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.
Prompted by signs that the current server is reaching its limits with the existing software, I am going to attempt to carry out the migration to the cloud (Amazon Web Services, aka AWS), that has been intended since we raised funds for this purpose early this year. For those who do not remember, we will be using an AWS account owned by (and billed to) the Topos Institute. Brendan Fong is our principal contact at the Topos Institute.
The full migration will take a long time. I do not really have time for it myself at the moment, but the time has come that it is necessary to do something. I am going to try to gradually build functionality up; the first goal is simply for the nLab to be viewable (i.e. it will exist in read-only mode to begin with).
I am now going to disable the current nLab server, to prevent writes to the database whilst I am migrating existing content. My apologies for the inconvenience, but with the tiny amount of time that I have, this is the only simple way to do things cleanly.
I hope to be able to get read-only mode up this evening European time; I will keep you updated. The nForum will remain up.
PS - This is not just a question of moving the software over. The old/original Instiki will be completely gone once the migration is complete. The nLab should look the same/similar on the surface, but the only traces of the original code will be the CSS.
A quick update that everything is now in place for switching to read-only mode in the cloud: a migration script has been written, an S3 bucket has been set up in the cloud, nginx has been set up to make the redirect on the old machine, etc. I am in the process of copying the rendered content of all pages from the old machine to my local machine, and this is taking a while. Once that’s done, things should go fairly swiftly.
Edit: it’s still copying, and I have to leave off now. Will try to get it into read-only mode in the morning.
Thanks, Richard!!
I see that the first entries are visible now. Hopefully the personal webs will come along, too? :-)
(NB: This crossed with #3).
Touch wood, I think that the nLab is now somewhat functional in read-only mode. There are some caveats:
Point 1. is easily solved, and I will do it as soon as I have a chance. Point 2. is more involved. I have never liked the way it was done client-side in Instiki; I have always felt it should be done server-side, in which case we would not have this problem. Thus I intend to not fix this for the moment, and rather implement it in the course of implementing editing of pages. Point 3. is deliberate; as I gradually implement editing and further functionality, the menus will come back. Point 4. will be solved once the ability to edit is added, but if the security warnings are a major problem for somebody, I can work on it sooner.
There may be other things that are missing or no longer work correctly/as before; let me know if so.
Some further notes:
I have changed the way the context menu works. In fact I have implemented what was requested here and in the thread linked to there. It was implemented by a convoluted CSS hack before which didn’t work properly, and it was easier for me to re-implement it than to migrate the hack.
I also refactored the javascript mostly written by Adeel for theorem numbering, to get rid of the dependency on jQuery. Ideally it should behave exactly the same way as before, i.e. the changes should only be minor ones under the hood, but if you notice anything weird, let me know.
For those interested, the ’nLab’ in its new read-only mode is a static website, served from an S3 bucket. This is the way I intend things to be in the end, too. You can actually access pages directly at the root URL http://nlab-pages.s3.us-east-2.amazonaws.com/nlab/show/. Currently, requests to ǹcatlab.org and similar are still landing at the nginx server on the old machine, which is redirecting to AWS.
The old nLab almost functioned statically after major changes I made a few years ago, but it was still not quite static, so some tweaks were needed to get this read-only mode to work, basically combining the static content of the pages with some surrounding stuff. This is what the migration script that I mentioned does.
To me, the new website seems very snappy, snappier than before, even though the traffic is currently is proceeding via nginx on the old server. This should be the case: AWS should be lightning fast (and essentially never go down).
The next step will be to get editing functioning. This is considerable work, with a number of sub-parts. I have done some of the work before in preparatory work for this migration, but there is still a lot to do. I will keep updating here, but getting editing functioning is likely to be a matter of several days, not several hours. Other functionality will come gradually after that.
Thanks again, Richard, that’s great. We appreciate the work you are investing!
Allow me to come back to #3 since it has a little bit of urgency for me and maybe others, too: Would it be possible to make the personal webs display, soon?
Hi Richard,
maybe before looking into implementing editing, maybe the next step might best be to make the display come out consistently? That would help the broad user experience while the migration is underway, and might be a necessary prerequisite for re-starting editing, anyways.
Regarding redirects, I am afraid that these are used quite abundantly on Lab pages (such as for plurals, but also more generally) so that them being all broken could be perceived as being somewhat of a pain. Maybe implementing the redirects could be made a priority?
Just for the record, allow me to mention other glitches I see in the display:
The literature referencing seems to display incompletely now and the hyperlinks are not working (this might be a general issue with links to page anchors, I can’t check with the source now, but am looking at the output of (infinity,1)-Grothendieck construction).
Also I see that numbering of Def/Prop/etc.-environments seems to only work only partially now (maybe just for those coded in the new syntax) and that their formatting seems to be off generally (there is a big space between the “Proposition”-header and the first line).
No time for more just now, but just to say that this is going to be a painful period where many things do not work. The current read-only mode is temporary, and hacked together to provide some kind of service while the migration takes place, although some things are being implemented along the way which will remain. It is that or turning everything off completely whilst I make the migration.
I will address personal webs etc later.
Guest: which page were you trying to access?
Many thanks for your work on this. I’m more than happy to put up with some bumps along the way to the new servers! I’m not the op from #7, but I get an error like that when trying to access the following:
However, pages from my browser history redirect and load just fine. For example, this one:
EDIT: The first batch of broken links was from me trying to get to the top level page; I now see that it is
http://nlab-pages.s3.us-east-2.amazonaws.com/nlab/show/HomePage
From there, some links in the side bar work and some don’t. Reading some of the discussion above, I guess at least one of the issues is the plural redirects that Urs mentioned. For example,
I’m not complaining; I know these things take time. Just noting it here in case other people are following along too.
Here’s a separate issue I noticed with the sidebars on the pseudofunctor page: With the sidebar completely expanded (both “2-category theory” and “Higher category theory”), the links for the lower one are not clickable. According to the inspect element feature of my browser, they are blocked by some big mathjax box. For example, in the link to categorification/decategorification, the latter half of the second word is clickable, because it hangs outside of the overlaying mathjax box. But the first word, and most others on the list, are not clickable. Closing the “2-category theory” part of the sidebar moves all the “Higher category theory” links up, and then they are clickable.
Update: Actually the part of the “Higher category theory” sidebar hanging below the end of the text on the page is still unclickable, so that seems to be a key part of the issue.
Re #9 and #10: thank you very much, this is helpful! I think that I’ve fixed (in a quick spare moment) the issue with the first four links of #9 now, as long as one accesses them via nginx, i.e. https://ncatlab.org or https://ncatlab.org/nlab/show. I would still recommend to use the standard ncatlab.org/nlab/show/... URLs, though there is no problem (for the moment: I do not guarantee that these URLs will be stable; for afficionados, the HTML code being used for the redirect is 302) with using the direct URLs if one wishes, except that there will be no redirect magic: as you noted, one has to use the exact URL, e.g. to HomePage.
The other access issues are to do with redirects, as you say.
Thank you for the note about the context bar, I will look into that. The whole code around the context bars was a mess, so it does not surprise me that there are still things to clean up :-).
Yes, that is deliberate: as noted above, the current mode is a minimal read-only one. I suggest to use Google (this is in fact what I almost always do, I almost never use the nLab one): just search for nlab + page name, e.g. nlab Kan fibration or whatever. If you are in a country where Google is not available and no other search engine serves an equivalent purpose, let me know.
Re #3 and #5: Personal webs are something which I was intending to bring up. I have for now activated your personal web on the old server, Urs, and can do this for others on request.
In general, I would like to suggest to remove the notion of multiple webs from the software. It adds complexity, and breaks down the modularity of it. It is also possibly questionable whether the funds raised for the nLab should be used for hosting personal webs. With regard to what to do with the old/current personal webs, my suggestions are:
Such are my suggestions. I realise that any change of this kind can be difficult to digest, but have a think about it.
Thanks, Richard! And I see that even the graphics and the files are there on the personal web, that’s a relief. Thanks!
I’d be happy with whatever solution there could be, re point 3.
Just to note that the personal webs here are not meant to provide generic personal webspace, but as that part of the Lab where regular Lab-contributors keep notes that have a clear personal touch to them; mostly personal research notes that go in parallel with edits on the main Lab but want an author name attached.
I believe all webs that were active (except Joyal’s, who never interacted with the Lab) have always been used this way, and that this has been useful for the main Lab, which has plenty of links to personal webs for further discussion.
In this sense I think it would be a loss for the Lab’s main content – and thus for the value that donors get for their money! – if regular contributors would not have a personal web to partially withdraw to.
That said, a workable solution as in your item 3 could be a good compromise, if a compromise is needed.
We can think some more about it, no need to make a decision now.
Just a quick note that I have now fixed point 1. from #4, i.e. images and other files are now working in the cloud.
Thanks! That’s excellent.
And it’s indeed amazing how fast the rendering is now.
The only two remaining display issues I see at the moment is (i) the redirects and (ii) the Def-environment spacing:
E.g. at simplicial classifying space, there is a Definition 2.1 etc. which appears with about 1cm of vertical whitespace after “Definition 1”, were there used to be no whitespace, and maybe not even a new line.
This is particularly apparent for Proof-environments, where this vertical whitespace is clearly undesireable, see e.g. the Proof of Prop. 2.2 at (infinity,1)-Grothendieck construction.
(The other problems at (infinity,1)-Grothendieck construction that I mentioned above might be problems in the source code which maybe I can fix once editing is back.)
There are “personal” webs that are definitely intended as collaborative wikis for multiple people, not just as hosting space for individuals. For instance, the homotopytypetheory web, and the (private) hottmuri web. However, for the long run it would probably be better to integrate the content of the homotopytypetheory web with the main nlab, and we can certainly find alternative solutions for the muri.
Would you mind reactivating the hottmuri web on the old server for now, at least? Please leave it private. Thanks, and thanks also for your efforts in doing this migration!
For my own personal web I would be happy with a static archived version; I haven’t used it for anything for a while and have no plans to use it any more in the future.
Yes, great speed. Science of Logic is the acid test. Thanks!
Could you please reactivate my personal web, Richard?
I agree with Urs in #15 that it would be a shame to lose the facility to write one’s own notes. I wonder how many personal web pages are referenced from pages of the main Lab. We may be generating a load of dead links, such as the one to Todd’s page from surface diagram.
I have for now re-activated all non-nLab webs on the old CMU server. It will be a while before I’m ready to do something with them anyhow. Nevertheless, noted regarding your personal web, Mike, I may in fact use this to test an early iteration of the cloud nLab syntax-to-HTML renderer, which is coming along.
If somebody would like to do some useful work related to the migration, it would be great if you could help out with removing all occurrences of redirects of the same term to two different pages. I began on this, but it is taking too long, as it is manual work; I need to focus on the actual functionality. As of now I will implement it just to look for the first occurrence. The list of redirects that the new cloud renderer will be using can be fetched at the following link.
https://nlab-redirects.s3.us-east-2.amazonaws.com/redirects
It is just a text file, colon-separated. The page name is on the left of the colon, the redirect term on the right. The list of duplicates can easily be obtained on a Unix system by doing something like the following.
awk -F':' '{print $2}' < redirects | sort | uniq -cd
The current list is copied at the end of this post.
What I would like to do is be able to run a script to delete one of the redirects in each case. I.e. the work would be to, for each of these entries, find the two pages in the above file where the redirect is currently used, and manually choose which page should be directed to (you will probably have to visit both nLab pages and see which one fits better, though in some cases it is fairly obvious that one is more specific than the other and should be chosen). For me to be able to run a script, I just need the data in some easily amenable format, e.g. you can provide me with a file (or just paste the contents here) whose lines are of the form
page name:redirect term to be deleted
or whatever. Just be absolutely clear about whether the data in your file refers to those which are to be deleted or those which are to be retained! Please make a comment to say that you will do this if you wish to volunteer, so that we do not have two people doing it, leading to a waste of time for one of them.
2 2-type
2 cocones
2 cofibrant replacement functor
2 cofibrant resolution
2 cofibrant resolutions
2 comonadic functors
2 completely regular space
2 complex analytic manifold
2 concrete objects
2 conformal invariance
2 ∞-connected (∞,1)-topoi
2 ∞-connected (∞,1)-topos
2 ∞-connected (∞,1)-toposes
2 connection on a principal bundle
2 connections on principal bundles
2 ∞-conversion
2 countably locally discrete sets of subsets
2 CPO
2 cpos
2 C-star dynamical systems
2 C-star system
2 C-star systems
2 C-star-systems
2 cup product in differential generalized cohomology
2 cup product in generalized differential cohomology
2 cup product on differential cohomology
2 cup product on differential generalized cohomology
2 cup product on generalized differential cohomology
2 De Morganisation
2 derived dg-categories
2 derived dg-category
2 diagonal map
2 Dieudonné's theorem
2 diffeological groupoids
2 differentiable (∞,1)-categories
2 differentiable (∞,1)-category
2 differentiable (infinity,1)-categories
2 differential 1-form
2 differential 1-forms
2 double groupoids
2 dual Hopf algebra
2 empty 154
2 ∞-equivalence
2 equivalences of infinity-groupoids
2 ∞-equivalent
2 equivariant bordism
2 error correction
2 error corrections
2 étale scheme
2 fibrant replacement functor
2 fibrant resolution
2 fibrant resolutions
2 fibrant types
2 fpqc-topology
2 frame of open subsets
2 frames of open subsets
2 ∞-functor
2 ∞-functors
2 Galois theories
2 gauge field
2 gauge fields
2 generalized Calabi-Yau spaces
2 ∞-group
2 ∞-groups
2 Haag-Kastler net
2 Hamiltonian n-form
2 Hamiltonian n-forms
2 Hedberg's theorem
2 Hermitean structure
2 Hermitean structures
2 Hermitian metric
2 higher homotopy van Kampen Theorem
2 Higher Homotopy van Kampen Theorem
2 history
2 HOL light
2 holomorphic functions
2 homomorphism of frames
2 homomorphisms of frames
2 homotopy 2-types
2 homotopy 3-types
2 homotopy localizations
2 hom-space
2 hom-spaces
2 HQFTs
2 hypothetical judgement
2 hypothetical judgements
2 idempotent complete category
2 idempotent-complete category
2 indiscrete space
2 indiscrete spaces
2 infinity-connected (infinity,1)-topoi
2 infinity-connected (infinity,1)-toposes
2 inner automorphism group
2 inner automorphisms
2 intersection numbers
2 invariant state
2 invertible magmas
2 Jacobi theta-function
2 k-monoidal ∞-group
2 Kuratowski-finite
2 Kuratowski subfinite
2 KU-theory
2 lattice in a vector space
2 lattices in a vector space
2 Lefschetz decompositions
2 Leibniz's law
2 Lie infinity-groupoids
2 local base
2 local bases
2 localization of a model category
2 locally connected sites
2 locally simply-connected space
2 locally simply-connected spaces
2 Lorentz Lie algebras
2 Luis Javier Hernandez
2 Mark Lawson
2 model structure on monoids
2 moderate categories
2 moderate category
2 moduli spaces of flat connections
2 moduli stack of flat connections
2 moduli stacks of flat connections
2 monadic category
2 morphism of frames
2 morphisms of frames
2 motivic spectra
2 multicospan
2 multicospans
2 opposite (∞,1)-category
2 path-connected space
2 pointed homotopy class
2 Poisson 2-algebra
2 Poisson 2-algebras
2 Poisson formula
2 polynomials are continuous functions
2 precompact spaces
2 predicativity
2 pre-n-plectic manifold
2 pre-n-plectic manifolds
2 presheaf categories
2 presheaf category
2 prestack
2 prestacks
2 projectively flat connections
2 projective spaces
2 proof of negation
2 pure morphisms
2 <put page name here>s
2 quotient Hopf algebra
2 ramified extension
2 ramified extensions
2 reduced spaces
2 reduced type
2 reduced types
2 ∞-reduction
2 Regge slope
2 Regge slopes
2 regular functors
2 relative (∞,1)-limits
2 representations up to homotopy
2 representation up to homotopy
2 ∞-rule
2 Selberg zeta function of a Riemann surface
2 self-adjoint
2 semi-algebraic set
2 semi-locally simply-connected space
2 sigma-locally discrete sets of subsets
2 singular chain
2 singular chains
2 skeleta
2 skeletons
2 smooth differential 1-form
2 smooth differential 1-forms
2 smooth homotopy types
2 ∞-space
2 special linear groups
2 sphere bundle
2 sphere bundles
2 Stone algebra
2 Stone algebras
2 strict 2-functors
2 subterminal
2 subterminals
2 subtoposes
2 super-commutative super-algebra
2 super-commutative superalgebra
2 super-commutative super-algebras
2 super-commutative superalgebras
2 super infinity-group
2 super infinity-groups
2 super smooth ∞-groupoid
2 symmetric monoidal functors
2 symplectic orbifolds
2 syntactic sites
2 torsion groups
2 total décalage
2 totally bounded spaces
2 transitive closures
2 tubular neighborhoods
2 tubular neighbourhoods
2 tubular neighbourhood theorem
2 twisted infinity-bundles
2 underlying
2 unit of measurement
2 units of measurement
2 vertex Poisson algebra
2 weak fibration
2 weak fibrations
2 weak multicolimit
2 weight lattice
2 weight lattices
Hi Richard,
could ask your machine to output that list expanded such that each item appears together with the name of the two entries that the duplicate redirect points to?
e.g. if it said
“skeleta” redirects to: (1) “skeleton” and to (2) “simplicial skeleton”
“sphere bundle” redirects to: (1) “spherical fibration” and (2) “sphere fiber bundle”
then I could easily go through the whole list in a few minutes.
