Thread: Re: [pgsql-www] We need an Advocacy wiki
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 > We need to set up an Advocacy wiki with more liberal permissions. I've > been trying to use the developer wiki for that purpose, but it's just not > set up for people to add themselves so that they can do things like sign > up for shifts at a booth. I'm still not sure why we need two wikis - we are a transparent organization, after all, and we have enough brains to separate the advocacy pages from the development pages, with Categories if need be. We can also discuss changing the wiki permission scheme, but opening it up to anyone is generally a bad idea unless, like Wikipedia, you a) are willing to put up with vandalism and general subtle mischief and b) have the critical mass of watchers to keep (a) to a minimum. The experience of the interactive docs indicate that we do not have that mass yet, and I am reluctant to go that route anyway for public-facing pages that represent the project via a postgresql.org address. > So we need a wiki which works the normal way; anyone registered can edit. Not sure what your definition of normal is; that depends on what you mean by "registered". If it means just creating an account via web form, that's harldy an impediment to vandalism. We can certainly give more people the power to grant write-access to wiki accounts, if that's the perceived hold up. - -- Greg Sabino Mullane greg@turnstep.com PGP Key: 0x14964AC8 200708040952 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iD8DBQFGtIVWvJuQZxSWSsgRA3UMAJ9oBfw9Poic/8LAgnbYg9/wvqloKgCgz1kk 6QDoA4w+zYfvrycIttwWLjc= =MWMq -----END PGP SIGNATURE-----
On Saturday 04 August 2007 09:56, Greg Sabino Mullane wrote:
> but opening it up to anyone is
> generally a bad idea unless, like Wikipedia, you a) are willing to put up
> with vandalism and general subtle mischief and b) have the critical mass of
> watchers to keep (a) to a minimum. The experience of the interactive docs
> indicate that we do not have that mass yet,
What experience is this?
> and I am reluctant to go that
> route anyway for public-facing pages that represent the project via a
> postgresql.org address.
>
This is more likely the problem, ie. your personal feelings.
--
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL
Robert Treat wrote: > On Saturday 04 August 2007 09:56, Greg Sabino Mullane wrote: >> but opening it up to anyone is >> generally a bad idea unless, like Wikipedia, you a) are willing to put up >> with vandalism and general subtle mischief and b) have the critical mass of >> watchers to keep (a) to a minimum. The experience of the interactive docs >> indicate that we do not have that mass yet, > > What experience is this? I imagine he's referring to the mountain of garbage that used to build up until Magnus and I had a monster session moderating a few thousand comments to get them back under control. >> and I am reluctant to go that >> route anyway for public-facing pages that represent the project via a >> postgresql.org address. I don't see any need to have any public facing pages on a Wiki - are we too lazy to write things up for the website when we want to present them to the world? It's not like it's difficult to do. For collaboration work however, the Wiki is important I think - but I agree with Greg, we shouldn't need a second one. Can't we have an area on the current one with looser permissions? /D
Dave Page wrote: >>> and I am reluctant to go that route anyway for public-facing pages >>> that represent the project via a >>> postgresql.org address. > > I don't see any need to have any public facing pages on a Wiki - are we > too lazy to write things up for the website when we want to present them > to the world? It's not like it's difficult to do. > > For collaboration work however, the Wiki is important I think - but I > agree with Greg, we shouldn't need a second one. Can't we have an area > on the current one with looser permissions? Agreed. The whole idea behind a wiki is "reasonably loose permissions", right? AS Greg already suggested, perhaps we just need a "better way" for people to request permissions? (For example, right now it just says "contact greg or neil", but it doesn't tell you how - not even an email address...) And a structure of the wiki that has a section for advocacy of course - but we already have that. /Magnus
On Sat, Aug 04, 2007 at 04:15:24PM +0100, Dave Page wrote: > For collaboration work however, the Wiki is important I think - but I > agree with Greg, we shouldn't need a second one. Can't we have an area > on the current one with looser permissions? I think this is being blown way out of proportion. We're not wikipedia. We have nowhere near the attention level, nor the type of content that's likely to attract vandals. And before someone brings up the doc comments, there hasn't appeared to be much of a flood of garbage there since we instituted the login requirement. It's also completely unfair to try and correlate keeping an open wiki clean with doing the same for docs, because unlike docs we've got hundreds of folks who could ensure that the wiki stays clean. Can we please just give the public wiki a chance instead of coming up with a bunch of reasons it won't work before we've even tried? It's not like it's hard to change things later if needed. (BTW, when I say public wiki I mean one where anyone with an account can edit, not one where you don't need an account.) -- Decibel!, aka Jim Nasby decibel@decibel.org EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)
