User talk:Proteins/User scripts archive

Workshop, scripts etc.

Hi there! I was going to write you about the workshop soon anyway... I knew it was coming up soon. Just let me know if I can help out in any way! And thanks for letting me know about the scripts project; I'll take a look at them later. (Though if workshop participants are truly Wikipedia novices, I think starting off with scripts might be confusing? Depends on what you're going to cover, I guess). Best, -- phoebe / (talk to me) 19:26, 22 October 2008 (UTC)[reply]

Hi Phoebe, it's nice to hear from you! I didn't mean for the scientists to write scripts, but rather to use them for easier editing and to make their user experience more enjoyable and productive. My first scripts were meant to improve Wikipedia's accessibility and to give me practice before writing scripts for WYSIWYG editing. I've started that already, as you can see from User:Proteins/hottext.js and related scripts. They aren't working just yet, but the idea is that you can add text, images, references, tables, etc. wherever you want to. I'll try to write the script to ensure that the scientists' contributions comply with the Manual of Style automatically.

Still, I would appreciate whatever feedback you have on the first three scripts. It was fun for me to hear some Featured Articles read aloud using Fire Vox, and to finally understand all those interwiki languages in the lefthand column. Proteins (talk) 19:42, 22 October 2008 (UTC)[reply]

Script assessment and development

Hey, Proteins! I'm excited to work on these scripts and think about ways to improve them. I have already added the Fire Vox one and started listening to articles. It is excellent. I was wondering if you were aware of the "readability" tool we use at FAC. See, for example, this analysis of History of a Six Weeks' Tour, my latest FAC. Awadewit (talk) 18:46, 24 October 2008 (UTC)[reply]

Thank you, Awadewit, I had seen those readability scores but I haven't taken the time to understand them yet. The server seems very slow, at least for longer articles such as your FAC. A script might be able to do the same readability analysis faster and section-by-section to localize readability problems. Unfortunately, I'm pressed for time at the moment on- and off-wiki. Here I'm pre-occupied with debugging two scripts that unindent Talk pages for accessibility; the scripts correctly identify the indent level and color the text appropriately, but the HTML surgery required to unindent the text is difficult (for me). Once I fix that problem, I'll work on a set of five scripts for pseudo-WYSIWYG editing (adding sections, text, images, tables and references directly into the article) that I hope to finish for the December ASCB workshop. I'm glad you liked the FireVox script and if you catch any bugs, please let me know. Thanks, Proteins (talk) 21:32, 24 October 2008 (UTC)[reply]

User:Proteins/Start new article

Hi Proteins, Have you seen the new article wizard? That should give you some ideas for making an easy-to-use article creation interface. The only HTML forms supported in MediaWiki are inputboxes - see Help:Inputbox for more info on that. Just a bit of friendly wikistalking ... Graham87 00:55, 6 November 2008 (UTC)[reply]

Thanks, Graham, that's helpful! I was just being confused by the documentation at mw:Extension:Inputbox on MediaWiki when your friendly message showed up. The new article wizard seems similar to what I wanted to write, but I would like to be stricter with the scientists. For example, I'd like to require that they explain the significance of their topic and give some references for it before the article ever gets created. After working on the AfD script and that dictionary of acronyms, I was impressed by how easy it is to get AfD'd after creating an article, like WP:A7, and I wanted to spare them that grief on their first article. I was also going to engineer the script so that their article is created in their user Sandbox, and then moved. That would forestall a few conflicts, I foresee. Ideally, the script would provide a form for entering scholarly references and maybe even open a convenient PubMed search tab. Proteins (talk) 01:16, 6 November 2008 (UTC)[reply]

Script idea

Could you write a script that translates American English into British English and vice versa and all of the variations? See WP:ENGVAR. That would be wonderful. If you could do an entire WP:MOS script, that would be even better, but I realize that would be much harder. Awadewit (talk) 19:44, 17 November 2008 (UTC)[reply]

That would be relatively easy, if it involves changing only the spelling of a defined list of words, or if a well-defined conversion rule could be stated. I've already written a script that analyzes an article into its component words; it would just be a question of looping over the words and doing the conversion.

As I'm sure you realize, the quality of the conversion will depend strongly on the accuracy of the conversion rule or dictionary. For example, you might propose the rule that every word ending in -our (neigbour) or -re (centre) should be converted to or and er, respectively. However, that rule leaves out "neighbourly" and "centres". The script could check for those special cases (-s and -ly), but is there a definite rule or list that the script could adopt? Proteins (talk) 19:57, 17 November 2008 (UTC)[reply]

PS. I am planning an MoS script, beginning with overlinking, as part of a set of scripts to aid in producing more FAC's with more light and less heat. ;) Ideas are welcome. Proteins (talk) 20:04, 17 November 2008 (UTC)[reply]

I'm sure you know about Brighterorange's dash script - you could integrate that. That is definitely a place to start. WP:MOSQUOTE seems like it would be relatively easy. There is also WP:MOSHEAD and WP:ACCESS. SandyGeorgia would have some good ideas as well. Awadewit (talk) 20:12, 17 November 2008 (UTC)[reply]