(The above two items I have guessed. The correct answer for those two should be (1) and (2), respectively.)
No problem, good idea. I think the following does what you asked for. The redirect term is on the left of the colon, the page on the right.
2-type:homotopy 2-type
2-type:homotopy n-type
cocones:cocone
cocones:cone
cofibrant replacement functor:fibrant replacement
cofibrant replacement functor:resolution
cofibrant resolution:fibrant replacement
cofibrant resolution:resolution
cofibrant resolutions:fibrant replacement
cofibrant resolutions:resolution
comonadic functors:comonadic functor
comonadic functors:monadic functor
completely regular space:regular space
completely regular space:Tychonoff space
complex analytic manifold:analytic manifold
complex analytic manifold:complex manifold
concrete objects:codiscrete object
concrete objects:concrete object
conformal invariance:conformal group
conformal invariance:conformal invariance and scale invariance
∞-connected (∞,1)-topoi:infinity-connected (infinity,1)-topos
∞-connected (∞,1)-topoi:locally n-connected (n+1,1)-topos
∞-connected (∞,1)-topos:infinity-connected (infinity,1)-topos
∞-connected (∞,1)-topos:locally n-connected (n+1,1)-topos
∞-connected (∞,1)-toposes:infinity-connected (infinity,1)-topos
∞-connected (∞,1)-toposes:locally n-connected (n+1,1)-topos
connection on a principal bundle:connection on a bundle
connection on a principal bundle:principal connection
connections on principal bundles:connection on a bundle
connections on principal bundles:principal connection
∞-conversion:beta-reduction
∞-conversion:eta-conversion
countably locally discrete sets of subsets:countably locally discrete set of subsets
countably locally discrete sets of subsets:locally discrete set of subsets
CPO:cpo
CPO:dcpo
cpos:cpo
cpos:dcpo
C-star dynamical systems:C-star dynamical system
C-star dynamical systems:C-star-system
C-star system:C-star dynamical system
C-star system:C-star-system
C-star systems:C-star dynamical system
C-star systems:C-star-system
C-star-systems:C-star dynamical system
C-star-systems:C-star-system
cup product in differential generalized cohomology:Beilinson–Deligne cup product
cup product in differential generalized cohomology:cup product in differential cohomology
cup product in generalized differential cohomology:Beilinson–Deligne cup product
cup product in generalized differential cohomology:cup product in differential cohomology
cup product on differential cohomology:Beilinson–Deligne cup product
cup product on differential cohomology:cup product in differential cohomology
cup product on differential generalized cohomology:Beilinson–Deligne cup product
cup product on differential generalized cohomology:cup product in differential cohomology
cup product on generalized differential cohomology:Beilinson–Deligne cup product
cup product on generalized differential cohomology:cup product in differential cohomology
De Morganisation:De Morganization
De Morganisation:De Morgan topos
derived dg-categories:derived category
derived dg-categories:derived dg-category of a dg-category
derived dg-category:derived category
derived dg-category:derived dg-category of a dg-category
diagonal map:diagonal function
diagonal map:diagonal morphism
Dieudonné's theorem:paracompact Hausdorff spaces are normal
Dieudonné's theorem:paracompact topological space
diffeological groupoids:concrete smooth infinity-groupoid
diffeological groupoids:diffeological groupoid
differentiable (∞,1)-categories:differentiable (infinity,1)-category
differentiable (∞,1)-categories:Goodwillie-differentiable (infinity,1)-category
differentiable (∞,1)-category:differentiable (infinity,1)-category
differentiable (∞,1)-category:Goodwillie-differentiable (infinity,1)-category
differentiable (infinity,1)-categories:differentiable (infinity,1)-category
differentiable (infinity,1)-categories:Goodwillie-differentiable (infinity,1)-category
differential 1-form:cotangent bundle
differential 1-form:differential form
differential 1-forms:cotangent bundle
differential 1-forms:differential form
double groupoids:double category
double groupoids:double groupoid
dual Hopf algebra:dual bialgebra
dual Hopf algebra:dual gebra
empty 154:Attempt to Introduce the Concept of Negative Quantities into Philosophy
empty 154:construction in philosophy
∞-equivalence:beta-reduction
∞-equivalence:eta-conversion
equivalences of infinity-groupoids:equivalence of categories
equivalences of infinity-groupoids:equivalence of infinity-groupoids
∞-equivalent:beta-reduction
∞-equivalent:eta-conversion
equivariant bordism:equivariant bordism homology theory
equivariant bordism:equivariant cobordism theory
error correction:error correcting code
error correction:error correction code
error corrections:error correcting code
error corrections:error correction code
étale scheme:Demazure, lectures on p-divisible groups, I.8, constant- and étale schemes
étale scheme:etale scheme
fibrant replacement functor:fibrant replacement
fibrant replacement functor:resolution
fibrant resolution:fibrant replacement
fibrant resolution:resolution
fibrant resolutions:fibrant replacement
fibrant resolutions:resolution
fibrant types:fibrant type
fibrant types:two-level type theory
fpqc-topology:fpqc site
fpqc-topology:fpqc topology
frame of open subsets:category of open subsets
frame of open subsets:frame of opens
frames of open subsets:category of open subsets
frames of open subsets:frame of opens
∞-functor:delta-functor
∞-functor:n-functor
∞-functors:delta-functor
∞-functors:n-functor
Galois theories:Galois Theories
Galois theories:Galois theory
gauge field:field (physics)
gauge field:gauge theory
gauge fields:field (physics)
gauge fields:gauge theory
generalized Calabi-Yau spaces:generalized Calabi-Yau manifold
generalized Calabi-Yau spaces:generalized Calabi-Yau space
∞-group:infinity-group
∞-group:Omega-group
∞-groups:infinity-group
∞-groups:Omega-group
Haag-Kastler net:causally local net of observables
Haag-Kastler net:Haag-Kastler axioms
Hamiltonian n-form:Hamiltonian form
Hamiltonian n-form:n-plectic geometry
Hamiltonian n-forms:Hamiltonian form
Hamiltonian n-forms:n-plectic geometry
Hedberg's theorem:decidable equality
Hedberg's theorem:h-set
Hermitean structure:almost Hermitian structure
Hermitean structure:Hermitian manifold
Hermitean structures:almost Hermitian structure
Hermitean structures:Hermitian manifold
Hermitian metric:Hermitian form
Hermitian metric:Hermitian manifold
higher homotopy van Kampen Theorem:higher homotopy van Kampen theorem
higher homotopy van Kampen Theorem:higher van Kampen theorem
Higher Homotopy van Kampen Theorem:higher homotopy van Kampen theorem
Higher Homotopy van Kampen Theorem:higher van Kampen theorem
history:history of mathematics
history:monadic > history
HOL light:HOL
HOL light:HOL Light
holomorphic functions:Dolbeault complex
holomorphic functions:holomorphic function
homomorphism of frames:frame
homomorphism of frames:locale
homomorphisms of frames:frame
homomorphisms of frames:locale
homotopy 2-types:homotopy 2-type
homotopy 2-types:homotopy n-type
homotopy 3-types:homotopy 3-type
homotopy 3-types:homotopy n-type
homotopy localizations:cohomology localization
homotopy localizations:localization at geometric homotopies
hom-space:hom-object
hom-space:(infinity,1)-categorical hom-space
hom-spaces:hom-object
hom-spaces:(infinity,1)-categorical hom-space
HQFTs:homological quantum field theory
HQFTs:HQFT
hypothetical judgement:judgment
hypothetical judgement:sequent
hypothetical judgements:judgment
hypothetical judgements:sequent
idempotent complete category:Cauchy complete category
idempotent complete category:Karoubi envelope
idempotent-complete category:Cauchy complete category
idempotent-complete category:Karoubi envelope
indiscrete space:codiscrete space
indiscrete space:discrete and indiscrete topology
indiscrete spaces:codiscrete space
indiscrete spaces:discrete and indiscrete topology
infinity-connected (infinity,1)-topoi:infinity-connected (infinity,1)-topos
infinity-connected (infinity,1)-topoi:locally n-connected (n+1,1)-topos
infinity-connected (infinity,1)-toposes:infinity-connected (infinity,1)-topos
infinity-connected (infinity,1)-toposes:locally n-connected (n+1,1)-topos
inner automorphism group:inner automorphism
inner automorphism group:inner autormorphism
inner automorphisms:inner automorphism
inner automorphisms:inner autormorphism
intersection numbers:intersection number
intersection numbers:intersection pairing
invariant state:C-star dynamical system
invariant state:C-star-system
invertible magmas:invertible magma
invertible magmas:invertible quasigroup
Jacobi theta-function:Jacobi form
Jacobi theta-function:Jacobi theta function
k-monoidal ∞-group:abelian infinity-group
k-monoidal ∞-group:k-monoidal infinity-group
Kuratowski-finite:finite object
Kuratowski-finite:finite set
Kuratowski subfinite:finite object
Kuratowski subfinite:finite set
KU-theory:K-theory spectrum
KU-theory:topological K-theory
lattice in a vector space:lattice (discrete subgroup)
lattice in a vector space:lattice (in a vector space, etc.)
lattices in a vector space:lattice (discrete subgroup)
lattices in a vector space:lattice (in a vector space, etc.)
