According to SiteMatrix we have 739 projects at the moment. There are three master partitions for the servers: s1 for enwiki only, s2 for 19 other projects and s3 for all the rest (that's 719 projects).
My homewiki is one of those 719 projects. And I feel a bit neglected. Replication is halted since 34 days. LuceneSearch 2.1 is active on enwiki since October and on dewiki and some other big wikis since December. Most other wikis have still no access to the new features. Even the "+incategory:" feature which is active on enwiki since April 2008 is not active on most wikis as of February 2009.
It seems, we are very low at the priority list.
Marcus Buck
Marcus Buck schrieb:
According to SiteMatrix we have 739 projects at the moment. There are three master partitions for the servers: s1 for enwiki only, s2 for 19 other projects and s3 for all the rest (that's 719 projects).
My homewiki is one of those 719 projects. And I feel a bit neglected. Replication is halted since 34 days.
You mean replication to the toolserver, i expect, so this is the wrong list, really. But anyway: this doesn't have anything to do with "second class" wikis -- replication broke when servers were moved, we are waiting for new hardware. It's due to arrive "really soon now".
Note that the third larges wiki, the French Wikipedia, on on s3 too.
-- daniel
Hoi, Can someone please explain why this is ? Thanks, GerardM
2009/2/1 Marcus Buck wiki@marcusbuck.org
According to SiteMatrix we have 739 projects at the moment. There are three master partitions for the servers: s1 for enwiki only, s2 for 19 other projects and s3 for all the rest (that's 719 projects).
My homewiki is one of those 719 projects. And I feel a bit neglected. Replication is halted since 34 days. LuceneSearch 2.1 is active on enwiki since October and on dewiki and some other big wikis since December. Most other wikis have still no access to the new features. Even the "+incategory:" feature which is active on enwiki since April 2008 is not active on most wikis as of February 2009.
It seems, we are very low at the priority list.
Marcus Buck
Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
Hi, If I read the original e-mail correctly there were two issues mentioned: long replication for s3 -- explained by the deleted logs -- and long delays before any new feature (search upgrades in particular) are enabled on what are perceived to be smaller wikis. I think it would be beneficial for everyone if a solution, explanation for the second issue was found, and we left the first issue be until the new servers arrive (or if they do not come for a long time it can be taken up again at a later time).
Thanks, Bence Damokos
On Sun, Feb 1, 2009 at 6:18 PM, Marcus Buck wiki@marcusbuck.org wrote:
According to SiteMatrix we have 739 projects at the moment. There are three master partitions for the servers: s1 for enwiki only, s2 for 19 other projects and s3 for all the rest (that's 719 projects).
My homewiki is one of those 719 projects. And I feel a bit neglected. Replication is halted since 34 days. LuceneSearch 2.1 is active on enwiki since October and on dewiki and some other big wikis since December. Most other wikis have still no access to the new features. Even the "+incategory:" feature which is active on enwiki since April 2008 is not active on most wikis as of February 2009.
It seems, we are very low at the priority list.
Marcus Buck
Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
On Mon, Feb 2, 2009 at 1:19 PM, Bence Damokos bdamokos@gmail.com wrote:
If I read the original e-mail correctly there were two issues mentioned: long replication for s3 -- explained by the deleted logs -- and long delays before any new feature (search upgrades in particular) are enabled on what are perceived to be smaller wikis. I think it would be beneficial for everyone if a solution, explanation for the second issue was found, and we left the first issue be until the new servers arrive (or if they do not come for a long time it can be taken up again at a later time).
For search features, I found this thread:
http://lists.wikimedia.org/pipermail/wikitech-l/2008-October/039861.html
I'm not clear on the exact reasons. My impression there is that the feature could not easily be enabled for all wikis at once (due to rack space issues?), so it was enabled one by one. Given that, it seems logical enough to start with the largest wikis to get the most feedback earliest in the process. This post refers to "hardware issues":
http://lists.wikimedia.org/pipermail/wikitech-l/2008-October/040022.html
This one says that some more RAM is being ordered:
http://lists.wikimedia.org/pipermail/wikitech-l/2009-January/040977.html
If this still isn't finished yet, obviously that's a problem, and I don't know what the answer is.
