Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
Anyone have any ideas about how to improve "latest changes"? It would be really nice if it were a threaded forum or something else with an RSS feed. But it should also be easy to link to nlab pages from it. How hard would it be to hack some forum or blog software (like here or the cafe) so that we could make links directly to the nlab, say just like this?
I thought about writing a simple filter that converted the current latest changes to an RSS feed. I've also thought that another option would be to dedicate an area of this forum (or some other forum) to it and link everything across. However, I've now learnt (thanks to Jacques) that Rails apps can be integrated in a closer fashion than Rails and non-Rails apps. I also was initially keen to base it in PHP so that it could run on a different server to the lab itself, but now that we've gone for a memory upgrade then I'm less bothered by this issue and think that the tighter integration would pay off.
If we wrote (or found) a simple forum/blog app in Rails that did what we wanted, then we could link it in to the nLab so that the latest changes page was essentially a front to the new app.
I've only just learnt about the possibilities with the integration so I'm not clear on what is and isn't possible, but the possibility of a close integration so that to the user it was all as if it were the same application gives me enough reason to investigate further and consider doing it this way.
The integration stuff, by the way, goes by the name of 'ActiveResource'.
That would be pretty neat. Could we also put a "comment" box on the "edit" pages on the nlab that would automatically post something to the latest changes forum?
I'm in two minds about this one. It would make it easier to log stuff to 'latest changes' but then would make it less likely that people would go over there and read everyone else's log (in particular, since not everyone is quite sure what RSS is for). Also, it may make it too easy to log stuff and 'latest changes' would fill up with crud. I think I'd rather put a link to 'latest changes' at the top of every page (perhaps in place of the 'recently revised' link) and maybe a short message next to the 'submit' button on the edit page saying "Don't forget to log significant changes at 'latest changes'".
Apart from that one, it feels to me as though 'latest changes' should basically be a forum-type thing. Anyone can post, but can't change posts (if we, like Instiki, don't have logins then you can't edit your own post either). Posts can 'reply' to other posts. There should be several ways to view the posts: threaded, chronological, today, last n days (specific values of n?), and, of course, RSS. The syntax should be that of Instiki, and wiki-links should link to the nlab.
Anything else?
re: boxes on edit pages, I can see your points.
Your summary of features sounds pretty good to me. Perhaps, though, we should think about how much "replying" we should encourage people to do at 'latest changes' versus actually having discussions at the pages themselves?
There's already too much replying at latest changes, in my opinion. Sometimes it works, but mostly it should go on the relevant page. (But then, you might want to log the fact that you made a comment! That depends on whether you expect the person that the comment is for to read the page anyway, really.)
We could also think about attaching discussion forums to individual pages, the way that Wikipedia does with its talk: pages. I'm not sure whether I think that's a good idea, just throwing it out there. But if we had a forum running with instiki syntax and nlab links, it might make some of the longer discussions easier to follow (one could subscribe just to the RSS feed for a particular discussion, for instance), and also easier to preserve and continue without cluttering up the pages themselves. We might even think about implementing a syntax for nlab pages to link to forum discussions. Just brainstorming.
Syntax to link back and forth would be very useful, I think. (And throw the Café in there too.)
Can we set up RSS feeds for specific threads in this Forum, so that it would be useful for someone who wants to comment on a page to simply throw up a link to a Forum thread, or (if there is already such a link) add to the existing discussion thread? All in all, I like having the discussion right there where it's relevant, but I can see how having an RSS feed would be useful.
I don't know how I survived without RSS feeds! In one window I can see if there are any new comments/discussions here, any new pages at the lab, any new comments/posts at the cafe, any updates to the instiki software, any complaints about my LaTeX plugins (any comments at all about those!) ... if I had to check all of those pages individually I'd never get any work done. Or something. In fact, one page has just lost its RSS and that means that I only check it when I remember - not very often! As it's a page I ought to keep an eye on (since it's the discussion thread on some software I wrote) then not having RSS is an issue.
I think that RSS works fairly well on this forum. I haven't tested all possibilities, and I had to tweak the plugin a little (you may remember that), but there's a link up above saying "Discussion Feed" so I think that that would give you the feed for this particular discussion. It looks like that can also happen by category.
So one quick option would be to have a dedicated category in this forum for 'latest changes'. New comments become new discussions, whilst replies become continuations of a discussion. Integration from this side wouldn't be a problem - it'd be easy to write a wikilinks plugin that took latest changes and made it point to the right page (actually, that sounds like a darn good idea anyway).
So we could do almost everything from this forum. The main disadvantage would be the integration with the nlab from the nlab side. Loose integration - a link to this forum - would be easy enough. Tight integration, where the relevant discussion gets inset into a page in the lab, would be harder, but easier in Rails than PHP.
Discussion pages and RSS would save people having to keep putting "more on discussion at ..." on the 'latest changes'. They'd just have to log that there was an interesting discussion and then leave everyone to bookmark them as they wish. Heck, we can even have a feed reader on the site that everyone can use and personalise ... the possibilities are endless! We could take over the world!
Or something.
While we're dreaming, I'd also like a 'lab elves' blog where I can make announcements like the fact that something might have gone wrong with unicode characters so let me know if you spot one.
Actually, if we're thinking big ... on one site there should be:
And if anyone says "google wave" then I will personally block their IP.
Do we really want discussions to start from latest changes like you say in #9, Andrew?
I want discussions to start from the relevant page. I would like to have a 'specific pages' category on the Forum, and then I can start a thread for graph if I want to, putting it in that category and titling it 'graph'. If people use the RSS feed, then I don't have to bother with latest changes at all when I continue a discussion; I don't even have to bother with latest changes when I start a discussion (just link it from the page that I'm discussing), although I might note it on latest changes anyway if I want to solicit some kind of feedback.
One problem with moving things to a Forum, however, is that there is more than one thing being discussed at graph. It's nice that we can keep each conversation in its own query box; to me, it's not an improvement to move to a single Forum thread, even if that does get me an RSS feed. Of course, we could put each conversation in its own thread, but then I no longer have an RSS feed that tells me about new discussions on graph … unless each page gets its own category … pretty complicated!
All in all, what I'm saying is that I find discussions in query boxes very easy to follow.
What if instiki-syntax included something like !discussion NAME, which we could insert into a page anywhere, and it would display the forum discusssion called NAME inline. This would look very much like current query boxes, but could be even better, e.g. threaded and expandable/collapsible. It would include a link to open the forum discussion alone, from which you could reply to the discussion; the inline widget could also support replying directly. The inline widget could also have a link to an RSS feed for that discussion, making it easy to subscribe to. And, of course, if you entered !discussion NAME where there is no discussion called NAME yet, then the inline widget would give you a link to create it, just like how you upload a file now. This would also have the advantage that a single discussion could be displayed on multiple pages; we've occasionally had query-box discussions that were relevant to more than one page.
Admittedly, this doesn't automatically give us a single RSS feed that tracks all discussions on a given page, unless we put discussions in categories. However, I'm a bit of an RSS newbie and I'm actually a little scared of the idea of having separate RSS feeds for each page on which there is a discussion. Presumably that means we'll have to subscribe individually to each such feed; that seems like (potentially) a lot of feeds! And can my feed reader handle hundreds of feeds without slowing to a crawl? Right now I read RSS with Thunderbird, and for the n-Forum I just subscribe to the "all comments" feed.
If we do end up with separate feeds for separate page/discussions, it would be nice to have some way of being notified when new discussions start---either they could be mentioned on the "latest changes" discussion manually, or there could be a separate "new discussions" feed.
Well, okay maybe there's a little too much dreaming going on there!
If we had a Rails app, then it would be relatively easy to embed stuff in pages since they're designed to talk to each other. However, that would entail writing (or at least adapting) our own program.
As a first step, we can simply create a 'latest changes' area on this forum and then link here instead of to the Instiki page. Given that you have to click a link anyway, we can make that point wherever we like. What would be more useful, but shouldn't be difficult, would be to have it on every page, for example instead of the 'export' link, or instead of 'recently revised'. That also shouldn't be too difficult (though I'd need to check with Jacques that this wouldn't break anything with updates).
Tighter integration will need hacking of Instiki so that would take a little more time. Thus it might be a good idea to try a simple link-up to see if people like the idea, and then when we've decided what we do and don't like, design our own app.
