User talk:Jayen466/Archives/2014/November
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
The Signpost: 29 October 2014
- Featured content: Go West, young man
- In the media: Wikipedia a trusted source on Ebola; Wikipedia study labeled government waste; football biography goes viral
- Maps tagathon: Find 10,000 digitised maps this weekend
- Traffic report: Ebola, Ultron, and Creepy Articles
The Signpost: 05 November 2014
The Signpost: 12 November 2014
- In the media: Amazon Echo; EU freedom of panorama; Bluebeard's Castle
- Traffic report: Holidays, anyone?
- Featured content: Wikipedia goes to church in Lithuania
- WikiProject report: Talking hospitals
Precious
languages
Danke, merci, thank you, Andreas, capable of European languages and writing for several Wikipedias and their newsletters, for your contributions to quality articles such as Siege of Godesberg and The Seduction of Ingmar Bergman, for quoting "Wikipedia is a porch light for the moths of bias" and "answers are dangerous, they kill your wonder", for a careful look at sources and freedom of language, - you are an awesome Wikipedian!
- Thanks, Gerda. --Andreas JN466 16:44, 27 November 2014 (UTC)
Your wikimedia-l post on the 2005 budget
I don't answer on wikimedia-l, because it's a cesspool, but i happened to come across your recent post comparing 2014 to 2005. A few points, that you are conveniently not taking into account:
- In 2005 we had 500 000 articles, now 4,6 million. A factor 9 increase
- Those numbers do not include discussion that is 6x times more than we have articles
- For each page there are on average 21 historic revisions, so that increase actually counts another 20 times.
- The above doesn't even account for all the other languages and sister sites.
- We had almost NO images included or uploaded, now we have some 35 million of them.
- Our codebase was 200 000 lines of code, now we are at some 3 million (hard to tell, too many extensions to accurately aggregate). The core alone (1.3 million LoC) would require 341 man years to build, at a 60 000 dollar salary (rather low for high tech industry actually), that is an estimated $ 20 million cost.
- Our devs still sort of understood all parts of our code, now 1 person hardly can understand 1 extension.
- Because of this everything now needs to be documented and maintainable (not really needed before 2005 and something we still haven't achieved because it wasn't needed back then).
- We need to be up around the world, 24/7, redundant and need to have backups of all of that data.
- We need to move forward with things like IPv6, HTTPS etc
- This complexity of the system is so large, that for volunteer developers it seems to be a lot more difficult to contribute in a significant a way as they did in 2005 (gut feeling). So comparatively we have a lot less volunteering on a technical level. (also for legal reasons, with regard to access to servers with private data for instance).
My point is: complexity (limited to technical issues) increased exponentially on all branches of the tree, not just the leaf called pageviews. It might not always be visible to many people, but it is reality and that complexity is mostly for very good and very logical reasons. It was wonderful what we were able to do back then with such limited resources, but it was not sustainable, even back then. That's WHY Brion got hired in the first place. His hiring was the first admission of that problem. Actually, a lot of the problems we have now, are actually the result of not investing enough back then. We have a file database structure, that all the devs would love to fix, but we can't because it's probably going to be a 10 month period of hurt to do so. Had we had enough time in 2006, it could have been fixed in a few days at that time. But 2014 is not 2006.
I'm sure that the legal department, the personnel department and all other various aspects of our organization have similar problems with increased complexity. Sure you can run something on a budget, that is no problem, but can you keep it alive in the years that follow, can you sustain it's natural growth over the years, let alone the growth that you desire by way of your mission. We are no longer adventurers blazingly running forward, we have become establishment and that comes with responsibilities, complexity and conservatism. Are there departments that run less than efficient ? I'm sure of it, show me a work place that doesn't. Should we strive to fix that ? Of course we should. Question what gets spend, how to spend and how to acquire those funds, but don't pretend that where we are now compares to 2005 in any way whatsoever, because they don't even come close to each other... We are in a different ballpark and pining for the old cozy days, where all the editors knew who Jimmy, Magnus, Tim and Brion were, ain't gonna change that. —TheDJ (talk • contribs) 14:31, 27 November 2014 (UTC)
- Thanks. The point of my post was simply that the fundraising banners create the impression that money is needed to keep Wikipedia online and ad-free. This makes people think that if they do not donate, Wikipedia will go offline, or will have to carry ads next year in order to survive. This simply isn't a reflection of the current situation. Looking at the maths, according to my reckoning the Foundation has taken around $165 million over the last four years (from summer 2010 to summer 2014), about ten times as much as it took over the first seven years of its existence. Wikipedia was a copiously illustrated, top-10 website in 2008/2009, when overall expenses for the year were $5.6 million. I'd rather the fundraising banners gave a more meaningful and transparent reflection of what the money is wanted for (which is mostly about professionalising its software engineering and outreach, rather than keeping Wikipedia online and ad-free). And, as an afterthought, it might be worth mentioning in the banners that Wikipedia is written by unpaid volunteers, people just like those prospective donors. On the one hand, it might help with recruitment, and on the other it would help counter the misconception that the Wikimedia Foundation's staff actually oversee content generation on Wikipedia. Regards, --Andreas JN466 16:08, 28 November 2014 (UTC)
The Signpost: 26 November 2014
- Featured content: Orbital Science: Now you're thinking with explosions
- WikiProject report: Back with the military historians
- Traffic report: Big in Japan