Talk:Strategy/Wikimedia movement/2018-20/2019 Community Conversations/Product & Technology

Current situation

Software developers are hard to integrate

It is not so easy to become an open-source developer (in MediaWiki). Software development on Linux seems more easy to do? MediaWiki seems a closed user environment? And there seems to be few progress on new projects/bug fixing is taking too much elapsed time? e.g. the QR codes interface has been broken for almost a full year now without being solved. Should be an easy one to fix? Geert Van Pamel (WMBE) (talk) 18:56, 25 March 2019 (UTC)

Mobile editing high cost benefit ratio

Editing software for mobile devices has a too hight cost/benefit ratio. Mobile editing us much more difficult relative to the laptop platform. Also the screen of smartphones is too small and the Android platform is not so handy to edit with. So I would propose to have more energy available for standard platform software development. Geert Van Pamel (WMBE) (talk) 18:56, 25 March 2019 (UTC)

Good points - both of them. Thanks for bringing this up. Stephane (talk) 16:27, 25 April 2019 (UTC)

Disagree. (1) While it is true mobile is not the optimal editing space, check the global ownership for laptop and mobile phone, you will realize in most part of the worlds people own mobile phone in stead of a laptop/computer. In which case mobile editing would be the only viable option. Other complex wikitext editing can be done by other editor and they can focus on just write. (2) Mobile editing tool can be used for iPad and similar tablet devices, which can be quite convenient in some editing environment, and the full power of a computer is usually not necessary to just edit wikipedia. Viztor (talk) 15:33, 11 July 2019 (UTC)

Why this scope

Funding?

I think that "Funding" belongs to Resource Allocation. Every aspect of Wikimedia needs a little or a lot of funding. But resource allocation has to be centralised. Parts have to ask, with properly reasoned documents. And then a central unit will have to distribute the limited resources (even if we had a trillon dollars, that would be a limit) to cover the needs of the diferent parts. That's nothing new: it's a definition of Economics. B25es (talk) 18:33, 11 April 2019 (UTC)

Hello @B25es - we put funding because at this stage of our discussions we weren't sure of how other groups (starting with RA but also and most importantly Revenue Streams) had included P&T in it. Cheers, Stephane (talk) 16:24, 25 April 2019 (UTC)

Key questions

CTO criteria

Dear Product and Technology working group:

Please approve use of the following criteria for CTO search, qualification, hiring, and evaluation. Please attract and retain a CTO who can:

(1) reinforce the Foundation's privacy infrastructure;

(2) move away from providing personally identifiable information to the dozens of researchers worldwide under nondisclosure agreements;

(3) find a fuzzing middle ground to provide approximate but non-personally identyding readership log information;

(4) explore alternatives to staying with PHP long term;

(5) build a strategy to combat censorship in Turkey and China;

(6) support IPFS with more than just dumps;

(7) execute on the industry-wide Encrypted-SNI project with devoted headcount and budget;

(8) commit to open source hardware, e.g. opencompute.org servers;

(9) ramp up Community Wishlist implementation as long term Foundation technology projects wrap up or transition to support mode;

(10) fix tools that have fallen into disrepair, e.g., Categorder which sorts the en:w:WP:BACKLOG categories by pageviews;

(11) produce a Course Management System for Wikiversity;

(12) produce a pronunciation tutor for Wiktionary; and

(13) remain competitive with other top-ten website compensation by paying SF-livable salaries.

Thank you for your kind help with this request. James Salsman (talk) 03:45, 24 April 2019 (UTC)

Hi James, thanks for the suggestion(s) but I think framing job descriptions for the Foundation is a little bit outside of our immediate remit. This being said, I agree that some of the questions you raise have an impact or certainly are important for the movement and we shall certainly take them into account when framing our recommendations. Stephane (talk) 16:24, 25 April 2019 (UTC)

Thank you for your quick reply, and for considering the points raised. I would hope that setting long-term strategic goals is as much within your remit as recommending that the next CTO, whose search is currently underway, be evaluated in accordance with their ability to achieve them. James Salsman (talk) 17:32, 26 April 2019 (UTC)

You can continue to hope that it is within their remit if you wish, but they have told you that it is not. --Deskana (talk) 08:10, 28 April 2019 (UTC)