Lefschetz decompositions:hard Lefschetz theorem
Lefschetz decompositions:Lefschetz decomposition
Leibniz's law:identity of indiscernibles
Leibniz's law:Leibniz algebra
Lie infinity-groupoids:Lie n-groupoid
Lie infinity-groupoids:smooth infinity-groupoid
local base:neighborhood
local base:topological base
local bases:neighborhood
local bases:topological base
localization of a model category:localization of a model category > history
localization of a model category:localization of model categories
locally connected sites:locally connected site
locally connected sites:strongly connected site
locally simply-connected space:semi-locally simply connected topological space
locally simply-connected space:semi-locally simply-connected topological space
locally simply-connected spaces:semi-locally simply connected topological space
locally simply-connected spaces:semi-locally simply-connected topological space
Lorentz Lie algebras:Lorentz Lie algebra
Lorentz Lie algebras:orthogonal Lie algebra
Luis Javier Hernandez:Luis Javier Hernandez > history
Luis Javier Hernandez:Luis Javier Hernández Paricio
Mark Lawson:Mark Lawson > history
Mark Lawson:Mark V. Lawson
model structure on monoids:model structure on monoids in a monoidal model category
model structure on monoids:monoid axiom in a monoidal model category
moderate categories:large category
moderate categories:small category
moderate category:large category
moderate category:small category
moduli spaces of flat connections:moduli space of connections
moduli spaces of flat connections:moduli space of flat connections
moduli stack of flat connections:moduli space of connections
moduli stack of flat connections:moduli space of flat connections
moduli stacks of flat connections:moduli space of connections
moduli stacks of flat connections:moduli space of flat connections
monadic category:algebraic category
monadic category:monadic functor
morphism of frames:frame
morphism of frames:locale
morphisms of frames:frame
morphisms of frames:locale
motivic spectra:motivic homotopy theory
motivic spectra:motivic spectrum
multicospan:multi-cospan
multicospan:multispan
multicospans:multi-cospan
multicospans:multispan
opposite (∞,1)-category:opposite (infinity,1)-category
opposite (∞,1)-category:opposite quasi-category
path-connected space:2009 June changes
path-connected space:connected space
pointed homotopy class:homotopy class
pointed homotopy class:pointed homotopy type
Poisson 2-algebra:Gerstenhaber algebra
Poisson 2-algebra:Poisson n-algebra
Poisson 2-algebras:Gerstenhaber algebra
Poisson 2-algebras:Poisson n-algebra
Poisson formula:group of ideles
Poisson formula:Poisson summation formula
polynomials are continuous functions:polynomials are continuous
polynomials are continuous functions:rational functions are continuous
precompact spaces:precompact space
precompact spaces:totally bounded space
predicativity:predicative mathematics
predicativity:predicative topos
pre-n-plectic manifold:n-plectic form
pre-n-plectic manifold:n-plectic geometry
pre-n-plectic manifolds:n-plectic form
pre-n-plectic manifolds:n-plectic geometry
presheaf categories:category of presheaves
presheaf categories:presheaf
presheaf category:category of presheaves
presheaf category:presheaf
prestack:(2,1)-presheaf
prestack:separated (2,1)-presheaf
prestacks:(2,1)-presheaf
prestacks:separated (2,1)-presheaf
projectively flat connections:flat connection
projectively flat connections:projectively flat connection
projective spaces:projective space
projective spaces:projective stack
proof of negation:proof by contradiction
proof of negation:refutation by contradiction
pure morphisms:pure morphism
pure morphisms:pure subobject
<put page name here>s:split inclusion of von Neumann algebras
<put page name here>s:Template page
quotient Hopf algebra:biideal
quotient Hopf algebra:quotient bialgebra
ramified extension:branched cover
ramified extension:ramification of ideals
ramified extensions:branched cover
ramified extensions:ramification of ideals
reduced spaces:reduced object
reduced spaces:reduced space
reduced type:reduced object
reduced type:reduced space
reduced types:reduced object
reduced types:reduced space
∞-reduction:beta-reduction
∞-reduction:eta-conversion
Regge slope:Regge trajectory
Regge slope:string length scale
Regge slopes:Regge trajectory
Regge slopes:string length scale
regular functors:coherent functor
regular functors:regular functor
relative (∞,1)-limits:relative (∞,1)-limit
relative (∞,1)-limits:relative (infinity,1)-limit
representations up to homotopy:infinity-representation
representations up to homotopy:Lie infinity-algebroid representation
representation up to homotopy:infinity-representation
representation up to homotopy:Lie infinity-algebroid representation
∞-rule:beta-reduction
∞-rule:eta-conversion
Selberg zeta function of a Riemann surface:Selberg zeta function
Selberg zeta function of a Riemann surface:zeta function of a Riemann surface
self-adjoint:self-adjoint morphism
self-adjoint:self-adjoint operator
semi-algebraic set:semialgebraic manifold
semi-algebraic set:semialgebraic set
semi-locally simply-connected space:semi-locally simply connected topological space
semi-locally simply-connected space:semi-locally simply-connected topological space
sigma-locally discrete sets of subsets:countably locally discrete set of subsets
sigma-locally discrete sets of subsets:locally discrete set of subsets
singular chain:singular cohomology
singular chain:singular homology
singular chains:singular cohomology
singular chains:singular homology
skeleta:simplicial skeleton
skeleta:skeleton
skeletons:simplicial skeleton
skeletons:skeleton
smooth differential 1-form:cotangent bundle
smooth differential 1-form:differential form
smooth differential 1-forms:cotangent bundle
smooth differential 1-forms:differential form
smooth homotopy types:smooth homotopy type
smooth homotopy types:smooth infinity-groupoid
∞-space:Gamma-space
∞-space:infinity-space
special linear groups:projective special linear group
special linear groups:special linear group
sphere bundle:sphere fiber bundle
sphere bundle:spherical fibration
sphere bundles:sphere fiber bundle
sphere bundles:spherical fibration
Stone algebra:De Morgan Heyting algebra
Stone algebra:Stone duality
Stone algebras:De Morgan Heyting algebra
Stone algebras:Stone duality
strict 2-functors:2-functor
strict 2-functors:strict 2-functor
subterminal:subsingleton
subterminal:subterminal object
subterminals:subsingleton
subterminals:subterminal object
subtoposes:geometric embedding
subtoposes:subtopos
super-commutative super-algebra:super algebra
super-commutative super-algebra:supercommutative algebra
super-commutative superalgebra:super algebra
super-commutative superalgebra:supercommutative algebra
super-commutative super-algebras:super algebra
super-commutative super-algebras:supercommutative algebra
super-commutative superalgebras:super algebra
super-commutative superalgebras:supercommutative algebra
super infinity-group:super formal smooth infinity-groupoid
super infinity-group:super infinity-groupoid
super infinity-groups:super formal smooth infinity-groupoid
super infinity-groups:super infinity-groupoid
super smooth ∞-groupoid:smooth super infinity-groupoid
super smooth ∞-groupoid:super formal smooth infinity-groupoid
symmetric monoidal functors:braided monoidal functor
symmetric monoidal functors:symmetric monoidal functor
symplectic orbifolds:symplectic manifold
symplectic orbifolds:symplectic orbifold
syntactic sites:syntactic category
syntactic sites:syntactic site
torsion groups:torsion group
torsion groups:torsion subgroup
total décalage:bisimplicial set
total décalage:decalage
totally bounded spaces:precompact space
totally bounded spaces:totally bounded space
transitive closures:transitive closure
transitive closures:transitive set
tubular neighborhoods:tubular neighborhood
tubular neighborhoods:tubular neighbourhood
tubular neighbourhoods:tubular neighborhood
tubular neighbourhoods:tubular neighbourhood
tubular neighbourhood theorem:tubular neighborhood theorem
tubular neighbourhood theorem:tubular neighborhood
twisted infinity-bundles:twisted ∞-bundles
twisted infinity-bundles:twisted infinity-bundle
underlying:forgetful functor
underlying:underlying set
unit of measurement:physical unit
unit of measurement:unit
units of measurement:physical unit
units of measurement:unit
vertex Poisson algebra:Coisson algebra
vertex Poisson algebra:Poisson vertex algebra
weak fibration:fibrations of quasi-categories
weak fibration:Street fibration
weak fibrations:fibrations of quasi-categories
weak fibrations:Street fibration
weak multicolimit:weak colimit
weak multicolimit:weak multilimit
weight lattice:root (in representation theory)
weight lattice:weight (in representation theory)
weight lattices:root (in representation theory)
weight lattices:weight (in representation theory)
Thanks! Will look into it tomorrow.
