Before the
ARCHITECTURAL AND TRANSPORTATION BARRIERS COMPLIANCE
BOARD
WASHINGTON, DC 20004-1111
In the Matter of
Electronic and Information Technology Accessibility
Standards
Section 508 of the Rehabilitation Act
Amendments
Notice of Proposed Rulemaking
36 CFR Part 1194
Docket No. 2000-01
Office of Technical and Information Services
Architectural and Transportation Barriers Compliance Board
1331 F Street, NW Suite 1000
Washington, DC 20004-1111
section508nprm@access-board.gov
Comments of Trace R & D Center
June 8, 2000
Gregg C Vanderheiden Ph.D.
Professor, Industrial Engineering
Director Trace R & D Center
University of Wisconsin - Madison
2107 Engineering Centers Bldg.
1550 Engineering Dr.
Madison, WI 53706
These comments are submitted by the Trace Research and Development Center, a Rehabilitation Engineering Research Center (RERC) at the University of Wisconsin-Madison, funded by the National Institute on Disability and Rehabilitation Research (NIDRR) of the US Dept of Education. The core focus of the Trace Center RERC is on access to information technologies. The Trace Center is also part of a joint RERC with Gallaudet University on Access to Telecommunications.
The Trace Center has worked in the area of disability access to standard information and telecommunication systems for the last twenty years. Early work focused on the development of access features (e.g., StickyKeys, MouseKeys), and work with Microsoft, Apple Computer, IBM, and others to help them incorporate the features in their standard operating systems (e.g., Windows, Mac OS, etc.). Recent work has included the development of a set of techniques for cross-disability access (called "EZ Access"), the development of a reference design to illustrate how a cell phone could be compliant with the FCC's accessibility regulations at a reasonable cost using today's technology, and co-authorship of the W3C's Web Content Accessibility Guidelines.
The opinions stated herein are those of the author and do not necessarily reflect the views of the funding agencies of the Trace Center or the University of Wisconsin.
We want to compliment the Access Board's work done so far on the guidelines. Although the task of creating the 508 standards appears to be complex, it is even harder when you actually try to sit down and figure out how to create a set of rules that will be effective today, make sense tomorrow and be applied across a very wide and continually changing range of technologies. We apologize for being late in turning in the comments, but it took much longer than we anticipated to formulate solutions to address the issues we identified without creating new problems.
We would like to compliment the Access Board for the following actions:
These represented a major step forward in the guidelines. These changes also help to highlight key issues that needed to be addressed.
In reviewing the guidelines, we identified a number of issues. Some of these originated with the Electronic and Information Technology Access Advisory Committee (EITAAC)document, while others are side effects of the improvements that were made. Some of these problems only became obvious recently as it became clearer how these guidelines need to be implemented within the federal procurement process and the constraints that this process places on the government. These, in turn, effect requirements on the accessibility standards themselves.
We have organized these comments into 4 parts
1. An over-generalization and under-specification of the (important) compatibility enhancement that was made to the guidelines.
One of the major advances introduced by the Access Board was the addition of the "compatibility with assistive technology " concept to the guidelines. The use of assistive technology, particularly when it comes to personal workstation computers, is an important strategy for providing accessibility. In fact, for more complex applications, compatibility with a screen reader generally provides better accessibility than if each application were to build or attempt to build its own separate accessible interface. Although this may or may not be true in the future, it is certainly true today and should be allowed and encouraged as an access strategy.
This approach is not without precedent. Buildings today are constructed and considered accessible to people in wheelchairs if they are usable by the people with their wheelchairs. There is no attempt made to require that people with wheelchairs be able to access and use the buildings while leaving their wheelchairs at the door.
As currently written, however, there is no differentiation in NPRM between environments where assistive technology would be present and environments or products where it is not practical to assume that the users will have assistive technology available to them. A requirement is made for kiosks and Information Transaction Machines (ITMs) that assistive technology should not be required for use of this class of device. However, there are many other classes of devices where assistive technology may not be present. In such cases allowing products to comply with the requirements through compatibility with assistive technology would not provide any real accessibility. In addition, for some disability groups, even ITMs will not be reasonably accessible without the presence of some assistive technology. For example, access to ITMs by individuals who are deaf/blind has already been demonstrated using a device such as a BrailleLite or Braille Note. Soon the capability to remotely interact with devices using an assistive technology will be standard. Yet, it is unlikely that it would be practical to build braille displays into ATMs or to require that they be usable by individuals with both profound hearing and vision impairments without their assistive technology.
Other examples include personal computers used as building directories, for registration, for card catalogs in libraries and other applications where users are not allowed to load their own software onto the products, and where the screen reader that an individual is familiar with cannot be assumed to be on the product. In these cases, making software compatible with, for example, screen readers should not be equated with accessibility.
