Invalid Login: Don't Explain Why

Most web sites that require a login of username and password return an ambiguous error message when login is invalid.

The common message is: "Invalid username / OR password."

Well, which was it?

It's very easy for the program to give you more specific information. Here's what happens when you login:

The program searches the database for a record containing that username AND password. If no record is found, then login is invalid. Most programs just spit out that message "Invalid username OR password" because one of those 2 was the culprit. (Actually, both inputs could have been wrong)

To clarify the matter, the program only needs to do an additional call to the database, looking for a record of the username.

If it finds a record with that username, then it knows that the password was wrong.

If it does not find a record with that username, then both the name and password were wrong.

Then the system could return a specific response.

For examples of LAZY programming / ambiguous response, see Yahoo and Hotmail. (Hotmail screen is atop this post, here is Yahoo)


Google does a better job on this, by giving the message "Username and password do not match. (and then it displays what name you entered)"

But it's easy to offer 2 specific responses when login is invalid.
1. If both the name and password is wrong: "There is no record of a xxxxxxx"
2. If there is a user of that name: "The password you entered does not match..."

The only reason to withhold those details is as a privacy measure. Since most usernames are now equal to an email address, if any person in the world wanted to know if a given email address was a member of a given site, he could enter it and a random password, and then receive a yes or no.

In that light, the Google method is tops.

But anyway, here are 2 screens from a system I made:

Undisclosed at LinkedIn

LinkedIn, a middling dowdy professional network web site, suffers from a number of shortcomings in its user interface. Today we address the issue of concealing your present employer. LinkedIn does not provide a systematic way to hide your present employer's name. Instead, you are left to use your own words to denote anonymity into the company name field. Some members choose "Undisclosed;" others, "N/A" or similar verbage. Thus, when you log in, and are informed of possible connections, you might see a message such as "50 new members at Undisclosed" because LinkedIn thinks that "Undisclosed" is a company!

All that is required is a setting, a switch, to show or hide the name of present company from Public and your network. HotJobs and Monster give members the option of such a screen.

Google Wises Up: Link to iGoogle

Google has updated its portal, there is now a new link "Back to iGoogle." As I noted in earlier post, the only way to get to iGoogle was by clicking on "Web" which also means 'web search results.'

In that earlier post, I describe how a simple search can produce errors and confusion. As well as Google has performed, it has probably lagged due to poor search user interface. If Google makes searching and navigation consistent, and does not generate errors, then it will enable an even larger share of searches. Meaning, more moolah.

Google Kind of Wants Us to Use iGoogle

Google won cheers this summer by adding a set of context links to its search page, but due to a major flaw--a single Link that does 2 different things, an omitted Link, lack of Navigation Labels, the user is often stranded and confused.

At issue is iGoogle, which can serve as a homepage to house short cuts, collections of links, and RSS feeds. I like iGoogle. I have added some content to it. I try to visit when I can... but how do I get there? Hmmm. Well, I just had a hunch, and typed in "igoogle.com," which worked!

Once you login to Google.com, you can use the link in the top right corner to go to "iGoogle."
Terrfic. I see my homepage. I also see a row of links on the left corner:
Web (unlinked, because I am there) | Images | Video | News | Maps | Gmail More (drop-down list of other services)

So, I have gathered that "Web" equals where I am.. And that's the error Google made: it denotes "Web" to mean 2 things: iGoogle, but this same link also means a type of Search (regular web search as opposed to those other types of Searches:images, video,maps). This is both unintuitive and leads to inconsistent results.

Here's is a typical sequence of events:
1. I've gotten to iGoogle.
2. I click 'Images'.
3. It leads to Image Search.
4. I choose not to Search. I click "Web" to return to iGoogle (the only link to iGoogle).
5. I click Books from the "More" drop-down list.
6. From the menu of Book categories, I choose "Linguistics."
7. This auto-populates the Search Input field with 'subject:"Linguistics"'
8. I want to return to iGoogle, so I click "Web."
9. This brings up an Empty Search Results page, as if I had meant to do a web search for "Subject:Linguistics." (this will always bring up an empty search because, as the page notes: subject:" Linguistics " was dropped from your search because it is not supported for this type of search.)

** See Bottom of this Post, to see Screen Shots of the Above Sequence **

That is not a good experience, and it is due to poor design.

The heart of the problem is failure to resolve a common issue wrested with by web site designers: should links be Verbs or Nouns or Adjectives? It is important to be consistent at least within a given group of links? Usually, a designer thinks through the logical ramifications, and chooses One and Sticks to it. Google, although guilty of omitting a heading for its disparate groups of links, does offer consistent logic/English in its Left top corner grouping IF VIEWED IN ONE CONTEXT.

But, amazingly, Google gives the word "Web" two meanings: a Place (iGoogle) and a type of Search, and it uses the very same Link to enforce both meanings. A dedicated link to "iGoogle" would go a long way to address this (the company has bought up half of the Carolinas to house its new data centers--but it doesn't have 40 pixels of space for an iGoogle link? )

If you look further at that Left Corner group of Links, just within the context of Search Type, it is not actually sensible. What is Google implying by these words?
Does "Web" imply "web pages?" Or is it not the 'parent' of the other search types?
The word used for each link within a group should be able to fit into One commonly structured sentence, which need not be displayed, but can be inferred from the group heading. At Google Left Corner, what would that sentence be?
"Search the Web for Web ".. no. What if we use them each as adjectives for the word 'search?'
"Web search".... "Images search" .... Wait! Pluralize image? Is that consistent? Is that hip? (Aren't only older citizens charged with inappropriate pluralizing of words?) Maybe the unstated sentence formula is: "Search All (of the world's known) X for ____". That would make sense, because "Video" does not need to take the "s" to become plural. But again, that word "Web" doesn't fit.

This brings up the need for a Group Label. It is important for designers to Label things not only because it helps the user, but it then helps the designer by keeping him mindful of what he is doing/putting together.

The Internet is still new, and different organizations and users have different ideas and views. This is accepted by all, and thus, we expect and even embrace different arrangements, viewpoints--so long as the viewpoint is Clear and consistent.

Google has built up heaps of goodwill. Most people defer to it, and will give it a break. If it wants to organize a set of Links, or Methods to view content, in a given way, then users will follow the way.. The problem is, the Way is not Understood.

There is a limit to how much can be Implied, rather than expressed, in web site navigation. There is a lot of information and communication made available by Google: it enables self-expression and labeling/tagging, and it itself develops and rolls out a lot of content. Further more, Google provides a lot of FAQs, and How-tos throughout its services, such as this Blogger. For some reason, they let things slide on the single most visited page that contains the single most important function.

Google's stated mission is to organize the world's information. But in the rush to do so, it hasn't properly organized itself.

1. from iGoogle, went to Image Search

2. Used the Left Corner "Web" link to return to iGoogle. No other link visible.

3. Used the "More" list to go to "Book Search"

4. Clicked a category from the Left, which auto-filled the search input field.

6. Clicked "Web" to return to iGoogle. But, in this case, since the Form was auto-filled, and since Google took it to be a Search Request, it brought up a No Search Results page. (and always will, because of the syntax of Book search is not recognized by Web Search).

Pandora: How Long Can We Sing this Song?

Pandora is a wonderful service. If you haven't already begun using this nifty Music Genome based personal radio player, you ought to.

Today I point out one glaring problem with the player :: it does not list the running time of the songs it plays. This information could appear within the individual song box which contains the album art, song title, album title and artist. Or it could appear in the Music controller, which is just a Play / Pause toggle button. Since most media players include total time, as well as elapsed AND time remaining, Pandora would be wise to do so. Hec, even to just mimic iTune's interface.


Another omission is Year of Release. It would be nice to know the year when each song was produced or released. For select songs, it appears in a Detail Page. But for some reason it doesn't appear for most songs. And it never appears in the song's box (above). Pandora invites users to "Guide Us." But, Pandora could do a little more guiding itself.

Talkin Wall Street Journal Blues

The WSJ Online has long been cheered for its ability to charge and collect subscription fees, and for its heaps of information. I'm glad to fork over $10 per month. The site refreshes and reloads every few minutes with breaking news that often directly effects my investments in Big Business.
There are a couple Interface Issues to call out::

1. iFrame on homepage doesn't Auto-Reload. While the homepage refreshes itself--and timestamps itself, its right column iFrame, which contains each user's "My Journal," a personalized collection of links and news, does not. My "My Journal", lists certain stock prices. When I see the time stamp on the homepage, and then look at my stock list, I am inclined to assume I have the latest prices. The iFrame should reload. Also, the list of stocks should be Time Stamped as well. No reason this important piece of info should not be included atop EACH quote.
Example: "price as of 3:24pm"
Some other news sources just give the vague, relative "up to 15 minute delay." Useless. Just give me the time at which that info was collected so I can compare it to present:
Example: "Time is 3:24pm EST. Price as of 3:18pm EST"

The WSJ does give the time stamp when you search a company's quote. Just not in the "My Online Journal" portfolios.