Incidentally, how do the nforum and nlab rate on performance? I could move the nforum over to the nlab host if that was felt to be a good move - this should only be considered on the basis of performance, not hostnames, since I can always (and probably should) make it seem as though the forum is hosted there without actually moving it at all.
And what "virtual hostname" should I pick? nforum.ncatlab.org? nforum.mathforge.org? I can pick anything.ncatlab.org or anything.mathforge.org. I'd rather not make it a subdomain of ncatlab.org since that domain is completely controlled by the instiki process so I'd have to add an exception. So much easier just to add a virtual domain.
Hmm, I just checked and it looks like vanilla (this forum software) can't do threaded discussions. However bbpress has a plugin that might be suitable. If threaded discussions seemed a good idea, that is. An alternative is to make a new category for each day and then each new entry on 'latest changes' is a new discussion in that category. I think that one can make categories a bit tree-like (sub categories and the like) and the RSS should be filterable so that you get one feed for everything "below" a certain level.
I just tried a mock-up. It doesn't quite work as fully as it ought to: one ought to be able to get a 'tree like' effect in the "categories" tab and that doesn't happen. Also, the parent 'latest changes' category should inherit all the discussions of the sub-categories. I'll investigate tomorrow whether it's easy to fix this.
For the moment, I created a parent category 'latest changes' and subcategories corresponding to the next four days. Actually, I forgot that for most of you it's still thursday (only 45 minutes left for me so I'm off to bed before I turn into a pumpkin). I think that if the inheritance can be made to work then this would be a reasonable system.
Thoughts?
(of course, ideally one could use itex here ...)
Okay, some hackery this morning has cleared things up a little. Multiple subcategories are possible, so one could create a tree a bit like:
Latest Changes
- September
- 1st
- 2nd
and so forth. The main hassle was getting the 'search' plugin to search multiple categories. This is needed because the RSS feeds work out of the search plugin so if one wants to bookmark all the 'latest changes' then one has to search in all the sub categories.
At the moment, the search doesn't have an automatic "search in subcategories" facility which would be necessary (it'd be a pain to have to redefine the 'latest changes' RSS each time we added a new day!). However, it took long enough to figure it out so far and I'd like to know if this is a reasonable idea before investing too much more time in it.
Re #13: My apologies if they should not have been removed. Here is the content. The pages can be put back if requested. The source compiles correctly in the actual nLab, the differences are due to the fact that only the old renderer is used on the nForum; I have lightly edited some syntax to make it display reasonably here.
An hyperanalytic function is a function that is locally given by a converging power series with a speed that corresponds to tetration.
Let $V$ and $W$ be complete Hausdorff topological vector spaces, let $W$ be locally convex, let $c$ be an element of $V$, and let $(a_0,a_1,a_2,\ldots)$ be an infinite sequence of homogeneous operators from $V$ to $W$ with each $a_k$ of degree $k$.
Given an element $c$ of $V$, consider the infinite series
$\sum_k a_k(x - c)^k$(a power series). Let $U$ be the interior of the set of $x$ such that this series converges in $W$; we call $U$ the domain of convergence of the power series. This series defines a function from $U$ to $W$; we are really interested in the case where $U$ is inhabited, in which case it is a balanced neighbourhood of $c$ in $V$ (which is Proposition 5.3 of Bochnak–Siciak).
Let $D$ be any subset of $V$ and $f$ any continuous function from $D$ to $W$. This function $f$ is hyperanalytic if, for every $c \in D$, there is a power series as above with inhabited domain of convergence $U$ such that
$f(x) = \sum_k a_k(x - c)^k$for every $x$ in both $D$ and $U$ and just rare $a_k$ are not equal zero. (That $f$ is continuous follows automatically in many cases, including of course the finite-dimensional case.)
{#Examples}
It is known that there is a fundamental connection between analyticity of the function and the convergence of its Fourier coefficients. The better the function, the faster its coefficients tend to zero, and vice versa. The power decrease of Fourier coefficients is inherent in functions of the $C^{k}$ class while exponential to analytical functions. Here there is a possibility of existence of the hyperanalytic functions, for which the decrease of the Fourier coefficients corresponds to tetration.
Natural hyperanalytic function occurs when considering reticulum with a step L, in which nodes there are not defined yet objects. The distribution of center’s objects can be described using the reticulum functions (RF). The definition of a one-dimensional RF is based on the following identity:
$\frac{1}{\sigma\sqrt{2\pi}}\int_{-\infty}^{\infty}e^{-\frac{1}{2}(\frac{x}{\sigma})^{2}}dx=\frac{1}{\sigma\sqrt{2\pi}}\int_{-\frac{L}{2}}^{\frac{L}{2}}\sum_{n=-\infty}^{\infty}e^{-\frac{1}{2}(\frac{x-nL}{\sigma})^{2}}dx=1.$From here RF is
$\mathbb{R}(x)=\frac{1}{\sigma\sqrt{2\pi}}\sum_{n=-\infty}^{\infty}e^{-\frac{1}{2}(\frac{x-nL}{\sigma})^{2}}.$It is obvious that the RF can not be laid out in the Fourier series because it does not have antiderivative that can be expressed as elementary functions. By virtue of this RF cannot be decomposed into even and odd functions, while an arbitrary analytic function $f$ can be only presented in the form of sum of odd and even functions in the interval $[a,b]$:
$f\left(x\right)=g\left(x\right)+h\left(x\right),$where
$g\left(x\right)=\frac{f\left(x-a\right)-f\left(b-x\right)}{2},$ $h\left(x\right)=\frac{f\left(x-a\right)+f\left(b-x\right)}{2}.$Due to this the RF can be laid out in an endless row of two primitive hyperanalytic functions by sequential attempts to decompose on even and odd functions. Thus, the RF can be decomposed by the simplest way, but such a series is not one like the orthonormal basis of Fourier series.
As it follows from (1.1) the mean value of RT is 1. However as will be seen from the further, it is expedient to choose the greater value of the decomposition’s constant member. Introduce the following definitions:
$\mathbb{R}\left(0\right)$ is
$\mathbb{R}\left(0\right)=\mathbb{R}_{max}=\frac{1}{\sigma\sqrt{2\pi}}\sum_{n=-\infty}^{\infty}e^{-\frac{1}{2}\left(\frac{-n}{\sigma}\right)^{2}},$$\mathbb{R}\left(1/2\right)$ is
$\mathbb{R}\left(1/2\right)=\mathbb{R}_{min}=\frac{1}{\sigma\sqrt{2\pi}}\sum_{n=-\infty}^{\infty}e^{-\frac{1}{2}\left(\frac{1/2-n}{\sigma}\right)^{2}}.$Then $A_{0}$ is the mean value of RF:
$A_{0}=\frac{\mathbb{R}_{max}+\mathbb{R}_{min}}{2}.$One can approximate first difference by the following way:
$A_{1}\left(x\right)=\frac{\mathbb{R}_{max}-\mathbb{R}_{min}}{2}cos\left(2\pi x\right).$Let introduce parameter of the fine structure $\alpha$ as function of $\sigma$:
$\alpha\left(\sigma\right)=\frac{1}{2}\frac{\mathbb{R}_{max}-\mathbb{R}_{min}}{\mathbb{R}_{max}+\mathbb{R}_{min}}.$Now $A_{1}\left(x\right)$ can be expressed:
$A_{1}\left(x\right)=\frac{\mathbb{R}_{max}+\mathbb{R}_{min}}{2}\left(2\alpha\left(\sigma\right)cos\left(2\pi x\right)\right).$The choice of the name and symbol of this parameter is due to the fact that
$\alpha\left(0.4992619105929628\right)=\alpha=\frac{e^{2}}{4\pi\epsilon_{0}\hbar c}$is the value known in physics as a fine structure constant.
Even differences are a primitive hyperanalytic function $\overline{\mathbb{V}}(2i\times2\pi x)$, which is quasisymmetric relative to the point $\text{x=0.25}$.
