From markw at illuminae.com Fri Jan 26 17:15:51 2007 From: markw at illuminae.com (Mark Wilkinson) Date: Fri, 26 Jan 2007 14:15:51 -0800 Subject: [MOBY-l] A message to ensure transparency Message-ID: Hi All, we're starting a research project in my lab that looks at better ways to discover appropriate web services and combinations of web services. One way we want to pursue is to log MOBY Central usage such that we can better prioritize the output of a findService call to MOBY Central (e.g. 9/10 people chose THIS service as their next step). To do this effectively I will be adding a "listener" to MOBY Central such that it records and catalogues the retrieveService requests from individual IP addresses in order to track usage behaviour and make better search outputs. I will hash the IP address so that I cannot easily go back and look-up who made the request - it isn't 100% private/secure, but frankly we're not interested in who requested the service, only in which service they requested immediately before and afterwards... besides, I suspect that anyone who was *that* concerned about privacy would have set-up their own mirror of MOBY Central :-) In any case, I wanted to make sure that there were no violent objections to me doing this before I do it. Please let me know, and if I hear no objections I'll add a line into the "Policy" document associated with the Vancouver MOBY Central on the homepage. Cheers all! M -- -- Mark Wilkinson Assistant Professor, Dept. Medical Genetics University of British Columbia PI Bioinformatics iCAPTURE Centre, St. Paul's Hospital ***CONFIDENTIALITY NOTICE*** This electronic message is intended only for the use of the addressee and may contain information that is privileged and confidential. Any dissemination, distribution or copying of this communication by unauthorized individuals is strictly prohibited. If you have received this communication in error, please notify the sender immediately by reply e-mail and delete the original and all copies from your system. From martin.senger at gmail.com Fri Jan 26 18:14:55 2007 From: martin.senger at gmail.com (Martin Senger) Date: Fri, 26 Jan 2007 17:14:55 -0600 Subject: [MOBY-l] [MOBY-dev] A message to ensure transparency In-Reply-To: References: Message-ID: <4d93f07c0701261514o71587961p71dd2afe9c86e986@mail.gmail.com> Hi, I do not see any problem with privacy, it is not my cup of tea, anyway. But I would like to mention that the findService() requests do not need to be done at all when somebody calls a service, and - vice-versa - many/many times a findService() request is done when Dashboard local cache is updated, without calling any service at all. What I am saying is that the findService() is a pure way how to measure service calls. Cheers, Martin -- Martin Senger email: martin.senger at gmail.com skype: martinsenger From gordonp at ucalgary.ca Fri Jan 26 18:19:40 2007 From: gordonp at ucalgary.ca (Paul Gordon) Date: Fri, 26 Jan 2007 16:19:40 -0700 Subject: [MOBY-l] [MOBY-dev] A message to ensure transparency In-Reply-To: <4d93f07c0701261514o71587961p71dd2afe9c86e986@mail.gmail.com> References: <4d93f07c0701261514o71587961p71dd2afe9c86e986@mail.gmail.com> Message-ID: <45BA8C8C.4000308@ucalgary.ca> Hi, I was just about to say the same thing (assuming you meant "poor", not "pure"). Seahawk caches findService calls, and make findService calls for data that isn't used (depending on the choice the user makes). To study this well, you need to embed this tracking in a client program. > Hi, > > I do not see any problem with privacy, it is not my cup of tea, anyway. > > But I would like to mention that the findService() requests do not need to > be done at all when somebody calls a service, and - vice-versa - many/many > times a findService() request is done when Dashboard local cache is updated, > without calling any service at all. What I am saying is that the > findService() is a pure way how to measure service calls. > > Cheers, > Martin > > From markw at illuminae.com Fri Jan 26 19:02:19 2007 From: markw at illuminae.com (Mark Wilkinson) Date: Fri, 26 Jan 2007 16:02:19 -0800 Subject: [MOBY-l] [MOBY-dev] A message to ensure transparency In-Reply-To: <45BA8C8C.4000308@ucalgary.ca> References: <4d93f07c0701261514o71587961p71dd2afe9c86e986@mail.gmail.com> <45BA8C8C.4000308@ucalgary.ca> Message-ID: Hi both, I had considered these scenarios also. We can likely distinguish a call from Dashboard (many calls over a very short time) from calls that reflect someone designing or running a service. Even then, the retrieveService call isn't always made by the client software... in fact, I believe Eddie told me that the Java libraries don't use the WSDL documents that MOBY Central generates, so they likely don't make a retrieveService call *at all*, which means that we can't track those usages. Still... even if we can only track the usages from gbrowse_moby, it's a start :-) M On Fri, 26 Jan 2007 15:19:40 -0800, Paul Gordon wrote: > Hi, > > I was just about to say the same thing (assuming you meant "poor", not > "pure"). Seahawk caches findService calls, and make findService calls > for data that isn't used (depending on the choice the user makes). To > study this well, you need to embed this tracking in a client program. >> Hi, >> >> I do not see any problem with privacy, it is not my cup of tea, anyway. >> >> But I would like to mention that the findService() requests do not need >> to >> be done at all when somebody calls a service, and - vice-versa - >> many/many >> times a findService() request is done when Dashboard local cache is >> updated, >> without calling any service at all. What I am saying is that the >> findService() is a pure way how to measure service calls. >> >> Cheers, >> Martin >> >> > > _______________________________________________ > moby-l mailing list > moby-l at lists.open-bio.org > http://lists.open-bio.org/mailman/listinfo/moby-l -- -- Mark Wilkinson Assistant Professor, Dept. Medical Genetics University of British Columbia PI Bioinformatics iCAPTURE Centre, St. Paul's Hospital ***CONFIDENTIALITY NOTICE*** This electronic message is intended only for the use of the addressee and may contain information that is privileged and confidential. Any dissemination, distribution or copying of this communication by unauthorized individuals is strictly prohibited. If you have received this communication in error, please notify the sender immediately by reply e-mail and delete the original and all copies from your system. From martin.senger at gmail.com Fri Jan 26 19:17:18 2007 From: martin.senger at gmail.com (Martin Senger) Date: Fri, 26 Jan 2007 18:17:18 -0600 Subject: [MOBY-l] [MOBY-dev] A message to ensure transparency In-Reply-To: References: <4d93f07c0701261514o71587961p71dd2afe9c86e986@mail.gmail.com> <45BA8C8C.4000308@ucalgary.ca> Message-ID: <4d93f07c0701261617s1218d2aayec7ce0495d0483d4@mail.gmail.com> Yes, Paul, I mean poor, not pure. Sorry... We can likely distinguish a call > from Dashboard (many calls over a very short time) Better, if we agree on a USER_AGENT string that Dashboard (and/or Seahawk) will use in the HTTP header, and that your software can spot and ignore. Cheers, Martin -- Martin Senger email: martin.senger at gmail.com skype: martinsenger