"Greg Sabino Mullane" <greg@turnstep.com> writes: > Not sure what your definition of normal is; that depends on what you mean by > "registered". If it means just creating an account via web form, that's > harldy an impediment to vandalism. We can certainly give more people the power > to grant write-access to wiki accounts, if that's the perceived hold up. Do we have any history of vandalism on the -hackers mailing list? There is no approval mechanism for people subscribing to the list. Would people be happy if every subscription to -hackers required someone to approve your membership? I went to do update the wiki recently, found I didn't have write access and gave up and went back to other things. I think any extra barriers are a bad thing. It ought to be open until there's a demonstrable problem rather than preemptively making it less useful because we anticipate problems we have no evidence of. -- Gregory Stark EnterpriseDB http://www.enterprisedb.com
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Dave Page wrote: > Robert Treat wrote: >> On Saturday 04 August 2007 09:56, Greg Sabino Mullane wrote: > I don't see any need to have any public facing pages on a Wiki - are we > too lazy to write things up for the website when we want to present them > to the world? It's not like it's difficult to do. Yes, in fact it is. It is a complete pain in the butt in comparison to editing a wiki. If I want a page added to the .Org I have to: A. Understand CVS B. Understand HTML C. Understand patch Worse, if I want to test my changes: A. Understand apache B. Understand mod_rewrite C. Install and configure PHP D. Figure out how everything is laid out in the htdocs structure for pgweb Now, *I* understand all these things (except mod_rewrite). I *can* contribute to the website and have. *I* and nobody else in this community should have to go through that much effort to add a page. As a comparison for CMDs website: (if the page exists) A. Login B. Edit page C. Review changes D. Save changes (if the page doesn't exist) A. Login B. Insert new row into pages and site_items table C. Edit page D. Review changes E. Save changes The only time anyone has to do any type of coding is if we want a new feature such as my blog. Compare to a wiki: A. Login B. Make change C. Review change D. Save change Now that I have written all of this. I am not suggesting that we change our web infrastructure. I am however suggesting that we stop insulting people and looking very arrogant about our, "It isn't like it is that hard", because it is indeed hard and much harder than it should be. Sincerely, Joshua D. Drake > > For collaboration work however, the Wiki is important I think - but I > agree with Greg, we shouldn't need a second one. Can't we have an area > on the current one with looser permissions? > > /D > > ---------------------------(end of broadcast)--------------------------- > TIP 3: Have you checked our extensive FAQ? > > http://www.postgresql.org/docs/faq > - -- === The PostgreSQL Company: Command Prompt, Inc. === Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240 Providing the most comprehensive PostgreSQL solutions since 1997 http://www.commandprompt.com/ Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate PostgreSQL Replication: http://www.commandprompt.com/products/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGtK/AATb/zqfZUUQRAhllAJ9h+B5K2AWAmnHzCcmnh9j9QwgvjwCfcYqb xBYMwCTHAhnrFgca8e06MKI= =Oata -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Magnus Hagander wrote:
> Dave Page wrote:
> Agreed. The whole idea behind a wiki is "reasonably loose permissions",
> right? AS Greg already suggested, perhaps we just need a "better way"
> for people to request permissions? (For example, right now it just says
> "contact greg or neil", but it doesn't tell you how - not even an email
> address...)
Or just have them email pgsql-www with the request and have more than
two people that can give the permissions?
Joshua D. Drake
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtK/1ATb/zqfZUUQRAlbrAKCGx3yhw7H+GB0XbPS4RJMR9FzlYwCglTmR
SPFq07j8CEbuszzGlfe1kig=
=E7OR
-----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Decibel! wrote:
> On Sat, Aug 04, 2007 at 04:15:24PM +0100, Dave Page wrote:
> It's also completely unfair to try and correlate keeping an open wiki
> clean with doing the same for docs, because unlike docs we've got
> hundreds of folks who could ensure that the wiki stays clean.
Not to mention cleaning a wiki is as easy as "revert". The real problem
here is people that don't want to accept what is now common technology
not only for collaboration but also for public facing pages.
