Thread: Dropping training events
I'd like to propose that we drop the ability to register training events on the www.postgresql.org site.
Reasons:
1. It's basically a system set up to game, where it helps to flood with a lot of instances and then cancel them (often without actually removing the registration on the website, and our moderators certainly don't have the time to go double-check which instances are actually happening etc)
2. It takes a lot of work for our moderators
3. Based on (1) it provides very little value to the end user the way it is now.
4. And based on (3), in the end it also doesn't provide much value to the companies providing training.
Instead, I propose that we long-term merge this functionality into the "professional services" part of the site, as a "this provider provides both support and training" or "this provider provides just training" kind of fields. Now, that section of the site *also* needs a complete redesign and re-thought, but the problems to consider (avoid what's basically spam and figuring out how to get a listing thats actually useful to end users) are very similar between things like support and training in this context.
But short term,let's just drop them because they're useless today.
Le mer. 7 nov. 2018 à 15:36, Magnus Hagander <magnus@hagander.net> a écrit :
I'd like to propose that we drop the ability to register training events on the www.postgresql.org site.Reasons:1. It's basically a system set up to game, where it helps to flood with a lot of instances and then cancel them (often without actually removing the registration on the website, and our moderators certainly don't have the time to go double-check which instances are actually happening etc)2. It takes a lot of work for our moderators3. Based on (1) it provides very little value to the end user the way it is now.4. And based on (3), in the end it also doesn't provide much value to the companies providing training.Instead, I propose that we long-term merge this functionality into the "professional services" part of the site, as a "this provider provides both support and training" or "this provider provides just training" kind of fields. Now, that section of the site *also* needs a complete redesign and re-thought, but the problems to consider (avoid what's basically spam and figuring out how to get a listing thats actually useful to end users) are very similar between things like support and training in this context.But short term,let's just drop them because they're useless today.
Totally agree.
--
Guillaume.
On 11/7/18 9:36 AM, Magnus Hagander wrote: > I'd like to propose that we drop the ability to register training events > on the www.postgresql.org <http://www.postgresql.org> site. > > Reasons: > > 1. It's basically a system set up to game, where it helps to flood with > a lot of instances and then cancel them (often without actually removing > the registration on the website, and our moderators certainly don't have > the time to go double-check which instances are actually happening etc) Yeah, I think this is the crux of it: there are some people who are using the system fairly for promoting GA trainings with high-quality content that do happen. And then there is the above, which unfortunately is not just speculation but has been confirmed in multiple instances. > 2. It takes a lot of work for our moderators I will certainly +1 this bullet point. > 3. Based on (1) it provides very little value to the end user the way it > is now. +1 for sure. Some of this is on us to clean up the UX (below). > 4. And based on (3), in the end it also doesn't provide much value to > the companies providing training. Well, it must provide some value because organizations keep submitting training events. > Instead, I propose that we long-term merge this functionality into the > "professional services" part of the site, as a "this provider provides > both support and training" or "this provider provides just training" > kind of fields. Now, that section of the site *also* needs a complete > redesign and re-thought, but the problems to consider (avoid what's > basically spam and figuring out how to get a listing thats actually > useful to end users) are very similar between things like support and > training in this context. +1, but I think with the correct interface it will make things much better for both our users and organizations that are helping to provide professional services around the community. > But short term,let's just drop them because they're useless today. +1. Jonathan
Attachment
On 2018-11-08 01:50, Jonathan S. Katz wrote: > On 11/7/18 9:36 AM, Magnus Hagander wrote: >> I'd like to propose that we drop the ability to register training >> events >> on the www.postgresql.org <http://www.postgresql.org> site. >> >> Reasons: >> >> 1. It's basically a system set up to game, where it helps to flood >> with >> a lot of instances and then cancel them (often without actually >> removing >> the registration on the website, and our moderators certainly don't >> have >> the time to go double-check which instances are actually happening >> etc) > > Yeah, I think this is the crux of it: there are some people who are > using the system fairly for promoting GA trainings with high-quality > content that do happen. And then there is the above, which > unfortunately > is not just speculation but has been confirmed in multiple instances. Would it be feasible to still have the training events, but blacklist (in an effective way) the abusers? Just from the point of view of rewarding those people who do create good value training + use the system fairly, and not punishing them because some places are abusing things. ? + Justin
On Wed, Nov 7, 2018 at 03:42:20PM +0100, Guillaume Lelarge wrote: > Le mer. 7 nov. 2018 à 15:36, Magnus Hagander <magnus@hagander.net> a écrit : > Instead, I propose that we long-term merge this functionality into the > "professional services" part of the site, as a "this provider provides both > support and training" or "this provider provides just training" kind of > fields. Now, that section of the site *also* needs a complete redesign and > re-thought, but the problems to consider (avoid what's basically spam and > figuring out how to get a listing thats actually useful to end users) are > very similar between things like support and training in this context. > > But short term,let's just drop them because they're useless today. > > Totally agree. +1 -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + As you are, so once was I. As I am, so you will be. + + Ancient Roman grave inscription +
On Wed, Nov 7, 2018 at 3:50 PM Jonathan S. Katz <jkatz@postgresql.org> wrote:
On 11/7/18 9:36 AM, Magnus Hagander wrote:
> I'd like to propose that we drop the ability to register training events
> on the www.postgresql.org <http://www.postgresql.org> site.