Its symmetrical part approximated in the following way:
$A_{2i}\left(x\right)=c_{2i}\left(cos\left(2i\times2\pi x\right)-1\right)$and
$\sum_{i=1}^{\infty}c_{2i}=\frac{\mathbb{R}_{max}+\mathbb{R}_{min}}{2}-1=2 * \sum_{k=1}^{\infty} \alpha^{4^{k}}.$Using the value
$\mathbb{R}\left(1/4\right)=\mathbb{R}_{1/4}=\frac{1}{\sigma\sqrt{2\pi}}\sum_{n=-\infty}^{\infty}e^{-\frac{1}{2}\left(\frac{1/4-n}{\sigma}\right)^{2}}$define the amplitude for $c_{2}$: $\frac{1}{2}\left(\frac{\mathbb{R}_{max}+\mathbb{R}_{min}}{2}-\mathbb{R}_{1/4}\right)=2\alpha^{4}$. This definition allows to select approximation $A\left(x\right)$ in the form:
$A\left(x\right)=\frac{\mathbb{R}_{max}+\mathbb{R}_{min}}{2}\left(1+2\alpha\left(\sigma\right)cos\left(2\pi x\right)\right)+2\alpha^{4}\left(cos\left(2\times2\pi x\right)-1\right).$Odd differences are a primitive hyperanalytic function $\mathbb{W}\left((2i-1)\times2\pi x\right)$, which is quasiantisymmetric relative to the point $\text{x=0.25}$.
Quasiantisymmetry of $\mathbb{W}\left(2\pi x\right)$ follows from the fact that the integral of $A\left(x\right)$ differs from 1:
$\int_{-1/2}^{1/2}\text{A}\left(x\right)\text{dx}-1=\frac{1}{4}\left(\mathbb{R}_{max}+\mathbb{R}_{min}\right)+\frac{1}{2}\mathbb{R}_{1/4}-1\simeq1.02E-34.$Thus function $\mathbb{W}\left((2i-1)\times2\pi x\right)$ should be decomposed in the even and odd function. Its even part is:
$\mathbb{W}^{\text{qs}}\left((2i-1)\times2\pi x\right)=\frac{\mathbb{W}\left((2i-1)\times2\pi x\right)+\mathbb{W}\left((2i-1)\times2\pi\left(0.5-x\right)\right)}{2}=\overline{\mathbb{V}}(2(i+1)\times2\pi x).$However, as shown above, $\overline{\mathbb{V}}(2i\times2\pi x)$ is not an even function.
The odd part of $\mathbb{W}\left((2i-1)\times2\pi x\right)$ is no longer a hyperanalytic function and is equal to:
$W^{\text{qa}}\left((2i-1)\times2\pi x\right)=\frac{\mathbb{W}\left((2i-1)\times2\pi x\right)-\mathbb{W}\left((2i-1)\times2\pi\left(0.5-x\right)\right)}{2}.$It can be approximated with any degree of accuracy following way:
$A(W^{\text{qa}}\left((2i-1)\times2\pi x\right))=\beta(cos\left(3(2i-1)\times2\pi x\right)-cos\left((2i-1)\times2\pi x\right)),$where $\beta$ is a normalizing multiplier.
Thus, the approximation of $\mathbb{R}(x)$ is:
$A\left(x\right)=\frac{\mathbb{R}_{max}+\mathbb{R}_{min}}{2}(1+2\alpha cos\left(2\pi x\right)) +2\sum_{i=1}^{\infty}\alpha^{4^{i}}\left(cos\left(2i\times 2\pi x\right)-1\right)+\frac{2}{\mathbb{W}_{max}}\sum_{i=1}^{\infty}\alpha^{9{i}^2}\left(cos\left(3 \times (2i-1)\times 2\pi x\right)-cos\left((2i-1) \times 2\pi x\right)\right),$where $\mathbb{W}_{max}$ is a normalizing multiplier.
Three-dimensional RF $\mathbb{R}\left(x,y,z\right)$ can be obtained from the definition $\left(1.2\right)$:
$\mathbb{R}\left(x,y,z\right)=\mathbb{R}_{max}^{2}\mathbb{R}\left(x\right).$Thus, the approximation of the three-dimensional RF is also the series of the fine structure constant $\alpha$ along any axis of the reticulum three-dimensional space, and the constant itself is a function of the dimensionless parameter $\sigma$, which is equal to quotient of the “diameter” of some physical object, located in each cell, to the grid step L.
To quantize the time the direct use of the lattice idea is too formal. It is therefore appropriate to use a definition of derivative with respect to time but without moving to the limit. Let $\mathbb{R}\left(t\right)$ is RF on a unit interval $\left[-\text{T/2},\text{T/2}\right]$ and $\tau=\sigma$ и $\text{T}=1$:
$\mathbb{R}\left(t\right)=\frac{1}{\tau\sqrt{2\pi}}\sum_{i=-\infty}^{\infty}\left[\exp\left(-\frac{1}{2}\left(\frac{t+\text{T/4}-i}{\tau}\right)^{2}\right)-\exp\left(-\frac{1}{2}\left(\frac{t-\text{T/4}-i}{\tau}\right)^{2}\right)\right].$By consistently subtracting sinuses, one can show that the approximation of the $\mathbb{R}\left(t\right)$ has the following form:
$A\left(t\right)=\sum_{k=0}^{\infty}\left(-1\right)^{k+1}a_{k}sin\left(2\pi\left(2k+1\right)t\right).$Let use $k+1$ equations with different values of $l$ to determine the coefficient’s values $a_{k}$:
$\sum_{i=0}^{k}\left(-1\right)^{i}a_{i}sin\left(\frac{2i+1}{2l+1}\frac{2\pi}{4}\right)=\mathbb{R}\left(\frac{1}{4\left(2l+1\right)}\right).$Given that $A\left(1/4\right)$ is numerically equal to $2\left(\mathbb{R}_{max}\left(\tau\right)+\mathbb{R}_{min}\left(\tau\right)\right)\alpha\left(\tau\right)$, equation can be written as follows:
$\alpha_{eff}\left(t,\tau\right)=\frac{1}{2\left(\mathbb{R}_{max}\left(\tau\right)+\mathbb{R}_{min}\left(\tau\right)\right)}\sum_{k=0}^{\infty}\left(-1\right)^{k+1}a_{k}sin\left(2\pi\left(2k+1\right)t\right).$$\mathbb{R}\left(t\right)$ is also a hyperanalytic function, as the next approximation takes place:
$\alpha_{eff}\left(t,\tau\right)=\sum_{k=0}^{\infty}\left(-1\right)^{k+1}\alpha^{(2k+1)^{2}}sin\left(2\pi\left(2k+1\right)t\right).$The theory of hyperanalytic function was constructed to some extent by A. Rybnikov (2014) http://www.gaussianfunction.com/.
Another point: I don't think that having each day in its own category removes the desire for threaded discussions, with indentations as in the Café. As far as latest changes goes, OK, that's probably good enough. But if you want to move individual pages' discussions and query boxes to the Forum, then I would want that before I considered it worth the change.
It seems reasonable to me, as long as we have wikilinks working here. If we don't want people to do much replying at latest changes, then lack of threading is probably not a big deal. But we might also want to try to get some input from other than the three (?) of us who seem to be participating in this discussion.
BTW, thanks so much (again) for all your work on this project (the nstuff in general)!! I am continually amazed at how much time you put into it.
Mike wrote:
BTW, thanks so much (again) for all your work on this project (the nstuff in general)!! I am continually amazed at how much time you put into it.
If you mean Andrew, then I agree; thank you Andrew!
If you mean me, then that's because I don't have a job. Hopefully my participation will drop off at the end of this month. (^_^)
I was thinking of Andrew, and I was thinking mostly about the technical stuff (for some reason I didn't see your last two comments when writing mine, forgot to refresh maybe) -- but I'm also always amazed at how much time you put into the nlab as well. (-: And I'm not sure why my opinion should matter more than yours.
Anyway, I agree that this isn't ready yet to replace query boxes, but it might be worth trying for latest changes, if the RSS/search and wikilinks can be gotten to work -- it seems like most of the complication may be hidden for the user. OTOH it'd be nice not to have to worry about finding the subcategory for the correct day to post in. Why is it important to separate out "latest changes" posts according to day? In a real forum like this one, all posts will automatically be dated with when they were made... so could we just have one discussion per month, say?
I'm not sure why my opinion should matter more than yours.
That's because you were the one asking for an RSS feed for latest changes. I'm happy enough without one.
I also agree with you about separate dates' being unnecessary; I'm not even sure that we need to break it up into months. I originally put dates in just like a date stamp; we organised them later. And I originally broke them up into months to keep the page from taking such a long time to load, which is not a problem here. On the other hand, we may find that some organisation is helpful, so it would be good if we could rearrange things if we find that we need to.