How to best return the result to you? Maybe like this:
2-type:homotopy 2-type
2-type:homotopy n-type
(1)
cocones:cocone
cocones:cone
(1)
cofibrant replacement functor:fibrant replacement
cofibrant replacement functor:resolution
(1)
cofibrant resolution:fibrant replacement
cofibrant resolution:resolution
(1)
cofibrant resolutions:fibrant replacement
cofibrant resolutions:resolution
(1)
(Hm, maybe if we just take the first one by default, we’ll have a good enough match without further struggle…)
I got an error like the one in #7 on http://nlab-pages.s3.us-east-2.amazonaws.com/nlab/show/lifting+property after following a link on http://nlab-pages.s3.us-east-2.amazonaws.com/nlab/show/fibration . The correct result does not appear when searching “nlab lifting property” on google. Other pages have been working fine. Not sure if this is helpful info but anyway there it is. Good luck with the migration and thanks for keeping the site up!
I wonder if we could put a banner at the top of all nLab pages notifying people about the migration in progress? Many people must be wondering about the missing functionality, missing webs, etc.
Re #28:
How to best return the result to you?
I think if you just either remove the lines that should not be there, or retain only the lines which should not be there, that would work well.
(Hm, maybe if we just take the first one by default, we’ll have a good enough match without further struggle…)
I don’t think so, it seemed a bit random for the ones I have already done.
Re #29: Thank you for getting in touch, yes, this is a ’redirect issue’. It will hopefully be solved fairly soon; I wish to handle these redirects properly server-side rather than write a client-side hack, so before I can fix it I need to have quite a bit of other infrastructure live. For some pages the fix should come quite soon, possibly this evening European time.
Re #30: I considered this, but decided against it for a few reasons that I’ll not go into now. I think these days people can probably find their way to this page or some other source which indicates what is going on; most people are only using the nLab in a way for which the current read-only mode should be satisfactory, I think. The main priority is to fix the redirects and certain relative links, and I’d rather just complete this asap. As mentioned, for a first group of pages this should come quite soon.
HI Richard,
okay, here the first bit of the list of redirects which I think should be there (having removed the ones which, I think, should not):
2-type:homotopy 2-type
cocones:cocone
cofibrant replacement functor:fibrant replacement
cofibrant resolution:fibrant replacement
cofibrant resolutions:fibrant replacement
comonadic functors:comonadic functor
completely regular space:Tychonoff space
complex analytic manifold:analytic manifold
concrete objects:concrete object
conformal invariance:conformal invariance and scale invariance
∞-connected (∞,1)-topoi:infinity-connected (infinity,1)-topos
∞-connected (∞,1)-topos:infinity-connected (infinity,1)-topos
∞-connected (∞,1)-toposes:infinity-connected (infinity,1)-topos
connection on a principal bundle:principal connection
connections on principal bundles:principal connection
∞-conversion:eta-conversion
countably locally discrete sets of subsets:countably locally discrete set of subsets
CPO:cpo
cpos:cpo
C-star dynamical systems:C-star dynamical system
C-star system:C-star-system
C-star systems:C-star-system
C-star-systems:C-star-system
cup product in differential generalized cohomology:cup product in differential cohomology
cup product in generalized differential cohomology:cup product in differential cohomology
cup product on differential cohomology:Beilinson–Deligne cup product
cup product on differential generalized cohomology:cup product in differential cohomology
cup product on generalized differential cohomology:cup product in differential cohomology
De Morganisation:De Morganization
derived dg-categories:derived dg-category of a dg-category
derived dg-category:derived dg-category of a dg-category
diagonal map:diagonal function
Dieudonné's theorem:paracompact Hausdorff spaces are normal
diffeological groupoids:diffeological groupoid
differentiable (∞,1)-categories:differentiable (infinity,1)-category
differentiable (∞,1)-categories:Goodwillie-differentiable (infinity,1)-category
The last item has an issue that needs attention: both differentiable (infinity,1)-category and Goodwillie-differentiable (infinity,1)-category seem to exist as separate pages, but have the same content. One of them should probably be removed and made a redirect to the other. Maybe the second one should be kept, as its more specific.
More in the following comments…
Continuing the list of redirects that should be in place:
differential 1-form:cotangent bundle
differential 1-forms:cotangent bundle
double groupoids:double groupoid
dual Hopf algebra:dual bialgebra
The following has an issue: Neither of these two redirects should exist:
empty 154:Attempt to Introduce the Concept of Negative Quantities into Philosophy
empty 154:construction in philosophy
to be continued…
This item I don’t know about without looking things up, some logician needs to check:
∞-equivalence:beta-reduction
∞-equivalence:eta-conversion
Continuing the list of redirects that should be there:
equivalences of infinity-groupoids:equivalence of infinity-groupoids
equivariant bordism:equivariant bordism homology theory
This item has an issue that needs attention:
error correction:error correcting code
error correction:error correction code
The stub entry error correction code should be removed and made a redirect to error correcting code
Continuing the list of redirects that should be there:
étale scheme:etale scheme
fibrant replacement functor:fibrant replacement
fibrant resolution:fibrant replacement
fibrant resolutions:fibrant replacement
fibrant types:fibrant type
fpqc-topology:fpqc topology
frame of open subsets:frame of opens
frames of open subsets:frame of opens
∞-functor:n-functor
∞-functors:n-functor
Galois theories:Galois theory
gauge field:gauge theory
gauge fields:gauge theory
The following item has an issue that needs extra attention:
generalized Calabi-Yau spaces:generalized Calabi-Yau manifold
generalized Calabi-Yau spaces:generalized Calabi-Yau space
The entry generalized Calabi-Yau space (which contains nothing but one reference item) needs to be removed and its title made a redirect to generalized Calabi-Yau space.
Continuing the list of redirects that should be there:
∞-group:infinity-group
∞-groups:infinity-group
Haag-Kastler net:causally local net of observables
Hamiltonian n-form:n-plectic geometry
Hamiltonian n-forms:n-plectic geometry
Hedberg's theorem:h-set
Hermitean structure:Hermitian manifold
Hermitean structures:Hermitian manifold
Hermitian metric:Hermitian manifold
The following item has an issue that needs extra attention:
higher homotopy van Kampen Theorem:higher homotopy van Kampen theorem
higher homotopy van Kampen Theorem:higher van Kampen theorem
Higher Homotopy van Kampen Theorem:higher homotopy van Kampen theorem
Higher Homotopy van Kampen Theorem:higher van Kampen theorem
The entries higher homotopy van Kampen theorem and higher van Kampen theorem look like they should be merged, with one title redirecting to the other. Probably “higher van Kampen theorem” is a good enough title.
Continuing the list of redirects that should be there:
history:history of mathematics
HOL light:HOL Light
holomorphic functions:holomorphic function
homomorphism of frames:frame
homomorphisms of frames:frame
homotopy 2-types:homotopy 2-type
homotopy 3-types:homotopy 3-type
homotopy localizations:localization at geometric homotopies
hom-space:(infinity,1)-categorical hom-space
hom-spaces:(infinity,1)-categorical hom-space
The following item has an issue that needs extra attention:
HQFTs:homological quantum field theory
HQFTs:HQFT
The stub entry homological quantum field theory should be merged into the more comprehensive HQFT and its title made a redirect.
Continuing the list of redirects that should be there:
hypothetical judgement:sequent
hypothetical judgements:sequent
idempotent complete category:Cauchy complete category
idempotent-complete category:Cauchy complete category
indiscrete space:discrete and indiscrete topology
indiscrete spaces:discrete and indiscrete topology
infinity-connected (infinity,1)-topoi:infinity-connected (infinity,1)-topos
infinity-connected (infinity,1)-toposes:infinity-connected (infinity,1)-topos
The following item has an issue that needs extra attention:
inner automorphism group:inner automorphism
inner automorphism group:inner autormorphism
inner automorphisms:inner automorphism
inner automorphisms:inner autormorphism
The stub entry with the misspelled title inner autormorphism needs to be deleted for good.