Try User:Proteins/mosheadcheck.js in this sandbox. Cheers, Proteins (talk) 22:20, 17 November 2008 (UTC)[reply]

Do you mind if I ask some FAC regulars to check it out as well? Awadewit (talk) 22:25, 17 November 2008 (UTC)[reply]

I've no objections, but please bear in mind that this script is grossly rudimentary, the work of an hour or two. I wouldn't want people to form a bad first impression of it — or of me! ;) — based on what the script cannot do (yet). On the other hand, I would welcome helpful suggestions, not to mention a clearly defined set of rules for WP:ENGVAR conversion. Thanks, Proteins (talk) 03:32, 18 November 2008 (UTC)[reply]

The major problem with automating Eng->US translation is that it would also translate the titles of references, you would need to tell it to ignore anything between ref tags. Tim Vickers (talk) 03:41, 20 November 2008 (UTC)[reply]

What MOS isn't

MOS is a guideline; it is not a code of laws, to be "violated" and enforced. A half-dozen of our learned and gallant colleagues think it is; and they thoroughly enjoy the process of enforcement.

They have turned much of FA into the sort of failed process which User:David Gerard discusses at length, in this essay, which I commend to you. In particular, they have made themselves into the sort of committee Gerard discusses, which judges articles by whether they comply with MOS hurdles, many of which they have invented, and are not required either by English idiom or good writing.

In the process, many FA and GA reviews have become jokes; it is always easier to quibble over the form of a header than read over over the prose for clarity; and either is easier than judging its accuracy, neutrality, and verifiability. Reviewers will tend to take the easy road, leaving the hard ones to others; too often all of them do.

Your bot is clearly a labor of love; but it will make the easy road easier still. It will make the encyclopedia worse. Please consider accepting our thanks, and crowning your work with {{db-author}}. Septentrionalis PMAnderson 06:04, 19 November 2008 (UTC)[reply]

Dear Dr. Anderson,

You might be surprised at the overlap between our thoughts, even if their expression takes such different forms. I would be very interested in your thoughts on improving FAC.

At the end of my talk at Wikimania 2008 (see my user page), I offered some deferential suggestions from the perspective of an outsider, one of which concerned FAC. I became interested in FAC's after my students and I found that most scientific FA's were excellent; I had assumed that the articles would undergo a quantum leap in quality while at FAC. Not so. In contrast, I found (in a casual survey) that the FAC process itself did relatively little to improve articles, at least those that went on to pass. Presumably, experienced authors know the hurdle they have to clear and write for that level.

To me, the emphasis on punctuation and other stylistic (MoS) issues at FAC seemed misplaced, and I said so in my talk. I tried to point out that when the outside world judges the worth of Wikipedia in published reviews, no one is going to mention whether the dashes were spaced. Basic copyediting and spell-checking are necessary, lest critics complain that the work on Wikipedia is "shoddy". But I believe that the chief hurdle to accepting Wikipedia is the reliability of its content. Well-written and well-punctuated claptrap is still claptrap; writing and copy-editing alone do not allay doubts.

Of course, my suggestion drew a little heat as "uninformed", and maybe it was; but I'm willing to learn better. I do see that FAC is one of the chief places where the MoS is enforced. I also agree with you that the MoS is not policy, but a guideline. I have strong feelings about WP:ACCESS, but otherwise I'm agnostic about dash spacing.

I believe that you and I differ on the solution. You would like to simplify the rules, so that they flow from fundamental principles; that seems sensible to me in general but impractical for MoS issues. My solution is rather to use a computer program to make MoS compliance trivial, consistent and open to everyone regardless of experience. Computer programs are also refreshingly free of human biases and the gradual creep of their standards. My wish is that, if an article can pass the script's standards, its author(s) can be confident that MoS criticism won't come up at FAC, allowing reviewers to focus on more essential elements. That in turn should expedite FAC's and make them more pleasant. Proteins (talk) 12:52, 19 November 2008 (UTC)[reply]

The problem is, as anyone who has ever turned off a spell-checker can tell you, programs are abominable editors.

My objection to MOS is not that it doesn't follow from fundamental principles; I'm an empiricist. My objection is that it doesn't describe the English language. I could go on at much more length, and less civility, on how this has come to pass; but I invite you to stick around WT:MOS (and WT:MOSNUM, which may be our worst subpage) to observe what happens. Septentrionalis PMAnderson 17:22, 19 November 2008 (UTC)[reply]

I'm sorry if I misunderstood your ideas on "fundamental principles". My impression was derived from my interpretation of your messages and of David Gerard's essay which you commended above.

I'm willing to spend time listening at WT:MOS and WT:MOSNUM. Since I'm a relative newbie, perhaps the discussions won't annoy me yet; I've been patient through worse pandaemonia. ;) More seriously, if you wished to e-mail me privately (or publicly) about your ideas for improving MoS and the FAC process, I would be grateful. Thanks, Proteins (talk) 18:19, 19 November 2008 (UTC)[reply]