Discussions of Mormons and Mormon life, Book of Mormon issues and evidences, and other Latter-day Saint (LDS) topics.

Saturday, November 02, 2013

Glitch at the Maxwell Institute: Website Redesign Breaks Hundreds of Links

UPDATE, Nov. 3, 2013: The good folks at the Maxwell Institute are aware of the website glitch (that's my little euphemism for disaster) and are working on a fix (a few hours of adding redirects on their site ought to do it, in my opinion). The promise of a fix means this: "If you liked your links, you can keep them." Now that's a promise I hope we can trust. I think they won't even need any superstars from Google for this little Mormon tech rally. And with luck, they won't be getting any government help for their website. But if the fix can't be done in a day or two, then it's back to doom and gloom since I'll need to spend a large chunk of my free time manually repairing all my links to articles on their website. 

UPDATE 2, Nov. 4, 2013: Sigh. I must abandon hope for a swift fix. My contact at the Maxwell Institute has informed me that the project to fix the numerous broken links with redirects is a "massive project" that will take much longer than the day or two that I expected. I cannot understand this, but apparently things at BYU complex and difficult. I'll still offer my humble suggestion that perhaps the old website can be resurrected until the knew one is ready to go live. But for now, I guess I need to start manually fixing links myself. Now that's a massive project.


Just in time for Halloween, something frightening rises from the cyber crypts over at the Maxwell Institute, with vaults of extensive scholarship of much interest to defenders of the LDS faith. Great website, rich legacy, and lots of good people make it possible--but an eerie website "glitch" over there threatens to leave many without access to vital information, and is making at least one webmaster and blogger howl.

Thousands of people have been aided by various documents available there, especially the scholarship in a great LDS publication, The Review of Books on the Book of Mormon published by the Maxwell Institute, originally called The FARMS Review of Books. Defenders of the LDS faith have found some of the most useful information in the many pages of those volumes, where many scholars have reviewed and responded to anti-Mormon publications as well as neutral and pro-LDS books. This blog (Mormanity) and my website at JeffLindsay.com,  especially the LDSFAQ area, have frequently found useful information at the Maxwell Institute and linked to it. Hundreds of links, actually. And now many have been broken in a puzzling redesign of MaxwellInstitute.com. It may be slicker in some ways, but if links are broken, it's a serious problem.

Please, don't break links to publications that many people use. [Again, they know this, of course, and are working on a fix. Wish it had been noticed though before the launch.]

Long ago the leaders at the Maxwell Institute (previous leaders) had expressed a commitment to keeping their online content available with links that would function even when the website was redesigned. They learned the importance of this over a decade ago when one of the early redesigns of their database caused links to break and forced webmasters like myself to manually repair broken links one at a time in a painful process. Many users explained the problem and we were relieved to hear that the leadership there was committed to never making such a mistake again. Well, here we go again. 

The old Review is now hard to access, in my opinion. Trying to pull it up brings not a page about the review or anything that one would expect, but partial search results that show a few volumes, obviously with a bug in the sorting algorithm. The volume/issue data are presented in this order: 1/1, 2/1, 3/1, 4/1, 5/1, 6/1, 17/1, 20/1, 7/1, 7/2, and 8/1. Huh? What are volumes 17 and 20 doing between vol. 6, no. 1 and vol. 6, no. 2? Three of the four search results pages have similar oddities, with only the last page of results having things in proper order. 

The real problem for me, though, is that past links to the content in the Review are now broken, and it isn't obvious how to fix them. For example, on my LDSFAQ page on the Book of Abraham,  I cite a 1992 article from John Gee in this way, but here showing my HTML: 
John Gee,  <"http://maxwellinstitute.byu.edu/publications/review/?vol=4&num=1&id=92"  target="_blank">A Tragedy of Errors</a>," Review of Books on the Book of Mormon, Foundation for Ancient Research and Mormon Studies, Provo, Utah, 4 (1992): 93-119.
It looks like something like this when viewed in a browser: 
John Gee, "A Tragedy of Errors," Review of Books on the Book of Mormon, Foundation for Ancient Research and Mormon Studies, Provo, Utah, 4 (1992): 93-119.
But that link, which has been in place for over a decade and has been helpful to many people, now fails. Clicking on it simply takes a person to the publications page of the Maxwell Institute where readers will be puzzled to see a blurb promoting Mormon's Codex (yes, a great book) but not the article they were looking for. The link is broken and there's no clue how to get to John Gee's article. A dedicated user can surmise that the "Review" link lower on the page should be pursued, and then can find volume 4 in the search results, and then can scroll down and see John Gee's article is actually there, thankfully. The new URL for the HTML version of the article is: http://publications.maxwellinstitute.byu.edu/fullscreen/?pub=1427&index=15. The numbering system and the address convention is quite different than the previous now defunct URL, http://maxwellinstitute.byu.edu/publications/review/?vol=4&num=1&id=92. That means a global search and replace won't help, and a simple algorithm for revising the URLs won't work. Lots of manual work (searching, copying, pasting) will be needed to fix the failed links. Aargh. 

