Hi, Maxim! See answer below
6/18/2010 1:45 AM, Maxim: > Hi, > > thank you for taking the time! I really appreciate! > Your answer sounds intuitive, but I have a conceptual problem: so far > I use such links: > > https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg18&position=chr8:128817498-128822855&hgct_customText=track%20type=bigWig%20name=testwig%20description=%22a%20bigBed%20track%22%20visibility=full%20bigDataUrl=http://log:p...@mydomain/testbrowser/bigwigfile.bw > > That means for the section > bigDataUrl=http://log:p...@mydomain/testbrowser/bigwigfile.bw > > I'll have to substitute with this text file listing all individual BigWig > files. NO Try setting hgct_customText= to the url-encoded url of a text page that lists all of the tracks to be loaded (or lists URLs to other text pages containing them). By the way, the idea was for you to use https://user:password@ in the bigDataUrl, not in the url to genome.ucsc.edu. Here are some things to think about: You can have one page that has one track. You can have one page that has several tracks. You can have one page that has one url to another page that has one or more tracks. You can have one page that has several tracks or urls to pages that contain several tracks or urls to pages that contain ... By the way, the various page(s) you might have can be stored in your own secure area, and the URLs to them are able to use https, user, password. The URL to genome.ucsc.edu is the only URL where you can't use https, mostly because of the our current cert problem. Although we are considering getting a proper cert for our site. -Galt > > But what about the stuff before the actual link to the file, telling > about the track, name and description? When I have 10 BigWig files > they will most likely be not characterized by a single argument, or > can I just leave the argument section? > > I tested the following: > 1. I left the bigwig specifications in the original link but listed > the actual links (in a file links.txt) to the files just like > > http://log:p...@mydomain/testbrowser/bigwigfile.bw > > Then I get an error: links.txt (the list with the links) is not a bigwig file! > > 2. I deleted all the arguments from the link that calls the links.txt > but listed the complete links for the bigwigfiles within links.txt: > https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg18&position=chr8:128817498-128822855&hgct_customText=http://log:p...@mydomain/testbrowser/links.txt > > does not work either. > > 3: > I just left the hgct_custom BigData statement: > https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg18&position=chr8:128817498-128822855&hgct_customText=bigDataUrl=http://log:p...@mydomain/testbrowser/links.txt > > and provided the links in links.txt like > http://log:p...@mydomain/testbrowser/bigwigfile.bw > > Then I get: > Unrecognized format line 1 of custom track: > > Now I'm really confused. I can't think about an additional alternative. > > Maxim > > > 2010/6/17 Galt Barber<[email protected]>: >> >> You can use https. >> We are probably going to get a trusted SSH certificate set up soon. >> You might have a way to create an exception for our URL in the meantime >> via configuration of your browser or OS. >> >> As far as loading mulitple tracks, >> simply create a text document that >> lists each track in turn like this, >> and then do your usual trick to >> cause that url to get loaded. >> >> The customtrack loader can include other >> URLs, somewhat analogously to the C include files. >> >> Simply list the ones you want to be loaded together: >> >> Here is an example of the text my page contains: >> >> https:/user:passw...@server/mycustomtrack1 >> https:/user:passw...@server/mycustomtrack2 >> https:/user:passw...@server/mycustomtrack3 >> >> The user and password are optional depending >> on your use. However the point is that you >> can simply make a list, and it will recursively >> fetch anything that starts with http://, https://, or ftp://. >> >> -Galt >> >> Ar 6/17/2010 4:31 AM, scríobh Maxim: >>> >>> Hi, >>> >>> I managed to display remote BigWig files hosted on my Server in UCSC >>> Browser. Right now I have a simple HTML document displaying the links. >>> As there are multiple files, the user has to click the individual >>> link, then go back and hit the others one by one. Recently I came >>> across a mailing list thread that explained that there is a way to >>> upload multiple files at once. As far as I remember, you have to >>> generate a text file that includes all the given URLs. Unfortunately >>> after many hours searching (I obviously have cleared my Browser cache >>> in between) I still cannot find this post. On the long run I'd like to >>> have a little Website with radiobuttons next to the respective >>> filenames present in the given folder so the user can just click what >>> he/she likes to see and batch upload everything selected. >>> >>> Is this possible at all? >>> >>> Additionally the files are in a password protected folder. I've >>> learned how to pass the login details with the URL link but wonder how >>> to protect the links. What will I have to do in order to use https >>> instead of http. Just providing https in the URL results in a browser >>> failure that the certificate cannot be trusted. >>> >>> Best regards >>> Maxim >>> _______________________________________________ >>> Genome maillist - [email protected] >>> https://lists.soe.ucsc.edu/mailman/listinfo/genome >> >> _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