It's unrealistic to expect identical treatment for enwiki and bxrwiki, of course, but I don't see gross inattention toward the smaller wikis. I do use at least one pretty regularly, too, mediawiki.org (although of course that's in English). Mostly features get rolled out for all wikis at the exact same time. Of course, which features get developed to begin with is a separate story -- there's clearly a bias toward Western languages there, and especially English (look at category sorting for an example of that).
I still don't know, why the search features are not activated on all projects.
Aryeh Gregor wrote:
For search features, I found this thread:
http://lists.wikimedia.org/pipermail/wikitech-l/2008-October/039861.html
There I read: "hopefully in few weeks time we will finish the whole migration". That was four months ago. Aryeh Gregor mentioned two other posts, that spoke about "some hardware issues" and "I *think* we are in process of ordering some more RAM". Why is it so hard to get some precise answers? What are the hardware issues, when was the RAM ordered, when will it be installed? Sometimes I am under the impression, that our developers (not the volunteer developers, but the paid tech staff who do the ordering and all that stuff) do not even read wikitech-l. According to http://wikimediafoundation.org/wiki/Staff: Brion Vibber, Tim Starling, Tomasz Finc, Ariel Glenn, Rob Halsell, Mark Bergsma, Michael Dale, Trevor Parscal.
And Aryeh Gregor said this: "it seems logical enough to start with the largest wikis to get the most feedback earliest in the process". Okay, granted. You could get the same amount of feedback if you switched all the small projects first, but of course communication would be harder. So that's a valid argument for that "discrimination". But how about some "positive discrimination" as compensation? Today for example a user on my homewiki asked me, why "Special:AncientPages", "Special:WantedPages" and some other special pages aren't updated, they would be useful for maintenance. They were deactivated at different times in 2006, 2007 and 2008 when the updates began to take too much time on the servers. But I'm sure it was none of the small projects that caused the problems. It should be unproblematic, if the scripts were activated again on the small projects.
Marcus Buck
On Thu, Feb 5, 2009 at 5:05 PM, Marcus Buck wiki@marcusbuck.org wrote:
There I read: "hopefully in few weeks time we will finish the whole migration". That was four months ago. Aryeh Gregor mentioned two other posts, that spoke about "some hardware issues" and "I *think* we are in process of ordering some more RAM". Why is it so hard to get some precise answers? What are the hardware issues, when was the RAM ordered, when will it be installed? Sometimes I am under the impression, that our developers (not the volunteer developers, but the paid tech staff who do the ordering and all that stuff) do not even read wikitech-l. According to http://wikimediafoundation.org/wiki/Staff: Brion Vibber, Tim Starling, Tomasz Finc, Ariel Glenn, Rob Halsell, Mark Bergsma, Michael Dale, Trevor Parscal.
Unfortunately, I just have no idea what the answer is to this.
Okay, granted. You could get the same amount of feedback if you switched all the small projects first, but of course communication would be harder. So that's a valid argument for that "discrimination". But how about some "positive discrimination" as compensation? Today for example a user on my homewiki asked me, why "Special:AncientPages", "Special:WantedPages" and some other special pages aren't updated, they would be useful for maintenance. They were deactivated at different times in 2006, 2007 and 2008 when the updates began to take too much time on the servers. But I'm sure it was none of the small projects that caused the problems. It should be unproblematic, if the scripts were activated again on the small projects.
Actually, I believe there are several special pages that are enabled only on the small wikis, not on larger ones. I don't know why WantedPages isn't among those. There are other things that smaller wikis can get but larger ones not, like getting DPL enabled.