In the case I cited above, at least the failed link brought the user to a publications page where further searching or exploring could be done at the Maxwell Institute. It wasn't a dead link, the kind that gives the dread 404 "page not found" error. Sadly, I wasn't so lucky on the second Maxwell Institute link that I tested today. Again on my Book of Abraham page, I had a link to another article by John Gee. The link is to a 2007 article in PDF form. Here is the URL I have been using for years: http://maxwellinstitute.byu.edu/pdf.php?filename=MzAwOTM3NTgzNy0xOS0yLnBkZg==&type=cmV2aWV3

Lest you think the mistake may be mine and that this link never worked, you can verify that the link was correct by using Archive.org. See, for example, Archive.org's 2009 archived document at that URL

Painfully, the previous Maxwell Institute URL that is in place at the moment on my website now gives a disastrous 404 error. It's the kind of error that Google and other search engines punish (having dead links with 404 errors is a cardinal sin, one that can lower your visibility in search results). It's an error that loses credibility with readers and makes them give up on what could have been a meaningful exploration. A fundamental tenet of web design is that if you must make changes that remove pages or kill links, don't let the reader be chased away with a "404 page not found" error, but with a custom 404 error page that redirects users to a useful page of some kind within the targeted domain. I'm shocked that many recently functioning links to Maxwell Institute pages will now give direct 404 errors. 

If you must break links, don't just surrender with a 404 error, but take the reader someplace useful. At least have a custom error page in place. Still not nearly as good as links that are properly redirected, but better than booting readers out into the cold. 

I'm disappointed. But this problem can be fixed.

16 comments:

Mike Parker said...

Jeff,

There is a way to programtically update all your existing MI links. It's a temporary fix, but it will keep you running until you can change your existing links (or MI redirects their old URLs).

Any instance of "http://maxwellinstitute.byu.edu", add "https://web.archive.org/web/20131015130030/" in front of it. This will redirect readers to the Internet Archive's cached version of the old MI site before they changed it.

Blair Hodges said...

Jeff, we're aware of this annoying problem and we're working on it. Check out the Frequently Asked Questions on the new MI site where Sandra Thorne explains:

"A system is being created to re-direct all the URLs from the old site to their locations in the new site. So all the favorites, bookmarks, URLs embedded in articles, footnotes, etc., will eventually be live links again."

http://maxwellinstitute.byu.edu/news-events/faq-new-maxwell-institute-website/

Mike, good advice on the work-around. Hopefully it won't be necessary for very long.





Blair Hodges said...

Just one more thing, to be more specific: We're aware of this problem and we're working to fully address it, so there's no need to send emails about this particular issue as Jeff suggests. Thanks!

Jeff Lindsay said...

Thanks, Blair. Timing? Readers need the info now, and Google's penalties can be swift, so if it's more than a day or two, I'll have to take up the manual yoke and begin plowing through those dead links one at a time, finding the mysterious new locations one by one.

In my opinion, either a script or a lot of manual redirects on the site ought to be achievable in a single day of staff time Monday. Let me know if that's an unrealistic hope. And certainly a redirect to catch 404 errors can be added in a few minutes.

Jeff Lindsay said...

YIKES - their FAQ dated Oct. 26 shows they've been aware of the dead links problem for over a week, certainly longer, and yet launched the changes anyway with the assurance that "all the favorites, bookmarks, URLs embedded in articles, footnotes, etc., will eventually be live links again."

Eventually? As in when? Redirects to every broken link could have manually be done in the time they've known of the problem without inconveniencing the public, if this was a priority. It clearly isn't. Or maybe it is but they have been getting government help on this. I think I need to just hit the panic button and start manually slogging through my links to fix them. Ugh.

Greg said...

If there was any sanity in the world:

1) Making links not break would be the top priority. FAIR is doing a new website with all volunteer labor, and we can manage it.
2) If the "new and improved" website broke things, you'd revert to the one that worked until you got it fixed, so other websites don't have problems, so it doesn't thrash SEO status, etc.
3) If you're in the business of providing info, if whoever you hired to do your upgrade did this, you'd fire them, and prob the heads of whoever was in charge would roll too if #2 wasn't invoked post haste.

And, how long do redirects really take? I built templates for the FAIRwiki for every FR and JBMS article, which included HTML and PDF links for each article. I did it over 2-3 days in my spare time.

Surely someone with paid employees who does this full time could do at least that well? Is it REALLY that hard? They can consult the FairWiki to see old links to each.

If the new website had been _designed_ to make past data hard to find, it could hardly have done a better job.

Not bad for 18 months's work. :-)

Anonymous said...

Students, undergrad and post-grad, with no real-world experience being led by academics and ivory-tower types with no real world experience. This amateurishness is much like the first iteration of lds.org and ldscatalog.org.

Anonymous said...

Oh, and the first iteration of familysearch.org, too.

Blair Hodges said...

