I added an option to output the data in correct nzb format, though you still will need to copy the data to a file. I noticed that quite a few (maybe all?) nzb files produced by alt.binaries.nl do not import correctly into NewsPro, because the subject lines do not end with a segment count (xx/xx). I added a check to make sure they're appended if not already present.
So now the bookmarklet is not tied to one newsreader (wasn't my intention, though. It's because Alex talked about dropping clipboard support.)
Mike
NZB support added to bookmarklets
As i am really interested in really understanding this nzb file subject, would it be possible to explain it at some future time, when it has crystallized out? I mean, this is a new topic, and subject to constant change at the moment.
I want to seriously use the nzb files when it does not change so fast.
I want to seriously use the nzb files when it does not change so fast.
naturally it is better to write generic things, still it is no problem to leave the clipboard.
thanks for noticing the subject issue, indeed, it is the case, i'll fix that. the problem with such approach - in .nzb the designer forgot to define the number of parts, so if there are 5 parts and only first 2 are present the message will be still imported as complete if there is no hint in the subject.
what is needed is .nzb2 - to drop xml and to add number of parts in a straightforward format
thanks for noticing the subject issue, indeed, it is the case, i'll fix that. the problem with such approach - in .nzb the designer forgot to define the number of parts, so if there are 5 parts and only first 2 are present the message will be still imported as complete if there is no hint in the subject.
what is needed is .nzb2 - to drop xml and to add number of parts in a straightforward format
I don't think alt.binaries.nl indexes incompletes. If I do indexing, I'll probably use nget as a backend, and get headers from EN, and nget can determine which ones are complete, though I'd probably also display the incompletes, but marked as such.
The thing I'm not liking about nzb being xml is that you have to html-entity-ize stuff that will never be displayed in a browser, like & to &.
About my previous thinking about bandwidth issues, I decided they would compress very well "as is" in a zip, so that would solve that issue without resorting to some "proprietary" scheme.
The thing I'm not liking about nzb being xml is that you have to html-entity-ize stuff that will never be displayed in a browser, like & to &.
About my previous thinking about bandwidth issues, I decided they would compress very well "as is" in a zip, so that would solve that issue without resorting to some "proprietary" scheme.
it is a bit a mess not to define the number of parts, all parts are needed to be read first before adding records. the format is clearly messy.
.zml format is better in this respect (and it is gzipped), but it is a bit cryptic (not human readable). should be something in the middle, but we are stuck in the war between the index service providers.
an example of an ambiguous .nzb from www.binaries.nl
------------
...... [2/3]
part 1
part 2
------------
if to judge by subject it has 3 parts since we don't know it is nzb from binaries.nl.
it is a major issue with the usenet - most problems are artificial.
it should work here:
http://www.netwu.com/newspro/npr3742.zip
.zml format is better in this respect (and it is gzipped), but it is a bit cryptic (not human readable). should be something in the middle, but we are stuck in the war between the index service providers.
an example of an ambiguous .nzb from www.binaries.nl
------------
...... [2/3]
part 1
part 2
------------
if to judge by subject it has 3 parts since we don't know it is nzb from binaries.nl.
it is a major issue with the usenet - most problems are artificial.
it should work here:
http://www.netwu.com/newspro/npr3742.zip