AW, Look What They’ve Done to Our Links, Ma

Péter's Picks & Pans Online 27(4) July/August, 2003

I know that many have shared my experience, and some even posted their thoughts in the What Were They Thinking thread on the JESSE listserv about the new ALA Website (AW).   ALA chose to celebrate the National Library Week  with the launch of AW. In the eyes of most visitors it must have been the Week of Frustrations, Failures and Fiascos.

I presume someone will write a very scholarly article about AW (which may get published as early as 2004), but as a die-hard educator, I would like to show you a little broader and deeper panorama with many illustrative screenshots about  the  new AW right now as you experience it, and  also to tell you what was I thinking about AW and the overall  digital attitude of ALA. Maybe, it can be used as a backgrounder  for creating those boilerplate petitions for members, to lobby ALA to reconsider AW and its digital strategy.

What Was I Thinking (and Hoping)?

It is quite discouraging that while ALA has published so many good books and journal articles in some of the best and most affordable scholarly journals of our profession about systems analysis, Web site design, information service and database evaluation, it does not practice what it preaches. I present a short summary first for those who battle with ADD, then I proceed with the detailed and illustrated review.  Links to external sites are marked in conventional blue, and they open in a new window. Jump links to sections within this document are marked with green for those who don’t want to read through this piece but would rather go cherrypicking  from this introductory part. 

Presenting the new AW struck me as if a dietetic nurse would pontificate about health issues over munching a Big Mac with extra cheese and mayo, or as the Hungarian saying puts it succinctly: preaching water, drinking wine. 

As a starter, I hoped that the new AW would finally offer the full text digital archives of  ALA journals for subscribers (maybe for some extra charge), and a database of their abstracts free for anyone as so many publishers have been doing  this (including all the major players in the scholarly field of library and information science except for the Haworth Press). You find among them  not only the non-profit ones but also the largest commercial publishers which are not exactly known for “donatious” attitude. 

What I saw at first, instead,  was the (hopefully temporary) disappearance of the free table of contents, abstracts and selective full-text articles  from 19 volumes of Information Technology and Libraries (ITAL),  all the back issues of Journal of Library Automation (JOLA),  and all the volumes of Reference and User Services Quarterly (RUSQ), except for a current issue with a grand total of three abstracts  - just to  mention some of my favorite ALA journals. Many more sources have not made the transition to the new AW.

For fairness, two of my other favorites, College and Research Libraries and College & Research Libraries News, mostly survived the reorganization (from content perspective), although the latter is missing the 1994 and 1995 volumes which were available earlier. The access perspective, however, is another question.

I thought and hoped that the new AW would sport digital object identifiers (DOIs), as do more than 8 million scholarly articles, published in primary documents of more than 200  publishers. The International DOI Foundation did an excellent job in spreading the gospel, and working with rival publishers, which is like herding cats. Using DOI could have offered an elegant and lasting solution instead of  the current chaos  which cripples tens of thousands of existing links to  previous AW destinations, and may happen at the next reorganization. 

My estimate is higher than the number floated around in messages last week, which have not considered the many links to ALA journals and other ALA sites from publishers' and aggregators' abstracting/indexing and full-text archives.  In the long run  this will annoy many more users of the new AW in lack of decent and comprehensive redirects with the exception of some of the high level ALA sites. H.W. Wilson which has done an utterly professional job in verifying, cleaning up and consolidating URLs in the journals it covers is probably the hardest hit by the below the belt punches (or should I call them Tysonian ear bites)  of ALA

I hoped that new AW will have a much better search engine than the old AW. Instead, the new search engine seems to experience a tsunami of senior moments at best.  The page which  could be seen the most often when clicking on an existing link, was an error message to which I just refer to for  a shorthand  as the Sorry! page. It counsels  users by telling them to use the advanced search page (which should be  called “disadvanced” search page) , or to navigate  using the site map (which is more like a site gazetteer), or call the help desk.

.
.

It’s not easy to decide which one of the three is the most frustrating. The first two options I will discuss and illustrate in detail below.  As for the third:  it is impossible to get through to a live person at the given extension, but a voice recording assures you that they will come back to you within 3 business days with an answer, although it may not be the  URL of the new location as many pages have been left behind or lost in the hectic move).

