DIFF method

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

DIFF method

Werner Donné

Hi,

Shouldn't there be a DIFF method? Calculating the difference between
two versions may be a very specialised operation for some document
formats. The software or resources to accomplish it may not be available
to clients.

Regards,

Werner.
--
Werner Donné  --  Re
Engelbeekstraat 8
B-3300 Tienen
tel: (+32) 486 425803 e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Geoffrey M Clemm

The only interesting thing aobut a DIFF method is the format of its result.
There was such a wide variation in the different formats that diff tools
use for different types of resources that it didn't appear to be something
on which we could get consensus.  There are some standards for diff formats,
e.g., http://www.ietf.org/rfc/rfc3284.txt?number=3284 (The VCDIFF Generic
Differencing and Compression Data Format) and some new ones being
proposed, e.g., http://www.ietf.org/internet-drafts/draft-ietf-simple-xcap-diff-03.txt

Cheers,
Geoff

Werner wrote on 03/20/2006 04:28:34 AM:
> Shouldn't there be a DIFF method? Calculating the difference between
> two versions may be a very specialised operation for some document
> formats. The software or resources to accomplish it may not be available
> to clients.
Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Werner Donné

Geoffrey M Clemm wrote:
>
> The only interesting thing aobut a DIFF method is the format of its result.

The diff calculation itself can require software that can't be distributed
to clients, because of licensing limitations, programming language, resources
or whatever. Note that it can be more than just a diff of text files. I
will use DeltaXML, for example, which is specialised in calculating the diff
between two XML documents. It inserts mark-up in its own namespace in the
resulting document.

The format can also be converted on the fly, depending on the "Accept"
request header. I could convert it to PDF, again with tools that are only
available on the server.

> Cheers,
> Geoff

Regards,

Werner.

>
> Werner wrote on 03/20/2006 04:28:34 AM:
>> Shouldn't there be a DIFF method? Calculating the difference between
>> two versions may be a very specialised operation for some document
>> formats. The software or resources to accomplish it may not be available
>> to clients.

--
Werner Donné  --  Re
Engelbeekstraat 8
B-3300 Tienen
tel: (+32) 486 425803 e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Geoffrey M Clemm

I wasn't questioning the desirability (from a client's perspective) of
having a DIFF method, but was just pointing out that the challenge is
to come up with a standard result format that more than one server will
be willing/able to produce.

Cheers,
Geoff

Werner wrote on 03/20/2006 08:24:55 AM:

> Geoffrey M Clemm wrote:
> > The only interesting thing about a DIFF method is the format of its result.

> > There was such a wide variation in the different formats that diff tools
> > use for different types of resources that it didn't appear to be something

> > on which we could get consensus.  There are some standards for diff formats,

> > e.g., http://www.ietf.org/rfc/rfc3284.txt?number=3284 (
The VCDIFF Generic
> > Differencing and Compression Data Format) and some new ones being
> > proposed, e.g., http://www.ietf.org/internet-drafts/draft-ietf-simple-xcap-diff-03.txt

>
> The diff calculation itself can require software that can't be distributed
> to clients, because of licensing limitations, programming language, resources
> or whatever. Note that it can be more than just a diff of text files. I
> will use DeltaXML, for example, which is specialised in calculating the diff
> between two XML documents. It inserts mark-up in its own namespace in the
> resulting document.
>
> The format can also be converted on the fly, depending on the "Accept"
> request header. I could convert it to PDF, again with tools that are only
> available on the server.
>
> > Cheers,
> > Geoff
>
> Regards,
>
> Werner.
>
> >
> > Werner wrote on 03/20/2006 04:28:34 AM:
> >> Shouldn't there be a DIFF method? Calculating the difference between
> >> two versions may be a very specialised operation for some document
> >> formats. The software or resources to accomplish it may not be available
> >> to clients.
>
> --
> Werner Donné  --  Re
> Engelbeekstraat 8
> B-3300 Tienen
> tel: (+32) 486 425803   e-mail: [hidden email]
>
Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Werner Donné

Geoffrey M Clemm wrote:
>
> I wasn't questioning the desirability (from a client's perspective) of
> having a DIFF method, but was just pointing out that the challenge is
> to come up with a standard result format that more than one server will
> be willing/able to produce.

I don't see why there should be a standard result format. There are by
definition many, because there are many content types. For some a diff
may even not be possible. There can also be several result formats
for the same content type. That is a matter of representation.

>
> Cheers,
> Geoff

Regards,

Werner.

>
> Werner wrote on 03/20/2006 08:24:55 AM:
>
>> Geoffrey M Clemm wrote:
>> > The only interesting thing about a DIFF method is the format of its
> result.
>> > There was such a wide variation in the different formats that diff
> tools
>> > use for different types of resources that it didn't appear to be
> something
>> > on which we could get consensus.  There are some standards for diff
> formats,
>> > e.g., http://www.ietf.org/rfc/rfc3284.txt?number=3284 (The VCDIFF
> Generic
>> > Differencing and Compression Data Format) and some new ones being
>> > proposed, e.g.,
> http://www.ietf.org/internet-drafts/draft-ietf-simple-xcap-diff-03.txt
>>
>> The diff calculation itself can require software that can't be distributed
>> to clients, because of licensing limitations, programming language,
> resources
>> or whatever. Note that it can be more than just a diff of text files. I
>> will use DeltaXML, for example, which is specialised in calculating
> the diff
>> between two XML documents. It inserts mark-up in its own namespace in the
>> resulting document.
>>
>> The format can also be converted on the fly, depending on the "Accept"
>> request header. I could convert it to PDF, again with tools that are only
>> available on the server.
>>
>> > Cheers,
>> > Geoff
>>
>> Regards,
>>
>> Werner.
>>
>> >
>> > Werner wrote on 03/20/2006 04:28:34 AM:
>> >> Shouldn't there be a DIFF method? Calculating the difference between
>> >> two versions may be a very specialised operation for some document
>> >> formats. The software or resources to accomplish it may not be
> available
>> >> to clients.
>>
>> --
>> Werner Donné  --  Re
>> Engelbeekstraat 8
>> B-3300 Tienen
>> tel: (+32) 486 425803   e-mail: [hidden email]
>>

--
Werner Donné  --  Re
Engelbeekstraat 8
B-3300 Tienen
tel: (+32) 486 425803 e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Julian Reschke

Werner Donné wrote:

> Geoffrey M Clemm wrote:
>> I wasn't questioning the desirability (from a client's perspective) of
>> having a DIFF method, but was just pointing out that the challenge is
>> to come up with a standard result format that more than one server will
>> be willing/able to produce.
>
> I don't see why there should be a standard result format. There are by
> definition many, because there are many content types. For some a diff
> may even not be possible. There can also be several result formats
> for the same content type. That is a matter of representation.
>
>> Cheers,
>> Geoff
>
> Regards,
>
> Werner.

Werner,

a standard DIFF method is of little use unless there is at least one
general purpose diff format that clients can rely on being present.

Note that last time people tried to define an HTTP PATCH method (I think
Nov 2004 on the HTTP mailing list), we came across the very same issue.

Best regards, Julian

Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Werner Donné

Julian,

What would clients do with the format? Present the result?
Servers could generate HTML.

Regards,

Werner.

Julian Reschke wrote:

>
> Werner Donné wrote:
>> Geoffrey M Clemm wrote:
>>> I wasn't questioning the desirability (from a client's perspective) of
>>> having a DIFF method, but was just pointing out that the challenge is
>>> to come up with a standard result format that more than one server will
>>> be willing/able to produce.
>>
>> I don't see why there should be a standard result format. There are by
>> definition many, because there are many content types. For some a diff
>> may even not be possible. There can also be several result formats
>> for the same content type. That is a matter of representation.
>>
>>> Cheers,
>>> Geoff
>>
>> Regards,
>>
>> Werner.
>
> Werner,
>
> a standard DIFF method is of little use unless there is at least one
> general purpose diff format that clients can rely on being present.
>
> Note that last time people tried to define an HTTP PATCH method (I think
> Nov 2004 on the HTTP mailing list), we came across the very same issue.
>
> Best regards, Julian
>
>

--
Werner Donné  --  Re
Engelbeekstraat 8
B-3300 Tienen
tel: (+32) 486 425803 e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Julian Reschke

Werner Donné wrote:
> Julian,
>
> What would clients do with the format? Present the result?
> Servers could generate HTML.

I would expect that the response from a DIFF could be applied by a
client to an existing copy, effectively patching it. But that requires a
standard format...



Reply | Threaded
Open this post in threaded view
|

Re: DIFF method

Werner Donné

This is focused on flat files. For that you can extract both and
perform diff and patch. In the context of a document repository
you would do a merge. The diff in the server is necessary because
in a versioning system users want to be able to inspect changes
in the version history. The result of a DIFF is no different than
that of a GET. Only the processing is different.

Julian Reschke wrote:

> Werner Donné wrote:
>> Julian,
>>
>> What would clients do with the format? Present the result?
>> Servers could generate HTML.
>
> I would expect that the response from a DIFF could be applied by a
> client to an existing copy, effectively patching it. But that requires a
> standard format...
>
>
>

--
Werner Donné  --  Re
Engelbeekstraat 8
B-3300 Tienen
tel: (+32) 486 425803 e-mail: [hidden email]