Joshua D. Drake
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtLCFATb/zqfZUUQRAnttAKCipxFYp+iL/nLQYcZ/IFkKexAJ+wCfZmiP
yp7wMgrWX19A9IxKjdDaA+Q=
=iJg2
-----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Gregory Stark wrote:
> "Greg Sabino Mullane" <greg@turnstep.com> writes:
>
> I went to do update the wiki recently, found I didn't have write access and
> gave up and went back to other things. I think any extra barriers are a bad
> thing. It ought to be open until there's a demonstrable problem rather than
> preemptively making it less useful because we anticipate problems we have no
> evidence of.
>
Yep, which is why it wasn't until OSCON that I had a wiki account that I
actually paid attention to.
Sincerely,
Joshua D. Drake
>
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtLEPATb/zqfZUUQRAjW1AKCVTzFCIWbAMFt5gDQ//5b/Zo4bawCgnjQW
8Bm9cJTDu7XL2ljvwZgqM/4=
=u6J4
-----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Oh, and have a great weekend! :)
Joshua D. Drake
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtLFMATb/zqfZUUQRAjWEAJ0YRfLXazHNXM9Hd13T6n89+8Gy9wCdHgjn
H0gcA8X3wxkx7HbBd14CavM=
=ccJa
-----END PGP SIGNATURE-----
Gregory Stark wrote: > "Greg Sabino Mullane" <greg@turnstep.com> writes: > >> Not sure what your definition of normal is; that depends on what you mean by >> "registered". If it means just creating an account via web form, that's >> harldy an impediment to vandalism. We can certainly give more people the power >> to grant write-access to wiki accounts, if that's the perceived hold up. > > Do we have any history of vandalism on the -hackers mailing list? There is no > approval mechanism for people subscribing to the list. Would people be happy > if every subscription to -hackers required someone to approve your membership? No, but we *do* have a history of vandalism/spamming on the website. This includes the interactive docs, news posts, event posts and professional services. That's pretty much every single part of the website that actually has a submit button, except for the bug reporting form - which goes through the majordomo moderation system so it's still moderated. But - it helped significantly when we started requesting community logins for these forms. It's still not gone - there are people advertising for UK hotels and a few other things that actually sign up for a community account with a temp email address and post from there. This is the main reason why we still have manual verification on all these things even though they require a login. That said, allowing people to sign up for an account in an automated way that does require email verification would work, as long as there is: 1) A way to revoke and ban addresses 2) Somebody to keep track of things, and remove spam and revoke/ban these users. (this would be the same level of verification that we have for the mailinglists - it's not like they're unvalidated) But the work required for (2) is a lot less than a completely open system of course. If there are a couple of people who are willing to take that upon them (which there seems to be, given the activity on the wiki) we could always give that a try? //Magnus
Joshua D. Drake wrote: > Dave Page wrote: >> Robert Treat wrote: >>> On Saturday 04 August 2007 09:56, Greg Sabino Mullane wrote: >> I don't see any need to have any public facing pages on a Wiki - are we >> too lazy to write things up for the website when we want to present them >> to the world? It's not like it's difficult to do. > > Yes, in fact it is. It is a complete pain in the butt in comparison to > editing a wiki. If I want a page added to the .Org I have to: Ever tried techdocs? ;-) > A. Understand CVS No, no need for that. You can just send your files to -www. > B. Understand HTML That, there is need for. Personally, I know a lot of people who find that easier than some of the weird wiki markup thingies around :-P I think the big problem for testing that is that our web *layout* (the CSS and div-ifying) is very complex. I'd really love to see that simplified - by someone who know it well enough. But the truth is, most pages would be perfectly fine written just using a couple of <h> and <p> tags, along with possibly a couple of <a href>:s. As long as you don't need to muck about with complex layout stuff, that's trivial. And AFAIK, you can't really muck around with said complex layout stuff in wikis either, without hard-coding CSS and HTML the same way. > C. Understand patch No need for that if you're adding a page. And not really needed if you're changing one either, as one of the web guys can take care of that. > Worse, if I want to test my changes: There should be no need to test your changes unless you're writing *code*. Which is not what we're talking about here, really. > The only time anyone has to do any type of coding is if we want a new > feature such as my blog. Same goes for postgresql.org - as long as you're just putting in text, there is no need for the majority of the points in your list. > Now that I have written all of this. I am not suggesting that we change > our web infrastructure. I am however suggesting that we stop insulting > people and looking very arrogant about our, "It isn't like it is that > hard", because it is indeed hard and much harder than it should be. If it really is that hard (and I honestly don't believe that it is, but I'm willing to accept that others find it that way), then we *should* fix it. The hard part is agreeing on *how*, because there are so many requirements... > Oh, and have a great weekend! :) You, too! //Magnus
