Wikipedia talk:WikiProject Mathematics/PlanetMath Exchange/Archive 1

From Wikipedia, the free encyclopedia

Archive 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.

Contents

Article categories

I reduced the # of list of article categories because once we look through all the content and copy it over we can easily regenerate a list of all PM entries and/or updates. There may be other good reasons for having additional categories and keeping them categorized for the long term that I'm not thinking of, though. CryptoDerk 18:06, Jan 28, 2005 (UTC)

So, are you planning to auto-generate a list of all Planet math articles (say, by region of math)? That should be a starting point. By the way, I am not sure an entry needs to be deleted from the list after being copied over. It never hurts to actually keep a map and a comparison between each PM article and WP articles. Especially since both places are dynamical, later some more transfers of info/comparisons could be made. Oleg Alexandrov | talk 18:28, 28 Jan 2005 (UTC)
Sure, if I generate the list I can break it up by subject. Maybe someone with some scripting ability would be better for generating such a list, though, as my natural inclination is to hack it out in C, heh. See http://planetmath.org/browse/objects/ for the classification system on PM. If we decide not to delete then we will definitely need an article that contains articles that have already been processed. CryptoDerk 18:37, Jan 28, 2005 (UTC)
Having a list somewhere of already processed articles may help reduce duplication - e.g. people "helpfully" adding more pages to be merged/moved when in fact this has already been done. Of course, this depends a bit as to how many pages are involved here. If we are talking 1000+ pages then it may be more efficient to work in batches (move analysis over, then algebra, then geometry, etc.) Terry 18:43, 28 Jan 2005 (UTC)
OK, let me try one more time. Probably this is not the way to go, but I just think I am being misunderstood. Say we are talking "Functional analysis", where there are say 300 articles (a resonable number). What I have in mind, is one single master list, of 300 entries. The first thing for each entry should be the Planet math article, the second one the corresponding Wikipedia articles if any (all these need to be links). The third one should be the status of the given entry. So, anybody can always take a look at this list, see if any given article was looked upon, if yes, if anything was done, if there is any opinion about what should be done, and so on. To me, it looks much better than adding and deleting things all the time. Oleg Alexandrov | talk 18:58, 28 Jan 2005 (UTC)
Example:
Works for me. CryptoDerk 19:03, Jan 28, 2005 (UTC)

Well, this is just one idea. Many other ways can be better, I don't know. Now that I made myself clear, looking forward to other suggestions, so that whatever is decided, is eventually community supported. Oleg Alexandrov | talk 19:09, 28 Jan 2005 (UTC)

Attempt at generating automated links to Planet Math articles

To manually create a list of entries of all topics of Planet Math is impossible, they have more than 4000 of them. I tried to automatically create a list of all articles in Functional analysis. See here User:CryptoDerk/planetmathproject/test_Functional_analysis (Note: newer version now at: Wikipedia:WikiProject_Mathematics/PlanetMath_Exchange/46-XX_Functional_analysis. Paul August ☎ 15:55, Jan 31, 2005 (UTC)) for the result. Things are not pretty, chiefly because they use PNG images in article titles, but those do not become good links. Oleg Alexandrov | talk 03:02, 29 Jan 2005 (UTC)

