You can publish the code to github and then post a link here ;) >From memory, I don't think it's quite as simple as changing couch_file. We write in blocks and reserve a bit of each block to indicate whether we're inside an attachment or not (to prevent us accidentally reading a couchdb database stored in an attachment). That plus the padding you'd need for a block cipher could explain your difficulty.
B. On 7 July 2011 10:43, Couchdb Couchdb <[email protected]> wrote: > The trace: > reserved slot index out of range > [error] [<0.625.0>] OS Process Error <0.32695.0> :: {os_process_error, > {exit_status,1}} > > reserved slot index out of range > [error] [<0.625.0>] OS Process Error <0.32689.0> :: {os_process_error, > {exit_status,1}} > reserved slot index out of range > [error] [<0.625.0>] OS Process Error <0.32692.0> :: {os_process_error, > {exit_status,1}} > > I can publish the code, but the server denies the mail > > 2011/7/7 Couchdb Couchdb <[email protected]> > >> Just run ./compile.sh script, to compile in debug modus(but change the >> paths) >> >> >> 2011/7/7 Couchdb Couchdb <[email protected]> >> >>> Here the trace: >>> >>> reserved slot index out of range >>> [error] [<0.625.0>] OS Process Error <0.32674.0> :: {os_process_error, >>> {exit_status,1}} >>> reserved slot index out of range >>> [error] [<0.625.0>] OS Process Error <0.32680.0> :: {os_process_error, >>> {exit_status,1}} >>> ... >>> >>> Encryption of couchdb is my bachelor thesis and also the organization is >>> interested in. >>> >>> >>> >>> >>> 2011/7/7 Jason Smith <[email protected]> >>> >>>> >>>> That is very interesting! >>>> >>>> It sounds like perhaps you are implementing encryption to meet a >>>> regulatory or other policy requirement. >>>> >>>> Obviously you have experience so I assume you have ruled out >>>> alternatives, such as running CouchDB on an encrypted filesystem? >>>> >>>> I wonder if it is possible for you to publish your work-in-progress? >>>> That will enable many people to help you investigate. If it is not >>>> possible, no doubt the community will do its best. (I see Benoit >>>> already sent a reply.) >>>> >>>> -- >>>> Iris Couch >>>> >>> >>> >> >