But the point is, arrange things how we think will be useful on a forum, not just to match what's on latest changes now.
Well, okay maybe a different subcategory per day is a bit overkill. The point was that in this particular forum software, one can organise stuff by category to arbitrary depth, but the actual discussions are linear. So the idea is that each significant 'latest change' starts a new discussion and any replies - and only replies - continue that discussion. How the discussions are organised is up to you. I picked days because that's what's on the current version. Maybe weeks? One can always shift discussions from one category to another so reorganising later is always possible (though it'd be a hassle to shift loads - may be easier with direct access to the database).
So what would be useful? Wikilinks shouldn't be too hard to implement. Presumably they should always be on (a bit like the 'vlatex' plugin that does the maths here) and shouldn't be "text formatter" dependent. A simple link should go to the nlab page of that name, whilst link (otherlab) should go to the other lab page (is that the right syntax? It should be like that on the Instiki software). I need to add 'search in subcategories' to the search page so that the RSS behaves correctly. Anything else? (Try to keep it in the realm of possibilities!)
I agree about the link syntax. That (and whatever you need to get the RSS working right) is enough to make it work for me.
Are there really enough replies that we need to have a separate discussion for every latest-change? There are really hardly any replies at all -- I expect that if we do it the way you're describing, we'll end up with lots and lots of discussions containing only one post.
Although we might also change the way we use it, for instance we might start posting additional changes to a single topic or group of topics in the same discussion as the original one. Actually, now that I think about it, that might be really useful, to have latest changes grouped by subject rather than by date. I often find myself skimming through the latest changes looking for ones I'm interested in, when there is a lot going on that is not so much up my alley.
We could have it so that there was one discussion per page, although that would add a layer of difficulty when figuring out where to post your update - you have to find the correct discussion to add it to.
I don't see that having lots of single-post discussions is a problem. There's the same number of actual posts, they're just organised slightly differently. What this would allow you to do is to hide follow-ups, but ensure that follow-ups don't get too complicated with threads branching off in all directions (if threads can do that!). So although we don't get many replies, this would make it obvious what were replies and what the replies were replying to (without going down the whole messy "threading" route).
By using, and bookmarking, searches one could narrow down one's field of interest in 'latest changes' so that you could easily see the changes to a particular topic. Maybe the subject of the post should always include the page title.
However, I wouldn't want to make it too easy to narrow down what is seen in the 'latest changes' section. We ought to be interested in finding ways to contribute to others' work and so keeping an eye on what's going on in other parts of the lab is important - even if we don't follow all of it, we may be able to suggest a reference, or correct a misconception that could save someone a lot of time and effort. That's something that we should be actively encouraging.
One other technical feature: I suppose that we ought to allow anonymous posting for 'latest changes'. Although it'd be nice if contributors registered, we shouldn't impose extra burdens on them.
I see that Toby's started trying it out! I'd better get on with the required changes!
I've enable anonymous posting (requires solving a 'reCaptcha' which I hope no-one minds). I've started hacking a wikilinks plugin (surprisingly, this wasn't already available). And I just have to add a 'search subcategories' to the search capability. Though if we decide not to categorise so finely then this isn't so urgent (that is, if everything just gets lumped in 'latest changes' category rather than subdividing by date). What's the prevailing opinion on this, by the way?
My opinion is to lump.
I also still prefer lumping it all together, perhaps grouped in some larger way like by month.
By the way, it is kind of confusing right now that we have three discussions called "latest changes." Does the forum software allow arbitrary numbers of discussions with the same name? That seems like a recipe for confusion to me...
Right, wikilinks should now work. The syntax is \[\[(type:)?link(|text)\]\]
where either 'type' or 'text' can be empty. If 'type' is empty then it defaults to 'nlab', if 'text' is empty then it defaults to 'link'. This gets replaced by a hyperlink with 'text' as the visible text. The hyperlink is determined by a template, which depends on 'type', in which a certain string (%s
) is swapped for 'link'. The default template is thus:
http://ncatlab.org/nlab/show/%s
If the 'type' is not known, then the following template is used:
http://ncatlab.org/%t/show/%s
where %t
gets replaced by the 'type'. Note that \[\[:link\]\]
is different to \[\[link\]\]
as the first has empty type whereas the second has default type.
The type can only be letters and is case insensitive, if other characters are used then they get subsumed into the 'link'. The 'link' gets urlencoded as an attempt to avoid nasty links.
At the moment, the available types are the default and catch-all described above, plus google and wikipedia. The matching occurs everywhere, even in code blocks, so to escape it use backslashes. Examples:
\[\[direct limit\]\]
-> direct limit
\[\[colimit|direct limit\]\]
-> direct limit
\[\[doriath:page one|An amazing page\]\]
-> An amazing page
\[\[google:nlab|Google search on nlab\]\]
-> Google search on nlab
\[\[wikipedia:calculus|Wikipedia's page on calculus\]\]
-> Wikipedia's page on calculus
I can add other templates, and the Wikipedia one is currently to the search which may not be what is wanted. To request a template, tell me the 'type' and the 'template' (where %s
should be replaced by the link text).
Hmm, escaping with backslashes doesn't work outside of code blocks when using the markdown filter. I suspect that the order of filters matters here: markdown converts \[
to [
which then gets matched by this filter. Probably not a lot I can do about that without hacking Markdown. But then, the '[' and ']' are so overloaded that they should almost never be used to produce actual brackets. Use entities instead (&#91;
[ and &#93;
], if I remember right).
So if we agree to lump things together in the 'latest changes' category then we now have all the functionality that we want, at least to start with.
I would suggest that we put the page name(s) in the titles (these won't link since the formatting only applies to the comment body, so also put the page name(s) in the comment bodies). I would suggest that only the page name(s) be in the titles, though I suppose it could be useful to have a single word indicating the level of severity of edit.
To answer Mike's question elsewhere, discussion titles are non-unique. It is the discussion ID that is unique. We can use this to our advantage if you like the above scheme since then a search for discussions matching 'limit' will fetch all the comments on changes to 'limit' (and probably 'colimit' and 'directed limit' ...).
Excellent!! You are awesome. Shall we suggest that everyone give it a try?
I'm a bit hesitant about just going for it without some indication that anyone other than us three is aware of the discussion! To make it really work (read: to give it a fair trial) we'll need to redirect 'latest changes' here. That's easy enough to do, but it might provoke complaints!
I should delete the spurious subcategories first as well since we're, in the first instance, not going for any deeper divisions. I'll do that later when I've worked up the energy to delete enough cookies to log out (need to eat some cookies first ...)
Hi,
sorry, I was busy with other things.
I am in favor of this idea. At least in as far as I can see what "this idea" is.
So once "this" is activated, the idea is that the link latest changes will take people to a forum thread into which they type their remarks essentially exactly verbatim as they are doing so far. Right?
What is that about categories? we are not going to have to assign date categories by hand to every entry for this to work, have we?
What is that about categories? we are not going to have to assign date categories by hand to every entry for this to work, have we?
No. It was part of the original plan simply because we started as close to the current system as possible. But it got thumped down fairly early on.
Great, so let us not stop Andrew from implementing this revolution then!
Right, I've removed the spurious date sub-categories so there's only the 'latest changes' category. It's clean, so we can switch over.
If someone else wants to make the switch, then feel free. Otherwise, I'll do it tomorrow. As I see it, the scheme should be:
Have I missed something?
<grumble>Since people still insist on adding new entries to latest changes to the middle the top where they will actually be seen, I would like this to be implemented as soon as possible!</grumble>
To help that, I will now test whether a Guest can start a new discussion.
Danger, Will Robinson! No, a Guest cannot start a new discussion.
To begin with, there's no link to http://www.math.ntnu.no/~stacey/Vanilla/nForum/post.php on the Guest's main page, just some stuff about how you need an account if you want to post anything. (There's some stuff about that at the top other pages too, such as this one.) What's more, when I go to that URI anyway, I am not allowed to start the discussion:
Some problems were encountered
You cannot take part in the discussions because you are not signed in.
You cannot take part in the discussions because you are not signed in.
<div>
<blockquote>
Very few people there pay attention to what is going on here.