Thus compatibility is a very powerful and useful concept that needs to be applied more broadly to the guidelines. As it is currently worded however it will create situations where it is applied where it should not be, and other situations where it is not allowed where it should be allowed.
2. The deletion of any specific regulations for providing access for people who are blind to any E&IT other than personal computers and the deletion of any specific regulations for people who are deaf to any E&IT other than telecommunications.
In compiling the NPRM, the Access Board took both the component specific-guidelines and all the specific items out from under EITAAC's performance guidelines and used them to construct the general and component requirements in the NPRM. Two key specific requirements, however, were omitted in this process. Specifically:
a)The guideline requiring that visual information should be provided in auditory form and tactile when appropriate and
b)The requirement that auditory information be provided in a visual fashion.
As a result, there are no specific requirements of any kind for providing access to E&IT overall for people who are deaf or people who are blind. There are extensive provisions for compatibility with assistive technology that would provide access for people who are blind to personal computers. There are also provisions allowing individuals who are deaf to use their assistive technologies with telecommunication products. However, E&IT is much broader than these two categories of devices. Although these categories have been two primary points of concern in the past, there are many more individuals with both types of disabilities who will be facing many additional barriers with other devices in the workplace.
Relying on the functional performance guidelines alone is not sufficient. Although they would require companies to make all of their products accessible to these two groups, they provide no guidance as to how this might be done or what would be considered sufficient. They also provide nothing that would be measurable or verifiable. In the itemized review below and in our proposed reorganization, we explore this problem and propose strategies for addressing it.
3. An imbalance between focus on providing access to the way products have been designed in the past and the way they will be designed in the future, which will lead to problems in interpreting and applying the guidelines within the next 18 months and moving forward.
The current guidelines are very good at providing access to personal computers for people who are blind and access to telecommunications for people who are deaf or have hearing impairments. They also work well for products that are designed the way personal computers and devices have been designed to date. However, they do not do apply well against newer technologies that are evolving. In some cases, this will result in unduly constraining industry. In other cases it will cause it to be ambiguous as to what is required or what would be a sufficient solution to the problem/requirements.
4. Problems created with the specific wording of some of the provisions which make it difficult to meet the different access needs of personal workstations versus public access or shared E&IT.
In a number of places, the way the items are phrased makes an assumption as to whether the technology would be a personal device or a public device. Some aspects of this problem were touched upon in the discussion above. For example, the discussion of compatibility.
Whether something is a personal workstation device or is a shared or public device is a key issue for accessibility, particularly as information technologies are dispersed into the environment and both employees and the public will encounter them in places other than their personal workstations. In the NPRM, an assumption was made that access to public kiosks and ITMs should not require the use of assistive technologies. This exception does not apply to any other E&IT. This poses two problems
The first problem here is that "kiosks and ITMs" is too narrow a definition for public or shared technologies. The second problem is that, for some individuals, access to public systems will need to be via assistive technologies. (For example, individuals who are deaf/blind.) In addition, access to products in one's personal workstation area can often be done through assistive technologies today but may sometimes need to be done through built in features. As mentioned previously, some of these issues were not generated by the Access Board but were inherited from the EITAAC.
5. Issues surrounding definitions and use of terms for which definitions are not provided.
Finally, there are a number of issues around definitions. In some cases a key term is used but not defined. In other cases there are words that are used two different ways in the guidelines in ways that could cause confusion. In these cases the definition needs to be changed or the term used in the guidelines should be differentiated.
For example, "fundamental alteration" can be used as a defense for not following the guidelines. But that term is not defined anywhere in the regulations. It is important that this term be defined since people have widely varying interpretations of the term. It is generally agreed that adding access shouldn't change the basic operation of the product in a fundamental way. Some interpret this, however, to mean that you shouldn't add voice, for example, to a product even if it wouldn't significantly affect the cost, size, shape, or battery life. Since added functionality and added output modes are the essence of how you make products accessible, a clearer definition is needed or adding access features themselves could be considered fundamental alteration, even if it did not have any significant effect on the way the standard mode of operation or its size, weight etc. This is discussed in more detail in the line-by-line discussion.
Another example is the term "alternative modes." The word 'mode" is used both here when discussing alternative ways of getting information to consumers as part of the support process and also in performance guidelines where it says that "products must be accessible through at least one mode". In one case, fax is a legitimate alternative mode. In another case, faxing information is not. Some clarification is needed here.
The definition of "TTY" also needs an adjustment. As currently written, industry could be responsible for providing compatibility to a wide variety of different signal formats and communication and protocols.
This and other related problems can be addressed through changes to the wording of items as suggested below.