If the help desk can tell you the new URL, have much patience, a clear phone line, and a large piece of paper ready. For example, what used to be http://www.ala.org/ascla/charl.html for a useful document about how to make the ACRL conference on a shoestring, has been transformed into the URL shown below when AW was launched, just when participants  were preparing for the conference trip. They must have had fun trying to jot it down while on the phone with the help desk people, then type it in with the correct  number of underscores

http://www.ala.org/Content/NavigationMenu/ACRL/Publications/College_and_Research_Libraries_

News/Back_Issues__2002/October/Charlotte_on_a_shoestring.htm

The query charlotte shoestring did not find it, but gave a not exactly warm and fuzzy error message shown below. (There is no help file about the search syntax to learn that space between words means exact phrase, but this is not the only reason why  it would be premature to consider AW for an ASCLA award for excellence in accessibility. In case you go to Charlotte for some other event click  here for that piece.)

.

.

Week of Festivities (and Frustrations, Failures and Fiascos)

ALA knew that the National Library Week is coming (as it has been for 55 years), and probably wanted to deliver a gift, but as will become obvious it  was way behind  with finalizing the new site. Still, it kicked off the week on April 6th with the launch of the new AW. I wish it had waited for the next anniversary. It’s not only that the wrapping is less than careful, it is the whole package that seems to be  a kick in the face of most visitors, and basic systems design rules.

I kicked off our 30th wedding anniversary celebrations the same day.  While I am often a dollar short and a day late with anniversaries and festive occasions, I knew that this day would be coming, so I ordered ahead of time the special kukui nut lei, sneaked out, picked it up early morning, and with some little other gifts and a personal card I could almost show up  on time with these and the breakfast tray. I was not perfect, I never am, but I think I was better than ALA with organizing the week of festivities. (Yes, I know that sweet pea is the “official” flower, pearl or ivory the appropriate jewelry, and Hallmark the authorized supplier of all kinds of cards with cliché sentences for 30th wedding anniversaries).

Sure AW was (or will be) a surprise gift for all of us who created the tens of thousands of  links to the old AW in the past few years, as well as for the users who will click on the now dead links. Sure, this reorganization was an ill-timed and very bad  (gift) idea, given the massively dysfunctional nature of the new AW.

.

I can’t give you exact numbers how many links there are to the various old AW sites, but the obvious check provides only a rough estimate, much lower than the real number of casualties. Google claims 22,900 links to www.ala.org, Alexa’s number is 18,340. AllTheWeb’s number is so off-the-wall that I don’t even quote it.

.

More to dead links than meets the eyes

There are  many more sources linking to the old AW than the above search results may suggest. For example, there are  a great number of articles in the digital versions  of LIS journals which provide links in the text or in the endnotes  to the selectively published  full-text  articles of  various ALA  Division journals, and/or to ALA websites. These are not visible at all for any of the Web-wide search engines as they are in password protected sites and displayed only in response to actual queries. Here is an in-text reference from the Journal of Academic Librarianship to an ACRL website, and the response to the link (after I removed the closing parenthesis that should have been done by the copy editors who probably were too busy checking the compliance with their punctuation guidelines where it does not matter, and time and again overlooked, or may have added lethal HTML punctuation errors, sacrificing access at the altar of publishers’ style sheets).

.

Click for larger screenshot!

.

This is what happens with an endnote  link from Serials Review to an article in the ALCTS Newsletter

.

Click for larger screenshot!

.

Aggregators are also hit hard by this reckless reorganization, and H.W. Wilson would be right to call it a below the belt hit. Why? Because H.W. Wilson had done the most of every content providers that I know to verify, correct and consolidate in a special field the URLs mentioned by the authors. Scholarly publishers could take a page or two from H. W. Wilson's book on how to cite URLs.  In the example below two of the three links from an article to ALA sites bring up the dreaded Sorry! page.

.
.

Bibliographies and reading lists of various LIS courses which link not only to the full-text articles mentioned above but -in order to enhance those reading lists- also to  abstracts scattered at various sites of the old AW. Until the launch of the new AW  such  links of mine as shown below worked fine, bringing up the appropriate abstract or full text available at various sites - ALA being one of them.

.

.