Continuing the list of redirects that should be there:
intersection numbers:intersection number
invariant state:C-star dynamical system
invertible magmas:invertible magma
Jacobi theta-function:Jacobi theta function
k-monoidal ∞-group:k-monoidal infinity-group
Kuratowski-finite:finite set
Kuratowski subfinite:finite set
KU-theory:topological K-theory
lattice in a vector space:lattice (in a vector space, etc.)
lattices in a vector space:lattice (in a vector space, etc.)
Lefschetz decompositions:Lefschetz decomposition
Leibniz's law:identity of indiscernibles
Lie infinity-groupoids:Lie n-groupoid
local base:topological base
local bases:topological base
localization of a model category:localization of model categories
locally connected sites:locally connected site
The following item needs special attention:
locally simply-connected space:semi-locally simply connected topological space
locally simply-connected space:semi-locally simply-connected topological space
locally simply-connected spaces:semi-locally simply connected topological space
locally simply-connected spaces:semi-locally simply-connected topological space
These are duplicate entries, whose titles differ only by a hyphen. They need to be merged, and one title redirected to the other.
Continuing the list of redirects that should be there:
Lorentz Lie algebras:Lorentz Lie algebra
Luis Javier Hernandez:Luis Javier Hernández Paricio
Mark Lawson:Mark V. Lawson
model structure on monoids:model structure on monoids in a monoidal model category
moderate categories:large category
moderate category:large category
moduli spaces of flat connections:moduli space of flat connections
moduli stack of flat connections:moduli space of flat connections
moduli stacks of flat connections:moduli space of flat connections
monadic category:monadic functor
morphism of frames:frame
morphisms of frames:frame
motivic spectra:motivic spectrum
multicospan:multi-cospan
multicospans:multi-cospan
opposite (∞,1)-category:opposite (infinity,1)-category
path-connected space:connected space
pointed homotopy class:homotopy class
Poisson 2-algebra:Gerstenhaber algebra
Poisson 2-algebras:Gerstenhaber algebra
Poisson formula:Poisson summation formula
polynomials are continuous functions:polynomials are continuous
precompact spaces:precompact space
predicativity:predicative mathematics
pre-n-plectic manifold:n-plectic form
pre-n-plectic manifolds:n-plectic form
presheaf categories:category of presheaves
presheaf category:category of presheaves
prestack:(2,1)-presheaf
prestacks:(2,1)-presheaf
projectively flat connections:projectively flat connection
projective spaces:projective space
proof of negation:refutation by contradiction
pure morphisms:pure morphism
The following item needs extra attenion:
<put page name here>s:split inclusion of von Neumann algebras
<put page name here>s:Template page
This is some glitch with the code, should be removed in the source.
Continuing the list of redirects that should be there:
quotient Hopf algebra:quotient bialgebra
ramified extension:ramification of ideals
ramified extensions:ramification of ideals
reduced spaces:reduced space
reduced type:reduced object
reduced types:reduced object
Regge slope:Regge trajectory
Regge slopes:Regge trajectory
regular functors:regular functor
The following item needs extra attention:
relative (∞,1)-limits:relative (∞,1)-limit
relative (∞,1)-limits:relative (infinity,1)-limit
These are duplicate entries that need to be merged, with one title redirecting to the other.
Continuing with the list of redirects that should be there:
representations up to homotopy:infinity-representation
representation up to homotopy:infinity-representation
Selberg zeta function of a Riemann surface:zeta function of a Riemann surface
self-adjoint:self-adjoint operator
semi-algebraic set:semialgebraic set
sigma-locally discrete sets of subsets:countably locally discrete set of subsets
singular chain:singular homology
singular chains:singular homology
skeleta:skeleton
skeletons:skeleton
smooth differential 1-form:cotangent bundle
smooth differential 1-forms:cotangent bundle
smooth homotopy types:smooth homotopy type
∞-space:infinity-space
special linear groups:special linear group
sphere bundle:sphere fiber bundle
sphere bundles:sphere fiber bundle
Stone algebra:Stone duality
Stone algebras:Stone duality
strict 2-functors:strict 2-functor
subterminal:subterminal object
subterminals:subterminal object
subtoposes:subtopos
Continuing the list of redirects that should be there:
super-commutative super-algebra:supercommutative algebra
super-commutative superalgebra:supercommutative algebra
super-commutative super-algebras:supercommutative algebra
super-commutative superalgebras:supercommutative algebra
super infinity-group:super infinity-groupoid
super infinity-groups:super infinity-groupoid
super smooth ∞-groupoid:smooth super infinity-groupoid
symmetric monoidal functors:symmetric monoidal functor
symplectic orbifolds:symplectic orbifold
syntactic sites:syntactic site
torsion groups:torsion group
total décalage:decalage
totally bounded spaces:totally bounded space
transitive closures:transitive closure
The following item needs extra attention:
tubular neighborhoods:tubular neighborhood
tubular neighborhoods:tubular neighbourhood
tubular neighbourhoods:tubular neighborhood
tubular neighbourhoods:tubular neighbourhood
Here tubular neighborhood and tubular neighbourhood is a pair of duplicate entries that need to be merged, and one title made a redirect to the other.
Continuing the list of redirects that should be there:
tubular neighbourhood theorem:tubular neighborhood theorem
The following item needs extra attenion:
twisted infinity-bundles:twisted ∞-bundles
twisted infinity-bundles:twisted infinity-bundle
This is a pair of duplicate entries that need to be merged. The one with the plural title should go and be made a redirect.
Continuing the list of redirects that should be there:
underlying:forgetful functor
unit of measurement:physical unit
units of measurement:physical unit
vertex Poisson algebra:Poisson vertex algebra
weak fibration:Street fibration
weak fibrations:Street fibration
weak multicolimit:weak multilimit
weight lattice:weight (in representation theory)
weight lattices:weight (in representation theory)
That’s all (according to #26).
Not to put any pressure on you at all, Richard, but could you give us some idea of when editing functionality on the Lab will be in place?
Re #47 etc: Thank you Urs! I will not be carrying out any of the manual page merges etc at this point, I will just link to one or the other of the pages, but we can come back to this once the migration is further advanced. Very good to clear this up!
Re #48: It is difficult to say, currently I am usually able to manage a couple of hours of work on it in the evening (or night, depending on one’s definition of evening). Progress is going well, we will end up I believe with much better (simpler, more maintainable, more performant, etc) software and infrastructure. The major thing is not editing itself, but rather generation of HTML from the nLab syntax (i.e. enhanced Markdown), i.e. rendering. I have come a long way with this, and in particular two of the major things are now more or less done (internal page links and parsing of theorem environments). I expect to be able to start generating pages using this new cloud renderer soon; exactly when is dependent upon how much time I find, but I think within a few days. Adding the ability to actually edit will come after that; it should be possible via API (everything will be possible via API in the new software, i.e. without a graphical user interface) quite quickly, and a graphical user interface will come fairly soon after that depending on what I prioritise within the migration as a whole.
By the way, I have been in touch with Brendan, and the costs so far look very promising. I am implementing a very modern, ’serverless’ architecture (there will be no permanent server in the traditional sense, rather we will have a static website + some ’lambda functions’ which fire on editing and for other functionality), and a consequence of this will be that the costs should mainly come from the amount of data we are transferring when people view pages; thus I expect the costs to be a bit more than since I started the migration a few days ago, but not very much more, since all the pages are viewable now, and will actually be significantly smaller in size once the new cloud renderer is in place. That the costs are completely fine so far is thus very encouraging.
Thanks for the update, and of course all the work you’re putting in it!
It’s like watching the James Webb telescope going through its deployment sequence.
Hi Richard,
sure, I can do the page mergers etc. once editing is back.
By the way, will we be able, eventually, to keep our usual page URL-s without all the AWS code in them?
I see that Google searches for our pages are beginning to produce results in the form (for example)
nLab D-brane – Amazon AWS
This would seem to be an undue amount of reference to Jeff’s little business, if it stayed that way.
It’s like watching the James Webb telescope going through its deployment sequence.
Hehe!
By the way, will we be able, eventually, to keep our usual page URL-s without all the AWS code in them?
Yes, the old URLs still work, and I recommend that people keep using the old URLs in their browser if they have the choice. Google should quickly re-adapt itself to the old URLs once we adjust the DNS to point to AWS directly, rather than via a redirect from the old server as now.
re #52 (Richard, probably):
Thanks. But I am wondering if we can stop the expanded AWS-URL-s from appearing? Maybe not a big deal, just wondering.
re #48 (David):
Just to highlight that editing on the personal webs still works! So if you feel like making notes that should eventually go into an Lab page, you can still write them into your personal Sandbox, and copy that over later.