6. Need for specificity to facilitate implementation
In recent discussions with those who will actually be implementing these guidelines (both in industry and the federal government), we have also learned a number of things that we were not aware of at the time of the original EITAAC meetings. In particular, the need for the guidelines to be more objective and either testable or verifiable has been brought to light. Many of us expected that once the guidelines were promulgated, simpler checklists and other similar mechanisms could be used at the purchasing level to make the process more concrete for individual technologies and to account for changes in what was available on the market. Subsequent discussions have indicated that, due to the nature of federal regulations and the FAR, it would not be possible for agencies or others to create checklists which could be used by either companies or purchasing agents to "interpret" the compliance with the guidelines unless such checklists went through the same process of approval as the guidelines themselves. No one is evidently empowered to create a document that would stand in front of the FAR regulations as an interpretation of compliance.
This has led to an understanding that we are likely to need guidelines that are as specific as possible. Procurement agents need something that is specific enough that they can have a chance of monitoring for compliance. Perhaps, more importantly, industry will have a set of criteria which are specific and objective enough that they can tell where they comply and have some confidence that others will concur. This is important to industry for two reasons. First, a vendor does not want to sign his or her name to something saying they comply only to be told later that their interpretation of compliance is not reasonable and that they, in fact, do not comply. Quality companies do not want to be left in such a situation. Secondly, vendors do not want to be in a position where they win a bid, but might later lose the contract due to a challenge by another company based on 508 compliance. Finally we have seen that industry is justifiably concerned about wasting a lot of time, effort, and money, wrestling with the question of what compliance means, instead of being able to focus those efforts on having their design team build the necessary flexibility and functionality into their products.
Specificity, where it can be done without compromising companies' options and ability to innovate, greatly facilitates companies ability to move forward confidently with their designs. It is also critical to any testability or verifiability measures by all parties.
7. Industry's need for flexibility and cost containment
At the same time that both government and industry need clarity in the guidelines, companies also need the guidelines to be as flexible as possible to accommodate innovation and to be able to address access in a fashion that best fits into the particular design parameters of their products. While industry is concerned about under-specification that leads to ambiguity, it is also very concerned about over-specification that might limit their ability to innovate and introduce new types, combinations and form factors (packages) for their products. They are also concerned about minimizing the manufacturing cost to incorporate the access features and any impact on the default or standard user interface and usability of their products.
8. Hybrid Approach
In order to try to address these often-conflicting needs while still maintaining a set of guidelines which will provide meaningful access for people with disabilities, we found that one is unable to take either a "pure product design" guidelines or a "pure functional" requirement approach. A pure functional requirement approach would yield guidelines that are short and simple but very ambiguous. They give industry maximum flexibility, but provide no guidance to industry or the government agencies as to what is expected, what works, or what is sufficient.
Therefore, in putting together our comments, we have built on the NPRM guidelines, working on extending the hybrid approach, while addressing the gaps or problems seen in the EITAAC/NPRM guidelines in order to better addresses the need for both objective yet flexible guidelines.
At the same time we are sharing this proposed reorganization of the guidelines with the Access Board, we are sharing it with our colleagues in industry and consumer groups. Although our comments are already based on their input, we seek additional thoughts and reaction from them to this proposed approach as part of our ongoing research efforts on accessibility and guidelines. As we receive their feedback, we will forward those comments to the Access Board. We have also encouraged people to respond directly to the Access Board if they like.
In order to facilitate compilation and correlation of the comments, we present the bulk of our individual comments on the current 508 NPRM in this section. Our comments are presented in the order in which the items appear in the NPRM.
NOTE: To facilitate moving between sections, each item will begin with the letters NPRM: followed by identifying text for that item. The comment then follows and begins with COMMENT:
NPRM: 1194.1 Purpose
No comment
NPRM: 1194.2 Application
Paragraph a - No comment
Paragraph b - No comment
Paragraph c - No comment
Paragraph d - comments on paragraph d included in the following
general comments
COMMENT: The NPRM stated that these provisions laid out in the NPRM were based upon the EITAAC report with certain items being rearranged and renamed. However, this doesn't reflect the fact that rearranging these provisions, in some cases, substantially changed their scope and applicability. Some provisions that were originally intended to apply to all E&IT had been changed to apply to only a particular subcategory of E&IT.
In general, we agree with a restriction for many of the provisions since guaranteeing their applicability across all E&IT may not be possible. As long as there are strong functional performance criteria to cover all E&IT, we feel this strategy of the Access Board will work.
We were troubled, however, by language that appeared in the NPRM which stated:
"Paragraph (D) explains how each section of this rule is to be applied. In general, the requirements in 1194.21, 1194.23, and 1194.25 are assumed to satisfy the functional performance criteria in 1194.27."