Joshua D. Drake wrote: > Decibel! wrote: >> On Sat, Aug 04, 2007 at 04:15:24PM +0100, Dave Page wrote: > >> It's also completely unfair to try and correlate keeping an open wiki >> clean with doing the same for docs, because unlike docs we've got >> hundreds of folks who could ensure that the wiki stays clean. > > Not to mention cleaning a wiki is as easy as "revert". The real problem > here is people that don't want to accept what is now common technology > not only for collaboration but also for public facing pages. I certainly know it's turning into a common technology for that. The unstructured nature of it has made it a *lot* harder to find anything on the homepages of many of the projects that have switched to it. It also in general makes it impossible to determine what information is official and what is not. //Magnus
On Sat, Aug 04, 2007 at 07:10:07PM +0200, Magnus Hagander wrote: > Gregory Stark wrote: > > "Greg Sabino Mullane" <greg@turnstep.com> writes: > But - it helped significantly when we started requesting community > logins for these forms. It's still not gone - there are people > advertising for UK hotels and a few other things that actually sign up > for a community account with a temp email address and post from there. > This is the main reason why we still have manual verification on all > these things even though they require a login. Actually, I rather doubt it's real, live people, because it's pretty easy for a bot to spoof a form that doesn't have captcha or some other means of verification. In any case, any wiki should have that in place, so it's essentially a non-issue. -- Decibel!, aka Jim Nasby decibel@decibel.org EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)
Decibel! wrote: > On Sat, Aug 04, 2007 at 04:15:24PM +0100, Dave Page wrote: >> For collaboration work however, the Wiki is important I think - but I >> agree with Greg, we shouldn't need a second one. Can't we have an area >> on the current one with looser permissions? > > I think this is being blown way out of proportion. > > We're not wikipedia. We have nowhere near the attention level, nor the > type of content that's likely to attract vandals. And before someone > brings up the doc comments, there hasn't appeared to be much of a flood > of garbage there since we instituted the login requirement. Correct. There's still some, but it's much better now. > It's also completely unfair to try and correlate keeping an open wiki > clean with doing the same for docs, because unlike docs we've got > hundreds of folks who could ensure that the wiki stays clean. Not sure that's a fair count. Looking at the wiki user list there are certainly 215 accounts. But by my untrained eye, a lot of those look like automated users created by spam-bots in order to see if they can create spam-pages. It could be that we have actual users named Zy9Yqd, Yx9Qbh and Xj0Y6g, but I seriously doubt it. And that's a clear indication that there are people (or rather, bots) probing the wiki already trying to post crap. > Can we please just give the public wiki a chance instead of coming up > with a bunch of reasons it won't work before we've even tried? It's not > like it's hard to change things later if needed. > > (BTW, when I say public wiki I mean one where anyone with an account can > edit, not one where you don't need an account.) As long as that holds, I'm absolutely up for giving it a try. Maybe part of the disagreement has been from a misunderstanding of what a "public wiki" is. In my book, a *public* wiki is one that doesn't need a verified account. (I assume that you refer to verified account above. If not, I don't agree until you add the word verified) //Magnus
On Sat, Aug 04, 2007 at 07:23:48PM +0200, Magnus Hagander wrote: > Decibel! wrote: > > On Sat, Aug 04, 2007 at 04:15:24PM +0100, Dave Page wrote: > >> For collaboration work however, the Wiki is important I think - but I > >> agree with Greg, we shouldn't need a second one. Can't we have an area > >> on the current one with looser permissions? > > > > I think this is being blown way out of proportion. > > > > We're not wikipedia. We have nowhere near the attention level, nor the > > type of content that's likely to attract vandals. And before someone > > brings up the doc comments, there hasn't appeared to be much of a flood > > of garbage there since we instituted the login requirement. > > Correct. There's still some, but it's much better now. > > > > It's also completely unfair to try and correlate keeping an open wiki > > clean with doing the same for docs, because unlike docs we've got > > hundreds of folks who could ensure that the wiki stays clean. > > Not sure that's a fair count. Looking at the wiki user list there are > certainly 215 accounts. But by my untrained eye, a lot of those look > like