Philippe Grosjean schrieb:
> > This is not true. Version 0.8-8 of SciViews-R > (http://www.sciviews.org/SciViews-R) is compatible with R 2.1.1. There > were several months of delay between R 2.1.X and the compatible > SciViews-R version... There were a lot of changes required in my code, > because of the internationalization of R. I though translating R in > French was a priority, so, I have been working on "R French" first. > Also, I took a little time with my baby that was born in April... life > is not limited to programming R! Congratulations I know the time when my boy was born. > > Remember that Open Source is a collaborative and voluntary approach of > developing software. Please don't misunderstand my question. We developed a couple of (commercial) programs for medical university use. On of the required things were compatible interfaces between different hardware and software parts. for example: There was a definition of a function call: function (Var1, Var2, ....., Version) begin case version of version1: ... version2 ... end {case} end {function} This functions and procedures did the communication between different versions of hard an software. And the open source community is the best development environment to develop and take care for such API. So back to my suggestion. You told me that you must do a lot of changes in your GUI. If there would be one API which is serving a couple of old R Versions, then all running systems of GUIs would be running even the code for the latest R-version is changed. And only one part of the code must be changed not all available GUIs. > If you want to use all the time the latest R version, just learn to > use its standard command line interface. Some of the user want to use the newest command line interface some of (the newbies) want to use GUIs > If you prefer SciViews, JGR, R Commander, etc., then you should expect > some delay before R releases and compatible GUI releases. Indeed, one > must say that R Commander has always been released on time with the > latest R version, as fart as I know, and JGR is very close to. and this is the point that I thought maybe there could be some improvement with an universal API for GUIs > Latest SciViews version was very, very late. I apologize for that. Please don't think that this was my intention - it was just the other side. I realized that there was a long time especially for your nice GUI and I thought: It seems to be wasted time if Mr Grosjean must always do such a lot work after there is a new version of R. And therefor I tough about something to do for the R-Community. But dear community please see this suggestion in a manner that I do not have any knowledge about the work of the core team but that I am fascinated about the work, but that I am not able to win user for R because of problems with missing or not working GUIs. Missing: I do not know whether you know Wordperfect Office. They want the user to change from MS-Office. They build in a function: Switch to Word Mode, or switch to Powerpoint Mode. and this is the point to get Statistica user and SPSS user away from their programs. medium universities in Germany have between 500 and 1000 licences for SPSS. What a lot of money which is not available for the researchers Conclusion. I did not want a just ready solution, but I thought about some steps for the next years. There is an API available if I understand Prof. Ripley right, but then I do not understand the problems of the GUI developer. And please again I do not want to blame anybody of any development team I want to understand the project. > ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