Since many of the specific requirements in the EITAAC final report were removed from the NPRM or made specific to only certain technologies, this language is troubling. For example, there are no general or component specific provisions for enhancing audio in sections 1194.21, 23, 25. Similarly, there are no provisions to prevent the use of controls that require simultaneous activation as long as it could be done on one hand. There are no provisions in sections 1194.21, 23, 25 that, if followed, would result in E&IT products, other than personal computers, to be accessible to for individuals who are blind. Provisions that would address these access issues are only found in section 1194.27, the functional performance criteria. If products complying with sections 1194.21, 23, and 25 are assumed to satisfy the functional performance criteria in 1194.27", as stated above, then it is possible to comply with 508 without having any real provision for hearing access outside of telecommunications or blindness access outside of personal computers.
It is imperative that this language be changed in the final rule so it is clear that the provisions in 1194.21, 1194.23, and 1194.25 are not meant to provide specific criterion which would necessarily satisfy the functional performance criteria. Rather it is suggested that language can be found elsewhere in the NPRM be used to describe the relationship. This language says:
"Therefore, when evaluating the compliance for any product, first look to the compliance with 1194.21, 1194.23, and 1194.25, then apply the performance criteria in 1194.27 to elements or technologies not covered in those sections and to the overall product functions."
Another remark that is confusing is the statement.
"Where there is overlap, the specific provisions in 1194.21, 1194. 23, and 1194.25, prevail over the general provisions in 1194.27."
After just stating that they should review 1194.21, 1194. 23, and 1194.25 and then apply the functional criterion as the final overall test, this statement says that 1194.21, 1194. 23, and 1194.25 override the functional criterion. If one looks carefully at the requirements in 1194.21, 1194.23, and 1194.25, one can see that they do not completely cover any of the disabilities or functionality covered in the Functional Performance items. This language might have made sense in the old EITAAC version because in the EITAAC report there were specific provisions in the Functional Performance section. But it does not make sense within the current NPRM. While 1194.21, 1194. 23, and 1194.25 do give specific means for achieving some aspects of the Functional Performance Criterion, (in which case there is no need for them to override), we do not see how a specific provision to achieve a particular function should override the general provision that the product be usable as a whole to the individual.
RECOMMENDATION:
It is recommended that the conflicting language be removed from the discussion and that a single set of instructions be provided which instructs the individuals to first apply the specific measures that apply to their technology and its components and then to use the functional performance criteria to evaluate the overall product accessibility.
SPECIFICALLY:
If it is felt that the specific guidelines conflict with the functional then one or the other is written improperly (or specific requirements are embedded in a functional performance criterion). A specific guideline should always fulfill or partially fulfill one or more of the functional guidelines. And there should be no specific design criteria in the Functional Performance Guidelines to override. Finally, a specific design spec should not override the overall functional goal of making a product accessible.
It is further suggested that the instructions for interpreting the provisions be included more clearly directly in 1190.2 The questions we have heard most often relate to confusion surrounding how to apply the different sections of the provisions and how they relate to each other.
NPRM: 1194.3 General Exceptions - Paragraphs A, B, C, D
No comment
NPRM: Paragraph E -Fundamental Alteration
COMMENT: The NPRM states that compliance "does not require a fundamental alteration in the nature of a product or its components." However, the term fundamental alteration does not appear in the list of definitions. In discussions around accessibility the term "fundamental alteration" often comes up. Whereas there is general agreement that including accessibility should not require the product to be fundamentally altered, there is a very wide range of interpretation as to what "fundamental alteration" means.
Some define a "fundamental alteration" as being a modification that fundamentally alters the ability of the product to function as it was originally intended. For example, as pointed out in the NPRM, adding a large display to a pager so that it would no longer fit in a pocket would change the way the product was intended to operate. Similarly modifying a free hand sketching program so that it was vector based so that it could be fully operated with a keyboard might fundamentally alter that application.
Others, however, have defined fundamental alteration as anything that changes the basic character of the product. For example, they consider adding voice to a product which did not previously speak to be a fundamental alteration, even if the addition of voice did not increase the product's size or significantly alter its price or battery life. Because the product did not previously speak and would now have speech, they argue, it should be considered to be fundamentally altered - even if the device continues to operate as it did before the addition of voice output.
Applying the latter interpretation of fundamental alteration could lead to many or most accessibility features being declared fundamental alterations.
Such an important concept should not go without definition.
RECOMMENDATION: It is recommended that a definition of fundamental alteration be included with wording similar to the following.
"A fundamental alteration is a change to a product which causes the product to be unable to carry out its primary mission or function or prevents the product from operating in its original manner. Adding an alternative presentation mode or operating mode shall not in itself constitute a fundamental alteration unless it meaningfully interferes with the original functionality, features, or operating modes of the product."
NPRM: Section 1194.5 Equivalent Facilitation
COMMENT 1: The idea of equivalent facilitation is a very important component of the 508 standards. It is essential in order to allow industry to innovate as well as to develop more effective accessibility strategies as new technologies evolve.
