FYI: ViewCVS

FYI: ViewCVS

[Tom Tromey]

ViewCVS is a reimplementation of cvsweb in Python:

    http://www.lyra.org/cgi-bin/viewcvs.cgi/viewcvs/

It might actually be actively maintained.  (My queries to the cvsweb
maintainers have gone unanswered.)  Maybe someday we should switch to
this.  I'm going to ask him about the `NONE' problem -- if ViewCVS
solves that then it might be reason enough to change.

Tom

Re: FYI: ViewCVS

[Jason Molenda]

On Sat, Aug 19, 2000 at 12:59:04PM -0600, Tom Tromey wrote:
> ViewCVS is a reimplementation of cvsweb in Python:

Yah, I'd seen that before - IIRC they were about on par with cvsweb's
level of features a while ago.

You can always install both of them.

BTW I believe there was a bugtraq sort of report about some possible
exploits in the older cvsweb version we're running on sourceware.  It'd
probably be good to grab the latest and install it some time - I'm using
it at Yahoo with no problems.  (I keep meaning to do it myself but have
yet to actually get around to it)

> I'm going to ask him about the `NONE' problem -- if ViewCVS
> solves that then it might be reason enough to change.

I kind of doubt it, unless it's parsing the RCS files on its own.
cvsweb uses rdiff to do the diff-without-a-checkout, and rdiff is
implemented within cvs and is missing many of the features of the
normal diff, making any kind of reasonable workaround for the NONE
diffs impossible.  (at least that's the conclusion I came to when
I looked at fixing this a couple of months ago)


Jason

Re: FYI: ViewCVS

[Tom Tromey]

Jason> I kind of doubt it, unless it's parsing the RCS files on its
Jason> own.  cvsweb uses rdiff to do the diff-without-a-checkout, and
Jason> rdiff is implemented within cvs and is missing many of the
Jason> features of the normal diff, making any kind of reasonable
Jason> workaround for the NONE diffs impossible.  (at least that's the
Jason> conclusion I came to when I looked at fixing this a couple of
Jason> months ago)

Me too, but maybe they are energetic enough to fix rdiff, or something
like that.

Tom

FYI: ViewCVS

[Tom Tromey]

ViewCVS is a reimplementation of cvsweb in Python:

    http://www.lyra.org/cgi-bin/viewcvs.cgi/viewcvs/

It might actually be actively maintained.  (My queries to the cvsweb
maintainers have gone unanswered.)  Maybe someday we should switch to
this.  I'm going to ask him about the `NONE' problem -- if ViewCVS
solves that then it might be reason enough to change.

Tom

Re: FYI: ViewCVS

[Jason Molenda]

On Sat, Aug 19, 2000 at 12:59:04PM -0600, Tom Tromey wrote:
> ViewCVS is a reimplementation of cvsweb in Python:

Yah, I'd seen that before - IIRC they were about on par with cvsweb's
level of features a while ago.

You can always install both of them.

BTW I believe there was a bugtraq sort of report about some possible
exploits in the older cvsweb version we're running on sourceware.  It'd
probably be good to grab the latest and install it some time - I'm using
it at Yahoo with no problems.  (I keep meaning to do it myself but have
yet to actually get around to it)

> I'm going to ask him about the `NONE' problem -- if ViewCVS
> solves that then it might be reason enough to change.

I kind of doubt it, unless it's parsing the RCS files on its own.
cvsweb uses rdiff to do the diff-without-a-checkout, and rdiff is
implemented within cvs and is missing many of the features of the
normal diff, making any kind of reasonable workaround for the NONE
diffs impossible.  (at least that's the conclusion I came to when
I looked at fixing this a couple of months ago)


Jason

Re: FYI: ViewCVS

[Tom Tromey]

Jason> I kind of doubt it, unless it's parsing the RCS files on its
Jason> own.  cvsweb uses rdiff to do the diff-without-a-checkout, and
Jason> rdiff is implemented within cvs and is missing many of the
Jason> features of the normal diff, making any kind of reasonable
Jason> workaround for the NONE diffs impossible.  (at least that's the
Jason> conclusion I came to when I looked at fixing this a couple of
Jason> months ago)

Me too, but maybe they are energetic enough to fix rdiff, or something
like that.

Tom