[Talk-GB] liam123's latest
Two changesets: http://www.openstreetmap.org/browse/changeset/2510163 reverted cleanly http://www.openstreetmap.org/browse/changeset/2510485 failed to revert "410 gone" - I suspect there was a node/way changed in the second changeset that was also in the first. The automatic reversion is getting a failure rate of abut half at the moment, so liam123 is steadily degrading map quality because nothing is being done to deal with the manual stuff. David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
[Talk-GB] liam123's latest
On 17 Sep 2009, at 13:23, David Earl wrote: > Two changesets: > > http://www.openstreetmap.org/browse/changeset/2510163 reverted cleanly > > http://www.openstreetmap.org/browse/changeset/2510485 failed to revert > "410 gone" - I suspect there was a node/way changed in the second > changeset that was also in the first. > > The automatic reversion is getting a failure rate of abut half at the > moment, so liam123 is steadily degrading map quality because nothing > is > being done to deal with the manual stuff. > Vandalism has already got to the top on my priority list (along with getting the license changed to avoid loosing too much data when it eventually happens). Who would join a 'talk-counter_vandalism' list or support its creation? Talk-gb and talk-transit (two lists I find very useful) are partly useful because they are focused. I believe a talk-counter_vandalism list would help provide the focus for the many many tools and approaches that could be developed to protect the work done in OSM to data and would result in better tools being created sooned. If we have 4-5 people in the UK then I will propose we request a list and promote it on other lists including talk. Regards, Peter > David > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
I've been thinking about this for a while. Is he doing anything (legally) wrong? Is he always connecting from the same IP address? Can the login script be amended to point suspect users to a "You've been dicking about with the map - we're going to let you continue in the hope that at some point you're going to mature into somebody who wants to be helpful rather than a pain in the backside" page? Presumably his IP address is known... Can somebody contact his ISP and get them to do something or forward a letter? Chances are that if it's a kid, mummy and daddy are paying for the Internet connection. Nick. ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17/09/2009 14:09, Peter Miller wrote: > Who would join a 'talk-counter_vandalism' list or support its creation? Yes. But can we call it something less judgemental: not all incorrect changes are vandalism, and people seeing their account names on such a list would be most depressing. talk-dubious-changes? or some such. David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17 Sep 2009, at 14:18, David Earl wrote: > On 17/09/2009 14:09, Peter Miller wrote: >> Who would join a 'talk-counter_vandalism' list or support its >> creation? > > Yes. But can we call it something less judgemental: not all > incorrect changes are vandalism, and people seeing their account > names on such a list would be most depressing. > > talk-dubious-changes? Sorry, I wasn't clear. The purpose of the list is to allow people to discuss and co-ordinate the creation of good counter-vandalism tools internationally, not to be a public flogging place for people who make mistakes or indeed who scribble on the map or do worse things. Vandalism and mistakes are different but need the same tools. If the list is focused on creating tools to respond to really bad malicious calculated vandalism then it will also provide the tools to recover from innocent mistakes as well. Possibly a different name would be clearer talk-Counter_vandalism_tools, but that is getting a bit long. Any other ideas or feedback? Regards, Peter > > or some such. > > David > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17/09/2009 14:30, Peter Miller wrote: > Possibly a different name would be clearer > > talk-Counter_vandalism_tools, but that is getting a bit long. Any other > ideas or feedback? talk-reversion-tools? ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Another one at 13:56, also failed to revert http://www.openstreetmap.org/browse/changeset/2511524 changeset upload failed: 412 Precondition Failed On 17/09/2009 13:23, David Earl wrote: > Two changesets: > > http://www.openstreetmap.org/browse/changeset/2510163 reverted cleanly > > http://www.openstreetmap.org/browse/changeset/2510485 failed to revert > "410 gone" - I suspect there was a node/way changed in the second > changeset that was also in the first. > > The automatic reversion is getting a failure rate of abut half at the > moment, so liam123 is steadily degrading map quality because nothing is > being done to deal with the manual stuff. > > David > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17/09/2009 14:35, David Earl wrote: > Another one at 13:56, also failed to revert > http://www.openstreetmap.org/browse/changeset/2511524 > changeset upload failed: 412 Precondition Failed Actually, on his page it says "still editing". Perhaps that's why and I can have another go as soon as it terminates. David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17 Sep 2009, at 14:30, David Earl wrote: > On 17/09/2009 14:30, Peter Miller wrote: >> Possibly a different name would be clearer >> talk-Counter_vandalism_tools, but that is getting a bit long. Any >> other ideas or feedback? > > talk-reversion-tools? Not very inclusive. I want it to also cover 'patrol' software to spot suspicious sorts of dicking around, such as a new contributor moving a bunch of nodes that have been in one place for a length of time. Looking for offensive phrases in tags. Users who have recently had work reverted who make new edits. Possibly methods of rating contributors with white lists for users who don't get reverted and have been editing for some time etc. Regards, Peter > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17/09/2009 14:38, David Earl wrote: > On 17/09/2009 14:35, David Earl wrote: >> Another one at 13:56, also failed to revert >> http://www.openstreetmap.org/browse/changeset/2511524 >> changeset upload failed: 412 Precondition Failed > > Actually, on his page it says "still editing". Perhaps that's why and I > can have another go as soon as it terminates. Now he