I don't think that looks too bad. The graphics in the title seem like semi-rare occurrences.. and as long as things are categorized and we have a link, they'll be looked at by users just the same, I imagine. CryptoDerk 03:39, Jan 29, 2005 (UTC)
I'm quite impressed with that output. So it doesn't come out perfectly; so what? That's what us users are for! :) It provides an amazing catalogue of things to check for -- as chosen by a (semi-) independent group with similar goals. If that could be included in the output, finding the relevant article should become easier. Nice work! PS: I'm sure this is easily fixed, but the links use the text & instead of &, and thus don't work. PPS: I take it back; I think PM must have had an error when I clicked on one of its links, and I jumped to a conclusion about the cause after changing the & to &, but they are working fine now. Ben Cairns 07:48, 29 Jan 2005 (UTC).
I think your list is great. Paul August ☎ 19:12, Jan 29, 2005 (UTC)
I don't quite understand what If that could be included in the output, finding the relevant article should become easier means. My code can of course be modified to include whatever other info is available in those PlanetMath pages.
So, if somebody has some other ideas, suggestions, they are welcome.
If we eventually decide that the machine generated version is the way to go, more thought needs to be given to where these lists need to be put, and how to tell people of their purpose. Oleg Alexandrov | talk 21:59, 29 Jan 2005 (UTC)
I realized I could include the alt of each image in the article title. Since the alt of a PNG which came from a math formula is the LaTeX code for the formula, now we know what each link is about.
This has its own problems. Wikipedia refuses to make a thing like this
[[ hi {there} ]]
into a link (no, the above was not in nowiki tags, it just refused to become a link. But oh, well, I will think more of this.
Also, below, some thoughts on the project name are needed. Oleg Alexandrov | talk 04:18, 30 Jan 2005 (UTC)
Re: my indecipherable comment -- sorry, that was from an earlier version and I neglected to remove it. Now struck-through. Regarding usage, I think a brief explanation at the top of such a list would be sufficient, and then it's just a matter of starting to use it. People will get the hint. If anyone goes too far wrong, they can always be set straight. As it develops, more detailed explanation can go on the project page. Cheers, Ben Cairns 04:25, 30 Jan 2005 (UTC).
I went through the first five articles on Wikipedia:WikiProject_Mathematics/PlanetMath_Exchange/46-XX_Functional_analysis, added some comments there, and now I would like to share my thoughts resulting from this. Firstly, I was rather skeptical about spending time on the WP/PM exchange but I have to say that it seems a good way to fill some gaps in WP and review some of the articles. On the list generated by Oleg: note that almost all articles occur multiple times; this leads to the risk that people comment on the same PM article in different places. Furthermore, what to do about the proofs in PM? I don't think it's useful to have them in WP. Of course, it wouldn't hurt to copy the proofs over, especially if they are put on separate pages, but I guess it should be low priority. -- Jitse Niesen 13:32, 31 Jan 2005 (UTC)
On Wikipedia:WikiProject_Mathematics/PlanetMath_Exchange/46-XX_Functional_analysis, I reformatted Jitse Niesen's comments, for what I think is better readability. I made them as indented lines following the item commented upon, leaving the status field for just a status entry, like:
  • needs to be copied
  • copied
  • needs to be merged
  • merged
  • WP article more complete
  • Not needed on WP
What do people think about this? Paul August ☎ 14:50, Jan 31, 2005 (UTC)
Good idea. -- Jitse Niesen 15:24, 31 Jan 2005 (UTC)
I went ahead and updated the Planetmath Instructions template to reflect the above. I also rewrote them a bit. If anyone doesn't like it feel free to change it back. Paul August ☎ 17:38, Jan 31, 2005 (UTC)
I think the multiple occurrence of items in the list is a problem. Do they also appear in multiple categories? How should we deal with this? Paul August ☎ 15:09, Jan 31, 2005 (UTC)
Ok I've done a bit of research on PM. Each article can, of course, have multiple "classifications" and so can appear multiple times within each top level category as well as appear in multiple top level categories. Should we rethink our organization scheme for our lists? Or should we deal with this manually, by asking individual editors who work on an article to check the PM article's classification, which appears at the bottom the PM page (perhaps this could be added to the auto-generated lists?) and if there are multiple classifications, and hence multiple entries in our lists, then, for example, pick one entry as "primary" and for every other entry, add a pointer to the "primary" entry? Paul August ☎ 15:47, Jan 31, 2005 (UTC)

Attempt at generating automated links (2)

You are right, an article show up multiple times because they put each article in several categories. My perl script could be taught to randomly pick one instance of the article and declare it primary, and make the other ones point ot it. But of course any automated generation can go only that far. Any other ideas? Oleg Alexandrov | talk 16:05, 31 Jan 2005 (UTC)

I think a randomly generated "primary" entry would probably be ok. Can your script keep track of duplicates across lists, as well as within lists? For an example of how I see a manual entry working/looking see the entry I made here: Wikipedia:WikiProject Mathematics/PlanetMath Exchange/46-XX Functional analysis#46F05 Topological linear spaces of test functions, distributions and ultradistributions. Paul August ☎ 16:24, Jan 31, 2005 (UTC)
I think that you needn't be bother with generating the links, as it would be quite okay to only use the first MSC class and disregard the other classes. For instance, the "T_f is a first order distribution" article would be listed under 46F05 and not under 46-00. -- Jitse Niesen 16:31, 31 Jan 2005 (UTC)
Yes using just the "first" entry listed in the PM article and omitting the rest would probably be ok. Although I could see some uses for having the other entries as well, since it would allow editors to approach the entries form different starting points, especially if this can be automated. Paul August ☎ 17:12, Jan 31, 2005 (UTC)
I understand Jitse's suggestion. But teaching my script to keep track of duplicates is quite easy, and can also work across lists, not only within them. So I will follow Paul's suggestion. I will get to this tonight, Los Angeles time. Then I will complete all the red links too, and not touch Functional analysis (whatever duplicates show up there I will fix by hand). Oleg Alexandrov | talk 17:44, 31 Jan 2005 (UTC)

planetmath template

The template: {{User:CryptoDerk/planetmath}} As it currently works, requires a "title" field, which is used as the title of the PM article, as well as to construct a link to the PM article in the form of [http://planetmath.org/encyclopedia/{{{title}}}.html. But since the "title" in the link needs to have the spaces and punctuation removed, this means that the title of the PM article will also have punctuation and spaces missing. For example for Potential theory, the template produces the following:

This article incorporates material from [[PlanetMath:{{{id}}}|PotentialTheory]] on PlanetMath, which is licensed under the GFDL.

Can anything be done about this? Paul August ☎ 19:04, Jan 29, 2005 (UTC) This has now been changed see below. Paul August ☎ 02:39, Jan 30, 2005 (UTC)

What about changing the template so that the link is generated using the object id of the PM article (like those given in User:CryptoDerk/planetmathproject/test Functional analysis) instead of the name.html link now used? Paul August ☎ 19:16, Jan 29, 2005 (UTC)

Feel free to change it. I was tentatively going for something with one variable instead of two, but I fully expected it to be changed by the time all was said and done. CryptoDerk 20:04, Jan 29, 2005 (UTC)

Ok, I've now changed the template and its instructions for use, to include two variables, "id" which is the PlanetMath pages's "object id" (which can be found at the bottom of the PlanetMath page, or in the PlanetMath links listed here: User:CryptoDerk/planetmathproject/test Functional analysis), and "title", which is the PlanetMath page's title. So now {{User:CryptoDerk/planetmath|id=6659|title=Potential Theory}} produces:

This article incorporates material from Potential Theory on PlanetMath, which is licensed under the GFDL.

Much better. Now in the auto generation of them the ID should be displayed as well. CryptoDerk 05:17, Jan 30, 2005 (UTC)

Project name

Time to think for an official name for the project. I would suggest:

The Wikipedia-PlanetMath exchange project

This for two reasons:

(a) I hope things will go both ways, once there is a nice list of correspondences, with comments comparing the two articles.

(b) This will make both projects stronger, not one a clone of the other. So like a synergy. The two web sites will never coincide, because they have very different purposes and philosophies. So there will be very very much content which is not shared. Oleg Alexandrov | talk 03:34, 30 Jan 2005 (UTC)

I think those are good reasons, and a good name. So, Wikipedia:WikiProject PlanetMath Exchange would seem an appropriate location for the project. That said, why not just create the page and we can make a start on the project proper? :) Ben Cairns 04:34, 30 Jan 2005 (UTC).
Since this is a rather localized project in terms of scope, wouldn't it be advantageous to have it as a subproject of the WikiProject in math? I suppose either way we will link to it from the WikiProject in math. Oh, also what about the template name? Template:pm would be good, cause it's short, although Template:planetmath might also be good since it's unlikely to collide with someone who may want to use "pm" as a template in the future. CryptoDerk 05:15, Jan 30, 2005 (UTC)
I think it is important to make it a subproject of WikiProject in math. And Template:planetmath could indeed be better, for the reason mentioned. Oleg Alexandrov | talk 05:45, 30 Jan 2005 (UTC)
OK. I moved the template into the template namespace Template:planetmath. I doubt anything's going to change with that. I made some changes to the planetmathproject page based on how the list you auto-generated looks. I also moved the page to Wikipedia:WikiProject Mathematics/PlanetMath Exchange which seems like a place and name that everybody agrees on. I'll put a note at the top saying that it's not "live" yet. CryptoDerk 15:39, Jan 30, 2005 (UTC)

We need to bring this to Mathematics project talk page

I think, we need to bring this discussion to Wikipedia talk: WikiProject Mathematics. So as to include all the members/readers of the Mathematics project in these discussions. Something should be announced there about what has been accomplished and "decided" so far. Paul August ☎ 16:35, Jan 30, 2005 (UTC)

It has been linked to from there, and has been discussed there a bit too :/ I'll post another notice though. CryptoDerk 16:47, Jan 30, 2005 (UTC)

Thanks. Paul August ☎ 17:27, Jan 30, 2005 (UTC)

I would further propose posting a summary of these discussion there as follows:

Some of the things which have been discussed (here: Wikipedia talk:WikiProject Mathematics/PlanetMath Exchange) and tentatively agreed upon and/or accomplished are:

  1. We should go forward with this project.
  2. The project name should be: "PlanetMath Exchange".
  3. It should be a subproject of this project with the project page at: Wikipedia:WikiProject Mathematics/PlanetMath Exchange. A first draft of that page now exists there.
  4. There should be an auto-generated list of all PlanetMath articles. The first auto-generated list of PlanetMath articles has been created here: Wikipedia:WikiProject Mathematics/PlanetMath Exchange/46-XX Functional analysis.
  5. There should be a template created to facilitate the creation of a link to the appropriate PlanetMath article in any newly created WP article based on a PM one. Such a template has been created: Template:planetmath.
Please comment on the above, either here, are on the talk page.

Is this agreeable? Paul August ☎ 17:56, Jan 30, 2005 (UTC)

Sounds fine to me. Since Oleg is apparently working on the list now, why not wait a bit so we can provide a link to that list instead of the test one? CryptoDerk 19:52, Jan 30, 2005 (UTC)
So, now that we have a permanent home, I need to do some work, as you guys mentioned. Tonight I will generate the list of mathematical topics which will be on the main page. I will also generate a list of articles, but just for one topic, to see again the feedback. So hold on, I will get to it in 2 or 3 hours or so. Oleg Alexandrov | talk 00:37, 31 Jan 2005 (UTC)

This all seems good to me. Oleg: did you see CryptoDerk's rquest to have the PM "Object ID" included in the list? This might facilitate using the Planetmath template, which now uses an id variable. Paul August ☎ 00:46, Jan 31, 2005 (UTC)

Yes, I noticed the request. Actually, when I get to it (soon today), I will do a trial run, and ask for another round of feedback. Then I will generate the final version. Oleg Alexandrov | talk 00:57, 31 Jan 2005 (UTC)

Done! I put a list on the main Wikipedia:WikiProject Mathematics/PlanetMath Exchange page, and created just one article in that list, namely "Functional analysis". I will do no more until all people express their opinions about the format, and what else is needed. Who will announce our project on Wikipedia:WikiProject Mathematics? Oleg Alexandrov | talk 02:33, 31 Jan 2005 (UTC)

Looks good. I made a couple of small fixes to Template:Planetmath instructions. The actual list looks fine though, as far as I can tell. Regarding putting it on the main page it doesn't matter so much who as where. I think it best fits under scope — what do other people think? CryptoDerk 04:25, Jan 31, 2005 (UTC)
I don't know. The list is long. Some instructions need to be before the list. Indeed, what do other people think?
I think you noticed that all elements in the list are red links, except for "Functional analysis", which is like a sample. I did this on purpose. I want more feedback before I commit myself to filling in all pages, because I did generate all the listings automatically, but clicking inside of each link and pasting the content in, still needs to be done by hand. So again, looking for feeback before final comittment.
You're the boss. Could you announce the project on the talk page of WikiProject Math as Paul August suggested? Thanks. Oleg Alexandrov | talk 04:39, 31 Jan 2005 (UTC)
I looked at both the PM Exchange main page and the Functional analysis page and both looked fine to me. Feel free to wait a while to get more opinions, though. I'll write something up about the Exchange and post it on the main WikiProject in math page. The project has been mentioned several times on the talk page, but when all the lists are filled in we should mention it again. CryptoDerk 04:49, Jan 31, 2005 (UTC)
So basically you want me to complete the lists first, right? No problem, I can do that tomorrow. But then there is no turning back. It will be quite painful to revisit around 100 pages again. Oleg Alexandrov | talk 05:11, 31 Jan 2005 (UTC)
That's not what I'm saying. I'm saying I don't have any issues with the lists. You may want to wait for other input though. CryptoDerk 06:32, Jan 31, 2005 (UTC)