COMMENT 2: There is some ambiguity in the NPRM text accompanying the example of equivalent facilitation. Is the example trying to say that having the information available by phone is an "equivalent facilitation" and acceptable alternative to making the kiosk accessible? If so, then it should be stated explicitly that the phone access to the same information available from the kiosk would have to be available in the same location as the kiosk, be free, provide all the same information and be available during the same hours the kiosk is available.
RATIONALE: If the information is available by phone to everyone - and they still see a value in paying to put a kiosk out in public, then there is also some purpose or value to having access to that information in that location. As a result, the person with a disability should have access to that information in that location as well. Being in the same location and being free, (not a pay phone call) would be especially important for kiosks serving as building directories, registration kiosks, etc. The text in the NPRM should also note that the information available by phone would have to be complete and as up-to-date as the information in the kiosk.
RECOMMENDATION: It is recommended that the text accompanying the discussion of equivalent facilitation include text such as the following
An equivalent facilitation would at least have to provide access to the exact same information (same completeness and freshness), same services or functions, in the same location, during the same hours and at the same cost as the device it is acting as an equivalent for.
If the Access Board does not want to list these specifics, it should as least specify that all these things were true for the telephone example. Otherwise, it implies that the phone access could be an equivalent facilitation without meeting these conditions.
If the NPRM is suggesting that a telephone-like handset could be added to a kiosk along with speech output from the kiosk - then that should be more clearly stated. The use of the words 'touch tone' make it sound like it must be a telephone, since that is the only place you find touch tones. If it is not meant to refer to telephone access to a distant computer then it might be changed from:
ORIGINAL TEXT:
"For example, an information kiosk which is not accessible to a person who is blind might be made accessible by having a telephone handset that connects to a computer that responds to touch-tone commands and delivers the same information audibly."
EDITED TEXT:
"For example, an information kiosk which is not accessible to a person who is blind might be made accessible by having a telephone handset that connects to the kiosk's computer and is operated from switches or a keypad on the computer and delivers the same information audibly."
COMMENT 3: A logical question that always comes up when equivalent facilitation is raised is "who decides whether something is a viable 'equivalent facilitation'?" The Access Board may wish to a comment explaining the process of evaluating equivalent facilitation.
NPRM: Definition of Accessible
Accessible. Electronic and information technology which complies with the requirements of this part.
COMMENT: It is common in accessibility guidelines of this type to define products as "accessible" if they comply with the guidelines. If this is done, however, great care must be taken in how the guidelines are crafted. As the functional performance criteria are currently worded in the NPRM, accessibility should not be allowed to be equated with compliance. This problem is fixable, and we later suggest how the wording of the provisions could be altered to correct this problem (in the functional performance criteria section.)
The problem in the existing NPRM language stems from the way compatibility with assistive technologies is tied into the language. Before we discuss the problem with the way that compatibility is handled in the current NPRM, let us repeat that we think the incorporation of compatibility into the provisions, in general, is a very good idea and a necessary change particularly in relation to personal workstations. Thus, we are not advocating that compatibility be removed as a functional component of accessibility for personal workstations. We think this is a valuable addition to the rules that was made by the Access Board staff and its affect needs to be retained in the final rule. The way it is incorporated into the wording of the provisions, however, needs to be adjusted.
Currently, the way the wording is structured, products can be declared accessible if they are merely compatible with assistive technology. According to the current wording, this is true even if it is known that the individual who will be encountering the product would not have the necessary assistive technology with them or would be barred from using it. There is no qualification in the rule requiring that the assistive technology actually be present or that there is any expectation of its being present before compatibility would be acceptable for compliance. For example, consider an (AT compatible) computer purchased or rented to serve as a shared email station at conferences and events where the AT needed to make it accessible will not be installed. This computer would compatible but would not be accessible or useable by anyone at the conference with a disability. Under the current structure of the NPRM this computer would "comply." If the definition of accessible is compliance that computer would be declared accessible because it complied.
This problem is addressable, as we discuss later, but it requires that compatibility be handled differently than as an "OR" clause on the functional performance criteria.
If compatibility is changed so that it is only fulfils compliance when AT can and will be present, then this definition of accessibility could stand. However, the way that "compatibility" is handled in the rules needs to be examined extraordinarily carefully before this definition of accessibility is placed into the regulations. This definition is likely to be taken as a legal definition of accessibility and applied far beyond the scope of the 508 regulations.
RECOMMENDATION:
If the guidelines are changed so that compatibility is only applicable when the assistive technology will be present when and where users try to use the E&IT then the current definition of "accessible = compliance" can remain. Otherwise we do not recommend this definition of accessible be codified in this regulation since it will have far reaching effects where products, work stations etc could be deemed legally "accessible" even in the absence of the assistive technologies needed to make them usable.
