Re: NCC#2003092385 [db-wg] uodate failure
Hi Randy, thanks for bringing this up! [ the "keyword noise" had me confused too, a short while ago :-( ] I am wondering whether "inline" error messages are such a good idea? I guess there's a good chance that they remain in the updated and re-sent message. That could give us another set of failure modes. _I_ would probably go for a more descriptive error message at the end, including the _attribute_ that made the reference, _and_ the _string_ used as a reference. -WW ______________________________________________________________________ From: Randy Bush <randy@psg.com> To: RIPE Database Administration <ripe-dbm@ripe.net> CC: db-wg@ripe.net Subject: Re: NCC#2003092385 [db-wg] uodate failure Date: Mon, 8 Sep 2003 06:07:48 -0700 btw, to be very clear, and as a once-upon-a-time compiler hacker who would have been severely excoriated for poor diagnostic reporting, what i would have liked to have received would be something along the lines of Modify FAILED: [aut-num] AS3130 ***Info: Authorisation passed ***Info: Syntax check passed aut-num: AS3130 as-name: RGNET-3130 descr: RGnet/PSGnet admin-c: RB366 ^ unknown object referenced tech-c: RB366 ^ unknown object referenced import: from AS1239 accept ANY import: from AS2914 accept ANY import: from AS3927 accept AS3927 import: from AS3970 accept AS3970 import: from AS4128 accept AS4128 import: from AS101 accept AS-PNW-GIGAPOP import: from AS11726 accept AS-ACUMEN import: from AS12111 accept AS-OEN import: from AS16713 accept AS-NOANET-PEERING import: from AS2044 accept AS2044 import: from AS3491 accept AS-CAIS import: from AS4513 accept AS-GLOBIX import: from AS4565 accept AS-EPOCH import: from AS6456 accept AS6456 import: from AS7610 accept AS-SINGAREN import: from AS8075 accept AS8075 export: to AS1239 announce AS-RGNET export: to AS2914 announce AS-RGNET export: to AS3927 announce ANY export: to AS3970 announce ANY export: to AS4128 announce ANY export: to AS101 announce AS-RGNET export: to AS11726 announce AS-RGNET export: to AS12111 announce AS-RGNET export: to AS16713 announce AS-RGNET export: to AS2044 announce AS-RGNET export: to AS3491 announce AS-RGNET export: to AS4513 announce AS-RGNET export: to AS4565 announce AS-RGNET export: to AS6456 announce AS-RGNET export: to AS7610 announce AS-RGNET export: to AS8075 announce AS-RGNET notify: rw@rg.net mnt-by: MAINT-RGNET changed: randy@psg.com 20030905 source: RIPE ***Info: Authorisation for [aut-num] AS3130 using mnt-by: authenticated by: MAINT-RGNET randy --------------------------------------------------------------------------------
I am wondering whether "inline" error messages are such a good idea? I guess there's a good chance that they remain in the updated and re-sent message. That could give us another set of failure modes.
do you send the diagnostic output back to a compiler?
_I_ would probably go for a more descriptive error message at the end, including the _attribute_ that made the reference, _and_ the _string_ used as a reference.
then have line and character position numbers. you _really_ need a completely unambiguous trail to the offending text randy
On Monday, Sep 8, 2003, at 18:57 Europe/Dublin, Randy Bush wrote:
do you send the diagnostic output back to a compiler?
If you call auto-dbm a compiler! 8-) One could use "###" instead of "***", and "#" instead of a blank line to protect against the kind of failure mode Wilfried suggests. Do ppl really correct (sic) and re-submit failure responses so carelessly? Seriously, though: it seems the developers have the "diagnostic message" issue taken well on board. The only detail I'm not too clear about is the timescale. Best, Niall
On Tuesday, September 9, 2003, at 10:50 AM, Niall O'Reilly wrote:
On Monday, Sep 8, 2003, at 18:57 Europe/Dublin, Randy Bush wrote:
do you send the diagnostic output back to a compiler?
If you call auto-dbm a compiler! 8-)
One could use "###" instead of "***", and "#" instead of a blank line to protect against the kind of failure mode Wilfried suggests.
Do ppl really correct (sic) and re-submit failure responses so carelessly?
Yes, absolutely. It is great fun to browse the DB logs on a rainy afternoon :)
Seriously, though: it seems the developers have the "diagnostic message" issue taken well on board. The only detail I'm not too clear about is the timescale.
I reckon the group manager is on holidays, so we should give them a few days to provide a response to this one, but it would indeed be good to have one. Joao
participants (4)
-
Joao Luis Silva Damas -
Niall O'Reilly -
Randy Bush -
Wilfried Woeber, UniVie/ACOnet