Oleg: Before you go to the trouble of generating all the lists, It might be good if we made the "announcement" I suggested above first, to make sure anyone who hasn't been following the discussion here, and wants to comment on it gets a chance to. Paul August ☎ 05:49, Jan 31, 2005 (UTC)

I agree. Oleg Alexandrov | talk 06:03, 31 Jan 2005 (UTC)

Crypto, do you concur? Paul August ☎ 06:07, Jan 31, 2005 (UTC)

Ok. Oleg, has made an announcement on the Mathematics project talk page, and I have and Crypto have added to it. Paul August ☎ 07:01, Jan 31, 2005 (UTC)

Oleg: Also, I don't know what's involved, but perhaps some of us could help you generate all those lists. I'd be willing to help if I could. Paul August ☎ 06:07, Jan 31, 2005 (UTC)

Generating the lists is as easy as typing the name of the perl script. Then I just need to click on each red link and paste the content in. It is the latter one which is kind of annoying, as there are many red links. So, I did not want to do this more than once.
See my comment in the "Attempt at generating..." section about the multiple occurnce of an item problem. Oleg Alexandrov | talk 16:08, 31 Jan 2005 (UTC)

Usage of Template:planetmath etc.

I was thinking that while in many circumstances the template for acknowledging PlanetMath will be a necessary addition to updated or new pages with PM content, in some cases it would NOT be necessary. For example, if someone sees something on PM that they think would go well in a WP article, but edit it into WP entirely in 'their own words', then I don't see any copyright purpose to including the template. After all, I'm sure people have been doing this all along. So, I'd propose the following recipe:

  1. If any text is copied from PM to WP and re-edited, then the template should be added to the bottom of the page. The template for should include a reference to the GFDL. This gives appropriate credit and satisfies copyright requirements. Furthermore, the copied text should be saved in its original form to (say) [[Talk:Article Name/PlanetMath Exchange]]. A WP-ready version can be assembled there from the original PM TeX.
  2. If a PM article 'inspires' the addition of content to a WP article but no actual text is transfered (even informally), then I propose that a link to the PM article under 'External References' is sufficient (perhaps using the web reference template or a new one specifically for PM references). Since the concepts (in contrast to the text that presents them) are not covered by copyright, it seems to me that this would give credit without attributing unnecessary copyrights to the text of the WP article.

I think guidelines for this should be established early on, because it would be a great pain to change these things later. Working out what came from where a long time after the fact could be quite difficult (unless the precise copied text is kept somewhere). Keeping a copy of the original text should also facilitate actual 'exchange' between the two projects. What do you think? Ben Cairns 00:09, 31 Jan 2005 (UTC).

Linking back to PM even if the text isn't used is what I'd hope people would be nice/smart enough to do in the first place, heh. Regarding the template, it can be changed to read "This article is based on whatever from PlanetMath, which is under the terms of the GFDL". Someone can probably think of a better wording there, so I'll hold off on changing it for now. As for the copying of the text into a subpage of the article, is that really necessary? We could provide a link back to the history of the article on PM. CryptoDerk 01:07, Jan 31, 2005 (UTC)
I think we should not get obsessed about things. Anybody can take a look at the page history for older versions, that's all. Not even an explicit link to the history should be necessary. Oleg Alexandrov | talk 01:52, 31 Jan 2005 (UTC)
Right. I think linking back to the article itself is sufficient — I was just presenting a link to the history as an alternative in case people overwhelmingly felt like linking back to it was a good idea. CryptoDerk 01:56, Jan 31, 2005 (UTC)
Yes, of course, people should be crediting and citing sources for all sorts of good reasons, but there's no need to attribute copyrights (even under the GFDL) when they are not present. Regarding the addition of the original text: fair enough. I thought it might be getting a little elaborate, but also thought that I'd put the idea out since at times it might be nice to know what exactly had been added. Ben Cairns 03:23, 31 Jan 2005 (UTC).