</blockquote>
<p>Making "latest changes" live here will drastically raise awareness of the forum here.</p>
<p>It will also inevitably change the nature of the nForum from a place purely for meta-discussion into a place for genuine discussion of nLab content, I suppose.</p>
<p>Which is probably a good thing in the long run. I kept hoping that the nCafe would be able to play that role, as it has an established supply of regular lurkers who might take interest, but likely the forum here has plenty of advantages for such as discussion compared to the blog. Not the least one being that everyone can start a top-level discussion.</p>
<p>With the highly appreciated new functionality of having wiki-links here on the forum, what are the prospects of having typeset math here on the forum?</p>
</div>
Right, I want double pay for this one.
I think that I've solved the problem with starting new discussions. It seems to work when I try it. Basically, if you are somewhere within the 'latest changes' category then you get a 'start new discussion' link which takes you to the right page. There's a recaptcha to solve before you can start your new discussion, and if you have an account here you can log in instead.
This is hard-coded to work only for the 'latest changes' category. If we do want subcategories then it'll take a little more coding to enable that so give me advance warning that that's what you want! (You have to do that anyway since I'll need to fix the search setup).
Please can people test this. I know it's annoying to have to delete cookies to log out, but it will be useful since it's only by testing before we "go live" that we'll know what the issues are.
Urs, by 'typeset math' I presume that you mean MathML (via the iTeX filter). Am I right? You can do latex->pictures here. I keep forgetting what I have the delimiters set to, I think it's double dollars: .
In reply to Eric's comment: take a look at the top entry for 'latest changes' on September 11th. What else would you like me to do?
I suppose that a summary would be useful, though.
Thanks, Andrew.
Let me just seconmd Eric in that as someone late to your long discussion about this here, it is very hard to figure out what exactly you have settled on and what exactly is going to be implmented. I am sure it is all to be found somewere atound here and plain obvious to those involved in it, but it is not easily recognized when coming from further outside.
So if you have the time it would be great, when moving "latest changes" here, if we could provide a link to a place where all this is summarized briefly.
The scheme is:
\[\[Lab Elf\]\]
produces: Lab Elf.Does that make sense? If not, please ask since if it doesn't make sense to you then it probably won't make sense to anyone else coming to this fresh and it'd be good to get it right.
When we shift someone (me if no-one else) will put all this both in the current 'latest changes' page on the n-Lab and in the first discussion in the 'latest changes' category here.
Yes, makes very good sense. Thanks for all your efforts, Andrew, this is really great.
Eric, your post justifies the move!
The problem is that there isn't somewhere that I can put an announcement about something like this. And this is, after all, really just about technical jiggery-pokery - it's a change to the infrastructure aimed at making the experience better for the users.
But you are right that now that it seems that those who participated in the discussion are happy with the new scheme (am I right, Mike and Toby?), that we ought to give a little notice before actually changing. I can't do anything about the Cafe but I can put a message somewhere on the n-lab. Howabout in 'latest changes'?!
Just to reasure you on one point, people won't have to register to post in the 'latest changes' category (though they will if they want to post in any other category) so it shouldn't be any harder to use than the current system. In fact, it should be easier.
The link to 'latest changes' on the n-Lab will now point here to the category 'latest changes'.
I don't understand what you mean by this.
It would be a Bad Thing if http://ncatlab.org/nlab/show/latest+changes simply HTTP 301 redirects to the forum, because then nobody can go back and see what used to be there; it breaks the wiki. (However, the page can be archived to 2009 September changes and then replaced with a link here for the transition.)
It would be a Bad Thing if http://ncatlab.org/nlab/recently_revised similarly redirects or otherwise goes away; I find it useful, now that it doesn't hang the rest of the Lab, and perhaps others do too.
But it would be a Good Thing if the link on the HomePage telling people to look at latest changes now tells them to look at the forum category.
Wikilinks automatically link back to the nlab, thus
\[\[Lab Elf\]\]
produces: Lab Elf.
Remember to make that with SGML numerical character entities when you post the instructions for newcomers!
Wikilinks automatically link back to the nlab, thus
[[Lab Elf]]
produces: Lab Elf.
You can do latex->pictures here. I keep forgetting what I have the delimiters set to, I think it's double dollars
You know that it's double dollars, since you produced an example. That's convenient, but the problem with double dollars is that you can't preview it.
But you can preview the iTeX if you use <latex>
…</latex>
! Good to know when you try something tricky.
Is there an obstacle to making double-dollarsign-math previewable? It would also be more comfortable if it were just single dollar-signs, since the math ends up going inline rather than getting displayed.
To spell out what I meant: the link to 'latest changes' on the homepage should point here. However, the wiki page 'latest changes' will not be redirected but will contain an appropriate message and point here. I propose that the 'latest changes' page be cycled so that it's clear that what is on the n-lab is archive material and shouldn't be added to.
As for the maths here, it was something I hacked together right at the beginning in order to get some functionality. It would be nice to be able to use iTeX here and have MathML output. There's several steps to doing that, and any offers of help would be gladly received!
So it doesn't work too well with previewing because I didn't take that into account when I wrote the code - and it was heavily based on the wordpress LaTeX plugins that don't have to worry about previews. The question is whether it is worth getting the current plugin more user-friendly, or simply developing an iTeX-system. I guess that people coming from the n-Lab would prefer to use iTeX.
To avoid distracting from the important topic of that thread, let me reply to Toby's comment here.
I'd made it so that guests could only start discussions in the 'latest changes' thread, but by default anonymous posting was enabled on the whole forum. I hadn't thought about the implications of this, but after seeing that particular post then I decided that I didn't like it. So I've disabled it outside the 'latest changes' category. I should also change the message to make it clear that the person's name won't be displayed unless they put it in themselves, so they ought to do so.
Over at the lab, Zoran posted the following:
I am strongly against shifting latest changes to the forum. I do not know how to quickly link and do other features like here, it requires more downloading capacity when on expensive network like mobile, it may require account, it does not get recorded when downloading the whole nlab etc etc. Logging to forum is anyway pain when on mobile network. It logges you off for example if you are idle for 30 minutes. I will not do it simply. I quit logging changes if it is to the forum. I will edit nlab without logging changes in that case. Nlab is nlab, and it should be self contained. Forum is about general policies, it is complicated enough to explain to the new userts that there is latest changes notificatiopn, notg in addition now that they have to have an additional account and additional web page with different software. I never use RSS feeds nor want to use them: I do not check latest changes unles sI am generally interested what is there. If I work on the item “jabberwocky” I WORK on it. If I want to see latest changes I look at them. It is very important that I can download the whole nlab including the latest changes histories. Forum is different system and it should not be mandatory to use it. I also find useful that I can link and cut and paste formulas and nlab links easily within nlab latest changes the same way as I do the rest, the forum has a bit different formatting and makes it harder. Also nlab item jabberwocky has down there a link that it was mentioned with link in latest changes what I also find useful. I also do not find any argument in the saying that if I log to forum for looking at latest changes I will also see be “informed on the new things”. Thjis is not a feature but a DISTRACTION when I work hard and follow references and try to format my mathematical text. Forum is about policies and politics, and software.- Nlab is about mathematics. I like to havce that CLEANLY separated. The state of my mind is the prerequisite for working on nlab. The alternative is that I work only in my personal nlab if you impose this new policy of mixing with forum.
Here's my response.
I do not know how to quickly link and do other features like here
You can use markdown syntax and wikilinks just like in the n-lab. The content of the 'latest changes' page has never yet used complicated syntax that isn't available in the forum, but if there is a need for it then I will enable it.
it requires more downloading capacity when on expensive network like mobile
Really? Let's test. Latest changes on n-Lab, total download size 260Kb, main page of forum, total download size 122Kb.
it may require account
Doesn't.
it does not get recorded when downloading the whole nlab
Given that you are using the wget
script that I wrote (at least, I hope you're using that or something like it and not clicking export_html
), it would be extremely easy to modify it to take a snapshot of the forum as well.
Logging to forum is anyway pain when on mobile network
Why? What's so awful about it? Anyway, you don't have to for the 'latest changes' category.
It logges you off for example if you are idle for 30 minutes.
Not if you click 'remember me'. Anyway, you don't have to log on for the 'latest changes' category.
it is complicated enough to explain to the new userts that there is latest changes notificatiopn, notg in addition now that they have to have an additional account and additional web page with different software.
Instructions: when you've finished editing a page, click on the link 'latest changes' on the homepage and fill in a brief description of what you've done.
