hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Enis Soztutar <enis.soz.nu...@gmail.com>
Subject Re: [jira] Updated: (HADOOP-800) More improvements to DFS browsing WI
Date Tue, 15 May 2007 11:51:47 GMT
Hi, i have moved (x) in it's own issue : 
http://issues.apache.org/jira/browse/HADOOP-1370

Unfortunatelly i again was not able to reproduce (c). The resulting html 
from so called file viewer(actually browseBlock.jsp) does not contain 
any directives to refresh the page, may be you are not using the latest 
version of hadoop? Arkady, can you please send me the html of the page 
so that i can check again? normally for html documents contain
<meta http-equiv="refresh" content="600"> or some javascript to refresh 
the page, but as far as i can tell, browseBlock.jsp does not.

Coming to (d), the text from the file is wrapped in a html <textarea> . 
The number of rows and columns are required attributes to be given to 
this tag. The html looks like :

<textarea cols="100" rows="25" wrap="virtual" READONLY>

//text of the file

</textarea>

we cannot make cols stretch as wide as the window width(because we don't 
know!). We can either increase it to say 120 or make it a parameter, but 
the latter will seem ugly i think.

Another option like you suggested is to delete the textarea tag, but IMO 
it clearly seperates the file's content from the rest of the page, and 
also it allows for more compact display of long content.

long story short, I suggest we increase cols to 120 to attack (d).


Enis Soztutar wrote:
> Andrzej Bialecki wrote:
>> Arkady Borkovsky wrote:
>>
>>> (x) While here:  it would be so nice to add the CURRENT TIME to all 
>>> the pages.
>>> For naive users like myself, understanding universal time is very 
>>> difficult.  So knowing what the cluster thinks about current time 
>>> makes it so much easier to understand when a job has actually 
>>> started or ended....
>>
>> I concur - I work with clusters located in wildly different 
>> timezones, and it would be of tremendous help to see in the UI the 
>> machine's notion of the current time (which sometimes differs even 
>> from the current time in that timezone ;) ).
>>
>>
> I think we can move (x) to a seperate issue, since it is a general 
> requirement not only for dfs.
>
> For (c) and (d), i will check again the code to see what can be done.
>


Mime
View raw message