>
> Reasons:
>
> 1. It's basically a system set up to game, where it helps to flood with
> a lot of instances and then cancel them (often without actually removing
> the registration on the website, and our moderators certainly don't have
> the time to go double-check which instances are actually happening etc)
Yeah, I think this is the crux of it: there are some people who are
using the system fairly for promoting GA trainings with high-quality
content that do happen. And then there is the above, which unfortunately
is not just speculation but has been confirmed in multiple instances.
> 2. It takes a lot of work for our moderators
I will certainly +1 this bullet point.
> 3. Based on (1) it provides very little value to the end user the way it
> is now.
+1 for sure. Some of this is on us to clean up the UX (below).
> 4. And based on (3), in the end it also doesn't provide much value to
> the companies providing training.
Well, it must provide some value because organizations keep submitting
training events.
> Instead, I propose that we long-term merge this functionality into the
> "professional services" part of the site, as a "this provider provides
> both support and training" or "this provider provides just training"
> kind of fields. Now, that section of the site *also* needs a complete
> redesign and re-thought, but the problems to consider (avoid what's
> basically spam and figuring out how to get a listing thats actually
> useful to end users) are very similar between things like support and
> training in this context.
+1, but I think with the correct interface it will make things much
better for both our users and organizations that are helping to provide
professional services around the community.
> But short term,let's just drop them because they're useless today.
+1.
PFA a patch that implements this.
Attachment
Hi, On 12/5/18 10:44 AM, Magnus Hagander wrote: > PFA a patch that implements this. Thanks for working on this! Overall it looks good, and testing the core functionality appears to work. A couple of comments: 1. In "templates/index.html" on line 128, there is still a reference to "training sessions" - given some events (e.g. conferences) have training sessions attached to them, I odn't think it's imperative we remove those words, but wanted to note it in order to solicit feedback. 2. In "templates/pages/download/recognition.html" There is a reference to this wiki page: https://wiki.postgresql.org/wiki/NewsEventsApproval#Approving_Events_and_Event-Related_News_.28excluding_training.29 First, we probably have to update the wiki as we have removed the training section (either remove the content or move it to a "Deprecated" section?) which would possibly cause use to update the title for readability purposes. Updating the URL would not break the URL, but it would break the anchor jump. Thanks! Jonathan
Attachment
On Thu, Dec 6, 2018 at 3:25 PM Jonathan S. Katz <jkatz@postgresql.org> wrote:
Hi,
On 12/5/18 10:44 AM, Magnus Hagander wrote:
> PFA a patch that implements this.
Thanks for working on this! Overall it looks good, and testing the core
functionality appears to work.
A couple of comments:
1. In "templates/index.html" on line 128, there is still a reference to
"training sessions" - given some events (e.g. conferences) have training
sessions attached to them, I odn't think it's imperative we remove those
words, but wanted to note it in order to solicit feedback.
I did notice that one as well, an considered it something we should leave.