We cannot stop the AWS URLs from appearing at the moment, but we will able to later in the process. Actually I think that Google should not have changed their links on the basis of a 302, but it could be due to the fact the page links on the nLab are relative, which means that when expanded they are the AWS URLs, not the original ones, so Google’s algorithms may detect a lot of links of this form to a certain page and decide that they must be correct. I am changing the links to be absolute with the new renderer (it makes no difference speed-wise, at least not when the DNS points to the same URL as the relative one).
Hi Leon, thank you very much for the kind offer and for getting in touch! Actually something that would be very useful would be if you could fetch the redirects file, which I’ll call F, that can be obtained at the link I gave in #24, and then go through Urs’ work in the comments following it, editing F to ensure that no term in F redirects to more than one page. Where Urs mentions that manual work is required, just pick one of the two possibilities to link to as you see fit; feel free to ask here if in doubt.
I went through Urs’s comments and made the edits in the ’redirects’ file. I didn’t do anything to any of the ones that Urs said that manual work is required. The list of duplicates is much smaller now. https://raw.githubusercontent.com/leon2k2k2k/nlab-redirect/main/redirects
Thank you very much Leon, I will make use of that soon! There will be a lot more work to do very soon. At the moment I’d like to do the main API programming work myself as I think a singularity of vision will be helpful, but there will no doubt be a lot of scripted work that will need doing to make things compatible, etc, where your help would be very useful.
As a quick update to everyone, I now have a renderer with decent functionality (not everything yet, but enough to render many pages, including fixing redirects on those pages) working locally. I need to test it a little more, but it will not be long now before we will start to see page generation, and after that the possibility of editing, using it.
Sounds great.
Just a thought: Might it be easily possible to make one auxiliary nLab page (e.g the Sandbox) be editable for the meantime? That would allow us to keep making notes and copy&paste them over later.
I have been doing this in the Sandbox of my personal web, but it would be better to be in the main Lab web, to see that links work.
Urs 50
You mean crosslinks with the Lab ? I suppose if we make links to external stuff the syntax from external web is the same and for the Lab links work now and we need to just erase nlab: from within the brackets when migrating paragraphs, so I am not sure what else scenario may bring problems. Or there is something else in mind ? (I am asking as I was planning to do few things the same way in coming days and I want to anticipate possible problems).
Just a quick note that the page outer automorphism has been rendered with the new renderer. It may not seem like much, but most of the hard work in getting a lot of things to work is now completed by the fact that that works. There are numerous major changes to the way the rendering was done before, which I’ll summarise at a later point rather than go into now; for now, please just check whether things look/work satisfactorily. Redirects should now work, for instance.
The next step will involve moving a large batch of pages to being rendered by the new renderer, and allowing editing of these via API. I have not been feeling too well for the past few days and am fitting this work into very late hours, but I don’t expect this next step to take very long.
(If you happen to have viewed this page whilst I was working on it, the Javascript files it relies on may have been cached, and if for example redirects do not work, this is the likely cause; let me know if so, and I can let you know how to force an update).
Thanks, looks good!
Let’s see, the pointer at outer automorphism to R-symmetry group does not redirect to R-symmetry at the moment.
(Maybe it’s my Javascript, as you say, but I think that I have not been on this page for ages.)
Hmm, that redirect works for me currently (there are a couple of others on the page too: ’0-truncated’, and ’inner automorphisms’, maybe more), but I did need to refresh the page for some reason first. That shouldn’t be necessary; let’s keep an eye on it and see if it stabilises.
I should say that trying to use redirects on the nForum will still not work. Eventually the nForum will be updated/replaced, but not until the migration of the nLab itself is completed.
Regarding the redirects:
I have tried again from the entry, refreshing first, but it doesn’t work for me. I also tried links at random other entries, for instance to “Drinfel’d” at chiral algebra, but it doesn’t work for me (XML error “AccessDenied”).
(I am on Win+Firefox, if that matters.)
Only outer automorphism has been rendered using the new renderer so far, so redirects will still not work on other pages. I’ll look into why it’s not working for you; if anyone else reading this can test as well and see if it works for them, that would be appreciated!
Edit: I think this may be a Firefox-specific issue, which I’ll address later. If it doesn’t work for somebody in a different browser, please let me know.
The “outer automorphism” page is misbehaving.
When I CTRL-click links, my expectation is that nothing happens in the current tab and a new tab is opened with the target of the link.
What is actually happening (e.g. by CTRL-clicking on the “group” link in the “outer automorphism” nLab page) is that both the current tab navigates to the target of the link and I get a new tab with the target of the link.
SHIFT-click (which should have the behavior ’open link in new window’) has the same problem.
I suppose this is just the new renderer; while the “outer automorphism” page is misbehaving in this fashion, the “group” page is not.
I’m running Microsoft Edge, version 97.0.1072.55 (Official build) (64-bit) on Windows 10.
Just to add that I checked the page on my phone now (Android+Firefox) and there the redirecting works!
I checked with the brave browser and the first one (R-symmetry group) failed, but then succeeded after a full refresh. But then I tried a different redirect link (0-truncation) and that one failed, but then worked again after a refresh. I also have the same unexpected behaviour reported in #66 for Ctrl-click and Ctrl-Shift-click. Weirdly, after further testing, a Ctrl-click on a redirect link (R-symmetry group) failed in the new tab, but succeeded in the original tab (which should not have changed).
Hyperlinks are completely broken for outer automorphism in the Firefox browser. Simple click, control+click, and shift+click do not work at all. The problem appears to be caused by the embedded JavaScript. Why is it necessary to modify the standard behavior of clicks in the first place?
Regarding #69, I have been experimenting with different things for the last couple of hours, and broke things at various points. I do have something which works to some extent now with regard to Ctrl+Click and Shift+Click, etc, but I am not happy with it. I wish to change the behaviour of clicks for important reasons that I do not have time to go into just now, but will explain at a later point if this can be got to work.
(Edit 13th of January: I am still working on this page, and at any given time until further notice things may not work on this page.)
I have now implemented redirect functionality across the whole of the (new) nLab, in a way which also works on the nForum, I believe. Quite a simple thing in the end, my apologies that it has taken so long.
Let me know if there any issues.
Fantastic! Thanks.
With that out of the way, and hoping not to get on your nerves,allow me to raise again the issue of the Definition/Theorem/Proof-environments: They have, since migration, ca 1cm of superfluous whitespace between their headline und their bulk text, and it looks weird. To my mind, this is the last issue now in between us and a perfect display functionality of the pages!
Hi Richard,
I just see one subtle problem in a redirect: In trying to look at look at
I am being redirected to
But the latter is (beyond its grammatically broken title) a puny stub, while the former is (or was) a well-developed entry.
Can we recover this?
Nevermind – just saw #4 already mentioned http urls.
It’s probably worth noting that google seems to already be indexing the new domain with http links; e.g. my first hit for “topos nlab” is the page http://nlab-pages.s3.us-east-2.amazonaws.com/nlab/show/topos . Could this cause problems down the road?
Migration update: I have now rendered comma category using the new renderer. This is a much more complex page than the other one which has been rendered using the new renderer, outer automorphism, and I am now in a position to render a large number of pages. There are a few pieces of functionality the renderer still needs, but the only significant one from a code point of view work are the Tikz/XyMatrix figures; I know how I am going to do this, but it is requires a little extra infrastructure.
Please take a close look at comma category and explore it to see if anything does not work, or if you’d prefer anything to change. There are numerous major changes to the way things are done from the old nLab, which I don’t wish to discuss now; once the migration is completed, I’ll try to summarise all of the major changes. However, I would like to draw attention to the context menu; I think that the issues from #10 are now fixed, but please check. I have also tweaked the CSS quite a bit from before, please let me if it is OK, or if there’s anything you’d like tweaking. The actual implementation of the context menus is completely different now: they now are not regular pages, but instead can be found at /context/show, or /context/source for the source, i.e.
At comma category, I think that the issue mentioned in #73 is fixed. I have changed the underlying HTML of theorem environments in the new renderer, and there is not an immediate way to fix the old pages, so I will leave them for the moment; once they are rendered using the new renderer, the problem should disappear.
Before I do a mass rendering, I wish to put basic edit functionality in place, at least via API, so that all new pages can be edited. This should not take too long, but in any case is what I will focus on next. Then there will come a mass rendering, and at that point things will really start to be taking shape.
Regarding #76, I’m afraid I’ll have to ask for your patience for the moment; history is coming soon, possibly at the same time as the edits (definitely at the same time for newly made edits at least), but things have to be taken step by step.
