Re: Please mention the CVS parameter has changed in the web page

[Chiaki <ishikawa@yk.rim.or.jp>]

Gerald Pfeifer wrote:
> On Wed, 21 Jan 2004, Kelley Cook wrote:
> 
>>I believe that our instructions for accessing Savannah are missing a few
>>steps which any GCC developers likely wouldn't have noticed, since they
>>were already required for our cvs write access.
>>
>>a) You must have SSHv2 installed.
>>b) You must set CVS_RSH="ssh".


The item (b) above was crucial.

The use of CVS_RSH="ssh" has finally resolved the
read-only access problem for CVS repository!

e.g. I use bash.

CVS_RSH="ssh"
export CVS_RSH
cvs -z 9 -d :ext:anoncvs@savannah.gnu.org:/cvsroot/gcc update
The authenticity of host 'savannah.gnu.org (199.232.41.3)' can't be 
established.
RSA key fingerprint is 80:5a:b0:0c:ec:93:66:29:49:7e:04:2b:fd:ba:2c:d5.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'savannah.gnu.org,199.232.41.3' (RSA) to the 
list of known hosts.
? check-num-percent.awk
? check-num-percent.sh
? typescript
    ... it is updating the local CVS source file finally! ...

(I DID apply for write access as Zack suggested, though.)

I should have known that the timing when my access began failing
was conincidental to the report of server(s) broken into, but
didn't realize / connect that I now needed "ssh" as the
underlying transport. Now come to think of it, it makes sense, but
it took me so long to update my local tree :-(

Thank you for pointing out this necessary step.

But I wonder now.

How many people WITHOUT CVS write commit priviledge have
tried the current CVS gcc source   *IF* they don't know
how to obtain it?

Does savannah.gnu.org host keep log of anoncvs login?
If so, I suspect the record would show that
the number of anonymous login might have
dropped drastically for the last 1.5 months.

Since there are dedicated people running the self-build of
GCC source tree, etc., I think the testing is at least done
with minimal assurance, but the testing against diverse
source may not have happened as it had done prior to the
switch to "ssh"-based access in the last 1.5 months.

Just a thought.

> This patch is mostly fine, thanks.
> 
> Some issues I noticed: Please do not link to any SSH implementation and
> use the term "SSH2 protocol".  (In fact, I'd simply omit that -- hardly
> any client these days does not support SSH2, so this may confuse more
> users than it helps.)
> 
> The update of the page you refer to happened in January this year.
> 
> Both items 3 and 4 are optional, in the sense that one doesn't need to
> do 3 to do 4.  I suggest to merge these two items and use the original
> wording (or something like that).
> 
> Gerald


-- 
int main(void){int j=2003;/*(c)2003 cishikawa. */
char t[] ="<CI> @abcdefghijklmnopqrstuvwxyz.,\n\"";
char *i ="g>qtCIuqivb,gCwe\np@.ietCIuqi\"tqkvv is>dnamz";
while(*i)((j+=strchr(t,*i++)-(int)t),(j%=sizeof t-1),
(putchar(t[j])));return 0;}/* under GPL */