On Thu, Feb 5, 2009 at 2:05 PM, Marcus Buck wiki@marcusbuck.org wrote:
http://lists.wikimedia.org/pipermail/wikitech-l/2008-October/039861.html
There I read: "hopefully in few weeks time we will finish the whole migration". That was four months ago. Aryeh Gregor mentioned two other posts, that spoke about "some hardware issues" and "I *think* we are in process of ordering some more RAM". Why is it so hard to get some precise answers? What are the hardware issues, when was the RAM ordered, when will it be installed? Sometimes I am under the impression, that our developers (not the volunteer developers, but the paid tech staff who do the ordering and all that stuff) do not even read wikitech-l. According to http://wikimediafoundation.org/wiki/Staff: Brion Vibber, Tim Starling, Tomasz Finc, Ariel Glenn, Rob Halsell, Mark Bergsma, Michael Dale, Trevor Parscal.
I'm not entirely sure why you think that the technical staff need to report operational minutiae to you. If you wanted to know the status of getting Lucene 2.1 on small projects, perhaps it would be prudent to send a mailing list post asking what the status was, rather than a rant about "discrimination" against smaller projects. If you make an explicit request for information, there's no reason to suspect it won't be answered in a reasonable amount of time. If you lecture people about ignoring small projects, and bury your questions in that lecture, you are less likely to get a response.
It isn't that the technical staff don't care, it's just that Lucene 2.1 on smaller projects is one of a billion other things that tech staff need to work on, and without minimising the importance of other languages to Wikipedia's core mission, some quick stats show that about 70-80% of all hits come in English, German and French (totally arbitrary, tainted with Western selection bias, three big languages).
It should also be noted that, of those staff, only Brion, Tim, Rob, Mark, and to an extent Tomasz are involved in these sorts of operations matters. The rest are software developers.
Therefore, while you should, by all means, request information as to the status of certain operations things like this -- but posting outraged lectures on the importance of small languages isn't at all productive. You should split the problem up into its constituent problems, which are all separate (the Toolserver, Lucene, and query pages on small wikis), and try to have each dealt with by itself, rather than trying to lump it all together as a laundry list of complaints. You will have much better success in achieving your aims if you present your problems in this way.
Andrew Garrett
Andrew Garrett hett schreven:
I'm not entirely sure why you think that the technical staff need to report operational minutiae to you.
Cause it's their job to serve the community. And they haven't served a big part of the community since ten months in regard to Lucene. They don't need to report to _me_ but to the community.
If you wanted to know the status of getting Lucene 2.1 on small projects, perhaps it would be prudent to send a mailing list post asking what the status was, rather than a rant about "discrimination" against smaller projects.
I didn't "rant" and I didn't use the word "discrimination". I only used it in quotation marks after somebody else misinterpreted my words in that way. That misinterpretation occurred on foundation-l (where this thread was forwarded to [not by me]), so if you don't read foundation-l you couldn't know about that. My original post only contained "I feel a bit neglected" and "It seems, we are very low at the priority list". That's far from ranting. It's rather a "fact" after ten months.
If you make an explicit request for information, there's no reason to suspect it won't be answered in a reasonable amount of time. If you lecture people about ignoring small projects, and bury your questions in that lecture, you are less likely to get a response.
It isn't that the technical staff don't care, it's just that Lucene 2.1 on smaller projects is one of a billion other things that tech staff need to work on, and without minimising the importance of other languages to Wikipedia's core mission, some quick stats show that about 70-80% of all hits come in English, German and French (totally arbitrary, tainted with Western selection bias, three big languages).
That's basically "Small projects are no priority", isn't it?
It should also be noted that, of those staff, only Brion, Tim, Rob, Mark, and to an extent Tomasz are involved in these sorts of operations matters. The rest are software developers.
Therefore, while you should, by all means, request information as to the status of certain operations things like this -- but posting outraged lectures on the importance of small languages isn't at all productive. You should split the problem up into its constituent problems, which are all separate (the Toolserver, Lucene, and query pages on small wikis), and try to have each dealt with by itself, rather than trying to lump it all together as a laundry list of complaints. You will have much better success in achieving your aims if you present your problems in this way.