has a second open changeset, while the first is still open: http://www.openstreetmap.org/browse/changeset/2511982 ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Thu, Sep 17, 2009 at 2:40 PM, David Earl wrote: > On 17/09/2009 14:38, David Earl wrote: >> On 17/09/2009 14:35, David Earl wrote: >>> Another one at 13:56, also failed to revert >>> http://www.openstreetmap.org/browse/changeset/2511524 >>> changeset upload failed: 412 Precondition Failed >> >> Actually, on his page it says "still editing". Perhaps that's why and I >> can have another go as soon as it terminates. > > Now he has a second open changeset, while the first is still open: > > http://www.openstreetmap.org/browse/changeset/2511982 > That's just because Potlatch doesn't close the changeset unless you manually close it. The server will shut down open changesets after 1 hour in inactivity. If you shut down Potlatch, then reopen it it'll start a new changeset. The 412 can happen any time you try to revert something which has been edited since... if he was still editing it's more than likely -- you might need to consider both changesets together to do a proper revert (assuming Frederick's tool lets you do that.. not sure it does). Dave ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
2009/9/17 Peter Miller : > > On 17 Sep 2009, at 14:30, David Earl wrote: > >> On 17/09/2009 14:30, Peter Miller wrote: >>> Possibly a different name would be clearer >>> talk-Counter_vandalism_tools, but that is getting a bit long. Any >>> other ideas or feedback? >> >> talk-reversion-tools? > > Not very inclusive. I want it to also cover 'patrol' software to spot > suspicious sorts of dicking around, such as a new contributor moving a > bunch of nodes that have been in one place for a length of time. > Looking for offensive phrases in tags. Users who have recently had > work reverted who make new edits. Possibly methods of rating > contributors with white lists for users who don't get reverted and > have been editing for some time etc. talk-moderation or talk-moderation_tools? Since this all falls under the umbrella of moderating the edits being made. -- Matt Williams http://milliams.com ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17 Sep 2009, at 14:48, Matt Williams wrote: > 2009/9/17 Peter Miller : >> >> On 17 Sep 2009, at 14:30, David Earl wrote: >> >>> On 17/09/2009 14:30, Peter Miller wrote: Possibly a different name would be clearer talk-Counter_vandalism_tools, but that is getting a bit long. Any other ideas or feedback? >>> >>> talk-reversion-tools? >> >> Not very inclusive. I want it to also cover 'patrol' software to spot >> suspicious sorts of dicking around, such as a new contributor >> moving a >> bunch of nodes that have been in one place for a length of time. >> Looking for offensive phrases in tags. Users who have recently had >> work reverted who make new edits. Possibly methods of rating >> contributors with white lists for users who don't get reverted and >> have been editing for some time etc. > > talk-moderation or talk-moderation_tools? Since this all falls under > the umbrella of moderating the edits being made. Sounds good to me. 'talk-moderation' is short, and makes it clear that it covers all aspects of both the technology and the associated social side of managing contributions. Some people may think it is the place to complain about every bad edit, but we can deflect those requests with a good wiki. The other question though:- 1) Would you support the existence of such a list? 2) Would you join it? If no one supports it or would join it from talk-gb then I won't progress the idea any more! Regards, Peter > > -- > Matt Williams > http://milliams.com > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Peter Miller wrote: > The other question though:- > > 1) Would you support the existence of such a list? > 2) Would you join it? Yes, I'd join. ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Thu, Sep 17, 2009 at 3:13 PM, Peter Miller wrote: > 'talk-moderation' is short, and makes it clear that it covers all > aspects of both the technology and the associated social side of > managing contributions. Some people may think it is the place to > complain about every bad edit, but we can deflect those requests with > a good wiki. > > The other question though:- > > 1) Would you support the existence of such a list? > 2) Would you join it? talk-moderation sounds like a good name. I like the idea of the list, so I'd support it's existence, but I doubt I'd join it, because I'm signed up to too many lists as it is. Russ ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
I'd join too -- but I doubt I would be of all that much help. JR 2009/9/17 Someoneelse > Peter Miller wrote: > > The other question though:- > > > > 1) Would you support the existence of such a list? > > 2) Would you join it? > > Yes, I'd join. > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Yet another one in progress: Dartford in Kent is being steadily ruined with random bridges and the like: http://www.openstreetmap.org/browse/changeset/2513003 He's spent essentially all day today making random edits across Kent. David On 17/09/2009 14:35, David Earl wrote: > Another one at 13:56, also failed to revert > http://www.openstreetmap.org/browse/changeset/2511524 > changeset upload failed: 412 Precondition Failed > > On 17/09/2009 13:23, David Earl wrote: >> Two changesets: >> >> http://www.openstreetmap.org/browse/changeset/2510163 reverted cleanly >> >> http://www.openstreetmap.org/browse/changeset/2510485 failed to revert >> "410 gone" - I suspect there was a node/way changed in the second >> changeset that was also in the first. >> >> The automatic reversion is getting a failure rate of abut half at the >> moment, so liam123 is steadily degrading map quality because nothing is >> being done to deal with the manual stuff. >> >> David >> >> ___ >> Talk-GB mailing list >> Talk-GB@openstreetmap.org >> http://lists.openstreetmap.org/listinfo/talk-gb >> > > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
... which also won't revert now its closed ("changeset upload failed:
409 Conflict")
On 17/09/2009 17:42, David Earl wrote:
> Yet another one in progress: Dartford in Kent is being steadily ruined
> with random bridges and the like:
> http://www.openstreetmap.org/browse/changeset/2513003
>
> He's spent essentially all day today making random edits across Kent.