automated users created by spam-bots in order to see if they can > create spam-pages. It could be that we have actual users named Zy9Yqd, > Yx9Qbh and Xj0Y6g, but I seriously doubt it. And that's a clear > indication that there are people (or rather, bots) probing the wiki > already trying to post crap. Well, my point is that if we allow our users to easily get accounts, we'll have a lot of eyes on this... > As long as that holds, I'm absolutely up for giving it a try. Maybe part > of the disagreement has been from a misunderstanding of what a "public > wiki" is. In my book, a *public* wiki is one that doesn't need a > verified account. (I assume that you refer to verified account above. If > not, I don't agree until you add the word verified) Well, at the bare minimum we need a captcha or something similar. Without that then yes, we're going to get all kinds of crap accounts. Do we just have that turned off, or does mediawiki actually not support that? -- Decibel!, aka Jim Nasby decibel@decibel.org EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)
Attachment
Decibel! wrote: >> As long as that holds, I'm absolutely up for giving it a try. Maybe part >> of the disagreement has been from a misunderstanding of what a "public >> wiki" is. In my book, a *public* wiki is one that doesn't need a >> verified account. (I assume that you refer to verified account above. If >> not, I don't agree until you add the word verified) > > Well, at the bare minimum we need a captcha or something similar. > Without that then yes, we're going to get all kinds of crap accounts. > Do we just have that turned off, or does mediawiki actually not support > that? I'd say we want email verification, so that we can contact the authors when needed. That said, I have no clue about mediawiki ;-) //Magnus
On Saturday 04 August 2007 11:15, Dave Page wrote:
> Robert Treat wrote:
> > On Saturday 04 August 2007 09:56, Greg Sabino Mullane wrote:
> >> but opening it up to anyone is
> >> generally a bad idea unless, like Wikipedia, you a) are willing to put
> >> up with vandalism and general subtle mischief and b) have the critical
> >> mass of watchers to keep (a) to a minimum. The experience of the
> >> interactive docs indicate that we do not have that mass yet,
> >
> > What experience is this?
>
> I imagine he's referring to the mountain of garbage that used to build
> up until Magnus and I had a monster session moderating a few thousand
> comments to get them back under control.
>
*sigh*
Doesn't this suggest we do have mass to prevent vandalism/mischief ?
--
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL
Gregory Stark wrote: > "Greg Sabino Mullane" <greg@turnstep.com> writes: > >> Not sure what your definition of normal is; that depends on what you mean by >> "registered". If it means just creating an account via web form, that's >> harldy an impediment to vandalism. We can certainly give more people the power >> to grant write-access to wiki accounts, if that's the perceived hold up. > > Do we have any history of vandalism on the -hackers mailing list? There is no > approval mechanism for people subscribing to the list. Would people be happy > if every subscription to -hackers required someone to approve your membership? > > I went to do update the wiki recently, found I didn't have write access and > gave up and went back to other things. I think any extra barriers are a bad > thing. It ought to be open until there's a demonstrable problem rather than > preemptively making it less useful because we anticipate problems we have no > evidence of. Well we should definately require a login, even if its just so that we can trace who made what changes so that we know who has taken responsibility over a given section. regards, Lukas
On Sat, Aug 04, 2007 at 07:39:20PM +0200, Magnus Hagander wrote: > Decibel! wrote: > >> As long as that holds, I'm absolutely up for giving it a try. Maybe part > >> of the disagreement has been from a misunderstanding of what a "public > >> wiki" is. In my book, a *public* wiki is one that doesn't need a > >> verified account. (I assume that you refer to verified account above. If > >> not, I don't agree until you add the word verified) > > > > Well, at the bare minimum we need a captcha or something similar. > > Without that then yes, we're going to get all kinds of crap accounts. > > Do we just have that turned off, or does mediawiki actually not support > > that? > > I'd say we want email verification, so that we can contact the authors > when needed. > > That said, I have no clue about mediawiki ;-) I have a suspicion that spammers are sophisticated enough to be able to handle simple email verification... :/ -- Decibel!, aka Jim Nasby decibel@decibel.org EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)
Attachment
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Magnus Hagander wrote:
> Joshua D. Drake wrote:
>> Dave Page wrote:
>> Yes, in fact it is. It is a complete pain in the butt in comparison to
>> editing a wiki. If I want a page added to the .Org I have to:
>
> Ever tried techdocs? ;-)
No. I don't like the layout (and yes that is an opinion thing and I
don't want to get into it).