Accessibility should be defined as - "accessible under the conditions that will exist as people with disabilities encounter the product."
Accessibility should not be defined as the - "potential of a product to be accessible under certain conditions."
NPRM: Definitions: "Agency"
COMMENT: We have no specific comments on the definition of "Agency", although the topic of agency comes up when talking about the availability of documentation and support. There is a concern that accessible documentation and training is being left to the agencies without any required support from companies.
NPRM: Alternative Formats Definition
COMMENT: "Accessible internet programming or coding languages" is somewhat ambiguous.
RECOMMENDATION: The Board might consider using "accessible internet formats"
NPRM: Alternate Modes Definition
Alternate modes. Different means of providing information, including product documentation, to people with disabilities. Alternate modes may include, but are not limited to, voice, fax, relay service, TTY, Internet posting, captioning, text-to-speech synthesis, and audio description
COMMENT: The use of the word "mode" in this definition is problematic. The intent here is to define the phrase "alternate modes" as it is used in the section talking about dissemination. However, by using the word "modes" here it creates confusion with the access specifications throughout the regulations where it is specified that, "at least one mode shall be available for operation of the product." Some people in the past have suggested that one mode for providing information from a product was to call somebody up and ask them on the phone. Sending somebody information by fax should not be an alternate mode for providing information from a kiosk or a computer.
RECOMMENDATION: That the term "alternate modes" be changed to "alternate methods" so that the alternate methods for providing documentation and training support is not confused or substituted for alternate modes of operation for the product itself.
NPRM: Assistive Technology Definition
No comment
NPRM: Definition of Electronic & Information Technology
COMMENT: We have no specific comments on the definition of "Electronic & Information Technology".
However, the example at the end of the definition states, "medical equipment where information technology is integral to its operation are not information technology." We would suggest that the word "direct" be included in front of the word "operation" so that people with different types of disabilities are not precluded from participating in general data processing tasks involving medical data from medical instrumentation where the data processing equipment is not involved in the medical devices direct operation.
NPRM: Operable Controls Definition
Operable controls. A component of a product that requires physical contact for a normal operation. Operable controls include but are not limited to mechanically operated controls, paper trays, card slots, keyboards, or keypads.
COMMENT 1: Since this includes things like slots and trays, a more general term than the word "controls" might be used. Perhaps "operable mechanisms" or "operable components".
COMMENT 2: Almost everyone we talked with who read this definition mistook the phrase "paper trays" to be referring to the large trays on a copier which are loaded up with reams of blank paper for copying. It is only in reading the comments that they discovered that the NPRM was referring to the input and output paper trays. It is suggested that the definition be changed slightly to say, "input and output trays."
NOTE: In our reorganization, we use the term "Operable elements" for a grouping category. It is broader than this but is not used in any of the rules and does not need to be defined.
RECOMMENDATIONS:
NPRM: Telecommunications Definition
No comment
NPRM: TTY Definition
An abbreviation for teletypewriter. Machinery or equipment that employs interactive text based communications through the transmission of coded signals across the telephone network. TTYs may include, for example, devices known as TDDs (telecommunication display devices or telecommunication devices for deaf persons) or computers with special modems. TTYs are also called text telephones.
COMMENT: We suggest a slightly altered definition of TTY since this definition seems to include a broader range of topics than are usually included under TTY. For example, there are text telephones now that are not TTYs and are not TTY compatible.
We suggest that the term "baudot" be used in the definition to clearly narrow the definition to just TTYs. This is important from both a consumer and an industry stand point.
It is the hope of both industry and consumers that we as a country will eventually migrate away from baudot TTYs. Until that migration is complete, however, ( and industry no longer has the requirement to be compatible with baudot TTYs), companies are not interested in being required to make their systems compatible with baudot TTYs as well as a variety of other telecommunication devices being designed for and used by people who are deaf and using other technologies.
RECOMMENDATION: Add the word 'baudot' in front of the word "coded" in the definition
NPRM: Undue Burden Definition
No comment
NPRM: 1194.21 General Requirements
1194.21 (A) Color coding shall not be used as the only means of conveying information, indicating an action, prompting a response, or distinguishing a visual element
COMMENT 1: In the EITAAC document there were two items under the color section.
One topic dealt with the use of color to convey information. That item is captured very well in this NPRM item and we have no comments or suggestions for changing that item.
The second topic has to do with whether or not the information being presenting would become unreadable if an individual had a color deficiency, - not because it was color coded, but because the colors used would lose their contrast with each other. That is, the colors used for the lettering and the background would be distinct enough if you could see color, but would blend together if you could not see color, making the letters unreadable.
Note that this is a separate topic from the one addressed under non-embedded systems which required that software applications have a variety of colors that the user could select from. The problem cited here is more of an issue of contrast than it is of color, but it is caused by color perception deficiencies.