For those who don't know, I should introduce myself. I'm Blair, I'm new at the Maxwell Institute. Redesigning a site with such a large back catalog has been a massive undertaking with a lot of unexpected difficulties. The four issues I'm most disatisfied about on the new site are 1) the search functionality, 2) the author list page, and 3) the url redirect issue, 4) iTunes synchronization (we're having problems syncing the new feed with our already-established iTunes channel). There are other minor issues that bother me and that we're going to improve, but I believe these three are the most problematic. We're working to get all of these things worked out. As we announced when the site went live, this is a beta launch, indicating it's still being worked on.

As for positives, I much prefer its design and aesthetic compared to the old site. It is especially designed to work well with mobile devices, ipads, etc. I like being able to quickly view articles in a variety of ways, and the quick print feature. The site is also a lot easier on the back-end to update and add new things than it was on the cobbled together older site. Adding new articles, books, etc. is loads easier. I'm also glad the site does a much better job of highlighting the Institute's various initiatives, including METI, CPART, the Willes Center, etc. The Maxwell Institute is more than the (still important) review journal, and we want the website to communicate that. I'm also pleased that we've added a new blog and podcast (and not merely because I'm in charge of them...).

All in all, once the website is functioning properly I think most folks will really like it. I'm interested to hear more feedback, you can send it via our "Contact us" page on the site, or email me at blairhodges [@] byu.edu. Tell me what you like, what you don't like, etc. and help us keep improving.

Jeff Lindsay said...

By way of clarification: the point of my post was a) to let the LDS community know that we've got a huge problem requiring that either we or NMI fixes all the broken links, b) to urge NMI to fix the problem ASAP and not do this to us again, and c) to let webmasters of pro-LDS sites remember in the future to be careful about changing format and breaking links.

Anonymous said...

Hmmm, something is odd. The Maxwell Inst. could revert the website today and fix all those links. They could and should have spent the few hours needed to provide redirects for every article, or at least for every article that had incoming links that would be affected by the change. I'll bet only a few articles in each edition of FARMS Review had significant incoming links. Find and fix those - how long could that take? A day? Something is VERY ODD about this.

Revert. Repair. I suggest you do it today, or you'll just fuel wild (?) suspicions about the Institute and its alleged distaste for apologetics.

rameumptom said...

Blair, you say this is a "beta launch". Well, most companies don't close down the original site until you have successfully gotten from the beta test to the fully up and running.
I'm hoping for a new and improved Maxwell Institute, but really am not impressed with what I've seen so far. You bought Salt Press, but nothing has come out of it. Still waiting on a journal or articles to come forth, and there has been almost nothing.

It's time MI stepped up and did things right. Interpreter has fewer resources, and yet have seemed to have found their niche. I hope you guys can too, otherwise you may end up being akin to Obamacare: a massive entity that provides no real value....

Blair Hodges said...

Reporting back, it looks like a temporary re-direct is already in place for most if not all of our articles, book chapters, and transcripts--for the html pages.

I noticed that the broken link example Jeff offered in his post is a pdf link. I discovered that the pdf links were somehow missed by the design firm in the general redirect process, so we're letting the design firm know.

All old links (html and pdf) should temporarily lead to http://publications.maxwellinstitute.byu.edu/. Now we are in the process of manually redirecting each link from its old version to its new, and we have to do this for more than 8,000 individual links. (That estimate is low because all old pdf links aren't included in that figure.) Here's an example of an old link that now redirects to the new site:

http://maxwellinstitute.byu.edu/publications/books/?bookid=41&chapid=178

In the coming weeks all of the old links should redirect as that one does. Given our current set-up it isn't a process that we can do in a mere matter of hours, although I wish it was. If you discover a broken link to an html article (not a pdf link) you can forward it to us through our "Contact us" link. We're also hunting them down.

(One more point: The Occasional Papers series hasn't been moved to the new site yet, but it will be.)

Thanks, everyone.

Blair Hodges said...

rameumptom: We're definitely glad Interpreter has found their niche. We don't view ourselves as being in a contest with them although I understand the desire to keep score.

As for what the MI is up to, we have a few new translations from METI and ITS coming out in the next two months or so. The new issue of the Mormon Studies Review is due out in December, and it's right on track. We have a really impressive lineup. You can subscribe here:

http://maxwellinstitute.byu.edu/subscribe/

We're almost ready to re-issue the Salt Press titles and publish several all-new titles acquired through Salt. And we're gearing up to release a bunch of new books in the next few months, so keep watching the blog for updates if you're interested in what the MI is working on.

In the meantime, there are 60 or so blog posts from this year if you missed those.

http://maxwellinstitute.byu.edu/blog/

And the MI podcast has 15 or so episodes you can download for free. The newest episode features Sam Brown, author of In Heaven as it is on Earth.

Anonymous said...

The MI site is not user friendly. I see reference to an article on a pro LDS site and that article is linked to MI, when I click on the link, the article is not there. I have to hunt for it, never find it, not very user friendly. This problem is not recent. I have had problems finding articles on the BYU site too.

Blair Hodges said...

The MI site is not user friendly. I see reference to an article on a pro LDS site and that article is linked to MI, when I click on the link, the article is not there. I have to hunt for it, never find it, not very user friendly.

Right, this is the problem we're addressing individually, link by link, so that they lead to the proper place. It will take time, but we'll get it done.