>
>
>> A. Understand CVS
>
> No, no need for that. You can just send your files to -www.
You mean the straight HTML? what about everything that resides around it?
>> Now that I have written all of this. I am not suggesting that we change
>> our web infrastructure. I am however suggesting that we stop insulting
>> people and looking very arrogant about our, "It isn't like it is that
>> hard", because it is indeed hard and much harder than it should be.
>
> If it really is that hard (and I honestly don't believe that it is, but
> I'm willing to accept that others find it that way), then we *should*
> fix it. The hard part is agreeing on *how*, because there are so many
> requirements...
O.k. perhaps this is a perception problem. Consider my involvement in
the community and the fact that I thought that I had to do all of the
above in order to contribute to the site... Now take into account
someone who isn't as involved as I?
Where is the "Contributing to the PostgreSQL.Org website HOWTO"?
Sincerely,
Joshua D. Drake
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtO0KATb/zqfZUUQRAuG0AJ0f1QezCADg4S5tcTu5SahgLpPGnwCbBBIw
23g2sFOiNzIgV+HDMEiFN4g=
=yDl/
-----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Magnus Hagander wrote:
> Joshua D. Drake wrote:
>> Decibel! wrote:
>>> On Sat, Aug 04, 2007 at 04:15:24PM +0100, Dave Page wrote:
>>> It's also completely unfair to try and correlate keeping an open wiki
>>> clean with doing the same for docs, because unlike docs we've got
>>> hundreds of folks who could ensure that the wiki stays clean.
>> Not to mention cleaning a wiki is as easy as "revert". The real problem
>> here is people that don't want to accept what is now common technology
>> not only for collaboration but also for public facing pages.
>
> I certainly know it's turning into a common technology for that. The
> unstructured nature of it has made it a *lot* harder to find anything on
> the homepages of many of the projects that have switched to it. It also
> in general makes it impossible to determine what information is official
> and what is not.
Not when managed correctly. Keep in mind that I have zero problem with
requiring accounts for accessing the pages.
Sincerely,
Joshua D. Drake
>
>
> //Magnus
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: In versions below 8.0, the planner will ignore your desire to
> choose an index scan if your joining column's datatypes do not
> match
>
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtO1HATb/zqfZUUQRAi15AKCjUq4DzjJz57Pi6FJy3WhE6GOQWQCeN0+q
DiYLl2X8ueprNJWO+49g0rs=
=/4Oz
-----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Magnus Hagander wrote:
> Decibel! wrote:
>>> As long as that holds, I'm absolutely up for giving it a try. Maybe part
>>> of the disagreement has been from a misunderstanding of what a "public
>>> wiki" is. In my book, a *public* wiki is one that doesn't need a
>>> verified account. (I assume that you refer to verified account above. If
>>> not, I don't agree until you add the word verified)
>> Well, at the bare minimum we need a captcha or something similar.
>> Without that then yes, we're going to get all kinds of crap accounts.
>> Do we just have that turned off, or does mediawiki actually not support
>> that?
>
> I'd say we want email verification, so that we can contact the authors
> when needed.
Yes that is what I would expect.