>
> David
>
> On 17/09/2009 14:35, David Earl wrote:
>> Another one at 13:56, also failed to revert
>> http://www.openstreetmap.org/browse/changeset/2511524
>> changeset upload failed: 412 Precondition Failed
>>
>> On 17/09/2009 13:23, David Earl wrote:
>>> Two changesets:
>>>
>>> http://www.openstreetmap.org/browse/changeset/2510163 reverted cleanly
>>>
>>> http://www.openstreetmap.org/browse/changeset/2510485 failed to revert
>>> "410 gone" - I suspect there was a node/way changed in the second
>>> changeset that was also in the first.
>>>
>>> The automatic reversion is getting a failure rate of abut half at the
>>> moment, so liam123 is steadily degrading map quality because nothing is
>>> being done to deal with the manual stuff.
>>>
>>> David
>>>
>>> ___
>>> Talk-GB mailing list
>>> Talk-GB@openstreetmap.org
>>> http://lists.openstreetmap.org/listinfo/talk-gb
>>>
>>
>> ___
>> Talk-GB mailing list
>> Talk-GB@openstreetmap.org
>> http://lists.openstreetmap.org/listinfo/talk-gb
>>
>
>
> ___
> Talk-GB mailing list
> Talk-GB@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk-gb
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Hi, We seem unable to contact Liam123 via normal channels. This means we have to think of other ways of contacting members who continually put mistakes into the dataset. Is it not possible to put a flag on his account, so when he logs in he is told what he has been doing and asked to explain himself? Preferably with a form box that forwards the message to this list? Otherwise we are going to go around in circles trying to undo his work, rather than solving the problem. We need to get him on our side rather than just having him toy with something we feel is very important. Ciarán ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 17/09/09 17:58, Ciarán Mooney wrote: > We seem unable to contact Liam123 via normal channels. This means we > have to think of other ways of contacting members who continually put > mistakes into the dataset. Which is why he has been referred to the DWG who have other means of contacting people at their disposal, and the power to take action if people do not respond. > Is it not possible to put a flag on his account, so when he logs in he > is told what he has been doing and asked to explain himself? > Preferably with a form box that forwards the message to this list? That is not an ability we have at the moment, no. Tom -- Tom Hughes (t...@compton.nu) http://www.compton.nu/ ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
David,
David Earl wrote:
> ... which also won't revert now its closed ("changeset upload failed:
> 409 Conflict")
I also hear you talk about "automatic reversion" which "gets a failure
rate of about half". Are you sure that you know what you are doing?
Because if not then there's a risk that you make things worse.
Pardon if I'm doing you injustice but it sounds a bit as if you were
just pressing buttons without having given the issue much thought. Have
you tried to combine the edits you want to revert and to sort them in a
meaningful way for reverting? Have you identified those that don't
revert and split up his changesets? etc.
It is certainly not a good idea to blindly run the revert script on
anything liam123 is doing and if that is what you're attempting then I
would urge you to stop that and develop a more sophisticated approach.
Bye
Frederik
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
I would be tempted to go for a name such as talk-quality or similar for the list as that isn't as prescriptive as talk-moderation I would subscribe, time to read and reply is a wholly different question :) Cheers Chris On 17/09/09 15:13, Peter Miller wrote: > On 17 Sep 2009, at 14:48, Matt Williams wrote: > > >> 2009/9/17 Peter Miller: >> >>> On 17 Sep 2009, at 14:30, David Earl wrote: >>> >>> On 17/09/2009 14:30, Peter Miller wrote: > Possibly a different name would be clearer > talk-Counter_vandalism_tools, but that is getting a bit long. Any > other ideas or feedback? > talk-reversion-tools? >>> Not very inclusive. I want it to also cover 'patrol' software to spot >>> suspicious sorts of dicking around, such as a new contributor >>> moving a >>> bunch of nodes that have been in one place for a length of time. >>> Looking for offensive phrases in tags. Users who have recently had >>> work reverted who make new edits. Possibly methods of rating >>> contributors with white lists for users who don't get reverted and >>> have been editing for some time etc. >>> >> talk-moderation or talk-moderation_tools? Since this all falls under >> the umbrella of moderating the edits being made. >> > Sounds good to me. > > 'talk-moderation' is short, and makes it clear that it covers all > aspects of both the technology and the associated social side of > managing contributions. Some people may think it is the place to > complain about every bad edit, but we can deflect those requests with > a good wiki. > > The other question though:- > > 1) Would you support the existence of such a list? > 2) Would you join it? > > If no one supports it or would join it from talk-gb then I won't > progress the idea any more! > > > Regards, > > > Peter > > > >> -- >> Matt Williams >> http://milliams.com >> >> ___ >> Talk-GB mailing list >> Talk-GB@openstreetmap.org >> http://lists.openstreetmap.org/listinfo/talk-gb >> > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Not a big contributor to the lists, but... 1) Yes I'd support it 2) Yes I'd join (but unlikely to contribute much) Jeni http://blog.jennystuff.com Peter Miller wrote: > The other question though:- > > 1) Would you support the existence of such a list? > 2) Would you join it? > > If no one supports it or would join it from talk-gb then I won't > progress the idea any more! > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
I'd support talk-dataquality ( or data protection, or data monitoring). I'm concerend about what might happen when this sort of thing starts occurring in the W Mids where I've spent many hours surveying and editing. I can contribute to the discussion about process - would be happy to participate in any eventual mentoring scheme or any other part of any evolving process - can't code tools though Regards Brian ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18 Sep 2009, at 08:51, Brian Prangle wrote: > I'd support talk-dataquality ( or data protection, or data > monitoring). I'm concerend about what might happen when this sort of > thing starts occurring in the W Mids where I've spent many hours > surveying and editing. I can contribute to the discussion about > process - would be happy to participate in any eventual mentoring > scheme or any other part of any evolving process - can't code tools > though Great. That sounds like enough interest to make the request on talk to bottom out an agreed name and an agreed brief for the list and then set it up. I have set up a wiki page here to continue the discussion: http://wiki.openstreetmap.org/wiki/Proposed_moderation_list Personally I think 'data_quality' (or 'dataquality' or 'quality') would attract people who wanted to talk about the correct use of tags and that is something I really really don't want to get mixed into this list. 'data protection', 'data monitoring' and 'data moderation' all seem to be good. I will now do a post to talk now. Regards, Peter > > Regards > > Brian > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
> ... talk-dataquality Hey people, Rome is burning - well Orpington and Bromley anyway - and we're discussing the name of a mailing list! Liam is busy again this morning wrecking most of south east England. His attentions have spread as far as the south coast now. Frederick doesn't want me to try reverting with spending much more time learning about how to manually splice changesets together, which I don't have the time to do. So this guy is steadily eroding large parts of SE England and no-one is doing anything about it. Weeks of slogging around places on the ground is being destroyed by this idiot. Tom - this is persistent continuous abuse. I really think we have to block his IP address until such time as we can work out how to deal with the edits. They are just so pervasive and destructive. We've discussed suspending his account and the chances of him switching to a new one. An IP block is not foolproof either, but it is better than nothing, which is what we've got at the moment. ***PLEASE*** PULL THE PLUG ON HIM! David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/09 10:33, David Earl wrote: > Tom - this is persistent continuous abuse. I really think we have to > block his IP address until such time as we can work out how to deal with > the edits. They are just so pervasive and destructive. We've discussed > suspending his account and the chances of him switching to a new one. An > IP block is not foolproof either, but it is better than nothing, which > is what we've got at the moment. Well (a) we don't know and can't easily find out what IP address he is using and (b) in all probability it changes fairly often. The first step will be to block his account, not his IP address. > ***PLEASE*** PULL THE PLUG ON HIM! I have repeatedly stated that I am not prepared to block people on my own. Get the DWG to order him blocked and I will happily do so. Tom -- Tom Hughes (t...@compton.nu) http://www.compton.nu/ ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
2009/9/18 David Earl : > > ***PLEASE*** PULL THE PLUG ON HIM! > Is it a good idea to remove the Live Change Feature in Potlatch for everyone. I'm thinking this is the cause for a lot of our problems. I can't see why anyone would want it any more anyway. Its a dangerous feature without a purpose. Peter. ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Tom Hughes wrote: > I have repeatedly stated that I am not prepared to block people on my > own. Get the DWG to order him blocked and I will happily do so. Not to Tom personally, but: Can the DWG then finally speak up to acknowledge publicly that they *are* working on this case? And about removal/deactivation/hiding of Potlatch's live editing mode: yes, please. We've had a case in Belgium as well, recently, of someone dicking about in live mode, apparently unaware of the destructive nature of their actions. He/she is not as persistant yet as liam123, and I hope it doesn't get to that, because I'd rather not clean up for 3 months without hints that he/she can/will be stopped by the DWG. -- Lennard ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/2009 10:53, Tom Hughes wrote: > On 18/09/09 10:33, David Earl wrote: > >> Tom - this is persistent continuous abuse. I really think we have to >> block his IP address until such time as we can work out how to deal with >> the edits. They are just so pervasive and destructive. We've discussed >> suspending his account and the chances of him switching to a new one. An >> IP block is not foolproof either, but it is better than nothing, which >> is what we've got at the moment. > > Well (a) we don't know and can't easily find out what IP address he is > using and (b) in all probability it changes fairly often. The first step > will be to block his account, not his IP address. > >> ***PLEASE*** PULL THE PLUG ON HIM! > > I have repeatedly stated that I am not prepared to block people on my > own. Get the DWG to order him blocked and I will happily do so. How? How does the clamour that's been made over three months actually get turned into action? Is there a general email address for the members (who I can see on the wiki, and that you are one of), or do I have to get the email addresses for each member individually - it doesn't say on the wiki how to make contact, though obviously I recognise all the names. But this has been going for three months and he has poisoned large parts of Kent. How long and how much damage do we have to sustain before action is taken? He's spent the last two days doggedly undermining people's work, and far from Andy's previous assertion that he'll get bored, he hasn't. Indeed I have wondered whether there is more to this than just idiocy(*). Re IP addresses, it depends on how he is connected - mine for example never changes so long as I am using the same Mac address to connect. It is cited on the DWG page as one course of action, and I think it would be more effective than banning the account, as we'll likely lose track of him. It is trivial to get hold of the IP address - every HTTP request carries it, though a serious hacker would forge or suppress it, I doubt he's doing that - if he was not just playing, he'd be using multiple accounts. David * could he be a plant from OS to demonstrate the vulnerability of crowd source maps? OK< enough conspiracy theories. ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/2009 11:17, Lennard wrote: > And about removal/deactivation/hiding of Potlatch's live editing mode: > yes, please. We've had a case in Belgium as well, recently, of someone > dicking about in live mode, apparently unaware of the destructive nature > of their actions. +1 But I don't think that's our problem with liam123 - he's doing this deliberately, I'm certain. He's editing in such a way that the changes are subtle and hard to spot visually (like moving bus stops a hundred metres, changing the number of lanes, one way status and so on), and persistently over several months. David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Fri, Sep 18, 2009 at 10:56 AM, Peter Childs wrote: > 2009/9/18 David Earl : >> >> ***PLEASE*** PULL THE PLUG ON HIM! >> > > Is it a good idea to remove the Live Change Feature in Potlatch for everyone. > > I'm thinking this is the cause for a lot of our problems. > > I can't see why anyone would want it any more anyway. Its a dangerous > feature without a purpose. > This case has bugger all to do with Potlatch's live edit mode. Please don't look for scapegoats -- it doesn't help anyone. Dave ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Peter Childs wrote: > Is it a good idea to remove the Live Change Feature in Potlatch > for everyone. > > I'm thinking this is the cause for a lot of our problems. > > I can't see why anyone would want it any more anyway. Its a > dangerous feature without a purpose. *shrugs* You've already made it clear you can't see plenty of things. I know plenty of very experienced users who prefer editing in live mode. It has some significant advantages for the experienced user. Clearly Liam123 knows exactly what he is doing. He has read and deleted messages to him asking him to stop. Live editing mode has nothing to do with it. If you were to ban it, he would just make his changes as per usual and click 'Save'. Then all you've achieved is disadvantaging a lot of existing users for no benefit at all. I'm up for adding a warning the first time a user enters live mode. And I don't anticipate including live mode in Potlatch 2, but that's entirely due to ease of coding. Removing it from 1.x would serve no purpose. I can guarantee Liam123 would continue editing, and you'd be back within the month demanding that some other feature you've just hit upon should be banned. Richard ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Fri, Sep 18, 2009 at 11:26 AM, David Earl wrote: > On 18/09/2009 10:53, Tom Hughes wrote: >> On 18/09/09 10:33, David Earl wrote: >> >>> Tom - this is persistent continuous abuse. I really think we have to >>> block his IP address until such time as we can work out how to deal with >>> the edits. They are just so pervasive and destructive. We've discussed >>> suspending his account and the chances of him switching to a new one. An >>> IP block is not foolproof either, but it is better than nothing, which >>> is what we've got at the moment. >> >> Well (a) we don't know and can't easily find out what IP address he is >> using and (b) in all probability it changes fairly often. The first step >> will be to block his account, not his IP address. >> >>> ***PLEASE*** PULL THE PLUG ON HIM! >> >> I have repeatedly stated that I am not prepared to block people on my >> own. Get the DWG to order him blocked and I will happily do so. > > How? How does the clamour that's been made over three months actually > get turned into action? When we last discussed it it was decided it was better to keep track of him than ban him and not know where he was. We may well re-evaluate that. > > Is there a general email address for the members (who I can see on the > wiki, and that you are one of), or do I have to get the email addresses > for each member individually - it doesn't say on the wiki how to make > contact, though obviously I recognise all the names. > d...