgen, it's very likely to confuse users because they don't realize that the had a selection and then they find that features are missing in the export. But I don't have a opinion on it - there's probably no save default for it. In that way having the two menu entries might have been better. Jürgen -- Jürgen E. Fischer norBIT GmbH Tel. +49-4931-918175-31 Dipl.-Inf. (FH) Rheinstraße 13 Fax. +49-4931-918175-50 Software Engineer D-26506 Norden http://www.norbit.de QGIS release manager (PSC) GermanyIRC: jef on FreeNode HR Wallingford and its subsidiaries uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them. If you have received this message in error please advise us immediately and destroy all copies of it. HR Wallingford Limited Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom Registered in England No. 02562099 ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Field Calculator permanence
Hi Andreas On 2014/11/08 18:34, Andreas Neumann wrote: Hi Zoltan, There are many other file formats besides shapes, including many databases. So people should stop thinking that the universe only circles around shapefiles. Yes. I know. That's exactly what I am banging my head about. Do you know that some 15 years back, I had to spend vast amounts of energy explaining to some of my clients that SHP file format does NOT support topology!! But I supposed I have been spoilt. In 1984 when I started Geograph, I had a Mainframe (you know what that is? :-) ) based GIS package (called SICAD) and in 1984 I could write scripts, crawl up and down the data structure and do almost anything because it had a fully topologically connected spatial data structure. I ditched that in 1989 in favour of Unix based Genamap (now Linux) [and gone bung, but the SW still works], also FULL topology and full script-ability, and I still use that unless my clients demand that I use their software (and then they provide the licensing for the project). So, I have mostly by-passed the rigorous requirements of conformity. With these other file formats it is not very common to ship other files along with the main file. Ha ha ha... tell that to the masses. Guess what format our friendly RSA mapping office gives out their Vector Data in? Why do you think that drag/drop or add layer is unintuitive. Those are the normal ways to load a layer in QGIS. Well, not so much the 'drag drop', but the way Layer Definition Files are loaded: Because I never thought of it as an option (yes, that's my failure), but also because intuitively it should be on the Add Layer button or list, and not as a separate item slightly further down. If it appeared on the addlayer button, it would be noticed, and people would be more likely to check what that option does (and learn). It is in fact just another way to load a map type. So, I'm not at all grumbling, just suggesting that maybe that way of loading a layer should be on the Add Layer button/dropdown. (and not separately on the layer dropdown list) Anyway, thanks for your input, Regards, Zoltan Andreas On 07.11.2014 18:49, Zoltan Szecsei wrote: On 2014/11/07 19:19, Zoltan Szecsei wrote: Thanks for the pointer to 'Save Layer Definition file' - I've never noticed that, so I'll play with that for a bit. On 2014/11/07 18:59, Andreas Neumann wrote: However, if you save a layer definition file along with the shapefile and re-add that, they are preserved. So you can just store the layer definition file along with the shape and you are fine to share the data with your colleagues, preserving virtual fields ;-) Hi, OK - not very intuitive, (to have to either drag and drop the qlr file into the layer panel, or to add layer from definition file) Any reason that QGIS can't be made to look for qlr files when opening the SHP file the normal way? Regards thanks for the help. Zoltan Andreas On 07.11.2014 17:32, Zoltan Szecsei wrote: Hi Andreas, I'll fiddle with the Postgis idea - thanks. The orange part below: I closed the shapefile, deleted from legend and re-opened it - but could not see my virtual field, nor the expression I used to create it. Have I missed a trick somewhere? Regards, Z On 2014/11/07 18:25, Andreas Neumann wrote: Hi, What you probably ask for are database triggers combined with storage. This is not what virtual fields are for. If they were stored - they wouldn't be called virtual ;-) You can do that f.e. with Postgis. With shapefiles not. However, if you share the project with the other PC, the virtual fields work fine. Andreas On 07.11.2014 17:21, Zoltan Szecsei wrote: Hi Matthias, Thanks for the quick answer. The virtual field acts just like I want it to, except that it does not permanently store the column in the attribute table. I tried getting clever by then adding a second 'Field Calculation' to update the existing 'Sheet' field from the virtual field, but that won't do it either. So, for me there are two shortcomings: 1. To be able to store and recall the 'Expression' and not lose it after closing the Field Calculator 2. To be able to add the virtual fields that you create, into the attribute table. 3. (And then of course to re-open the Shape file on another PC and have these Field Calculations still functional) Any further possibilities I can try? Regards thanks again, Zoltan On 2014/11/07 18:01, Matthias Kuhn wrote: Hi Zoltan, Since QGIS 2.6 this possibility exists. It is called Virtual Field. You must not create the attribute when creating the shapefile. Just open the field calculator, create a new field in there and check the virtual field checkbox. Regards, Matthias On 07.11.2014 16:54, Zoltan Szecsei wrote: Hi, I'm using QGIS 2.6.0 on Win 7 x64 - installed via OSGEO4w Is there a way to store your field calculator expression to always process and update that specific field, should geometry be
Re: [Qgis-developer] Changes in API documentation?
thank you Martin regards, Luigi Pirelli On 7 November 2014 18:55, Martin Dobias wonder...@gmail.com wrote: On Tue, Nov 4, 2014 at 1:53 AM, Luigi Pirelli lui...@gmail.com wrote: Hi martin I almost agree, but qgis is not only a lib but kind of framework. It's use is different tespect the use i van do with qt or gdal. Not always I've the code available to investigate... but I generally have connection. In my point of view, api doc wasn't clutter Good news - due to popular demand I have reverted the change in master and release-2_6 branch. Cheers Martin ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer