On 03-Jul-2008, Raphael Geissert wrote: > On 03/07/2008, Raphael Hertzog <hertzog@debian.org> wrote: > > On Thu, 03 Jul 2008, Ben Finney wrote: > > > Why is the PTS page claiming there is an error processing the > > > watch file, and linking to a page that shows there was no such > > > error? > > The linked page says it could not check upstream. Which is still not an *error*. If upstream is not checkable, it's not an error if it's not checked. > Because it is "empty" (it does have some comment lines, but nothing > for uscan/DEHS). So DEHS thinks it has to check the watch file, but > since it is empty uscan doesn't complain and DEHS just reports that > it could not check it. Surely, if 'uscan' does not complain when using the watch file, that should satisfy DEHS. > DEHS is not intended to actually examine the watch files It should use 'uscan' for this, instead of inspecting the file itself. > so I will add a lintian check for empty watch files (hoping Russ > approves it) This would conflict with what lintian already does in the case of a *missing* watch file: it recommends that a watch file be created, and if upstream cannot be scanned, explain this in comments (making it "empty", if I understand you correctly). Instead of DEHS checking for empty watch files, it should rely on 'uscan' which already knows how to interpret them. > and prevent DEHS from listing packages with no errors from uscan in > the file being grabbed by the PTS. That would be good also. Thanks. -- \ “Holy knit one purl two, Batman!” —Robin | `\ | _o__) | Ben Finney <ben@benfinney.id.au>
Attachment:
signature.asc
Description: Digital signature