Regarding #74, thank you for raising this; this happened precisely because of the existence of a redirect with the same name as the page. This was essentially a bug in the old Instiki; it will not be possible to create redirects in this situation with the new software. I will run some script at a later point to try to find all such occurrences; for now, please continue to raise them if you notice them, and I’ll fix them manually. I have fixed this one manually now, rendering it with the new renderer to group actions on spheres, although there might be some caching in nginx/your browser which prevents this from working correctly for a little while.
Regarding #75: no, Google changed very quickly (within a day) when we began the migration, and I expect it will change back very quickly once the DNS points directly to AWS rather than being redirected there via nginx on the old server.
Apologies that it took so long between the previous update and this one, I was completely tied up with other things for some days, and was slowed down for some other reasons in addition.
The entry comma category looks great! Thanks.
Regarding group actions on spheres: This still shows an old stub version. The latest version had a long well-developed discussion with several subsections (you and me had talked about it somewhere, recently). Hopefully we didn’t lose the page?
No, the nLab in the state it was in when I started the migration is sitting safe and sound on the old server (backed up on various people’s machines). There was a bug in how I was extracting the content from the old server which omitted some of the content before feeding it into the new renderer; I have now fixed that bug, but I see now that this page is quite large and complex, and I need to work a little on the new renderer before I can render it. No time now, but I’ll do so as soon as I can, probably later today.
I see by the way that the CSS of the context menus can be improved a bit on a mobile phone, I’ll address that when I get a chance.
Thanks! No rush. But it’s good to hear that this page I wrote is still with us.
If xy-pic is being migrated, I’d like to advertise a suggestion left at GitHub (basically to change “UseTwocells” to “UseAllTwocells” in the preamble, allowing to draw more kinds of shapes). Not a priority, really.
Sorry for not having addressed that issue; I had seen it, but had not an opportunity to do it before the migration. I’m happy to do that and it is essentially trivial, so I just need to remember; just remind me if I forget!
I have now rendered group actions on spheres using the new renderer. I needed to add some further functionality to the latter and to robustify some aspects of it to get it to work, so things are coming along. Please let me know if anything looks wrong on this page. There are a number of things to check: an equation reference for instance. I had particular trouble trying to repair the fact that Instiki, or more probably its underlying Markdown renderer Maruku, was extremely permissive with regard to list syntax, in a way that is not really in accordance with the Markdown spec (albeit the latter is somewhat loosely defined and often somewhat loosely interpreted), and which is not permitted by the new underlying Markdown renderer (a Python library called mistletoe). Thus please especially check that all the numbered lists render correctly.
I am aware of one small issue, namely that any LaTeX in the context menu is not rendering; I will fix this when I get a chance.
Looks good to me. Thanks!!
One minor issue:
It used to be the case that, when following a link to an anchor inside a page, the target paragraph would be highlighted by a gray background box.
(We had introduced this years back when we found that, otherwise, it can be hard for the reader to figure out where exactly they are being pointed to. This is particularly relevant for pointers to reference items, but I found it useful also more generally.)
Could this feature be brought back?
Found a tiny issue: kernel is redirecting to kernle instead.
Thank you Amélia! This was the same kind of bug as that discussed in #77 (at ’Regarding #74…’), namely the old Instiki had allowed a redirect to be created for a term which was already used as a page name. I have corrected it now, and have rendered kernel using the new renderer (this was again quite a useful page to try the renderer on, allowing me to fix a few small things). You may have to clear your browser cache to see it.
As a general update, I now have an edit page that loads page source, and am working on completing the editing process (submission, interaction with the nForum, page history, etc). Apologies for the slowness, as usual, I have had limited time and have again been briefly ill. Nevertheless, it is not so far from being deployed. There are so many corner cases in the nLab page sources that I am reluctant to carry out a mass migration of pages to the new renderer yet; the more ’tricky’ individual pages I am able to try it on to catch corner cases, the better. Thus I would like please people to suggest some pages that they would like to work on in the immediate term (with the caveat that it might take me a week or more before I’m ready to deploy the edit page); let us say max 20 pages per person for the moment. I will then allow editing only for these to begin with.
I’d enjoy if the entry Knizhnik-Zamolodchikov equation could be edited. Not because there is much on it at the moment, but because I’d like to work on it.
Jast a propos, not of particular urgency:
The wave of edit announcements this morning was puzzling [edit: not just for me, as I see here] until I opened the edits and saw that they pertained all to the HoTT sub-web, which (I didn’t think about this before) must be editable because I asked you to keep sub-webs editable, but also is the only sub-web that automatically sends edit announcements to the nForum.
It’s more confusing than it ought to be since many of the subjects (like “monoid”) exist on the Lab and would be expected to direct to the Lab and to provide, among various other perspectives, that of HoTT.
I find it unfortunate and ill-conceived that the HoTT web decided for this parallelism instead of integrating into the Lab. While that’s not in our control, could we maybe make the distinction clearer by prefixing edit announcement from the HoTT-web by “HoTT”?
So I am suggesting that instead of just “monoid” the thread for their edits should be titled “HoTT:monoid”.
I am aware that the “Category” of these threads is already distinguished as “HoTT: Latest changes”, but (at least on my browser) this is in such small font and faint coloring compared to the title that it’s bound to be missed on scanning over the main page.
The AWS hyperlinks are now spreading beyond Google: https://mathoverflow.net/questions/415469/breen-deligne-packages-and-the-liquid-tensor-experiment.
Yes, because if someone wants to link to an nLab page, they cut and paste what’s in the browser bar. This problem has been happening for weeks, and it’s going to be annoying to update all these temporary link people are using.
Perhaps a notice at the top of each page giving a ’share’ link like many services do, so that people can get the real (currently redirecting) URL?
Just a quick clarification that the links are not temporary; they will continue to work even when the DNS for ncatlab.org has been switched to point to the new site. I wished to reserve the right to change the AWS root URL earlier on in the process, but I’m now happy with it, and it will stay.
I have, however, contacted Adeel now to initiate the process of changing the DNS. This will break personal web URLs, but we can probably work around that once I have a chance to look at it. It also has the disadvantage that history links, etc, will result in a generic 404 rather than a message referring to the migration, but this should not affect many people. The DNS for the nForum will continue to point to the old machine.
I am on track with regard making editing live, several components are in fact live already, but I do not wish to indicate how to use them until everything is in place.
Not sure if this has already been spotted but when the target of a nLab redirect has non-ascii characters, as appears to happen in for example
http://ncatlab.org/nlab/show/limit+in+a+quasi-category (target (∞, 1)-limit)
http://ncatlab.org/nlab/show/etale+space
the redirect doesn’t happen but ends in error NoSuchKey.
Thank you very much for spotting this Jesus, I’ll look into it as soon as I get a chance.
On a different note, I am in the process of changing the DNS. There are a number of components involved in this, and I think there is no way to avoid some temporary downtime for at least some things. It is possible that everything, including the nForum, will go down. Personal webs will definitely go down, but I will bring them back up at the latest tomorrow (Monday) evening, European time. Apologies for the inconvenience, but there seems to be a kind of vicious circle in that in order for SSL to work, I need a certificate whose verification step requires that DNS is working, and without the SSL certificate I cannot make the redirect back to the Saunders machine needed for personal webs to work. The nForum ’redirect’ back to the Saunders machine is purely DNS, so will hopefully work without downtime, but we’ll see.
To access personal webs temporarily until the redirect is in place, I have set things up on the old machine so that you can use the IP address of the CMU machine directly, e.g. http://128.2.67.291/schreiber/show/HomePage. It only works with http; the proper redirect once it is in place should work with https.
Hi, thank you very much for the offer! The DNS is now live and working, I believe. Redirects to the old server for personal webs are not yet working, but this is not a DNS matter; I will fix this later today. For now, hopefully using the IP address as in #96 provides a temporary workaround.
We are (now) using Route 53, but not letsencrypt; we are instead using Amazon’s certificate manager. The ’vicious circle’ I mentioned was due to the fact that I was setting up an AWS CloudFront distribution, and needed an SSL certificate for that in order for https to work.
On that note, to go back to something mentioned already in #1, https should now work. I have, for now at least, allowed http as well, as there are no significant security risks that I see in doing so for the nLab itself.
It may take a while before Google re-indexes to use https://ncatlab.org again, because the internal page links generated by the old Instiki software were relative, and thus the preference https://ncatlab.org may not readily become apparent to Google (it depends upon how their algorithms work), but it should happen once the pages are rendered using the new renderer.
I think that redirects to the old server for personal webs are now working, e.g. HomePage (schreiber). I briefly brought down the nForum in the process due to some misconfiguration of nginx on the old server, but I think everything is now working. In other words, the DNS migration is now complete, and the new nLab now has SSL :-).
Typing ncatlab.org into a browser still redirects to https://nlab-pages.s3.us-east-2.amazonaws.com/nlab/show/HomePage rather than https://ncatlab.org/nlab/show/HomePage