Covers the old system and the new, I think.
Since you (later) mention Jabberwocky, maybe you'll get this reference:
What I tell you three times is true: you don't have to have an account to post to the 'latest changes' part of this forum.
I never use RSS feeds nor want to use them:
I do and I do. I think that they are a fantastic idea. This proposal came about because someone else thinks they're worth having too (Mike, it happened to be). Just because you don't use them, should we be hampered too?
I do not check latest changes unles sI am generally interested what is there
I'm sorry, I can't resist this open goal: How do you know what is there?
Forum is different system and it should not be mandatory to use it.
It isn't. But then it isn't mandatory to log things to latest changes either. The point of it is to make it easy for others to know what you have done. This idea is to make it even easier.
I also find useful that I can link and cut and paste formulas and nlab links easily within nlab latest changes the same way as I do the rest,
Most of what you do on the nlab can be done here. Okay, so the mathematics is slightly different but the links are the same. There is very little actual mathematics on the latest changes page in the nlab so I'm not convinced that this is a particular problem.
the forum has a bit different formatting and makes it harder
I agree that it's a bit different. I intend working on that if it proves necessary. It may have escaped your notice, but iTeX itself is evolving. If you had an RSS reader, you'd know that there are loads of new arrows that can be used. I, personally, find it extremely irritating that I can't use \(
and \)
on the n-lab, so I have to learn a different syntax for writing on the n-lab as I do in papers.
Also nlab item jabberwocky has down there a link that it was mentioned with link in latest changes what I also find useful.
But when did that link appear? Was it recently, was it a week ago, have you already noted the change that that referred to, or is it a new change? The presence of that link is a one-shot system. Once it's been mentioned once, the link is there. If you had an RSS feed reader, you'd be able to see exactly when each page that you are interested in is mentioned. You can even set up a feed for specific pages! The possibilities are endless.
I also do not find any argument in the saying that if I log to forum for looking at latest changes I will also see be “informed on the new things”.
Here's an argument for you: if you don't pay attention to what's being discussed at the forum then one day you may turn up to the lab to find the place unrecognisable because those who have been paying attention have been discussing ways to make it better and their opinions have carried the day because their opinions were heard. If you don't tell us what you think, we can't figure it out for ourselves.
(To be continued ...)
Thjis is not a feature but a DISTRACTION when I work hard and follow references and try to format my mathematical text.
Fortunately, you can ignore it if you like. If you only go to the 'latest changes' category page then you can safely ignore the whole rest of the forum and not learn about useful things like how to download a copy of the n-lab without it grinding to a halt.
The alternative is that I work only in my personal nlab if you impose this new policy of mixing with forum.
I guess that this is the bit I really don't understand. Surely you have two windows or tabs open, one at the page you are editing, and one at the 'latest changes' page. When you're done with your edit then you flick over to the other and put a brief summary in there. Where in that does it matter that the two systems are using the same underlying software? If I made a mock-up of the 'latest changes' page so that the two looked exactly the same, would that matter?
But no-one is imposing anything. And you make it sound as though we've gone behind people's backs. The issue was raised, it was discussed, when we thought we had a reasonable idea then I put something on the latest changes page asking for comments, we didn't get many but Eric and Urs at least came and had a look, so we figured we'd give it a go.
Forum is about policies and politics, and software.- Nlab is about mathematics. I like to havce that CLEANLY separated.
Nlab is nlab, and it should be self contained.
This is the bit I really disagree with. The rest is just technical stuff.
Firstly, the separation that you ask for is impossible. The mathematical content of 'latest changes' is pretty small, and yet it's probably one of the most important pages on the lab.
But mainly, why is this separation so desirable? I see the n-lab as a way to help us do our research. To the extent that something helps with that, it's useful and worth trying. To the extent that something doesn't, then it should be changed. To say that the n-lab should be self contained just seems plain daft to me. There's discussion going on about how best to do references and bibliographies. Wouldn't you like it if typing \cite{arxiv:0505077}
produced a link to the correct arxiv record, and a properly formatted BibTeX output on TeX export? That sounds quite useful to me, and may be possible. But it's not going to be internal to the n-lab, at least not in the first instance. The real reason why is that there are already programs that do part of the work so why on earth would we spend time rewriting something that already exists just so that it is internal to the n-lab? Surely one of the great things about the internet is the fact that we can take lots of different systems, all doing different things, and chain them together via hyperlinks to get a coherent whole.
There may be other ways of doing this. It may be possible, for example, to embed a forum-like system inside Instiki. Would that be acceptable to you? It may be possible to have RSS for individual pages. However, those take more work to implement than this system which has the benefit of using something that's already lying around. If this doesn't work, we haven't lost too much - a couple of days work for me, a little for everyone else. Also, if it doesn't work then it'll tell us a bit more about what we really do want for latest changes.
I'm sorry if experimenting the n-lab doesn't feel appropriate to you. But I'm investing a fair amount of time in getting the infrastructure of the lab working and sometimes that requires experimenting. One thing's for sure: the n-lab infrastructure is not static.
To end on a lighter note, here's a quote from the Vicar of Dibley (can't remember which character said which and 30s on google didn't help):
A: There's good change and bad change.
B: Prawn cocktail crisps
A: They're bad change.
B: But traffic lights - if they never changed... they're good change
A: The important thing is to know which is which.
Okay, thanks Andrew and Zoran for all the input.
So we had thesis and antithesis, here is the synthesis that I draw from this
In summary two central points of the discussion are:
a) a discussion forum for latest changes has lots of desirreable advantages for those regular users who care (I do)
b) but for various purposes -- such as the needs of newbie and sporadic users, such as mirroring or downloading etc -- it is desireable to have the forum functionality be more an add-on than an integral part of the nLab.
The LaTeX comparison that Zoran makes is maybe a useful metaphor:
there is a rudimentary basic system that has been coherent and hasn't changed for years (or decades I guess), with all the advantages that brings with it
while at the same time everybody actually using it is free to choose various add-ons, like editors, discussion forums, etc, to actually access this.
So, that all sounds reasonable. Can we implement the compromise? Can we have discussion of latest changes at the forum, while those who choose for whatever reason to be rather not bothered with that to remain within the nLab and see the main logging information there, too?
Maybe there is some way to keep a page "latest changes" on the nLab and have its content be synchronized automatically with some content produced at the forum? Not the full forum discussion, maybe just the top-level notifications about new entries?
Such that, indeed as Zoran suggests, anyone taking a snapshot of the nLab alone would actually still have pages with the main latest-changes-logging?
Maybe the top-level entry anyone would post to the "latest changes" section at the forum logging some change could automatically be pasted into the nLab page "latest changes" on top of the list?
I am not sure, does that sound like a weird idea?
@ Eric:
I like having discussion right there on the page. I find it much more convenient than at Wikipedia. That's the same reason that I don't like discussion at latest changes.
@ Andrew:
Perhaps this thread should be moved to the 'latest changes' category so that Guest users can post here? Anybody who objects, like Zoran does, because they don't want to bother with a new login account, can then voice their objections.
@ Zoran:
Just some technical notes, really; although some of these are mild arguments against you, I'm still somewhat sympathetic to your position (but I was happy to go along with this change for those who use RSS readers, and maybe try that out myself).
two passwords
There are no passwords at the nLab. There are also no passwords in the 'latest changes' category at the forum if you post as Guest, although you have to sign your name by hand.
When I get message "Firefox has an update, do you want to download it" i click on NO and find that ad annoying.
You can turn off that ad! Open up Preferences/Options, go to Advanced, then Update, and uncheck the boxes to search for updates. You can always turn them back on again later.
If I say to somebody I meet at a conference that we have nice nlab and that its contribution involves two different web sites
Even if this goes through, don't tell them that. Tell them about the nLab, then later someone (even you) can tell them: ‘It's nice if you click on this link and log what you've done, since many of us find it helpful. You can also get an RSS feed of the log, if you're into that sort of thing.’. And if they don't like that, then they don't have to do it, although it does help the rest of us if they do.
latest changes page preamble has lots of preaching about usefulness and need for logging even small changes and about bad effects if one does not do it
I confess: I don't log everything. I logged a bit more when Recently Revised was down, but even then not every added link or comma. I log something if I think that anybody is reasonably likely to care about it. (That includes all new pages.)
not creating new versions if I update within 30 minutes incrementaly and nobody intervenes in between
On the forum, you can edit your posts without creating a new one. (Look at one of your comments here; under your name should be ‘edit’ and ‘delete’.) There is no time limit, and you can do it even if somebody else has written another post in the meantime.
On the forum, you can edit your posts without creating a new one.
However, it looks like the Guest cannot edit Guest comments. (’_‘)
So, the nlab actually already has an RSS feed for all changes: http://ncatlab.org/nlab/atom_with_headlines . It seems that mostly what is added beyond this by "latest changes" is information about which edits are "important" enough for their authors to bother mentioning them there, along with whatever comments about their changes that the authors feel like making. So if there is substantial feeling against moving latest changes to the forum, how hard would it be to hack instiki to have a major/minor-edit checkbox and a "comments" text entry when you edit a page, which would then show up in the RSS feed? Then we wouldn't need "latest changes" at all.
I feel like I've mentioned this before and some people were against it, but I don't remember why.
Here's my idea: you lot all discuss what you want. Then when you've decided, formulate a proposal and I'll tell you whether or not it's implementable. If it isn't (which will almost surely be the case), you can try again. In the meantime, rather than trying to figure out a way to satisfy everyone, I'll get on with something else. I fully expect to have plenty of time since the likelihood of you lot deciding on anything is also almost surely nil (witness the "what is the nlab about" discussion). In the time it takes you I may even have learnt a little about Ruby and Rails.
Moving this discussion to the 'latest changes' category seems like a reasonable idea. However, as Mike started this discussion then it's up to him to move it (I could override it as admin but that would involve logging out and logging back in again - too much effort, frankly, and I prefer not to abuse my superpowers).
Eric, looks like you were wrong about the pitchforks!
Andrew,
I fully understand that you are feeling frustrated now, having spent lots of energy into implementing this new feature and then hearing such opposition against it.
I am not sure what the best thing to do is. One thought that comes to my mind is this heretic one: you are currently the one in control of the software. Had not Eric convinced you to post a request for discussion for the latest-changes-2.0 implementation, you'd by now have already implemented it without discussion. And none of us could really have stopped you from doing it :-)
I am not sure what that means, it's just a thought that came to mind.
I am also not sure how we could have a more efficient decision mechanism. The nLab lives from having many different contributors. It is for instance not my private wiki on which I want to push people around to follow my guide. On the contrary, I'll find the nLab a success when the group dynamics has developed to a point where any administrative intervention on my part, such as these long discussions here, have become superfluous.
Now, it is hard to find a fool-proof road to that goal. But likely goingg through the trouble of openly discussing all kinds of things here is a necessary ingredient.
Finally, in view also of Mike's recent comment: it seems that no opposition whatsoever is to be expected from any changes to the nLab that come as voluntary add-ons. Wouldn't it be easy to have -- at least to begin with -- the new latest-changes system that you imagine in parallel to the existing latest changes page? Not the least this would give everyone the opportunity to see exactly what the new system actually will be like in practice. You kept pointing out that lots of points that were criticized in this discussion were actually not going to be as imagined by the critique. So maybe a test run in parallel to the existing system will be helpful.
I am not sure what that means
It means that Andrew is a decent and honourable person.
It may also mean that Eric will see the value of a formal decision-making process. (^_^)
So maybe a test run in parallel to the existing system will be helpful.
It might also be confusing to have latest changes posted in two different places, probably neither of them with all of them. But I also thought of this; it may be worth trying anyway.
And in my outline, we would go back to Step 3 until all concerns have been addressed or withdrawn.
Which possibly they have, since Zoran seems to have stepped aside. So maybe we should ask:
Are there any outstanding concerns?
I agree with you about airing on all three sites. But in my outline, the question ‘Are there any outstanding concerns?’ does not mean simply ‘Are there any new concerns?’; it is a question such that, if nobody answers Yes, then there is no need for a vote because we have reached consensus. (^_^)
My experience is that while seeking consensus is good, being unwilling to act without it is dangerous, since inaction is often a form of action.
If there were a number of people with various concerns, I would propose that maybe we try out the new system for a period of time (perhaps in parallel with the old one, but probably just on its own), and then revisit the question of whether to move permanently afterwards. But since Zoran is the only one who's voiced any objection to the new system, and he's said that he just wouldn't use it, it's not clear how much that would actually accomplish. Although it's also possible that some of us would discover unexpected things we don't like about the new system after using it for a while... so maybe such a "trial run" would be a good idea.
A trial run seems like a good idea in any case.
Concerning Eric's request for posting announcements to all three places, blog, nLab, nForum let me just notice that we actually did that.
Much as I regret it, it doesn't really seem as if there were a huge crowd of nLab contributors who would frequent the blog but not this forum here. You (Eric) kept emphasizing that only a handful of people are participating in the discussion on the forum here.
But also only a handful of people (the same handful) are regularly contributing on the nLab. So it would seem to me that this forum here might just as well be declared the official place for nLab announcements to be made. Contrary to what I expected, the blog regulars are not endorsing the nLab as a place to show activity. It's a bit sad, but it seems we tried everything to change it and have to face the reality.
Conversely, potential contributors here at MPI, the participants of those seminars that I kept wiki-fying, they are aware now well of the nLab but seem to feel unsure about what role the blog plays with it. So maybe we will just have to face that the nLab/nForum live a life independent of the blog, eventually.
I'm not sad about them being independent, in the sense of not dependent (i.e. a person can participate in one without contributing to, or especially understanding, the other). But I would hope that they would remain connected, and the core of us who frequent both may be enough for that. We frequently link to the lab in our posts to the cafe, and some lab pages conversely link to cafe discussions—so I think people who are initially drawn into one or the other may eventually learn about, and figure out, the other.
This is getting a little off-topic. Discussions on decision-making should be taking place over at the relevant thread, otherwise there's lots of potential for getting confused.
Let me sum up the state of play.
Some people (me included) think that latest changes as it is is not ideal. A particular feature was identified as being missing, namely RSS feeds.
Various ideas were batted back and forth as to how to add this feature. Eventually, a proposal to move latest changes here was proposed. It is worth making a couple of comments about this proposal that have not been made explicit before, though there have been hints of them:
At some point, a request for comments was made on the latest changes page. Looking at the comments on this page, this brought Eric and Urs over here.
Somewhere around here, I started hacking this forum to make it fit the desired features of latest changes so that the proposal would work.
Once this was done (or at least, thought to have been done), we announced the migration, as per Eric's suggestion (despite the outcome, still a very good suggestion!).
At this point, to coin a phrase, everything went pear-shaped.
We are now at an impasse. The technology is all in place. All that needs to happen is the link on the n-lab to latest changes to point to the category over here, and an manual redirect on the latest changes page itself (i.e. a note saying "click here" and explaining why). Anyone can implement that, it doesn't have to be me.
But I'm not going to implement that, so if you are waiting for me to do it then you'll be waiting a while. Before I do anything then I want the descision-making discussion and the 'scope of the n-lab' discussions sorted out.
The thing is that I'm in a little bit of an unusual situation regarding the n-lab. Urs and Toby remarked on it but whilst they may have thought through the implications themselves, they didn't share them here so I'll do so.
I'm simultaneously tech support, sysadmin, financier (technically in full, but morally only in part), and legal (note the adjective!) owner of the n-lab. I'm also a contributor (at least, when I get the time!) but as a contributor then I'm definitely a minor one. So me making a decision like this represents a huge potential conflict of interests!
As an example, as sysadmin, I'm prepared to make decisions to make the n-lab as smooth as possible within the remit that those don't have a noticeable detrimental effect on the majority use of the n-lab. So disabling export_html was okay, as almost no-one uses it (except robots), the functionality can be replaced with a wget
script, and it significantly improves the use of the lab for everyone (by removing a huge resource hog). But going beyond tweaks like this gets me into dangerous waters where optimising one bit degrades another and knowing which path to go down is a high-level decision. As sysadmin, it is definitely not my role to say to someone "Thanks for your input, but I'm going to ignore what you've said.".
So until the decision-making proceedure is sorted out, and I don't think that that can be done without also resolving the "what is the scope" discussion, I'm not going to implement any major changes.
On the other hand, I think we've made a fair bit of progress discussing this and I wouldn't like to see that go to waste. So just because I think that those discussions need clearing up, it doesn't mean that we can't get the proposal for latest changes finished.
