Jump to content

User talk:Jimbo Wales

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by A1candidate (talk | contribs) at 21:59, 24 May 2015. The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.


    Page Weight Matters

    Background: Page Weight Matters, by Chris Zacharias

    "Three years ago, while I was a web developer at YouTube, one of the senior engineers began a rant about the page weight of the video watch page being far too large. The page had ballooned to as high as 1.2MB and dozens of requests. This engineer openly vented that “if they can write an entire Quake clone in under 100KB, we have no excuse for this!” Given that I agreed with him and I was excited to find a new project, I decided to champion the cause of getting the YouTube watch page to weigh in under 100KB. On the shuttle home from San Bruno that night, I coded up a prototype. I decided to limit the functionality to just a basic masthead, the video player, five related videos, a sharing button, a flagging tool, and ten comments loaded in via AJAX. I code-named the project “Feather”.
    "Even with such a limited set of features, the page was weighing in at 250KB. I dug into the code and realized that our optimization tools (i.e. Closure compilation) were unable to exclude code that was never actually used in the page itself (which would be an unfair expectation of any tool under the circumstances). The only way to reduce the code further was to optimize by hand the CSS, Javascript, and image sprites myself. After three painstaking days, I had arrived at a much leaner solution. It still was not under 100KB though. Having just finished writing the HTML5 video player, I decided to plug it in instead of the far heavier Flash player. Bam! 98KB and only 14 requests. I threaded the code with some basic monitoring and launched an opt-in to a fraction of our traffic.
    "After a week of data collection, the numbers came back… and they were baffling. The average aggregate page latency under Feather had actually INCREASED. I had decreased the total page weight and number of requests to a tenth of what they were previously and somehow the numbers were showing that it was taking LONGER for videos to load on Feather. This could not be possible. Digging through the numbers more and after browser testing repeatedly, nothing made sense. I was just about to give up on the project, with my world view completely shattered, when my colleague discovered the answer: geography.
    "When we plotted the data geographically and compared it to our total numbers broken out by region, there was a disproportionate increase in traffic from places like Southeast Asia, South America, Africa, and even remote regions of Siberia. Further investigation revealed that, in these places, the average page load time under Feather was over TWO MINUTES! This meant that a regular video page, at over a megabyte, was taking more than TWENTY MINUTES to load! This was the penalty incurred before the video stream even had a chance to show the first frame. Correspondingly, entire populations of people simply could not use YouTube because it took too long to see anything. Under Feather, despite it taking over two minutes to get to the first frame of video, watching a video actually became a real possibility. Over the week, word of Feather had spread in these areas and our numbers were completely skewed as a result. Large numbers of people who were previously unable to use YouTube before were suddenly able to.
    "Through Feather, I learned a valuable lesson about the state of the Internet throughout the rest of the world. Many of us are fortunate to live in high bandwidth regions, but there are still large portions of the world that do not. By keeping your client side code small and lightweight, you can literally open your product up to new markets."

    Source: [ http://blog.chriszacharias.com/page-weight-matters ]

    (Emphasis added, capitalization in original.)

    (Reproduced under fair use: "The first factor is regarding whether the use in question helps fulfill the intention of copyright law to stimulate creativity for the enrichment of the general public, or whether it aims to only 'supersede the objects' of the original for reasons of personal profit.")

    Given the above, I think that we should start a project -- a real project with measurable goals and a schedule -- to reduce page weight. --Guy Macon (talk) 22:53, 21 May 2015 (UTC)[reply]

    I would support this 100%. The amount of stuff that comes from bits.wikipedia.org seems to be very large, for example. However WP does well by not having dozens of external gadgets and widgets and trackers like many sites. All the best: Rich Farmbrough, 15:15, 22 May 2015 (UTC).
    I think our situation is a bit different now, because of the use of mobile. Folks in Siberia and Africa are likely now viewing Wikipedia on mobile (likely somewhat faster) than on just a slower version of our desktops. How weight affects mobile is something I don't really know. Perhaps the question might be whether we want different versions for regular vs. mobile articles? BTW, in general I do think about 5% of our articles go on too long - even if you can quickly skim a long article, there is a limit to that.
    I do want to ask a related question - I'd asked at the village pump and editors didn't have much to say there. I'm working on a very long list, List of municipalities in Pennsylvania. There are 2562 munis in PA (likely the largest number in the US) and were 10 fields in the table. It sorts wonderfully and is very useful. I was quite surprised by this because at WP:NRHP we always would break up lists of 250+ because they loaded too slow and caused other problems like long times in saving the page. That was about 4 years ago though.
    Someone suggested putting photos on the list, and we're now doing that as an informal project. Less than 1,000 pix to go. I don't see the size with the pix causing any slow loading or saving. The article size is now 320,686 bytes about twice the size of today's featured article on Rodents. My connection (in the US) is not particularly fast. But is this load going to cause other people problems? (Not too many Siberians will want to see the list of PA munis, but you never know) Will all the pix cause problems on mobile? Not that I'm thinking of doing it on this list, but how does video affect things?
    --Smallbones(smalltalk) 16:48, 22 May 2015 (UTC)[reply]
    Note that I removed the hatting on my related question. I had asked it before at the Village Pump with no real help forthcoming. I don't think that the Help Desk would do any better because it is a mixture of technical and policy questions. To spell it out, it is now possible to make a 3,000 item list/table on Wikipedia and include a few bells and whistles like photos and wiki-links. 1) How big can a table like this get, and how many bells and whistles can be added, before people start having difficulty in loading or saving it? 2) How well does an article like this work on mobile? Do we want to have different versions of lists like this for mobile? Smallbones(smalltalk) 21:56, 22 May 2015 (UTC)[reply]
    Please ask your totally unrelated question in a new section. Hijacking my discussion is very rude, and I would advise others here to not reward your rude behavior by responding to you. --Guy Macon (talk) 05:13, 23 May 2015 (UTC)[reply]
    Optimizing for mobile (limited CPU power, small screen) is a different engineering challenge than optimizing for slow / expensive per-byte connections. I do a lot of work involving third-world countries, and in the very remote areas it is far more common to see a hundred laptops connecting through a single WIFI connection to a single DSL than it is to see a hundred smart phones connecting to a nearby cell tower. --Guy Macon (talk) 20:14, 22 May 2015 (UTC)[reply]

    Thanks Guy that was a fascinating story and very very relevant for us.--Jimbo Wales (talk) 17:31, 22 May 2015 (UTC)[reply]

    I have made a couple of proposals to try to get such a project started (or at least considered), but they have gone nowhere:
    I could, of course, post a policy RfC and get several hundred people to tell the WMF that yes, the English Wikipedia wants this to happen (and three people who support longer page load times...) and then spend a year pounding in the basic concept that unless there are published, measurable goals and a schedule it isn't a real project, but I would much prefer the WMF to support it because it is an obviously good idea. --Guy Macon (talk) 20:14, 22 May 2015 (UTC)[reply]
    There's at least 1 question hidden in here that I don't think is obvious, though I'm not likely to be one of the 3 peeople who support longer load times. 1st I do think there is a lot of brick-a-brack in some articles like long infoboxes, and templates that link to every member of some group of topics (perhaps 3 topics of 80 items each). Those could obviously be removed - but how much "weight" is thereby removed? Is it just the article size (that you see in the article history page) that matters, or is it calls on computing power from templates that most matters? Rephrasing: do we want to minimize the number of characters typed in the file or the number of templates?
    Also, there is material like photos and videos that I think are underused on Wikipedia. "A picture is worth a thousand words" may not apply to every article, but it probably does in articles on artists, museums, wildlife, and many other subjects. Our almost complete lack of video makes us one of the "least modern" websites around. I've met folks here who just don't see video as being an educational tool - everything that they consider to be educational can be written in words. I think that's dead wrong, consider which of the following gives more useful information a) an IPA phonetic transcription of how a person's name is pronounced, b) an audio recording of his or her voice pronouncing their name, or c) a 15 second video of the person introducing themselves? It's obvious to me that c is a clear winner.
    So does this "minimizing weight" project mean returning to a low image, no video encyclopedia? That's a tradeoff that is more a policy issue than a technical one. I couldn't buy into that. Smallbones(smalltalk) 21:56, 22 May 2015 (UTC)[reply]
    Actually, the number of characters is usually a very small portion of page weight. You can have a very large page that is pure text that has a very small page weight. It can be annoying for a human to read all of that, but it still loads very quickly. Templates don't matter other than the extra text and/or images they add to a page. At the level I am talking about, the only things that matter are the number of HTTP requests a page makes and the total number of bytes sent to the browser (the page weight).
    Here are some page weight resources for the curious, but of course the Wikipedia developers I am trying to reach already know all about page weight.
    --Guy Macon (talk) 05:13, 23 May 2015 (UTC)[reply]
    I'm not an expert, but I think a simple way to lose "weight" is to lose Javascript. This can of course be done with a browser option, but not everyone knows how to use it. Also, even on the en.m. version of this page with scripts disabled, I still see script loaders in the source code. I know that shutting off scripts improves speed and has very few downsides - you lose sortable buttons, and no video I know of but animated gifs works without it, hidden text isn't hidden by default, and the Lua editor doesn't work. But 99.9% of the time this is of no importance, and there's some extra peace of mind about security. I think a Javascript-free version of each page (en.nojs.wikipedia?) would be useful. Wnt (talk) 15:16, 23 May 2015 (UTC)[reply]
    Again, how to reduce page weight is not the issue. Any competent engineer can do it without suggestions from us, and the WMF engineers have already done a lot of the easy stuff. What is left are things like this:
    http://en.wikipedia.org/wiki/Main_Page has GZIP compression enabled.[1]
    https://en.wikipedia.org/wiki/Main_Page does not.[2]
    The question is, is it possible to enable compression for HTTPS without opening ourselves up to a BREACH (security exploit)?
    This requires research by a good engineer, and is not an issue that asking for opinions on Jimbo's talk page is likely to solve. As interesting as the topic is, we need to stop wasting our time talking about how to reduce page weight and concentrate on my proposal, which is to get someone at WMF assigned to working on page weight with with measurable goals and a schedule. They already know how to do it. --Guy Macon (talk) 17:00, 23 May 2015 (UTC)[reply]
    Are you sure that website works with HTTPS? I get my secure pages with gzip without issue. -- [[User:Edokter]] {{talk}} 17:22, 23 May 2015 (UTC)[reply]
    I am using a customer computer at a remote site so I only have online tools at the moment. I looked up several more online tools. http://www.gidnetwork.com/tools/gzip-test.php shows the https as compressed, several others call the https version bad URL, and http://checkgzipcompression.com/ shows it as uncompressed. I will update this when I get home and am able to use tools that I trust. Until then I will assume that the gzip-test page is correct.
    At the risk of straying even farther from my own topic, it would be interesting to see why the WMF engineers don't think BREACH (security exploit) is a problem. I am expecting something like "we looked into that and determined that using compression is safe in our situation". As I have often observed, they are pretty sharp and are unlikely to have missed something like this. --Guy Macon (talk) 22:39, 23 May 2015 (UTC)[reply]
    I think checkgzipcompression.com is broken. Another test of richardstoolbox.com (verifysslcertificate.com) does not even recognize that Wikipedia has an SSL certificate. My info comes directly from the HTTP heades I receive. -- [[User:Edokter]] {{talk}} 23:11, 23 May 2015 (UTC)[reply]

    Quixotic plea

    You are invited to join the discussion at Wikipedia talk:Wikipediholism test. Thanks. — {{U|Technical 13}} (e • t • c) 06:31, 23 May 2015 (UTC)Template:Z48[reply]

    RfC on acupuncture

    Since you've previously made some strongly worded comments on alternative medicine, I feel that your input over here might be useful. -A1candidate 06:38, 23 May 2015 (UTC)[reply]

    I will quote the substantive part of Jimbo's comment, which is indisputable: "If you can get your work published in respectable scientific journals - that is to say, if you can produce evidence through replicable scientific experiments, then Wikipedia will cover it appropriately." His well-known "lunatic charlatans" comment about people who operate outside that scientific framework was clear, and in my opinion, does not need to be reiterated. If Jimbo chooses to elaborate, then fine. Cullen328 Let's discuss it 06:56, 23 May 2015 (UTC)[reply]
    My concern is that many editors, including at least one administrator, are misusing Jimbo Wales' comments. Notice the way in which a longstanding administrator quoted this essay during a recent arbitration request. -A1candidate 07:25, 23 May 2015 (

    It's probably worth noting that the OP has just been given a 0RR restriction on acupuncture and 1RR on alt med more widely, by an uninvolved admin, and that he has tried several times to use Wikipedia processes to silence opponents, and failed every time. The "clean hands" doctrine is relevant here. Guy (Help!) 21:57, 24 May 2015 (UTC)[reply]

    I will appeal the restriction - because I have "clean hands". -A1candidate 21:59, 24 May 2015 (UTC)[reply]

    With regards to this...