The big problem is not that we have to change the reasonable URLs to the  anaconda  URLs of the new AW (although some browsers just crash when encountering  such monsters like this one

http://www.lita.org/Content/NavigationMenu/Our_Association/Divisions/LITA/LITA_Publications4/ITAL__

Information_Technology_and_Libraries/Volume_21,_No__2,_June_2002.htm#anchor261182

The big problem is to identify all our links to all the old AW sites in various sources we have access to with modification rights, and then to find their new location.   Tracking down your high school flame who moved to the Middle East is easier. If you believe that a few quick search at the new AW would quickly locate the new address of the linked documents, you obviously have not tried the search engine of the new AW (more about it later). Of course, I am just an amateur, but if it were easy, wouldn’t you expect that the designers could have replaced at least those links from within ALA announcements which link to ALA sites not valid anymore? Here is an example from an ALA News Release:

.
.

and here is the result which you will see ad nauseam, on the

infamous Sorry!  pages, many off them triggered by other ALA

sites.

 

 

Click for larger screenshot!

.

An even less obvious problem is that the estimates  of 20,000 dead links do not include either the links to the  ALA sites which start not with www.ala.org but with some other primary domain names affected by the reorganization. For example, all the issues of ITAL (and its predecessor, JOLA) start with www.lita.org. At first blush it is not that many (about 1,600 which are visible for Web-wide search engines).  

There is an even bigger problem. Nothing can be done with the many in-text and endnote links in the articles and conference papers which have been published in the publishers' archives, and in the abstracting and indexing and full-text databases of Gale, UMI, EBSCO and H.W. Wilson. They can't be expected to go back and change them.

Now, start identifying all these other (non-www.ala.org) URLs, and determine the number of links to them in order to get another estimate for a more realistic number of dead links. Then, guessstimate the number of links from the publishers' and aggregators archives which are not visible for the Web-wide search engines, to have a sense for the real loss, and e-mail me in the morning. To soften the blow, I can tell you that some sources escaped the reorganization mess, such as the ALA bookstore.

 

The Little Search Engine That Couldn’t

The launch of the  new web site at the beginning of the National Library Week and its condition   throughout the week  was akin  to your spouse coming home most of the days drunk, first just navigating through the rooms erratically, then searching endlessly for but not  finding even his wedding ring, let alone the gifts, then sobering up for a short time, followed by a period of painfully long pauses before answering questions, then not responding to inquiries, then just conking out like teenagers after binge drinking and messing up their hotel room on a Spring Break in Cancún. Not exactly one’s dream of a week of festivities, especially after we were promised a rose garden by the ALA Executive Director.

.

.

At the beginning of the National Library Week, the search engine could not return much useful information. The simple search then wised up a little bit, although it remained enigmatic, not revealing, for example,  how many hits it found. Neither made it any sense why  many hits appeared twice or more times in the result list when they seemed to take you to the very same page. Yes, the second item one has the acronym in the page title, but no difference otherwise.

.

.

The URLs are so long that they always run off the address cell, and can’t be read, but after I could capture them and paste into a page, I could see a  difference between the two URLs, but not a justification for getting to the seemingly same target page.

.

.

.

Often, the software  did not respond for so long that the system got  timed out repeatedly, and I started to make snapshots while waiting and waiting and waiting.

.

I have a nice collection of being refused access to http://www.ala.org using different browsers at different times. For space reason I show here only my Mozilla and Opera efforts.

.

MIA or DOA

Once I could get through, the software simply refused to give an answer to a dozen of known item searches when I tried to estimate the damage and check some of my links to one of my test journals, ITAL:  Information Technology and Libraries

After much trials and tribulations, I learned that Volumes 1-19 of ITAL are not yet available on the site. They were in the old AW, and I have links to several of the editorials and other full text articles, as well as to  informative abstracts. They are now dead, and even if I navigate to the new location of ITAL they are not there, so I can't replace the dead links.

.

JOLA’s Tables of Contents and selective abstracts were on the old AW (as shown by the Google cache result), and the new AW is "pleased to be able to offer the Tables of of Contents and many abstracts" but you would not be pleased because it ain't  as I write this. JOLA may seem to be ancient but many of its articles from the late 1970s when I was up to my neck in library automation projects are still more relevant and insightful than some of the most current articles.

.

.

.

I have seen significant omission also for another journal's abstracts which I use in my links, the Reference and User Services Quarterly (RUSQ). In the old AW all issues were accessible through the RUSA page from the first issue after the title change from RQ

.

.

