Humanist Discussion Group

Humanist Archives: Sept. 19, 2021, 7:04 a.m. Humanist 35.249 - Institutional Support for DH Websites

              Humanist Discussion Group, Vol. 35, No. 249.
        Department of Digital Humanities, University of Cologne
                      Hosted by DH-Cologne
                Submit to:

        Date: 2021-09-18 17:32:37+00:00
        From: Martin Holmes <>
        Subject: Re: [Humanist] 35.246: Institutional Support for DH Websites

Hi John,

I think it's true that very few people do understand the scale of the
challenge involved in preserving and maintaining not one or two but
potentially hundreds of web applications, all created with different
backend tools and frameworks, all of which age and obsolesce steadily,
and all of which will eventually fail one way or another.

Even a relatively simple WordPress site requires a backend database
(MySQL or whatever), a web server that supports PHP (whose versions age
out and become unsupported at a steady rate, just like the db server),
various plugins (because site-designers, who give little thought to
long-term maintenance, will inevitably pick up and install lots of
"useful" third-party plugins which are typically unvetted,
badly-supported and potentially dangerous), and lots of other JavaScript
and PHP libraries and tools. Your site will have a different
constellation of these dependencies and vulnerabilities than the next
one; each project will require continual hands-on expert maintenance,
with that burden growing rather than diminishing over time, until at
some point it will be impractical to keep it going at all.

So I don't think any library could possibly get access to the scale of
funding they would require to provide this sort of long term support.
Your library quite rightly said that they would happily archive your
static data, but they could never agree to maintain your custom website
in perpetuity. It's just not possible.

That's why the Endings project has been creating recommendations and
blueprints to enable us to build fully-functional interactive websites
that are as close to static data as possible; that have no backend
databases and no requirements for PHP or Java or anything else that
needs constant tending. This is perfectly practical and not very
difficult, and it can be done even at the end of a project, but it's
certainly better to start as you mean to go on and build your web
application this way from the beginning. I know it's a bit late to be
saying this to you now, but perhaps one approach you might be able to
take is to apply for a further grant, maybe in collaboration with the
library, to convert your existing resource into something which the
library can happily ingest, archive and continue to make available,
because it presents them with no maintenance burden.

I know you're familiar with the Map of Early Modern London project:


It's entirely static -- no database, no PHP, no WordPress, just HTML,
CSS and JavaScript -- and I think it works pretty well. The only
dependency we have is on Open Layers / Open Street Maps for our modern
maps, and that dependency can be sacrificed in the long term because
those modern maps are not really the main focus of the project.

I think we have to accept that the onus is on us, as web application
developers, to create products that libraries are able and willing to
archive, rather than on libraries to volunteer to take on maintenance of
countless various and fragile projects in perpetuity.


On 2021-09-17 10:11 p.m., Humanist wrote:
>                    Humanist Discussion Group, Vol. 35, No. 246.
>          Department of Digital Humanities, University of Cologne
>                               Hosted by DH-Cologne
>                  Submit to:
>      [1]    From: John Wall <>
>             Subject: Re: [Humanist] 35.245: Institutional Support for DH
Websites (291)
>      [2]    From: Alasdair Ekpenyong <>
>             Subject: Re: [Humanist] 35.245: Institutional Support for DH
Websites (45)
> --[1]------------------------------------------------------------------------
>          Date: 2021-09-17 17:43:51+00:00
>          From: John Wall <>
>          Subject: Re: [Humanist] 35.245: Institutional Support for DH Websites
> Matt, Gioele,
> Thank you for your extremely helpful and thought-provoking interventions!
> There is much here for me to understand and digest. Gioele, I will explore
> your strategy with our Special Collections folks when we next meet (at the
> moment, they keep postponing our next meeting, whatever that means!).
> Further suggestions are welcomed.
> I did, in fact, have a long talk with our Library's DH staff at the
> beginning of this phase of the project and came away with a page-long data
> management plan. This requirement is now a standard part of NEH grant
> proposals. But that plan only included their pledge to archive the data
> going into the site, not the site itself.
> What we submitted made the NEH happy enough to give us the money. But all
> of our discussion was about what they would promise to do at the end of the
> project, not how we could help prepare the project for storage as we worked
> on it. That would have been very helpful.  We developed the website in
> WordPress because that is what the university provided us to use.
> I must say that all this feels a bit like a problem of scale and
> imagination. Our Library (all research libraries?) understands the
> necessity of conserving data and knowledge as long as it comes in the form
> of books, journals, etc. They spend millions to house and maintain
> knowledge stored on paper. They hire staff and pay salaries to people to do
> this. They have gotten accustomed to doing this so they take it for
> granted.
> Now, we have a new technology for storing data and knowledge, with its own
> distinct challenges. One might think they would see this as an opportunity
> to ask for larger budgets, rather than talk about the impossibility of
> doing their basic job. But maybe I don't understand the scale of the
> challenge.

Humanities Computing and Media Centre
University of Victoria

Unsubscribe at:
List posts to:
List info and archives at at:
Listmember interface at:
Subscribe at: