The TrAC Internet Column: a status report

Stephen R. Heller
USDA, ARS
Beltsville, MD 20705-2350 USA

srheller@gig.usda.gov

It has been about 1.5 years since this Internet column for chemistry was initiated. It seems that this is a good time to pause and reflect on where the use of computers and the Internet has gone in this time and what the near future is likely to hold.

One concern I have about this column is, while it is being read by many (both in the printed and electronic form), there has been virtually no feedback (letters or e-mail to TrAC or the authors) on these columns. I have said on many occasions that the use of computers by chemists is low (which I define as less than 25%) [1]. I still think this is the case. The reasons for this are probably twofold. First there is the general tendency to not change the way one works. Second is the lack of real value or content which would induce individuals to change their work habits and start using computers and computer systems more often. For people to change there must be some strong incentive. Certainly something that will help you in your everyday work should be a strong motivator.

Before discussing the pros and cons of using the Internet as a tool for chemists, let's first look at the usage of the Internet version of this column for the articles published in its first year (1995). One of the very valuable benefits of electronic access is the ability to record every access (although one must add that it is not clear what each access really means, so, for reasons beyond the scope of this column, don't believe each access means someone read the article). While keeping privacy at a high level, the following summary usage information has been gathered:

Author Date publically available # Accesses
Heller 10/4/95 565
Dessy 15/6/95 280
Rzepa 15/6/95 857
Rzepa 28/7/95 351
Bachrach 10/4/95 776
Tissue 28/7/95 603

Universities in the USA and companies in the USA accessed these articles in about equal numbers. Outside the USA the access from the UK, Germany, The Netherlands, and Canada were the highest from the 63 different countries which accessed this column. In total (over 300 days of tracking statistics) there have been an average of 57 requests per day of the TrAC site, not all of which were for these computer columns.

Let's now list the pluses and minuses of chemistry on the Internet. On the plus side there are a lot of catalogs [2], instruments [3] and software product descriptions [4], tutorials for using computer systems [5], patents [6], societies [7], newsletters [8], and lots of bulletin boards [9]. (Please note that the list of Internet resources in references 3-8 are by no means complete, but should be a useful starting point for finding information in these areas.) On the negative side there are many Internet addresses (URL's - universal resource locators) that either disappear or change between the time you hear or read of them and the time you try to access them. There are many Internet sites which contain little or no useful information. (How many descriptions of analytical chemistry departments, their instrumentation, and their latest publications do you need?)

There are many, many Internet sites which point to other Internet sites, which, in turn, point to other Internet sites, which, in turn, point back to the first site. There is also the speed of the Internet, especially outside the USA. Over the past year, with the large numbers of users on services such as Compuserve [10] and America Online [11] the system is getting clogged with traffic. Will it become total gridlock, or will there be sufficient computer network bandwidth and switching capacity to keep the system functioning?

What we have then is typical of anything new. First there is unsuspecting euphoria. This is where I think we are with respect to Internet and the information highway. Soon there will an overreaction to this new technology reaching a low point of deep cynicism. This will be followed by a great improvement as real benefits are shown. Then things will settle out and this new technology will become understood and used on a routine and realistic basis.

For those of you who do not use the Internet on a regular (daily) basis, let me offer comments on what benefits you can find today. First there is e-mail (which, in my definition, includes bulletin boards). Second there is e-mail. Third there is e-mail. Perhaps mundane, but still, by far, the most useful thing around. E-mail allows you to easily communicate with colleagues next door or around the world. Telephone tag is mostly a thing of the past. E-mail, via a bulletin board, allows you to ask a question of perhaps 1000-3000 or more people, and most of the time, get a useful answer in a matter of hours. It's like having your own personal librarian, working 24 hours a day, 7 days a week. (Although if too many people are asking silly and very simple questions of these bulletin boards the system will disappear, no doubt to be reincarnated using a restricted access capability.) There are, in a real sense, no time zones on the Internet. It is prime, working time, 24 hours a day, somewhere on the Internet. Need to know something about a new chromatography column - then just ask. Want to know if others are having trouble with their new AA lamps - just ask. The next most used aspect of the Internet for chemists is probably access to information systems, such as the CAS STN system [12] and the Knight Ridder DIALOG system [13].

In writing this column I spent a considerable amount of time "surfing the net" [14] (that is, looking at various web sites around the world) for useful information in the area of analytical chemistry. From the many sites, many of which are listed in the references, I feel I can make a reasonable list of what is needed for chemists to really begin to use the Internet on a regular basis.

First, someone needs to develop an intelligent search engine for chemistry on the Internet. That is, a computer program that requires more than just one or two keywords for locating what you really want (if it is even on the Internet - see my experience in the next item). Some initial work in this area is being done by Gasteiger [15] and his co-workers in Erlangen, Germany in the area of protein data files.

Second, there needs to be a meaningful rating system for every chemistry site. There are organizations (e.g., Point [16] and Magellan [17]) which are doing this; however their rating schemes leave much to be desired for chemistry. In searching for "chemistry" on both these rating services, I found many strange results. The Point web site had 34 rated sites. The Magellan web site had 322 rated sites. While the Point web site had Virginia Tech (see TrAC article) as the highest rated site (1.0000 - an interesting number of significant digits), while the ACS web site was rated 0.8244. It also rated "Buzz Online", a magazine for those "yearning for that L.A. state of mind" highly at 0.8014, owing to an article in the magazine entitled "New Line's Casting Chemistry". Magellan's web site had its own idiosyncracies, such as rating the University of Rochester and Penn State Chemistry departments with 4 stars, but giving only 3 stars to Henry Rzepa's excellent site at Imperial College [18], and 2 stars to Jan Labanowski's widely used and highly rated OSC Computational Chemistry list [19]. One other limited rating scheme now being tested is that being organized by Rzepa [18]. He is asking chemists to "vote" on their favorite chemistry web sites and will present the results of this vote at the ACS Internet symposium in Orlando in August 1996.

Third, there needs to be, at the top of the first page, the date the site was last updated. Some sites are updated daily, others infrequently, others have never been updated. If one wants the periodic table, frequent updating is not a major issue. However, if you want the latest software bug correction, a one year web site would not bring a smile to your face. It would also be a good idea that the name of the person responsible for the web site (the webmaster) be clearly displayed.

Fourth, there should be a counter to indicate how many people have accessed that site. This would give a chemist going to the site some feeling for the usefulness and value of the site. (Again, as these counters are subject to all sorts of problems in trying to meaningfully assess what is a real hit, such numbers are meaningful now only in a qualitative sense.

Fifth, for those sites which require registration (most of which are free) there needs to be the ability for the chemist to provide his or her own account name and password. While I understand and appreciate the need for security, I am absolutely convinced that no one will ever be able to keep track of 10-50 different account names and different passwords. In my case, I often don't even remember if I even registered at a particular web site, let alone remember where I wrote down the account name and password.

Sixth, sites which have databases (such as the table of contents of a journal, or journal article abstracts) must have a mechanism to search that information. Paging through a year of titles and authors of Applied Spectroscopy is painful. I could do that in the library with the printed versions.

Seventh, those people who create web sites should, at the least, check to see what other web sites may exist in that specific area. Having a set of incomplete and random sites associated with a particular area of chemistry is not likely to bring people back to that site. For example, the ACS Analytical Chemistry Division web site has a link to the University of Florida NMR web site, but does not have a link to the University of Akron NMR web site which is much more extensive in its content.

In summary I would say there is progress in developing valuable tools for chemists on the Internet. Besides e-mail, bulletin boards, and access to online systems, there is a growing resource of up-to-date catalogs, software bug corrections, and newsletters which make it critical for every chemist to be on the net and use it on a regular basis. At present the Internet for chemists is a very good and very inexpensive way to start to look for new information and new ideas. The Internet is becoming the starting point for searching database since one can often get a start on finding something from a free Internet source and then cross over to a fee-based service (e.g. STN and DIALOG). With all the downsizing occurring in companies, universities, and the government, being a knowledgeable Internet user could make a big difference in your job status in the future.

References

1. S. R. Heller, "TrAC/Internet Column", Trends Anal. Chem., 14 (1995) 93-94or point your browser to: http://ch.ic.ac.uk/internet/heller.html

2. Catalogs:

Aldrich - http://www.sigma.sial.com/

Argus Chemicals - http://www.textnet.it/argus/argus.html

BOC Gases - http://www.science.com.au/boc/welcome.html

Calgon - http://www.calgon.com/

ChemSOURCE - http://www.chemsource.com/

Dow Chemicals - http://www.dow.com/specialty/index.html

Eastman Chemicals - http://www.eastman.com

Fisher - http://www.fisher1.com

Fluka - http://www.sigma.sial.com/

Glas-Col Lab Products - http://www.glasco.com/storefront.html

InterLab - http://www.interlab.co.uk

Jepson-Bolton - http://www.jepbol.com/

Labglass - http://pages.prodigy.com/labglass

Mettler-Toledo - http://www.mico.mt.com/

Sigma - http://www.sigma.sial.com/

Technical Glass - http://www.csn.net/~rbihler/

TrippNT - http://www.tripnt.com/

Whatman - http://www.whatman.com/html/products.html

WWW Chemicals - http://www.chem.com

3. Instruments/Equipment:

ATI-Mattson - http://netopia.com/ati-inst/matson.htm

ATI-Unicam - http://netopia.com/ati-inst/unicam.htm

Beckman - http://www.beckman.com/

Brucker - http://www.ir_soft.brucker.de

Coherent - http://www.techexpo.com/firms/coherent.html

Dionex - http://www.dionex.com/

Finnigan - http://www.finnigan.com/

Hitachi - http://www.hii.hitachi.com/

HP - http://www-dmo.external.hp.com:80/apg/products/products.html

Jasco - http://www.jasco.co.jp/

Jeol - http://www.jeol.com

Keithley Instruments - http://keithley.com/

Kratos Analytical - http://www.kratos.com

Millipore - http://www.millipore.com/

National Instruments - http://www.natinst.com/

Nicolet - http://www.nicolet.com

Perkin Elmer - http://www.pe.com

Rheodyne Valves - http://www.rheodyne.com

SGE - http://www/sge.com/

Shimadzu - http://www.shimadzu.com

Spectracell - http://www.netaxs.com/~gerryms/spctrcll.html (Magellan 3 star site)

Teknivent - http://www1a.inetr.net/teknivent

Thermo Jarrell-Ash - http://www.tja.com

Varian - http://www.varian.com/inst/index.htm

VG Organic - http://www.vgorganic.co.uk

VG Isotech - http://www.vgorganic.co.uk/vgi/vgi.htm

Waters - http://www.waters.com/

4. Software:

Beilstein - http://www.beilstein.com

Biosym - http://www.biosym.com

CambridgeSoft - http://www.camsci.com

Chemical Concepts - http://www.vchgroup.de/cc/

ChemInnovation - http://www.cheminnovation.com

HyperChem - http://www.hyper.com

Infometrix - http://www.halcyon.com/infometrix/welcome.html

Jandel Scientific - http://www.jandel.com

Labtech - http://www/labtech.com/

Mathsoft- http://www.mathsoft.com/

SoftShell - http://www/softshell.com

Synopsys - http://synopsys.co.uk

WindowChem - http://www.windowchem.com

5. Databases:

Beilstein - http://www.beilstein.com

Chemical Concepts - http://www.vchgroup.de/cc/

Daylight - http://www.daylight.com

6. Patents:

Derwent - http://www.derwent.co.uk/

Questel-Orbit - http://www.bedrock.com/patents/

7. Societies:

ACS - http://www.acs.org (Magellan 2 star rating)

ACS/Analytical Division - http://nexus.chemistry.duq.edu/analytical/analytical.html

AOAC International - http://www.aoac.org

Electrochemical Society - http://electrochemical.org/ces/ecs/html

FACSS - http://facss.org/info.html

IUPAC - http://chemistry.rsc.org/rsc/iupac.htm

PittCon - http://www.pittcon.org

Royal Society of Chemistry - http://chemistry.rsc.org/rsc/

Society for Applied Spectroscopy - http://esther.la.asu.edu/sas/society.html

8. Newsletters/Magazines:

Chemical Week - http://www.chemicalweek.com

NetSci - http://palms.awod.com/netsci/

Labtrader - http://www.magic.mb.ca/~econolab/

New Scientist - http://www.newscientist.com

9. Bulletin Boards/Listservers

To join the Analytical Science Discussion Group contact: analysis-l@fs4.in.umist.ac.uk

Computational Chemistry - http://www.osc.edu/chemistry.html

To join the Laboratory Interface Discussion Group contact: interface-l@f4.in.umist.ac.uk

To join the Process Sampling and Analysis Group contact: process-l@fs4.in.umist.ac.uk

10. http://www.compuserve.com

11. http://www.aol.com

12. http://www.cas.org

13. http://www.dialog.com

14. No animals were harmed in the preparation of this manuscript, however a number of WWW sites were hit.

15. e-mail: gasteiger@eros.ccc.uni-erlangen.de

16. http://www.point.com

17. http://www.mckinley.com

18. http://www.ch.ic.ac.uk/

19 A. Pisanty and J. K. Labanowski, Trends Anal. Chem., 15 (1996) 53-56 or point your browser to http://www.osc.edu/chemistry.html

Comments to the paper

According to the late US President Harry Truman, "the only thing new in the world is the history you don't know". During his term as President his science advisor, Vannevar Bush, wrote in the Atlantic Monthly, what is now viewed as the first Internet hypertext paper [1]. Included in this article, along with many insightful thoughts and ideas, is the notion that as one reads a manuscript the Bush memex system will enable him to "add marginal notes and comments".

In an effort to experiment with possible "new" methods of electronic publication and communication the TrAC Internet column will, when permission is given, append reviewers comments (edited to remove extraneous information) to the Internet version of an article as part of the article. Additional comments are welcome. What follows are three sets of comments about this article: TrAC - Internet Column by Stephen R. Heller.

1. Henry Rzepa <h.rzepa@ic.ac.uk>

Regarding the number of hits for papers published

Actually, that is pretty amazing, given the average readership of 2-6 often cited as a result of CAS surveys of printed articles. Mind you, I wonder how many of the above hits were in fact robots such as Infoseek, Lycos, Harvest, etc. And can you call a hit "readership"?

Regarding computer network capacity

As I understand it, its not just the capacity, but the more complex switching required to cope with the very large number of addresses, that is bringing it to its knees. Many routers in the USA are beginning to fail because of the load of transit traffic on its way to the Pacific basin. I believe the situation is far more serious than most people realise. The move to IPv6 might solve the problem, but it is going to be painful folks! My fervent hope is that it does not kill interest in the very people who we need to improve the quality.

Regarding prime, working time, 24 hours a day...

I think the golden goose is in danger of being killed here. Its precisely such questions on listservers that turns a lot of people OFF the Internet. By and large bulletin boards where such questions dominate do not have a large subscription list.

Regarding intelligent search engine

Several software houses (MDLI spring to mind) are beginning to develop genuinely new products, such as Chime, and I anticipate such agents eventually becoming available. I first wrote about such agents in the "Internet Guide for chemists" some two years ago now!

Regarding a meaningful rating system for every chemistry site

A "best of web" selected by CHEMISTS was introduced at the ACS meeting in 1995. People can vote for the 1996 awards via http://www.ch.ic.ac.uk/infobahn/boc.html#vote

The ACS has awards for synthetic chemistry, etc, so why not prestigious awards for Internet Effort. One big prestigious award, and watch the effect it will have!

Regarding the date the site was last updated

YES YES YES. And the author or "webmaster" identified properly so they can be contacted. As for where it should be, I always try to put the most important stuff at the top. I am not certain the mod date is quite in that category.

Regarding counters

Well, careful. There are statistics, and there are lies! What we need is a simple, but widely accepted counter. Actually, I am beginning to favour "uniquely different computers", i.e. pretty much corresponding to delegates at a conference. This is not the same as "this page has been visited 245 times", but even that is better than "this site had 57,000 hits" which could simply mean that one page had 37 gifs embedded in it (by the way, our hits do not include gifs). I do favour trends however. Over a period of many months, the sites that do not constantly improve will drop off dramatically in terms of whatever counter one uses.

Regarding those sites which require registration

I endorse that enthusiastically!

Regarding those sites which have databases

I fully support keyword searches as an intrinsic feature of a Web page. Oh for keyword searches! Some of the largest commercial sites do not even offer this! You do get superb graphics, but I almost always switch these off. It is an irony is it not, that the most effort often goes into the least viewed content of a page! I might mention that advances in server technology (Hyper-G) will solve this problem. Hyper-G provides an index even if the author has not set one up. It even separates author searches from title content searches. A great product.

Regarding web sites checking to see what other web sites may exist

Don't know how to solve that one. Basically, there is no easy "natural selection" mechanism on the Internet. Sites do go extinct, but not necessarily the worst!

Can I mention one further item not discussed (and for which facts are difficult to garner). I would like a survey of those sites which have attracted funding in their preparation or subsequently. I am aware of several Web sites that have been successful in attracting funding. However, in general, Web sites are not yet seen as productive in these terms (and hence popular with chairmen of departments). Yet grants ARE being written, funds are arriving, and eventually this WILL reflect in quality, permanence etc. Look at Lycos (now commercial), Yahoo (now commercial) and so on. I also know of at least two "Internet providers" who focus on the chemical industry, with the intent of also providing quality and permanence.

Finally, the issue of feedback on articles. Well Steve, all my feedback occurs as a result of refereeing these things for you. Somehow, once its published, its difficult to repeat it all. So if you wish, why not publish all our various comments now (mildly edited).

2. Steven Bachrach <smb@smb.chem.niu.edu>

The article is an interesting, personal perspective of the internet circa 1996. While I believe that Heller is correct on many issues, I am not so pessimistic as he is, am not so sure that his suggestions are the "best" way to see improvement on the Web.

First off, while the general content of the Web is clearly lacking, there has been significant growth and development of Internet tools that give much hope for the future. Many database providers are now incorporating the Web as the means for designing their front ends. There are many very useful educational sites on the Web, etc.

"(How many descriptions of analytical chemistry departments, their instrumentation, and their latest publications do you need?)" Well, if you are looking for graduate schools to attend, this info would be quite useful and if you are looking to attract students, you want this out there!

I am not so sure what to make of the hits numbers, and am not convinced what purpose they would serve if more of them were made public. For example, the Playboy web site has more hits per day than (probably) all the chemistry sites put together! The fact that Rzepa's paper has more hits than Bachrach's means what? Henry's paper is "better" than Bachrach's? Since more hits are there, that means I should read it? (Or perhaps, I should access Bachrach's paper 100 times today and drive up those number?)

I think that Heller is also overstating the importance of e-mail relative to other Internet services. E-mail is critical and will be influential over the long haul, but the Web is the future (well, maybe its here already - there are more www packets being moved than e-mail today!)

On the other hand, I completely agree with the need for a review apparatus. The McKinley rankings just stink!

So, what to do with the paper - publish it, but put in some way to directly allow comments to be sent off, and then attached to the paper. Start a www based discussion!

3. Ray Dessy <RDESSY@chemserver.chem.vt.edu>

Provocative material! It is nice to see some pessimism among all the hype concerning the Web. In browsing the electronic world in subjects ranging from encryption technology, through music, to philosophy one is struck by the fact that the quality and veracity differs greatly. With books I have at least the knowledge that, except for vanity press publications, someone (or many) reviewed the material. With the Net I know that some materials have never been reviewed or even proof-read. The danger is that you are dealing with a media that can impact the learning process in a negative direction if the readers are not discriminant. And many of the users do not have the experience to perform that function.

But quis custodiet ipsos custodes? And is "counting" really a good indicator? I recall an interview that I had at Texas A&M when Clarence Zener (of the diode) was A&S Dean. He showed me that my citation index rating was better than most of his Chemistry Department. My response was that I had a colleague whose rating was higher yet, merely because he had written a review article that was easier to cite than seeking the original literature. In that sense, I have on my desk a copy of "The Gutenberg Elegies" by Sven Birkerts (Faber and Faber, Boston 1994). It deals with the fate of reading in the electronic age. Its circulation will be small, but it is a very fine piece of craftsmanship, and an erudite view of the consequences of using a medium incorrectly. He points out that the electronic web may be fine for "naturwissenschaften" but very poor for "geisteswissenschaften"- natural sciences vs. cultural matters.

I am struck, and deeply concerned, that many people who tout the new media have little understanding of the cognitive psychologies involved.

You might also try asking questions of corporations who have installed 20-40,000 seats of groupware Intranet software about their experiences. The difficulty will be that their public responses will be different to their private responses. They have all experienced a less than satisfactory melding of minds due to the factors that you so nicely bring out. The technological problems may be solved, but I am very concerned about the human factors.

Reference

1. V. Bush, "As We May Think" (http://www.theAtlantic.com/atlantic/atlweb/flashbks/computer/Bushf.htm)

Would you like to add your comments to the above? If so, please respond to the Column Editor directly: srheller@gig.usda.gov