You are not logged in.
Pages: 1
hmmmm, this seems to be the best place for this.
i've been building packages per abs for quit sometime now for my amd k6-500 (i586).
the problem i am having for the past few versions of archstats is it compiles ok and archstats -r works good. i'm able to make the archstats.conf file ok. i am now running archstats-3.7-1
but when i run archstats -u i get this message:
Corrupted ArchStats Client
The ArchStats Server claims that this program (your ArchStats client) is
not valid. This can occur if you modify the client code, or are running a
very old version. It is recommended that you update your client and try
again.
(gauranteed no modification to the client code from me)
Also the top of the archstats.conf file looks like this:
# unique system id, do not edit!
#
SYSTEM_UID = ''
# on-line public reference to your system
#
SYSTEM_REF = ''
# Expected ArchStatsServer version number
#
SERVER_VERSION = ''
# Expected ArchStatsServer MD5 hash
#
SERVER_MD5 = ''
# system's hostname
#
SYSTEM_HOSTNAME = 'k6er-2'
seems to be something missing in the id dept.
thanks,
slyski
Joey: If a cow passes away from natural causes can I eat it.
Pheobe: Not if I get there first.
Offline
why not email Eric about this as he is responsible for the code for archstats?
AKA uknowme
I am not your friend
Offline
he told me that this is a non-i686 architectures problem and at this time no-one with an i586 or amd64 or other port will be able to participate in archstats.
this is what eric wrote:
You're running up against a 'security feature' in the
client. When you tell the client to update (send your
system info to the server), it computes its own md5sum
and version number and transmits that to the server as
part of its upload data. The server then checks the
client version against what it expects as the md5sum
and then allows the update to proceed, or sends back
a fail code. Your client receives the fail code and
that is what is causing the error message you're seeing.
Even though you haven't changed any source code, when
you recompile against another architecture, i586 in this
case, the resulting executable will possess a different
md5sum than the one that is shipped as an Arch package.
Admittedly, this is a weak attempt at security, but a
little is better than nothing, IMHO.
For now, I think you'll have to skip participating in
ArchStats with the i586. Once the port is more stable
and supported, then we can work on producing a
pre-compiled i586 version of archstats.
Thanks for bringing this to my attention. I'll have to
update the main archstats page and point out this problem
for non-i686 architectures. I know there are folks out
there trying to make an amd64 port too.
Best regards,
Eric
:cry:
slyski
Joey: If a cow passes away from natural causes can I eat it.
Pheobe: Not if I get there first.
Offline
bah archstats.
AKA uknowme
I am not your friend
Offline
Pages: 1