Planetmath reference to planetmath ref

It's a relatively minor thing, but I think that Template:Planetmath reference should be at Template:Planetmath ref. It's User:Bjcairns creation, so I didn't want to move it outright, plus I wanted to get some more opinions. I think a redirect would work as well. CryptoDerk 05:19, Jan 31, 2005 (UTC)

I don't mind; as they say, be bold. I went with 'reference' over 'ref' because of Template:Web reference, and I figured I could live with typing 'reference' in full once in a while. Ben Cairns 06:31, 31 Jan 2005 (UTC).
OK. Just didn't want to step on your toes. I won't move it until some other people chime in, though.CryptoDerk 06:33, Jan 31, 2005 (UTC)

Copying text from PlanetMath to Wikipedia

In the section "Converting text from PlanetMath to Wikipedia formats", the instructions say:

PlanetMath generates pages from a LaTeX source. To copy the contents of an article, just select the article text with the mouse, copy, and paste in a Wikipedia article (then each image will be replaced with its alt text, that is the formula which generated the image).

However when I try to copy the text, I don't seem to get any of the "images". For example see my attempt at copying Cauchy sequence to here: User:Paul August/Subpage 11. Am I doing something wrong? Paul August ☎ 20:05, Jan 31, 2005 (UTC)

Good point. I guess it depneds on what browser you use. It works for me, with Firefox on Linux, see my additions to your page. I will modify the instructions accordingly. Oleg Alexandrov | talk 20:44, 31 Jan 2005 (UTC)

Ok I had tried it using both Safari, and IE on Max OSX and they both did the same thing. I didn't think to try FireFox. When I did try FireFox, it worked fine, just as it did for you. I don't think this much of an issue though, although it might confuse people like it did me ;-) Paul August ☎ 21:15, Jan 31, 2005 (UTC)

So I changed the wording back to saying one needs to go to the LaTeX source, and mentioned in parantheses that for some browsers copying straight from the page will work too. We can remove the latter part altogether, or reword it, what do you think? Oleg Alexandrov | talk 23:02, 31 Jan 2005 (UTC)
If Firefox works, might as well mention that... IE certainly doesn't in a Win2K environment. Otherwise, I think it's fine as it is for now, or just reword as you see fit. As I see it, these things can be expanded on as necessary when the project goes 'live'. Ben Cairns 23:38, 31 Jan 2005 (UTC).
I added Firefox and reworded it. CryptoDerk 23:56, Jan 31, 2005 (UTC)

This all seems fine now. Paul August ☎ 00:57, Feb 1, 2005 (UTC)

Shortcut

I created a shortcut at WP:PMEX for the project page. CryptoDerk 00:49, Feb 1, 2005 (UTC)

Upload files status

I uploaded all the files now. The duplication notice seems to work, even across lists. One problem I noticed is that foreign accents on words did not get represented well (some encoding issues I guess). But I don't have any experience with that, so did not attempt to fix it. One we need to see if there are no bugs (if there are, I will do the uploads again, I think there is a way to automate that). Oleg Alexandrov | talk 03:15, 1 Feb 2005 (UTC)

The special (accented) character problem seems to be one of encoding. The English Wikipedia is apparently in the (non-Unicode) ISO-8859-1 encoding (see m:Help:Special characters. They appear correctly when I force Unicode encoding (in IE), and can possibly be copied and saved this way since ISO-8859-1 appears to have common characters like é as native. Other than that, the Probability list seems pretty good -- no apparent bugs. Excellent work! Ben Cairns 04:52, 1 Feb 2005 (UTC).
Yes well done Oleg! Paul August ☎ 05:09, Feb 1, 2005 (UTC)
Excellent job. I'll post a notice over on WP:VP. CryptoDerk 05:19, Feb 1, 2005 (UTC)

Talk pages

I'm going to redirect all talk pages of the article lists here. It seems unfair to ask people to keep all the list pages on their watchlists, and the conversation generated in one might be applicable to all the lists, the subproject, or the WikiProject in mathematics. Case in point, the comment that Ben Cairns posted in category 60 that I moved below... and yes I have all the lists on my watchlist (shut up!). CryptoDerk 15:30, Feb 1, 2005 (UTC)