Hi Ede, Michael, Matthias following also Michaels comment in the other email: My suggestion would be a proper (english) error message (= what you call a note). No stack-trace or so. Just a simple one-liner (plus weblink?) (using try/catch/...?)
Is that possible? Though - I see you did already an RC3. So at least we could change it for 1.5.3. Btw: didn't thought about the licence note. And it seems to me even more important if it is webstart. Or is licence anywhere else noted in webstart? stefan Am 14.05.12 05:16, schrieb edgar.sol...@web.de: > On 14.05.2012 03:00, Stefan Steiniger wrote: >> Hi Ede I have a question, >> >> In the commit comment Matthias wrote also about another cleanup/change: >> * removed the output of the Stacktrace, if the "readme.txt" was not >> found. With Java Webstart this file cannot be found. It's better to view >> nothing. A Stacktrace confuse the eunduser. >> * Small GUI fixup, the about and infotab content are noc alwas on top. >> >> with your revert - is that undone as well? If so - Why? - i.e. was it >> not useful or directly related to his other changes? > > i added links to the changelog comment when reverting them > http://jump-pilot.svn.sourceforge.net/viewvc/jump-pilot/core/trunk/ChangeLog?r1=2869&r2=2868&pathrev=2869 > > btw. the snapshots of his change are still online.. compare for yourself ;) > >> The discussion below just focuses on the "acknowledgement" of a missing >> resource. > > yes > >> My opinion actually would be: having a stack trace exposed to a user in >> a GUI is nothing that we should do. For that we have, as Matthias wrote, >> exactly the console message/log window. > > i fear users won't see/acknowledge the error there and hence won't come back > with a report to us. > >> However, I also see that the >> problem here is rather a packaging (testing?) issue. > > absolutely. i actually strongly feel that delivering OJ without readme.txt > and licenses/* violates some licenses we or our components are under. > > i argue (below) that > > a) webstart could live with the status quo for the time being until it can > ship readme.txt > b) for desktop version having readme.txt missing is an error, so there should > be a visible error. we could of course agree on showing a note instead of the > stack there. > > another alternative could be offering a link to an online readme.txt . but > again this should only be shown in webstart as desktop users cannot be > expected to have internet access. > > ..ede > >> PS: checked the new RC2 - looks good so far. Except that I had 2x >> download errors (i.e. only 26MB downloaded instead of 29/31). Not sure >> what happend - but thats probably a network or SF issue. > > Re-PS: probably an issue with the poor internet connection of your new > residence, rather than an sf.net issue at all. > >> Am 25.04.12 15:17, schrieb edgar.sol...@web.de: >>> On 25.04.2012 20:08, Matthias Scholz wrote: >>>> Hi ede >>>>> On 23.04.2012 21:16, Matthias Scholz wrote: >>>>> >>>>>> Hi ede, >>>>>> >>>>>>> On 22.04.2012 21:07, Matthias Scholz wrote: >>>>>>> >>>>>>> >>>>>>>> Hi, >>>>>>>> >>>>>>>> I've commited a small fix in the AboutDialog. If the "readme.txt" file >>>>>>>> was not found, the Stacktrace was shown. This is bad if you run OJ as >>>>>>>> Webstart. Now nothing is viewed. >>>>>>>> >>>>>>>> >>>>>>> actually not finding the readme.txt right now is an error, except when >>>>>>> developing. that's why i left the stack there. >>>>>>> >>>>>>> >>>>>> That's right, it's an error. But to display this directly in the GUI is >>>>>> not the best way. This should be the task of the logging framework or >>>>>> the error dialog window. >>>>>> >>>>> >>>>> having the user to acknowledge the error is a useless exercise as he's >>>>> merely meant to notice it, not to be disturbed in his workflow. using the >>>>> log framework would probably lead to the error not being noticed by >>>>> anyone. >>>>> >>>> With "logging framework" i don't mean only the classic ones, such as >>>> log4j. I also meant the error dialog, which seems to be the better >>>> option than a raw stacktrace. (My subjective opinion) >>>> Can you life with the current code, that nothing is displayed if no >>>> readme.txt is found? At least until i found a solution for the webstart. >>>> >>> >>> sorry, i can't. not having the readme.txt there is an error. having the >>> user to have to acknowledge this in a dialog with ok button is no option >>> (see "disturbs work flow" above). >>> >>> please find a solution that differentiates between webstart and desktop oj. >>> you could as well simply live with it for the time being until you have >>> webstart in a fashion that's able to load external files? if not, then >>> simply write a routine that integrates readme.txt into your jar file on >>> webstart releasing. >>> >>> ..ede >>> >>> ..ede >>> > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Jump-pilot-devel mailing list > Jump-pilot-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