@osmfoundation.org > But this has been going for three months and he has poisoned large parts > of Kent. How long and how much damage do we have to sustain before > action is taken? He's spent the last two days doggedly undermining > people's work, and far from Andy's previous assertion that he'll get > bored, he hasn't. Indeed I have wondered whether there is more to this > than just idiocy(*). You seem to think his edits are irreversible -- they are not. They are not hard to find, and they are not particularly hard to fix -- the problem is we lack the necessary tools to do it efficiently (so that one of us actually has enough time to do it before he starts again). If someone somewhere actually develops those tools then we can come back and make sure there's no lasting damage. > > Re IP addresses, it depends on how he is connected - mine for example > never changes so long as I am using the same Mac address to connect. It > is cited on the DWG page as one course of action, and I think it would > be more effective than banning the account, as we'll likely lose track > of him. Except that that's not how most ISPs work -- much of the UK is on a dynamic IP, and you might find yourself banning an entire exchange to stop one user. We can do that, but it really is an extreme measure. Dave ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/09 11:26, David Earl wrote: > On 18/09/2009 10:53, Tom Hughes wrote: >> On 18/09/09 10:33, David Earl wrote: >> >>> ***PLEASE*** PULL THE PLUG ON HIM! >> >> I have repeatedly stated that I am not prepared to block people on my >> own. Get the DWG to order him blocked and I will happily do so. > > How? How does the clamour that's been made over three months actually > get turned into action? As far as I can tell Liam123 was first referred to the DWG on 7th August. Our next meeting was on 25th August and I assume the case was discussed there although I'm having a hard time finding any record of it - most likely somebody was tasked to send him a direct email. The next DWG meeting is next week when I expect it will be discussed again and a decision made on what action to take. Yes we know this is slow, which is why we are working on setting up a ticketing system to allow more work to be done between meetings and why there is work being done to allow more fine grained blocking of users so that temporary suspensions and things can be used while detailed investigation is done. Right now the only things we have are sledgehammers so we have to be careful about how we use them. > Is there a general email address for the members (who I can see on the > wiki, and that you are one of), or do I have to get the email addresses > for each member individually - it doesn't say on the wiki how to make > contact, though obviously I recognise all the names. The address of the DWG is clearly listed on the vandalism page in the wiki although it doesn't currently seem to be on the DWG page for some reason. > Re IP addresses, it depends on how he is connected - mine for example > never changes so long as I am using the same Mac address to connect. It > is cited on the DWG page as one course of action, and I think it would > be more effective than banning the account, as we'll likely lose track > of him. Well bully for you. Now if the entire world is using the same ISP as you then everything will be fine. IP blocks are fragile in a number of ways, both because of the tendency for users to move IP and because they have to be manually configured on each server and therefore have a high chance of getting lost over time. User blocks are much more robust because they act at the rails level by changing the user's record in the database. They are definitely the first choice at the moment. > It is trivial to get hold of the IP address - every HTTP request carries > it, though a serious hacker would forge or suppress it, I doubt he's > doing that - if he was not just playing, he'd be using multiple accounts. I'm not a complete muppet thank you. I know full well that every HTTP request has an IP address associated with it. The problem is working out which HTTP requests are his! The web server access logs do not record the authenticated user for each request for the very simple reason that the web server has no idea as that is a rails level issue. The rails logs also do not log the user details, although they probably could be made to. It would be on a separate line to the IP address however which makes pulling them out much harder. Tom -- Tom Hughes (t...@compton.nu) http://www.compton.nu/ ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
>We seem unable to contact Liam123 via normal channels. This means we >have to think of other ways of contacting members who continually put >mistakes into the dataset. >Is it not possible to put a flag on his account, so when he logs in he >is told what he has been doing and asked to explain himself? >Preferably with a form box that forwards the message to this list? >Otherwise we are going to go around in circles trying to undo his >work, rather than solving the problem. We need to get him on our side >rather than just having him toy with something we feel is very >important. Couldn't we just kill his account? I guess he could sign up again but in principle it seems reasonable to take this action for blatant abuse. Nick ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Dave Stubbs wrote: > You seem to think his edits are irreversible -- they are not. They are > not hard to find, and they are not particularly hard to fix -- the They would be easier to fix if people didn't start reverting his changes by hand the moment he does them, which is what currently prevents a clean revert of his latest changesets. -- Lennard ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