2. In "templates/pages/download/recognition.html" There is a reference
to this wiki page:
https://wiki.postgresql.org/wiki/NewsEventsApproval#Approving_Events_and_Event-Related_News_.28excluding_training.29
First, we probably have to update the wiki as we have removed the
training section (either remove the content or move it to a "Deprecated"
section?) which would possibly cause use to update the title for
readability purposes. Updating the URL would not break the URL, but it
would break the anchor jump.
We should definitely update the page, yes. I don't think we need to create a deprecated session -- there is version history on the page after all.
TBH, I don't much care about anchor jumps inside the page breaking. As long as the main URL doesn't change (which it won't, as you say), it's fine.
On 12/6/18 9:57 AM, Magnus Hagander wrote: > On Thu, Dec 6, 2018 at 3:25 PM Jonathan S. Katz <jkatz@postgresql.org > <mailto:jkatz@postgresql.org>> wrote: > 1. In "templates/index.html" on line 128, there is still a reference to > "training sessions" - given some events (e.g. conferences) have training > sessions attached to them, I odn't think it's imperative we remove those > words, but wanted to note it in order to solicit feedback. > > > I did notice that one as well, an considered it something we should leave. I don't feel strongly either way, I'm ok with leaving it. > > > 2. In "templates/pages/download/recognition.html" There is a reference > to this wiki page: > > https://wiki.postgresql.org/wiki/NewsEventsApproval#Approving_Events_and_Event-Related_News_.28excluding_training.29 > > First, we probably have to update the wiki as we have removed the > training section (either remove the content or move it to a "Deprecated" > section?) which would possibly cause use to update the title for > readability purposes. Updating the URL would not break the URL, but it > would break the anchor jump. > > > We should definitely update the page, yes. I don't think we need to > create a deprecated session -- there is version history on the page > after all. > > TBH, I don't much care about anchor jumps inside the page breaking. As > long as the main URL doesn't change (which it won't, as you say), it's fine. I care enough about the jump that I'd fix it after the edits are made, but would not hold up this patch because of it. Anyway, +1. Thanks! Jonathan
Attachment
On Thu, Dec 6, 2018 at 4:25 PM Jonathan S. Katz <jkatz@postgresql.org> wrote:
On 12/6/18 9:57 AM, Magnus Hagander wrote:
> On Thu, Dec 6, 2018 at 3:25 PM Jonathan S. Katz <jkatz@postgresql.org
> <mailto:jkatz@postgresql.org>> wrote:
>
1. In "templates/index.html" on line 128, there is still a reference to
> "training sessions" - given some events (e.g. conferences) have training
> sessions attached to them, I odn't think it's imperative we remove those
> words, but wanted to note it in order to solicit feedback.
>
>
> I did notice that one as well, an considered it something we should leave.
I don't feel strongly either way, I'm ok with leaving it.
>
>
> 2. In "templates/pages/download/recognition.html" There is a reference
> to this wiki page:
>
> https://wiki.postgresql.org/wiki/NewsEventsApproval#Approving_Events_and_Event-Related_News_.28excluding_training.29
>
> First, we probably have to update the wiki as we have removed the
> training section (either remove the content or move it to a "Deprecated"
> section?) which would possibly cause use to update the title for
> readability purposes. Updating the URL would not break the URL, but it
> would break the anchor jump.
>
>
> We should definitely update the page, yes. I don't think we need to
> create a deprecated session -- there is version history on the page
> after all.
>
> TBH, I don't much care about anchor jumps inside the page breaking. As
> long as the main URL doesn't change (which it won't, as you say), it's fine.
I care enough about the jump that I'd fix it after the edits are made,
but would not hold up this patch because of it.
Anyway, +1.
Applied, thanks for the review!
On Thu, 8 Nov 2018 at 13:18, Justin Clift <justin@postgresql.org> wrote:
--
On 2018-11-08 01:50, Jonathan S. Katz wrote:
> On 11/7/18 9:36 AM, Magnus Hagander wrote:
>> I'd like to propose that we drop the ability to register training
>> events
>> on the www.postgresql.org <http://www.postgresql.org> site.
>>
>> Reasons:
>>
>> 1. It's basically a system set up to game, where it helps to flood
>> with
>> a lot of instances and then cancel them (often without actually
>> removing
>> the registration on the website, and our moderators certainly don't
>> have
>> the time to go double-check which instances are actually happening
>> etc)
>
> Yeah, I think this is the crux of it: there are some people who are
> using the system fairly for promoting GA trainings with high-quality
> content that do happen. And then there is the above, which
> unfortunately
> is not just speculation but has been confirmed in multiple instances.
Would it be feasible to still have the training events, but blacklist
(in
an effective way) the abusers?
Just from the point of view of rewarding those people who do create good
value training + use the system fairly, and not punishing them because
some
places are abusing things.
Somewhat surprised to not see a reply to Justin's reasonable comment.
Why would we not just block everyone who acts unreasonably? Just as we do in other cases.
Will we remove Hackers next because people post spam to it? That would save moderator time also. But obviously not a serious suggestion.
Blocking all training helps those who do NOT provide scheduled training, so the people who provide scheduled training have now been penalised for no reason.
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
On Fri, Dec 7, 2018 at 10:41 AM Simon Riggs <simon@2ndquadrant.com> wrote:
On Thu, 8 Nov 2018 at 13:18, Justin Clift <justin@postgresql.org> wrote:On 2018-11-08 01:50, Jonathan S. Katz wrote:
> On 11/7/18 9:36 AM, Magnus Hagander wrote:
>> I'd like to propose that we drop the ability to register training
>> events
>> on the www.postgresql.org <http://www.postgresql.org> site.
>>
>> Reasons:
>>
>> 1. It's basically a system set up to game, where it helps to flood
>> with
>> a lot of instances and then cancel them (often without actually
>> removing
>> the registration on the website, and our moderators certainly don't
>> have
>> the time to go double-check which instances are actually happening
>> etc)
>
> Yeah, I think this is the crux of it: there are some people who are
> using the system fairly for promoting GA trainings with high-quality
> content that do happen. And then there is the above, which
> unfortunately
> is not just speculation but has been confirmed in multiple instances.
Would it be feasible to still have the training events, but blacklist
(in
an effective way) the abusers?
Just from the point of view of rewarding those people who do create good
value training + use the system fairly, and not punishing them because
some
places are abusing things.Somewhat surprised to not see a reply to Justin's reasonable comment.
I agree he should've received a response. Somehow that fell between the crack.
Why would we not just block everyone who acts unreasonably? Just as we do in other cases.
Mainly because (1) there would be a lot of effort to dealing with that (there was *already* a lot of effort to it), (2) there would be very very little content left at that point anyway, and (3) that content would still not be very useful.
Will we remove Hackers next because people post spam to it? That would save moderator time also. But obviously not a serious suggestion.
If hackers provided no substantial value to the community then yes, we would remove Hackers to save moderator time.
Blocking all training helps those who do NOT provide scheduled training, so the people who provide scheduled training have now been penalised for no reason.
The system that we had did NOT help companies that provided scheduled training, really. What it helped were those people who used the system to flood with training that most likely would not happen, in order to draw clicks to their website.
And in the end, the website is there to help our *users*, not the companies providing training. And it provided very close to zero values to those.
Should we have a way to connect our users with those companies providing training? Absolutely! But it should not be the one we had, because it simply did not work. The time spent on trying to manage that broken system can much more efficiently be spent on creating a new system that actually helps *both* our users *and* those companies that are serious about providing training services.
On 2018-12-07 22:06, Magnus Hagander wrote: > On Fri, Dec 7, 2018 at 10:41 AM Simon Riggs <simon@2ndquadrant.com> > wrote: <snip> >> Somewhat surprised to not see a reply to Justin's reasonable comment. > > I agree he should've received a response. Somehow that fell between the > crack. No worries. Wondered what happened. :) > Should we have a way to connect our users with those companies > providing > training? Absolutely! But it should not be the one we had, because it > simply did not work. The time spent on trying to manage that broken > system > can much more efficiently be spent on creating a new system that > actually > helps *both* our users *and* those companies that are serious about > providing training services. Sounds like whatever the next/new/better approach should be, maybe going with a whitelist approach? eg known good people/orgs have first access to it. Further access given on a case by case basis. Hmmm, similar to the way the wiki system is done currently. Bad behaviour gets places dropped from the whitelist. eg don't screw up (badly) ? That being said, this is just mostly general principles stuff and may not really add anything useful. :) + Justin