This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
Review request for Viewer.
By MartinRJ Fayray.
Description
Testing
Bugs:
VWR-23741
Diffs
|
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
Review request for Viewer.
By MartinRJ Fayray.
|
Updated March 6, 2012, 10:31 p.m. Summary (updated)
|
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
Review request for Viewer.
By MartinRJ Fayray.
|
Updated March 6, 2012, 11:15 p.m. |
Description
Testing
Bugs:
VWR-23741
|
Diffs (updated)
|
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
Review request for Viewer.
By MartinRJ Fayray.
|
Updated March 24, 2012, 5:33 a.m. Changes
Summary (updated)
Description (updated)
|
|
Testing
|
Bugs:
STORM-1818
Diffs (updated)
|
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
It seems to me that the better fix would be to support the URLs, as you say some TPVs do, that are currently causing trouble.
- Jonathan
On March 24th, 2012, 5:33 a.m., MartinRJ Fayray wrote:
Review request for Viewer.
By MartinRJ Fayray.
Updated March 24, 2012, 5:33 a.m. |
Description |
Testing
Bugs:
STORM-1818
Diffs |
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
On March 24th, 2012, 7:22 a.m., Jonathan Yap wrote:
It seems to me that the better fix would be to support the URLs, as you say some TPVs do, that are currently causing trouble.
You can still login to non-Linden Lab grids via the --login URI parameter https://wiki.secondlife.com/wiki/Viewer_parameters But the grid selection combo doesn't support adding custom grids with a 'corrupt' format at all, as it seems the whole viewer is NOT designed for non-Linden Lab grids, see here: https://jira.secondlife.com/browse/VWR-28570?focusedCommentId=315423&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-315423 The hack to support these non-Linden Lab grid-format-urls as suggested here is an attempt to try parsing input data of any given format - be it valid or not - and a guessing game what the users input might mean, and that's in my honest opinion more of a 'political' question for the Linden Lab managment, and not a crash-fix-issue: https://jira.secondlife.com/browse/STORM-1818?focusedCommentId=314515&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-314515 Please test the fix/ship it. Thanks in advance. Kind regards, MartinRJ
- MartinRJ
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
I would like to point out that having all those linden-internal grids in there is utterly useless unless you are inside the linden lab network. How about adding a proper grid manager that supports actually accessible grids instead?
- Lance
On March 24th, 2012, 5:33 a.m., MartinRJ Fayray wrote:
Review request for Viewer.
By MartinRJ Fayray.
Updated March 24, 2012, 5:33 a.m. |
Description |
Testing
Bugs:
STORM-1818
Diffs |
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
On March 24th, 2012, 8:08 a.m., Lance Corrimal wrote:
I would like to point out that having all those linden-internal grids in there is utterly useless unless you are inside the linden lab network. How about adding a proper grid manager that supports actually accessible grids instead?
@Lance this is only a review of my crash fix, no public discussion. It is only about this severe bug (abnormal program termination). This is not a public user group meeting where we discuss features, please bear in mind that you just demanded me - a volunteer contributor - to add a proper grid manager: "How about adding a proper grid manager"? The answer is: no! Of course not!
- MartinRJ
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
Thanks for looking at this issue, MartinRJ. The only thing that give me pause for thought is - if the crash is coming from an uncaught exception, the 'right' thing to do seems to be to catch the exception and show the notification you've created then, aborting the login. Any clear reason why this isn't the approach? Does pre-validation give a better user experience? Cheers.
- Tofu
On March 24th, 2012, 5:33 a.m., MartinRJ Fayray wrote:
Review request for Viewer.
By MartinRJ Fayray.
Updated March 24, 2012, 5:33 a.m. |
Description |
Testing
Bugs:
STORM-1818
Diffs |
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
On March 25th, 2012, 3:13 a.m., Tofu Buzzard wrote:
Thanks for looking at this issue, MartinRJ. The only thing that give me pause for thought is - if the crash is coming from an uncaught exception, the 'right' thing to do seems to be to catch the exception and show the notification you've created then, aborting the login. Any clear reason why this isn't the approach? Does pre-validation give a better user experience? Cheers.
@Tofu: you want to remove the un-supported input string from the combo-box and have a valid grid selected instead. Otherwise, if you'd do what you've mentioned what would be the 'right' thing, and only catch the exception, if the user presses Login after that notification he'd get the next error.
- MartinRJ
This is an automatically generated e-mail. To reply, visit: http://codereview.secondlife.com/r/563/ |
Ship it!
Ah, that makes things much clearer. Yes, that would be lame.
- Tofu
On March 24th, 2012, 5:33 a.m., MartinRJ Fayray wrote:
Review request for Viewer.
By MartinRJ Fayray.
Updated March 24, 2012, 5:33 a.m. |
Description |
Testing
Bugs:
STORM-1818
Diffs |