    ...and with the choice of "Fight the power"... do you intend to declare war on your homeland? :-P

    Other than that, considering British politicians were also previously interviewed, couldn't there have been more talk about their proposals to restrict and monitor the Internet? Not to mention banning HTTPS... -- Mentifisto 11:04, 24 May 2015 (UTC)[reply]

    Your comment in Radio Times

    "When I first launched Wikipedia on 15 January 2001" [3]. This is the truth, but not the whole truth, and highly misleading. It was Larry Sanger who launched the wiki, under the domain Nupedia.com, on January 10 2001. This was immediately followed by a sort of mutiny among the Nupedia volunteers. Carl Anderson, Professor of Classical Studies, Michigan State University, called the wiki 'silliness'. Gaytha Langlois, Professor of Ecology, Bryant College, Smithfield, thought it was merely a “variation on structured chat rooms”. Nearly all the academics objected, apart from John Horvath.

    To resolve this, you suggested installing 'the wiki' under a totally different brand name. "That way, we separate the wiki from the Nupedia brand name. It is very important to all of us who have an emotional stake in Nupedia that we not harm the reputation of Nupedia". Shortly after (January 12th) you dismissed the wiki as a sort of chat room, and not even that. " The wiki software, in its current incarnation, is so wide open that it is hard to see it’s [sic] purpose other than as a chat room mechanism of sorts. Even then, I don’t know ". Larry came up with the name ‘Wikipedia’, and the wiki was re-launched under Wikipedia.com, January 15 2001.

    In your closing speech at Wikimania 2014, you said that truthfulness is something that is 'very, very valuable' . "Obviously truthfulness is very important if you're writing an encyclopedia, because you're not supposed to just make stuff up." Well truthfulness starts at home, right?

    Peter Damian (talk) 13:12, 24 May 2015 (UTC)[reply]

    So he told the truth but did not include the entire history of Wikipedia? What is your point? Chillum 13:21, 24 May 2015 (UTC)[reply]
    See the whole truth. Peter Damian (talk) 13:26, 24 May 2015 (UTC)[reply]
    While many people find Wikipedia a great place to spread truth, that is not really why we are here. Chillum 13:48, 24 May 2015 (UTC)[reply]
    "Obviously truthfulness is very important if you're writing an encyclopedia, because you're not supposed to just make stuff up." Peter Damian (talk) 13:57, 24 May 2015 (UTC)[reply]
    • Comment - I think this is more about giving due credit to other people who've also played a key role in creating Wikipedia in its original form. It is not about "spreading the truth", but simply setting the record straight. I doubt Jimbo Wales had any intention to hide the truth, but I have to say that unless one is the sole creator of something or some idea, it's always more ethicial to share credit with others ("We launched Wikipedia") rather than crediting yourself only ("I created Wikipedia"). I'm not too familiar with the history of Wikipedia, but if Peter Damian's statement is correct, then I tend to agree with him that "truthfulness starts at home" and we should focus on being more truthful instead of holding the moral high ground and going as far as to accuse others of deceit. -A1candidate 14:34, 24 May 2015 (UTC)[reply]
    • In fairness, please read Jimbo's entire sentence instead of the OP's selected snippet: "When I first launched Wikipedia on 15 January 2001 there was only a very small number of people giving their time to write articles, but we had a big ambition: to give everybody in the world free access to the sum of all human knowledge through an online encyclopedia edited entirely by volunteers." He is talking about other people also, and uses the word "we". Cullen328 Let's discuss it 17:15, 24 May 2015 (UTC)[reply]
    • This is still not consistent with Jimbo's statement at the time that "The wiki software, in its current incarnation, is so wide open that it is hard to see it’s [sic] purpose other than as a chat room mechanism of sorts. Even then, I don’t know ". I don't see how that is consistent with the "we had a big ambition" bit. Or was he referring to Nupedia in the BBC interview? Peter Damian (talk) 19:08, 24 May 2015 (UTC)[reply]
    • When you were unblocked, it was obvious that most of the people who (ill-advisedly in my view) voted for it thought you would spend your time more productively in future trying to improve the encyclopedia rather than engaging in the drama and attention-seeking that got you kicked out in the first place. So far you don't seem to be doing a great job proving them right. Prioryman (talk) 21:26, 24 May 2015 (UTC)[reply]
    Yes, this might be a case of trying to right a perceived great wrong. Liz Read! Talk! 21:59, 24 May 2015 (UTC)[reply]
    Retrieved from "https://en.wikipedia.org/w/index.php?title=User_talk:Jimbo_Wales&oldid=663861387"