RECOMMENDATION: No change to this item. A separate item is proposed later to handle the contrast problem both in general and as it relates to color deficiency problems.
NPRM: 1194.21 (B) Products which are freestanding, non-portable, and intended to be used in one location and which have operable controls shall comply with the following:
No comment on any of the four components to this item. All fine.
NPRM: 1194.21 (C) When flashing or blinking text, objects, or other elements are displayed, the flash rate shall not exceed two Hertz
COMMENT 1: The EITAAC recommended that frequencies of 3 Hz or lower or 60 Hz or higher be used. In the NPRM this was changed to 2 Hz.
RECOMMENDATION: We recommend that the provision be three hertz or lower in agreement with the research and recommendations of Jeavons and Harding who have conducted the most extensive research on this topic.
COMMENT 2: We note that the Access Board did not choose to set an upper limit. If it chooses to not do so, we at least recommend that they issue a warning to avoid any flicker in the range of twenty hertz, plus or minus fifteen, if at all possible since this is the frequency of maximum susceptibility to photo epileptic seizures.
RECOMMENDATION: We recommend that the NPRM require that products avoid any flickering between (but not including) 3 and 55 hertz (which is the power frequency for Europe).
NPRM: 1194.21 (D) If a timed response is required, at least one mode which does not require users to respond within a timed interval or allows users to adjust the timing and repetition of those intervals to at least 5 times the default setting, shall be provided.
It is noted that the board also seeks information on whether this would cause security problems and whether there are commercially available systems which would allow adjustments of timed intervals.
COMMENT: There are a number of different systems which allow users to adjust timing factors. On standard personal computers this is usually done through the control panel. On products using EZ Access, it can be done through a menu option as well as being invoked automatically when the EZ Access features are used. On a number of ATMs and other products, the user is prompted to see if they need additional time.
In some situations, the manner in which timing adjustments are implemented can cause a problem.
From a user perspective, for example, if a product asks if more time is needed and then doesn't provide enough time for the user to respond, it can leave the user unable to complete his or her task.
From a product standpoint, timeout periods that are too long could pose security risks. For example, many ATMs are designed to time-out when the individual does not perform any action for a given period of time. The timeout is intended to cause the ATM to reset before the next person in line steps up and uses the ATM while the original person's sign-in was still active. Increasing the time before the ATM resets would increase the ease with which someone else could step up to the ATM and find it still registered to the previous user.
In order to address these important issues, it is suggested that a variety of techniques be allowed in this area including the strategy of asking an individual if they need additional time.
RECOMMENDATION 1: That the language in the NPRM be extended somewhat to provide an additional option to manufacturers. The revised wording would be.
"At least one mode shall be provided which a) does not require the user to respond or read within a timed interval or b) allows users to adjust the timing and repetition of those intervals to at least five times the default setting or c) offers the user the opportunity to extend their time and allows at least ten seconds to respond to the offer."
RECOMMENDATION 2: That an advisory be inserted that alternate strategies to timing be used for resetting systems. For example,
ADVISORY: Consider using other techniques instead of time-out to reset a system or prevent a new user from using a previous user's sign on. For example, some ATMs are now using an inexpensive motion sensor (such as used on faucets) that will detect immediately when an individual departs. Some banks such as CitiCorp also use a technique where any additional transactions, which involve removal or transfer of cash, require the person to re-enter their PIN to ensure that no "tailgating" occurs.
This is particularly important on machines which use a card to swipe or dip where the individual can grab their cash and step away from the terminal since they already have their credit card. It should be noted that even the usual reset delay does not provide any real protection since it is often only seconds after you step away from the terminal before the next individual steps up.
NPRM: 1194.21 (E) Where biometric forms of user identification or activation are used, an alternative form of identification or activation, which does not require the user to possess particular biological characteristics, shall also be provided.
We note that the Access Board (in question five) requests additional information on this topic including the use of multiple biological forms of identification and whether alternate forms could be added at a later date. The Board also asks for differentiation between workstations and other applications.
COMMENT 1: Providing an option to offer multiple biological identification techniques would greatly reduce the probability that an individual would be prevented from gaining access while still maintaining the basic biological form of access. If one accepted fingerprint, voice or iris identification as options, for example, one would be pretty sure of providing access. The provision of just two of three, however, might be a problem for victims of burn or flash explosion. At that point, it is probable that some accommodation could be made for those extreme cases. It is an interesting question, however, if the government went to a system where all systems used and required a biologic sign-on with no non-biologic alternative. More likely, however, is a situation where there is a standard biological system which is used to sign-on to most systems and a different, perhaps more expensive, one is used as an alternative for those individuals unable to use the original.