**
The second 'issue' regards the online journal using the exact copy as the print version. It is common for the Editorial Board to refer to an accompanying column. They accomplish this, in print, with a parenthetical "See Above" or "See Below." Example: ("for more on why lower taxes are always best, see Mitch Wagner's piece above")

Ummm, hello? On the Internet, that is exactly the kind of reference that could be LINKED. The accompanying piece is not "above" or "below"--it simply is where it is, and can be goten to through a hyper link. By leaving it as is, the Online Journal turns the paper's helpful hint into a confusing challenge. Perhaps the Journal could hire someone to vet the articles for possible hyperlinks.

**

The third problem is a style 'issue.' At the top of each page, there is a link to see a Print Friendly version. This helpful feature is cleverly sponsored by "Toshiba" which apparently is hoping to make inroads in the printing 'space.' (Hey, good for them. Why desert an entire industry to HP just because HP is kicking butt. DONT give up!) Congrats to the WSJ account exec who brought in that deal... Anyhoo: the problem with the Print Version of a Page is that it needlessly includes some Elements that SHOULD NOT PRINT!!

Prime Example: Video Player. Many articles have an embedded Bright Cove video player to play a clip related to the 'piece.' If the user aims to Print the Piece, then she does not need to see a picture of a video player. This element should not be displayed.



Maybe the new owner of the WSJ will demand a new style sheet. Or, maybe the web team will insist on complete Independence from Ownership, as a condition for signing onto any deal. Yeah, that seems fair-- a CSS oversight committee.

Navigate Blogs via Labels

Imagine if your favorite website was only navigable via the Dates that the pages were updated. So, if you were at NY times, you wouldn't see 'Sports', 'Business'... you would see 'April 5, 9:30am', 'April 5, 9:42am'..... Ludicrous, right? And yet most bloggers leave a Year – Month archive as the only way for readers to find content (past posts).

Blogger now allows its bloggers to insert a module that displays "labels", the descriptive keywords that the blogger attaches* to his post. I just added it to this blog, this past week. Labels are more relevant to a reader's aims than just the Numbers which are Dates. Sure, a date can imply content. A December 06 post is a likely spot to find rambling about, oh, Christmas... snow. But there is no need to limit the imagination and interests of our dear readers! Labels aka tags (I'll write a post about those synonyms and various syntax of same Friday) cut right to the subject.

Hopefully, the Tag Module will catch on. Today, a random survey of 15 blogs found just 2 which offered Label Navigation. Blogger should promote it upon Login. Heck, otherwise, the blogger only finds it by clicking on Template and browsing the possible Page Elements. And heck, as long as I'm in 'tell Blogger what to do' mode, add this:

* Let Readers label posts, as well as the author. When readers come to a blog with Label Navigation, they would be able to view, together, both the author and other readers' labels.

Help Me, Blogger (Yeah)

When logged in users browse BlogSpot (which displays published view of users’ blogs), the site’s Header links to the user’s Dashboard, and to Sign Out. That is helpful. So, if you are reading someone else’s blog, you can jump right to your control panel aka ‘dashboard’ which lists and links to edit each of your blogs and profile. Here’s a screenshot:


But if you are viewing a BlogSpot page that you own (or to which you contribute), the Header links to “New Post,” “Customize,” and Sign Out. A link to Dashboard would be desirable here. Also, instead of, or next to, “New Post” there should be a link to “Edit Posts.” It is this link, which exists elsewhere in your dashboard, that displays the titles and previews and the # of comments of each of your posts, and has a button to ‘add new.’ A more suitable term is: ‘Manage Posts’ which conveys that you can edit/view existing posts as well as make an addition (The term 'Mangage: Links' exists on Dashboard).


Even if you contribute to just one blog, it would be convenient for the Header of the your blogspot page to link to ‘Manage Posts.’ And if you contribute to more than one blog, the link to ‘Dashboard’ would be of great help