The RUSA page  on the new AW offers a grand total of 3 abstracts from a current issue. That's it. Another missing in action journal even after the National Library Week.

.

.

I found out these –hopefully temporary- omissions and MIAs when the last item from my  checklist for ITAL (the one about Multimedia Research Support shown earlier) finally was found by the search engine.  I could recognize it from the single item hit list only because of the familiar volume and issue number, not because of the summary excerpt shown by the software. It is not from the abstract of the article I was searching for.

.

It turned out to be the Table of Contents and Abstracts page, and knowing what I was looking for (a known item search), I scrolled down on the page until I got to the entry I needed. As the unit record is not the abstract but the table of contents page, and the result list shows excerpt from the top of the page not from the context of the matching words, results will often be very enigmatic. (More about a better solution later).

.

Advanced or “Disadvanced” Search

The advanced qualifier is the one that should be put in quotations mark because it is anything but advanced when you want to use its only advanced feature, filtering a subject search by one or more ALA divisions and/or offices or subject areas. The search below, was 'guaranteed' to find at least the abstract of the Multimedia Research Support article from the LITA journal shown earlier. Did it?

.
.

Well, it returned this message, unfortunately without echoing the query, but you probably trust me that I did not change the above query to supercalifragilisticexpialidocious just to trigger this response. A variety of subject terms combined with a variety of divisions never yielded any results.

It is also a sign of the brain damage of the software, not just senior moment, that when I search for the Fast Fact column edited by Ann Viles in College & Research Libraries News. Using her full name (yes, the first name has no 'e' at the end), the software finds only two items.

.
.

Given the fact that this journal was the least neglected during the transition (though 1994 and 1995 issues were not transferred), there should be much more hits when searching by her name which deservedly appears with her column. Indeed, the columns are there but the search engine can't find them. When searching for her last name alone (trying both the simple and advanced search), one more record is retrieved, the ALA Handbook, where her name appears on the 101st page of the PDF file as Ann Viles. Why was this not retrieved when searching for Ann Viles is beyond me, but I am much more concerned about the large scale omission of her columns in the search results.   

Navigating and browsing

One may say that when looking for a known item I should browse rather than search, navigating down to the journal, volume and issue.  Trust me,  I know the importance of browsing before searching. I am the one who always reminds students that if you don’t browse before searching then you are like the guy who dives in a pool without first looking to see if there is water in it.

Anyhow, with this site I would have a hard time selling  browsing and navigation before or instead of searching. Yes, there a few milestones and landmarks to use, but they can be frustrating on AW for navigation. I wanted to get to CHOICE magazine as quickly as possible, so I clicked on the Products & Publications button, and got a nice list of ALA journals and newsletters. Unfortunately, CHOICE (and many other) was not among them. It should have come next to  Booklist, not only alphabetically, but also because CHOICE is also a review publication, targeted for academic libraries. The third entry on the page  did not make much sense and was not hot linked anyhow. With a hindsight, the title of this page and the heading in the outline should be Selective List of Periodicals.  

.

.

I frantically tried some other categories and subcategories, like Other Products, but I saw only ALA tchotchkes, no CHOICE. First I thought CHOICE may have been acquired by Information Today, Inc. but I checked my sources and got a negative reply.  Other companies which went on a shopping spree like Imelda Marcos in her salad days also flashed through my mind but I decided to go back to navigate the new AW.     

 

Site Map or Site Gazetteer 

Yes, there is something called a site map, for navigation but it is more like a site gazetteer. Using it is like navigating the streets of San Francisco  with the Columbia Gazetteer of North America in hand instead of the Dorling Kindersley city guide book. When you click on the site map button you get an enormously long list after an unusually long  wait, until the 800K site list is displayed. The screenshot below can’t do justice to the intimidation power of this list.

Suffice it to say that I had to scroll and scroll and scroll to get to the entry of  College and Research Libraries. But I can’t complain as I found CHOICE. It was hiding between College & Research Libraries News and College and Research Libraries, like Iraqi soldiers between civilians, but my consternation about filing order rules was washed away by the joy of locating these periodicals. Then again, my pleasure was tempered when from the home page there was no link to the individual issues, a problem if you find this page not through browsing but searching when you don’t see the context as in the site gazetteer approach. You see, I've come to recognize advantages of  the gazetteer's approach.

Interestingly, there is a button to print the site map. However, I would strongly advise you against it, unless you want to create a printout on a dot-matrix printer with that endless folded paper to see if it is long enough to wrap around the Equator.

Click for larger screenshot!

.

Basic System Design Issues

It is a good thing that some of the ALA journals have  well compiled indexes for each volume. These served very good purpose in the print world. However, it is not appropriate to use the same indexes without enhancements in the digital version.  Here is an excerpt from index of one volume of the C&RL.

.

.

The user who does not have at hand the printed issues has no idea about the page ranges of the individual issues within the volume (except for some extreme cases, like a page range of 3-21, which is obviously in the first issue). As of now, users often must guess in which issue is an article. These indexes should have and could have been enhanced by a direct link to the article's abstract through the issue's table of content page. It is not the best solution but better than leaving the users in the dark, and it could be made perfect if articles would become the unit records instead  of the table of contents and abstracts pages, which is very pre-21st centurish approach.   

Seeing a  result list produced by the new AW software is always a pleasure by virtue of getting some other feedback than the Sorry! pages and the  No Search Results pages.

Unfortunately, the abstracts and their bibliographic citations are not in their own record even after the emphatic reference in the press release to using database technology in the new AW.  The problem is that the result list shows some excerpt from the first item on the Web page, as that is the unit record for AW. That excerpt is unlikely to have anything to do with -say-  the seventh article on the very same Table of Contents page which matches the user’s query. The user would hardly understand why an item about Pinyin romanization was retrieved for the query about multimedia research support as shown earlier. 

This is a problem for practically all the results when searching for abstracts or full-text versions of articles except when the matching article is the first one in the journal and its excerpt is the one partially displayed in the result list.

This is not merely an academic question. Creating a record for each article (including the many free full text editorials and book reviews in some ALA journals), would provide much better result list for the users. There is a good reason for using individual works as a unit record in every journal archives. From these unit records it is easy to create even virtual issue records, and from those ones virtual or real volume records by the software.

This could also pave the way for the next level, which has been reached by most other publishers in the LIS arena, where ALA should take the lead in digitization instead of lagging years behind its peers: the use of Digital Object Identifiers. 

 

Digital Object Identifiers

There have been many ALA conference papers and journal articles about the digital object identifiers for a good reason. They allow unambiguous and layered access to individual articles for users depending on their status and thus authorization for access. It could also allow ALA directly to offer free abstracts and pay-per-view full-text articles to everyone, and "free" access to subscribers instead of not getting involved apart from licensing the full content to aggregators as is the case now. Other publishers have long recognized the advantage of getting in the ring of direct digital publishing.     

Very importantly, digital object identifiers also can solve most of  the problems of changing URLs, by keeping the DOI as the standard identifier, and behind the scene changing its pointer to a new URL which replaces the URL in the DOI-URL pair.  If ALA had been a trail-blazer, or at least a follower of the pack, it could have avoided this current chaos. 

There are many large publishers who joined the DOI Foundation, and more than 200 publishers signed up to the largest DOI registration agency, CrossRef. About 8 million articles have now DOI, which many publishers -rightly- flaunt. Among the many advantages of DOIs is their brevity, and much less error-prone nature. The more authors start knowing about the DOIs, the more will use them in their citations. It already offers incredibly smooth access to articles for qualified users  (and to  abstracts for anyone) in the CrossRef-enabled, DOI-endowed archives of participating publishers, as well as following citations not only within the publishers' own journals, but also among the journals of  rival publishers. It will also revolutionize in the long run, citation indexing and citation searching by offering very convenient and instant access to both cited and citing journals. In turn, this will enormously strengthen the value of the citation indexes, and will implement almost perfectly what Eugene Garfield has envisioned, and dreamt of for decades. It would be good if ALA would not stay out of the loop. The first step is to get acquainted with DOI in real.

The International DOI  Federation (IDF) offers the option to experiment with DOI for a mere $1,000. Better to start now as the next National Library Week is only 362 days away. If I were enough Javascript-savvy I would have included here a variable which would automatically keep counting down the days to the next festive occasion. Then again, I would declare any day a festive day when I see the first issue of an ALA journal sporting a DOI. The excellent content of most ALA journals would deserve reliable and efficient hyperlinking. So would librarians and other information professionals.          

Back to the eXTRA menu