Andrew Garrett
"outraged lectures" Where did I give outraged lectures? I'm puzzled. I'm really are. And I don't know what the purpose of "splitting up" the problems would be. If Brion, Tim, Rob, Mark, and Tomasz do not read the list, that will make no difference. If they do, they can answer in response to this post:
Is it correct, that the delay in rolling out LuceneSearch on all wikis is due to RAM shortage? If so, is additional RAM ordered or when will it be ordered? If it is ordered, when will it be installed and when will the newest version of Lucene go live on all wikis? If it is not yet ordered, when will it be ordered, what is the expected time frame until Lucene will be fully active on all projects? If RAM is not the reason for the delay, what is the reason?
Will it be possible to activate regular updates for all special pages on smaller wikis? If so, when will this happen? If not, what's the reason?
Marcus Buck
Marcus Buck wrote:
"outraged lectures" Where did I give outraged lectures? I'm puzzled. I'm really are. And I don't know what the purpose of "splitting up" the problems would be. If Brion, Tim, Rob, Mark, and Tomasz do not read the list, that will make no difference. If they do, they can answer in response to this post:
I elected to answer this thread on foundation-l, three days ago, because I think that's a better forum for unproductive trolling. I'd rather keep this list for useful technical discussions.
-- Tim Starling
Tim Starling hett schreven:
I elected to answer this thread on foundation-l, three days ago, because I think that's a better forum for unproductive trolling. I'd rather keep this list for useful technical discussions.
-- Tim Starling
In that post you addressed Gerard's concerns, but none of the questions I asked.
Marcus Buck
Marcus Buck I don't agree with you.
The developers work for the Foundation and have to report to the foundation, the don't work for our community and don't have to report back to us.
Best regards,
Huib
Huib Laurens wrote:
Marcus Buck I don't agree with you.
The developers work for the Foundation and have to report to the foundation, the don't work for our community and don't have to report back to us.
How can everybody be so right, and still the sum is wrong?
Marcus has a valid concern, regardless of his clumsy way of presenting it. People have come to rely on Wikipedia, not necessarily on the factual content, but on the service being available, including database dumps being produced, the API functioning, including the various scripts on the toolserver. Many people don't realize how much of this still depends on voluntary efforts. Perhaps we've failed to tell them this. The annual fundraiser was a success and brought in $6 million. But that doesn't pay for the toolserver (does it?) or the s3 replication. When such things fail, it is very frustrating.
We could go further: People complain about uncivilized admins, scaring newcomers away. But the fundraiser doesn't pay for keeping Wikipedia civilized. Perhaps it should, and that could need an annual budget of $60 million rather than $6 million, a staff of 200 rather than 20.
Should we put up red warning signs on the services that are not provided by paid staff? Or the opposite, a small logo telling which services are run by paid staff, paid by donations to WMF? It would make the WMF tech staff look more like a corporate IT department. If it's not on the list of approved services, paid by the annual budget, we shouldn't expect it to work.
On Fri, Feb 6, 2009 at 7:32 AM, Lars Aronsson lars@aronsson.se wrote:
We could go further: People complain about uncivilized admins, scaring newcomers away. But the fundraiser doesn't pay for keeping Wikipedia civilized. Perhaps it should, and that could need an annual budget of $60 million rather than $6 million, a staff of 200 rather than 20.
That nothing is done about incivility (unfortunately - I really believe it is chasing some good contributors away), is not because of financial issues, but because nobody really has an idea _what_ to do about it.
On Fri, Feb 6, 2009 at 3:36 AM, Andre Engels andreengels@gmail.com wrote:
On Fri, Feb 6, 2009 at 7:32 AM, Lars Aronsson lars@aronsson.se wrote:
We could go further: People complain about uncivilized admins, scaring newcomers away. But the fundraiser doesn't pay for keeping Wikipedia civilized. Perhaps it should, and that could need an annual budget of $60 million rather than $6 million, a staff of 200 rather than 20.
That nothing is done about incivility (unfortunately - I really believe it is chasing some good contributors away), is not because of financial issues, but because nobody really has an idea _what_ to do about it.
-- André Engels, andreengels@gmail.com
Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
No, people have come up with ideas to try and deal with it. See enwiki's massive dispute resolution process. They just haven't found a way to successfully deal with it. In any case, this is veering off topic for this list.
-Chad
wikitech-l@lists.wikimedia.org