@The other Kiwix guy: could you say a little bit about why such recommendations wouldn't fall within the first sentence of your WG's scope, "focus and prioritize the development of structures for continuous communication and ongoing connections between product / technology teams, groups, communities and other stakeholders...."? James Salsman (talk) 20:41, 28 April 2019 (UTC)

Hi James, re: your comment below that these should be considered as strategy suggestions regardless of the CTO search, in my personal opinion:

• (1)-(3) lack context - do you see some strategic threat or risk that you would heed off this way? Wikimedia already has pretty high standards of privacy and does publish anonymized data, and the drawbacks of providing access to non-anonymized data to select researchers are not obvious to me.
• (4) seems like a better question for an expert forum such as TechCom.
• (5) is something we intend to look at on a more abstract level, see our last scoping question ((6)-(7) are related but too specific to be useful as long-term strategy);
• (8) lacks context - do you see some strategic threat or risk that you would heed off this way, or some opportunity to make use of? Or is this a matter of values?
• (9) I understand as saying that the wishlist has more mission impact and resources should be redirected there as immediate projects run out (e.g. when the Growth team's current project ends, they should focus on wishlist tasks instead of a new editor retention project) - is that what you meant? If so, that basically implies that a vote is better at priority setting than experts such as product managers, which seems like a questionable claim. We do intend to look at how task priorization can be made more inclusive, though. ((10)-(12) are related but too specific to be useful as long-term strategy);
• (13) lacks context - do you see some strategic threat or risk that you would heed off this way? (FWIW as someone living in SF on a WMF salary, I'd say having a salary competitive with the tech top10 and having a livable salary are two very different things, and the latter is already the case.)

--Tgr (talk) 15:07, 6 June 2019 (UTC)

Please approve A/B mobile user interface design test

Dear Product and Technology Working Group members, please approve Community Wishlist Survey 2017/Archive/Replace or supplement mobile pencil icon with "edit" in square brackets and A/B test editing uptake. Thank you for your kind help with this request. James Salsman (talk) 04:18, 24 April 2019 (UTC)

Well I'm all for canvassing and stuff, but replacing a pencil icon certainly hasn't much to do with a 2030 strategy, does it? :-) Also why the heck post a link to a request that was denied in 2017? Stephane (talk) 16:24, 25 April 2019 (UTC)

You're right that the request itself is tactical matter, but I hope it can serve as an example of why the strategy of measuring user experience choices is better than the alternative of proceeding from ignorance. The request was originally denied because it required clearance from the design team, which was not part of product, and who I haven't been able to reach. I hope that by raising this example of a question which would have been answered under an ongoing strategy of design by measurement, that it will be easier for the organization to commit to the strategy. I'm completely convinced that the idea has merit and will result in more editing. While the pencil icon is widely understood to represent textual changes, the "anyone can edit" motto is more psychologically coupled with "[edit]". James Salsman (talk) 17:32, 26 April 2019 (UTC)

Privacy, tracking, and research partners

Today the assumption that Wikipedia readers are not tracked was stated by a subject matter expert in a different forum, and obliquely confirmed by a senior Foundation official. So (and partially in regard to the CTO criteria I requested above, in case they are disregarded wholesale) I want to remind the Working Group that dozens of the Foundation's research partners world-wide, who are covered by nondisclosure agreements that impose no specific data security requirements and who along with their coworkers are subject to subpoenas and national security letters, hold "the complete ... server logs from Wikipedia ... about 14 terabytes of raw logs per month" including the "IP address, proxy information, and user agent" of "Wikimedia’s full server logs, containing all HTTP requests to Wikimedia projects," as per pp. 19-20 here. James Salsman (talk) 16:53, 5 June 2019 (UTC)

I'd suggest taking this up with the Partnerships group, it's outside our focus. --Tgr (talk) 19:45, 5 June 2019 (UTC)

Done. If the first three items I listed on the #CTO criteria above happened, then this kind of data sharing could continue without involving privacy concerns. I'm sure that's a Product and Technology concern. I'd like to reiterate that list is strategically important to the Movement whether CTO criteria are considered in scope or not. James Salsman (talk) 00:20, 6 June 2019 (UTC)