Joshua D. Drake
>
> That said, I have no clue about mediawiki ;-)
>
> //Magnus
>
> ---------------------------(end of broadcast)---------------------------
> TIP 6: explain analyze is your friend
>
- --
=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/
Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGtO2AATb/zqfZUUQRArHwAKCDmdtrqmpwlDhSgTZTDDT/pwfZHACfQNjO
9K6sAq40IWH7M0UHd1f/bXQ=
=I1ET
-----END PGP SIGNATURE-----
Magnus Hagander schrieb: ... > Not sure that's a fair count. Looking at the wiki user list there are > certainly 215 accounts. But by my untrained eye, a lot of those look > like automated users created by spam-bots in order to see if they can > create spam-pages. It could be that we have actual users named Zy9Yqd, > Yx9Qbh and Xj0Y6g, but I seriously doubt it. And that's a clear > indication that there are people (or rather, bots) probing the wiki > already trying to post crap. > > >> Can we please just give the public wiki a chance instead of coming up >> with a bunch of reasons it won't work before we've even tried? It's not >> like it's hard to change things later if needed. >> >> (BTW, when I say public wiki I mean one where anyone with an account can >> edit, not one where you don't need an account.) > > As long as that holds, I'm absolutely up for giving it a try. Maybe part > of the disagreement has been from a misunderstanding of what a "public > wiki" is. In my book, a *public* wiki is one that doesn't need a > verified account. (I assume that you refer to verified account above. If > not, I don't agree until you add the word verified) Maybe the users could be created by referral or invitation? This way you would form a little web of trust instead of having almost alien people (or bots) trying to write something related to postgres. Regards Tino
Decibel! schrieb: ... > Well, at the bare minimum we need a captcha or something similar. > Without that then yes, we're going to get all kinds of crap accounts. > Do we just have that turned off, or does mediawiki actually not support > that? No Captcha please. These s*ck horribly and do not fight spambots if they really want to get in. Please see my other mail on alternative proposal. Regards Tino
On Sat, 4 Aug 2007, Magnus Hagander wrote: > AS Greg already suggested, perhaps we just need a "better way" for > people to request permissions? (For example, right now it just says > "contact greg or neil", but it doesn't tell you how - not even an email > address...) Fixed; there's now a new page at http://developer.postgresql.org/index.php/Editing_Guidelines that explains what to do (which our moderaters should look at now that I've made them more public--I left the e-mail addresses somewhat obfuscated similarly to the mailing lists). Like Greg Stark's story, this was a sore point for me. The week I wanted to start editing I tried just e-mailing GSM for approval, but it was during a time when he was unavailable. I waited a few days while unsure if I'd even contacted him correctly, then e-mailed Neil, then finally got in; had I not been really motivated I would have just given up long before getting permissions. Despite all that, I waste enough of my time cleaning up after spammers, vandals, and idiots on other wikis that I'm still on the side of those here suggesting this particular resources should stay controlled in this fashion. Clearing up the instructions solves most of what bugged me. What I'd suggest is turning those who can approve edit rights into a mailing list (so the note on the new page I made can say "e-mail dev-wiki-edit-request@postgresql.org" or something instead of mentioning multiple names) that forwards the request to everyone who has approval permissions. Then expand that list a bit so that's it's more likely it will hit someone who can do the approval in a timely fashion; first person to grant the rights cc's the list and the requester saying it's done, and barring the occasional harmless race condition dupe the whole thing would be simple enough. If Josh or others really need a true open wiki without such an approval process, I'd suggest popping that into another database and create another Wikimedia instance for it. I think having that all mixed in with the content on the developer's wiki will just make tracking edits harder for both groups. Having a "Recent changes" page that's small enough to browse easily is helpful for the scale of people involved in these pages at this point, and I do browse that section of the Developer's Wiki to see what's been going on. I know I'd be bothered if that got filled with booth work edits instead--and the booth workers would have an easier time policing their area if the developer edits weren't in their way. Plus, if it gets nailed hard you can just save the important stuff and nuke the whole temporary wiki rather than be compelled do a time-intensive cleanup; losing the history isn't as good of an idea for the developer's wiki. -- * Greg Smith gsmith@gregsmith.com http://www.gregsmith.com Baltimore, MD
On Sat, 4 Aug 2007, Decibel! wrote: > Well, at the bare minimum we need a captcha or something similar. > Without that then yes, we're going to get all kinds of crap accounts. > Do we just have that turned off, or does mediawiki actually not support > that? There are a couple of levels of captcha you can setup. For example, one wiki I work on a regularly allows edits by any account as soon as it's created. But if you add a reference to a URL that's outside of the Wiki itself, committing that edit requires completing a captcha (where the data you type is a combination of two common words appearing in the wiki). This makes it so anyone can add regular content almost instantly, but since the spammers can't automate linking to their sites it makes them less likely to target you. -- * Greg Smith gsmith@gregsmith.com http://www.gregsmith.com Baltimore, MD
On Sun, Aug 05, 2007 at 10:11:50AM +0200, Tino Wildenhain wrote: > Magnus Hagander schrieb: > Maybe the users could be created by referral or invitation? This way > you would form a little web of trust instead of having almost alien > people (or bots) trying to write something related to postgres. That's not really much better than what we've got right now... you're still relying on a human somewhere to get things done, which adds a lot of latency. -- Decibel!, aka Jim Nasby decibel@decibel.org EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)