I still have a problem seeing Zoran's point. I don't think that we have heard an 'antithesis' at all. Maybe someone more discerning than I am can sum up his points for me. The LaTeX analogy is a mess. LaTeX is currently at version 2e, with version 3 in the pipeline. TeX is somewhere around the 3.1415 mark and took 10 years to get there. I think that there's a confusion between backward compatibility, which is good, and freezing development, which is bad. Whilst one can say that TeX itself has been frozen, the TeX system has not. Saying that I cannot use 'pdflatex' or the 'tikz' system just because I didn't use them 10 years ago is nonsensical. Of course, I want my old papers to compile and produce what they used to. But I can have that and have the recent improvements.
As far as I can see, the only loss of actual functionality that Zoran identifies is the fact that pages get a link when they have been mentioned in the latest changes page. I agree that this would be hard to implement in the proposed system. However, in so far as this is a genuine feature, I would consider it severely broken: it only registers the first mention, it takes no account as to whether that content was significant or not, it gets wiped every month, you have to be on the page to see it anyway (so you can just look at the diff to see what changed), you have to go to latest changes to see why it was mentioned, in which case you may as well take a quick look to see if any other pages that you are interested in have changed.
The other issues that he raises have, I think, been dealt with. The two-sites issue is silly. Downloading the lab occurs via a script anyway so I can easily fix that to include the latest changes here. I think that the case is stronger for newbies getting dissuaded by not making latest changes feedable, and thus making it difficult to be a 'casual user': with the current system you're either all-in (checking every day) or not at all.
All the suggested compromises would involve a lot of hard hackery. The proposed solution did not actually take a lot of time to set-up. As I said above, it is a step along the way to improving the whole n-experience. I think it's a good step and I'm having a hard time finding anything bad about it.
Andrew, it seems there was general agreement here on a test run. So we should insert a link at latest changes that points to the new latest changes page and says something like "for a trial period, you are encouraged to log your latest changes here".
I will insert that link if you don't. But I am not sure what precisely the url has to be!
it seems there was general agreement here on a test run.
Two comments:
Everything is a test run! And as we discovered with the migration, for us, a test run needs to be live to be a proper test.
"general agreement". Errr, that's the basic problem, isn't it? The "general agreement" seems to be between the initial discussion group with you and Eric for good measure. What about the various Davids? What about Todd and John? Do we assume that they are sufficiently aware and therefore are indifferent? Given that we made two announcements on latest changes, we could have assumed that for the first (admittedly the second is a little hard to miss!).
But whatever decision system is decided upon, I guess there's no problem with you making an "executive decision" (indeed, I would like to ensure that whatever system there is has that built in). So here's the instructions on what to do:
Make the link in the 'contents' side bar point to http://www.math.ntnu.no/~stacey/Vanilla/nForum/?CategoryID=5.
On the current latest changes page, add a link to that url, including a statement about what it is, what it's for, and how to use it.
Also add instructions as a post in the 'latest changes' category here (I'll make that "sticky" so that it's always at the top). Name it something sensible like 'Instructions'.
The instructions are (I think, someone else may want to add to these):
Log latest changes by 'Start new discussion', include the page name (or a short version thereof) as the page title.
If you don't have an account, using the reCaptcha will mean you don't have to create one, but please be sure to put your name in your log message.
You can reply to posts by others (if as "guest" then please put your name in again).
You can use basic markdown syntax (make sure that the 'markdown' radio tag is checked under the comment box), wikilinks point back to the n-lab, and mathematics should be between double dollars.
You can subscribe to an RSS feed for the whole category. It is also possible to have "smart feeds" whereby any discussions matching a certain pattern are flagged (for help on this, just ask in the 'n-forum' category - account required - or email 'tech support' (i.e. me, but don't put my email in, figuring out my email is a small hurdle to ensure that I only get emailed as a last resort!)).
If you create an account (just takes a few seconds), the system will keep track of which discussions you have read for you.
By a "test run" I meant that we would commit ourselves to reopening the discussion about whether the new system is better at some predetermined point in the future after the switch-over. But given how much we like to talk, perhaps something formal like that is unnecessary.
Regardless, I think it's past time to try the switch-over.
Zoran's last comment said ‘It is simpler that you all ignore my complaints.’. Let's give it a test.
Since the changes are currently being broken up monthly, it might be convenient to say that our test is for October. Put a standout box at the top of latest changes announcing this fact (I know, another announcement, but the old one's been up so long that people are probably ignoring it now), and I'll follow Andrew's steps Wednesday afternoon (as I see it), when the date changes UTC.
But given how much we like to talk, perhaps something formal like that is unnecessary.
That's partly what I meant. The other part is that, for me, the whole n-lab has the feel of a "test run".
I've cleaned out the 'latest changes' category. Regular readers will be pleased to know that I moved the discussions that were there to a new category (Trash) rather than delete them. This discussion stays, obviously. The others were all people trying out the guest posting.
Once someone writes a post with instructions then I'll make it "sticky" which means it will always be at the top of the page.
By the way, I'll try to get around doing it. But if anyone else feels to do what it takes to "set up the new system", don't feel obliged to wait for me...
Urs, Toby volunteered to do it already with the sensible suggestion that it happen at the usual monthly rollover. Of course, if you are at a loose end I'm sure he'd be happy for you to do some of it!
OK, I made another announcement, this time at the top of latest changes. I will move most of that to a new ‘Instructions’ post (which Andrew will make sticky, I guess) in about 23 hours, so edit it now if you find it unclear. And then I will archive the 2009 September changes, giving us a new latest changes page that basically just tells people to come to the Forum.
Actually, maybe now this discussion should be moved back to a different, more meta-, category?
Probably, in which case it's your job, Mike! I think it'll be useful to have a specific feedback discussion in the latest changes category (mainly to keep all such feedback in one place and avoid cluttering up the latest changes category), but there's no reason why someone wanted to comment on the implementation should have to wade through all the discussion that lead to it if they don't want to.
So I see that Urs has started using this a day early! Now that I see it in use, I can think of two minor improvements:
There's a plugin for the second, but I think the first might take a little hackery, which is why I'll not bother if only I think it's worth it.
Yes, I felt like going ahead. I also thought that some role model might be helpful. :-)
I am already enjoying the new system: it's quicker to use. I don't have to retype my name, don't have to worry about formatting, I just go new-entry, didthisandthat, done. Nice.
Since the first post in a discussion is likely to be fairly short, I'd like to be able to see all the unread ones in one go.
I don't know what you mean by that, can you elaborate?
I'd like to be able to quickly mark a load of discussions as read.
The "read or not read" flag on the forum is totally out of whack for me since I read most things just via the rss feed, which doesn't mark them as read by me. Whereas my feed reader does keep track of what I've read there. There's probably no hope of synchronizing those, so this one doesn't matter much to me.
By my first point, I meant that as I tend to work on the fringes of the n-lab (content-wise), most of what goes by on latest changes doesn't concern me. But it's still good to glance through it, occasionally I learn something and even more occasionally I can contribute somewhere that I didn't think that I could. At the moment, I have to click on every discussion to see what's going on. If there was a view that simply showed the first comment in each discussion then that would fix it.
I don't think that there's any way to synchronise a feed reader with the forum. It would require information going in both ways, whereas at the moment it goes in just one. On the other hand, a "mark everything as read" might help you since if it did annoy you to have things out of synch, you could read all the ones you wanted to in your feed reader and then just click the "mark everything as read" over here. But if you primarily use your reader, it probably doesn't matter at all.
If you care to make the Instructions thread sticky, Andrew, here it is.
Consider it stuck.
I've also modified the view for the 'latest changes' category in line with what I said above. In the category view, and only for 'latest changes', it displays the first comment of each discussion. One could consider truncating these, but none seem to be too long. Also, maybe they could be hidden if they are already marked as read - that shouldn't be hard to do.
Looking at the content of most of them, I suspect that having a separate discussion for each page is a little over the top when most of them are just "created X". Perhaps if all that is wanted is to flag new pages, one could just write a "New Pages" discussion and list all the new ones that you've added in the past hour/day/whatever (since titles can be non-unique, and the person's name and the date are automatically added, just "New Pages" would be enough).
These are just thoughts from the first day. I'm sure I'll change my mind many times over the month!
I like the current situation with one discussion per "created X". It is quick and easy to type and quick and easy to follow.