I may be being a simpleton but can't we just disable write privileges for this user to the database? Then he can continue editing but it all has no effect ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Lennard wrote: > They would be easier to fix if people didn't start reverting his changes > by hand the moment he does them, which is what currently prevents a > clean revert of his latest changesets. That could be difficult - it would mean saying to people in potentially a wide area "please stopping mapping for a while". Certainly, when a problem with "inappropriate changes" occurred locally I didn't spot that there had been a large number of other similar changes had been made, just that a particular road was classified wrong. ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/2009 11:52, Tom Hughes wrote: >> Re IP addresses, it depends on how he is connected - mine for example >> never changes so long as I am using the same Mac address to connect. It >> is cited on the DWG page as one course of action, and I think it would >> be more effective than banning the account, as we'll likely lose track >> of him. > > Well bully for you. Now if the entire world is using the same ISP as you > then everything will be fine. Well, it seems a reasonable assumption to me that if my ISP does it this way it might be quite a common practice. But apparently not, so I stand corrected. Doesn't mean it's not worth looking at though - Virgin Media is widely used. >> It is trivial to get hold of the IP address - every HTTP request carries >> it, though a serious hacker would forge or suppress it, I doubt he's >> doing that - if he was not just playing, he'd be using multiple accounts. > > I'm not a complete muppet thank you. I know full well that every HTTP > request has an IP address associated with it. I didn't think you didn't, but you were the one who said it was hard. > The problem is working out which HTTP requests are his! The web server > access logs do not record the authenticated user for each request for > the very simple reason that the web server has no idea as that is a > rails level issue. > > The rails logs also do not log the user details, although they probably > could be made to. It would be on a separate line to the IP address > however which makes pulling them out much harder. I was thinking more along the lines of recording the IP address along with the other changeset information (but not, presumably, exposing it in the API) - after all, that's what the information is wanted in relation to, and it would allow us to see how the address is changing for any particular user. I'm sorry if I seem frustrated by this, but it is because I am. We've all spent thousands of hours each on this, and this guy is undermining everything we've all done. Even though it's not my area (though close), it completely destroys any confidence anyone might have in what they see everywhere. David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Fri, Sep 18, 2009 at 11:58 AM, Brian Prangle wrote: > I may be being a simpleton but can't we just disable write privileges for > this user to the database? Then he can continue editing but it all has no > effect > If somebody writes the code to enforce such a block then we'd have that option. It sounds like an interesting option to have. Dave ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/09 12:13, Dave Stubbs wrote: > On Fri, Sep 18, 2009 at 11:58 AM, Brian Prangle > wrote: >> I may be being a simpleton but can't we just disable write privileges for >> this user to the database? Then he can continue editing but it all has no >> effect > > If somebody writes the code to enforce such a block then we'd have that > option. > It sounds like an interesting option to have. Matt is working on some stuff along those lines at the moment I believe, which is what I was alluding to in my earlier message. Tom -- Tom Hughes (t...@compton.nu) http://www.compton.nu/ ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Someoneelse wrote: >> They would be easier to fix if people didn't start reverting his >> changes by hand the moment he does them, which is what currently >> prevents a clean revert of his latest changesets. > > That could be difficult - it would mean saying to people in potentially > a wide area "please stopping mapping for a while". Certainly, when a No, it would mean saying "please don't revert single changes made by liam123 until a full clean revert has been attempted". The changes I alluded to were part of changesets commented* with 'de-liam123isation', but upon closer inspection I think it reverted all liam123's changes. I thought earlier that he only reverted a few of the changes, which prompted my remark. http://www.openstreetmap.org/browse/changeset/2521914 http://www.openstreetmap.org/browse/changeset/2521863 -- Lennard ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18/09/2009 12:13, Dave Stubbs wrote: > On Fri, Sep 18, 2009 at 11:58 AM, Brian Prangle > wrote: >> I may be being a simpleton but can't we just disable write privileges for >> this user to the database? Then he can continue editing but it all has no >> effect >> > > If somebody writes the code to enforce such a block then we'd have that > option. > It sounds like an interesting option to have. I agree. I think it would be even more useful to be able to quarantine particular users changesets for manual review so the could be let through in the end - though there's all the problems of conflicting changes building up that that entails (or blocking further changes to those objects until the changeset is decided one way or the other), so I'm not under any illusion that this is easy. David ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
> I'm sorry if I seem frustrated by this, but it is because I am. We've > all spent thousands of hours each on this, and this guy is undermining > everything we've all done. Even though it's not my area (though close), > it completely destroys any confidence anyone might have in what they see > everywhere. > Yeah, it really is frustrating. Unfortunately we're currently trying to fight this with a pair of tweezers and a sieve, backed up by a cruise missile and a B52 load of cluster bombs. Dave ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
David Earl wrote: > ***PLEASE*** PULL THE PLUG ON HIM! Does anyone know if there are any discussions ongoing about what restrictions might be put in place on newly registered user who e.g. haven't uploaded any tracks and suddenly start editing in a wide area, to avoid "whack-a-mole" problems? (If there are I'll shut up now). ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
