What is up with the multiple dataset selection for workflows
Hi, There seem to be some undocumented changes to the way multiple datasets are provided to workflows. There seems to be some loss of functionality, since I cannot figure out how to select ALL files in the window... I can select individual files, but not all files that I can see.. It used to be possible to simply use the SHIFT click manner to select the whole list, but even the <PRESS ENTER> to select all does not seem to work...I also don't seem to get the search function to work... There is also a mysterious link icon that appears that looks like a chain of some sort, but is too small and does not have any mouse over help or antything... I suspect this is there for supporting paired end sequencing (WAY overdue) so you can match _R1 with _R2, but there is nothing in the news brief about this, but this option seems broken to me... Thanks Thon
To filter and select all, just click the select2 box with the dataset list, type to filter the list, and then either press once for the first match or hold enter for more/all. This isn't optimal, but as you'd guessed it's incremental work towards having more control over matching multiple inputs for workflows (and individual tools, eventually). On Wed, Apr 3, 2013 at 9:44 PM, Anthonius deBoer <thondeboer@me.com> wrote:
Hi,
There seem to be some undocumented changes to the way multiple datasets are provided to workflows. There seems to be some loss of functionality, since I cannot figure out how to select ALL files in the window... I can select individual files, but not all files that I can see..
It used to be possible to simply use the SHIFT click manner to select the whole list, but even the <PRESS ENTER> to select all does not seem to work...I also don't seem to get the search function to work...
There is also a mysterious link icon that appears that looks like a chain of some sort, but is too small and does not have any mouse over help or antything...
I suspect this is there for supporting paired end sequencing (WAY overdue) so you can match _R1 with _R2, but there is nothing in the news brief about this, but this option seems broken to me...
Thanks
Thon
___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
OK...I see what is going on here... But if you have to do 100 paired end sequencing runs, you'd be pressing enter 100 times.. You can't just hold ENTER since the 101st ENTER will actually launch the workflow (I am using CHROME) and since I have not set the _R2 data set, it will launch incorrectly and I have to cancel the whole 100 histories... Is there a way to turn this feature off? It is worse than not having it at all if it is implemented like this.... Thanks Thon On Apr 04, 2013, at 05:16 AM, Dannon Baker <dannon.baker@gmail.com> wrote: To filter and select all, just click the select2 box with the dataset list, type to filter the list, and then either press once for the first match or hold enter for more/all. This isn't optimal, but as you'd guessed it's incremental work towards having more control over matching multiple inputs for workflows (and individual tools, eventually). On Wed, Apr 3, 2013 at 9:44 PM, Anthonius deBoer <thondeboer@me.com> wrote: Hi, There seem to be some undocumented changes to the way multiple datasets are provided to workflows. There seems to be some loss of functionality, since I cannot figure out how to select ALL files in the window... I can select individual files, but not all files that I can see.. It used to be possible to simply use the SHIFT click manner to select the whole list, but even the <PRESS ENTER> to select all does not seem to work...I also don't seem to get the search function to work... There is also a mysterious link icon that appears that looks like a chain of some sort, but is too small and does not have any mouse over help or antything... I suspect this is there for supporting paired end sequencing (WAY overdue) so you can match _R1 with _R2, but there is nothing in the news brief about this, but this option seems broken to me... Thanks Thon ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
Is there a way to turn this feature off? It is worse than not having it at all if it is implemented like this....
Not sure I get what you want to turn off. You could just use the dropdown like normal? Like I mentioned before, I know the current implementation isn't ideal, it's an intermediate step attempting to deal with the select2 changes while we work on a true linked input widget.
I don't know, these are the same reservations I initially had as well. It seems that select2 could simply be disabled for multiple select boxes or for these multiple select boxes. My thought was with the fixes from pull request 136 that select2 was at worst a side-grade for multiple selection, but Anthonius makes a semi-compelling case that it is actually a downgrade - I have done the accidental submission thing myself. -John On Thu, Apr 4, 2013 at 12:22 PM, Dannon Baker <dannon.baker@gmail.com> wrote:
Is there a way to turn this feature off? It is worse than not having it at all if it is implemented like this....
Not sure I get what you want to turn off. You could just use the dropdown like normal? Like I mentioned before, I know the current implementation isn't ideal, it's an intermediate step attempting to deal with the select2 changes while we work on a true linked input widget.
___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
participants (3)
-
Anthonius deBoer -
Dannon Baker -
John Chilton