2009/9/18 Dave Stubbs : >> I'm sorry if I seem frustrated by this, but it is because I am. We've >> all spent thousands of hours each on this, and this guy is undermining >> everything we've all done. Even though it's not my area (though close), >> it completely destroys any confidence anyone might have in what they see >> everywhere. >> > > Yeah, it really is frustrating. > Unfortunately we're currently trying to fight this with a pair of > tweezers and a sieve, backed up by a cruise missile and a B52 load of > cluster bombs. Note: A new version of UserActivity was just announced on talk: http://wiki.openstreetmap.org/wiki/UserActivity. It may be of use for detecting future vandals. -- Matt Williams http://milliams.com ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Fri, Sep 18, 2009 at 12:19 PM, David Earl wrote: > I agree. I think it would be even more useful to be able to quarantine > particular users changesets for manual review so the could be let > through in the end - though there's all the problems of conflicting > changes building up that that entails (or blocking further changes to > those objects until the changeset is decided one way or the other), so > I'm not under any illusion that this is easy. > > David > I'd suggest locking everything a changeset touches for xx hours, with xx variable by user. In the quarantine period the changeset can be reverted, but no other changes are allowed. If it's not reverted in the quarantine period, everything gets unlocked. So the user sees their edit get rendered, but anyone can follow them round hitting the undo button. Richard ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Fri, Sep 18, 2009 at 2:22 PM, Richard Mann wrote: > > > On Fri, Sep 18, 2009 at 12:19 PM, David Earl > wrote: >> >> I agree. I think it would be even more useful to be able to quarantine >> particular users changesets for manual review so the could be let >> through in the end - though there's all the problems of conflicting >> changes building up that that entails (or blocking further changes to >> those objects until the changeset is decided one way or the other), so >> I'm not under any illusion that this is easy. >> >> David > > > I'd suggest locking everything a changeset touches for xx hours, with xx > variable by user. In the quarantine period the changeset can be reverted, > but no other changes are allowed. If it's not reverted in the quarantine > period, everything gets unlocked. So the user sees their edit get rendered, > but anyone can follow them round hitting the undo button. > Revert === Edit which makes it quite hard to make an edit, and would also block people from reverting reverts, or something. Plus it makes DOS attacks really really effective because new users have the biggest lock impact. The problem needs fixing with better tools for sorting out mess, not more weird and wonderful metrics for getting in people's way. Dave ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Peter Miller wrote: > 'data protection', 'data monitoring' and 'data moderation' > all seem to be good. > > 'data protection' might be confused with 'Data Protection Act', ie, legal data issues ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
>> > > 'data protection', 'data monitoring' and 'data moderation' > > all seem to be good. > > > > >'data protection' might be confused with 'Data Protection Act', ie, >legal data issues This discussion reminds me of Douglas Adams' Restaurant at the End of the Universe. The 3rd ship of hairdressers and management consultants who are marooned on prehistoric Earth form a committee to invent things to make life better. They are incapable of getting round to inventing the wheel, "the single simplest machine in the entire universe", because they spend all their time arguing about what colour it should be. :-) ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On 18 Sep 2009, at 17:33, Godfrey Bartlett wrote: >>> >>> 'data protection', 'data monitoring' and 'data moderation' >>> all seem to be good. >>> >>> >> 'data protection' might be confused with 'Data Protection Act', ie, >> legal data issues > > This discussion reminds me of Douglas Adams' Restaurant at the End of > the Universe. > The 3rd ship of hairdressers and management consultants who are > marooned > on prehistoric Earth form a committee to invent things to make life > better. They are incapable of getting round to inventing the wheel, > "the single simplest machine in the entire universe", because they > spend > all their time arguing about what colour it should be. :-) > Given that we are going to live with the name for 2+ years I think it is worth a brief discussion before selecting the name. I have done a post on talk this afternoon and have had a bunch of people add their names to the wiki and commented a bit more about names there so I think we are ready to choose one. But can I suggest we carry on this thread on talk or on the wiki page. http://wiki.openstreetmap.org/wiki/Proposed_moderation_list Regards, Peter > > > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
Hi, Dave Stubbs wrote: > The problem needs fixing with better tools for sorting out mess, not > more weird and wonderful metrics for getting in people's way. +1 Whenever a vandal pops up, we have tons of people coming up with tons of cool measures to shoot oneself in the foot, and if they had their way then the damage done by these measures would far outweigh anything vandals have done until now. Bye Frederik ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
I'm not entirely convinced Dave read my posting, because it wasn't a metric (but no matter). When I tidy up after my children, I just put the bits back in the box, I don't try to undo every single move they made in reverse order. And it's a lot easier if the pieces from different jigsaw puzzles don't get mixed up. Now it may be that it's difficult to distinguish between the batch-edit that is reverting a changeset and any other type of edit. Fair enough. But if we're finding that separating jigsaws is impractical, then maybe trying to keep them separate would be a sensible step to consider, while we're waiting for the magic jigsaw-separator to arrive? Richard On Fri, Sep 18, 2009 at 6:03 PM, Frederik Ramm wrote: > Hi, > > Dave Stubbs wrote: > > The problem needs fixing with better tools for sorting out mess, not > > more weird and wonderful metrics for getting in people's way. > > +1 > > Whenever a vandal pops up, we have tons of people coming up with tons of > cool measures to shoot oneself in the foot, and if they had their way > then the damage done by these measures would far outweigh anything > vandals have done until now. > > Bye > Frederik > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] liam123's latest
On Friday 18 Sep 2009 11:33:59 Richard Fairhurst wrote: > Removing it from 1.x would serve no purpose. I can guarantee Liam123 > would continue editing, and you'd be back within the month demanding > that some other feature you've just hit upon should be banned. Ok, here's an idea. Have a sandbox. Once you identify a user who is vandalising things, give him his own database to play with. Don't delete the account, or anything other than ask him to stop. All his changes are made to the sandbox, and not the live DB. Update the sandbox from the main DB every night, so it looks (to him) as though we are trying to correct his vandalism, without any effort on our part. Whether he gets bored and moves on, or not, is then his problem. Alexander. ___ Talk-GB mailing list Talk-GB@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-gb