COMMENT 2: A key issue around this topic, however, is how the biometric identification process is being applied. If it is being used to sign-on to a personal workstation, then it is unlikely that it should be required that all workstations have multiple biological identification equipment on them. It should be sufficient that the manufacture would agree to place one type of biological identification system with another acceptable and appropriate system at any time in the future should it become necessary or that they provide a mixture so that the alternate type would be available should it be needed for new or transferring employees as well as any current employees.
Where the biological ID system is being used to control access to physical areas (such as at a doorway) or to shared equipment, then it would probably be necessary that the multiple forms be on the device or that an alternate mechanism for operating the system be available to those who would need to access and use the space or resources now or at some time in the future.
NPRM: 1194.21 (F) Where touchscreens or touch-operated controls are used, such controls shall be operable without requiring body contact or close human body proximity, or all of the operations and functions that are available through such controls shall be made available through an alternate mode that does not require body contact or close human body proximity.
No comment.
NPRM: 1194.23 (A) (1) Mechanically Operated Controls, Keyboards, or Keypads
(1) Controls and keys shall be tactilely discernable without activating the controls or keys.
COMMENT: As written, this provision requires that all controls and keys be tactilely discernable. This is meant to apply to all mechanical controls. Unfortunately, this would outlaw all (legal) touchscreen controls, large and small. There is a later provision that requires touchscreens to be operable without requiring human body contact. Most of the touchscreens that comply with this provision use a mechanical mechanism not unlike that used on elastomeric keyboards. As a result, they would be technically covered under this provision and could not comply.
In addition, this provision requires that the buttons on a mouse be tactilely discernable since they are mechanical as well. However, there is no particular purpose in making mouse buttons tactilely discernable since the mouse is not usable by an individual who is blind. (Elsewhere in the component specific requirements, it is required that products which have keyboards allow complete functionality from the keyboard, so mouse access is not needed).
We would suggest that rather than saying that all controls must be tactilely discernable, only those controls that are needed for full operation of the device (in non-visual mode) need to be tactilely discernable.
RECOMMENDATION 1: It is, therefore, recommended that this provision be reworded slightly to say:
At least one mode shall be provided where all functionality of the product can be achieved using operating mechanisms which are tactually discernable without activation.
RECOMMENDATION 2: It is recommended that an advisory be provided which would recommend that products be made in a way to make it easy for individuals with low vision to be able to use their vision to locate keys. It is particularly important since individuals who are not blind tend to use their remaining vision rather than their tactile senses and therefore will often not find keys which are tactilely distinct but visually invisible (because they don't search in areas where there doesn't appear to be something). It is therefore recommended that the following advisory be added.
ADVISORY: Keys should also be visually discernable from the background. This could be accomplished by providing a high contrast between the key and the device body or with a dark edge on the key or a (contrasting) gap of sufficient width around the key (as is found on many keyboards).
Note: Although provision (1) "Controls and keys shall be tactilely discernable without activating the controls or keys." is correct, the rationale that goes with it misinterprets the provision. This provision simply says that the key should be discernable, meaning that they should be able to be found. The accompanying discussion seems to imply that discernable also means being able identify the key or distinguish it from other keys. In fact, being able to tactilely discern the keys, means that one can simply find and correctly count the number of keys that are present. Being able to identify the key is a separate task entirely. Thus, the references to using different shapes, spacing, or tactile markings suggest ways to identify the keys, not discern them from each other. Similarly, putting nibs on the "j" and "f" or grouping the keys has to do with providing mechanisms for identifying keys rather than making them tactually discernable. Although it is important to be able to identify keys, that requirement is not included in this provision's language. In fact, there is nothing in the current NPRM that discusses any strategies to allow an individual to identify keys if they have low vision or blindness.
RECOMMENDATION:
That the language regarding control identification be removed from the discussion for provision (1).
ADDITIONAL COMMENTS: In our recommendations for reorganization, we have included some provisions that were in the EITAAC report that would help address the issue of control identification such as providing all visual information in audible form.
NPRM: 1194.23 (A) (2) The status of all locking or toggle controls or keys shall be visually discernible, and discernible either through touch or sound
No comment
NPRM: 1194.23 (A) (3) Controls Shall Be Operable with One Hand and Shall Not Require Tight Grasping, Pinching, or Twisting of the Wrist. The force required to activate controls shall be five pounds (22.2 N) maximum.
COMMENT 1: The discussion for this provision refers to the need to be able to operate products without requiring fine motor control or requiring two hands or fingers or simultaneous operation. However, the provision itself does not include simultaneous operation or fine motor control. It is recommended that the phrases "fine motor control" and "simultaneous actions" be added to this provision.
COMMENT 2: Many products have different ways of carrying out an activity. For example, a product may have a mouse and a keyboard, both of which can be used to carry out activities on the screen. Whereas, the mouse would require fine motor control, the keyboard does not. In order to prevent having a rule which would seem to eliminate the ability to use fine pointing controls for those that have fine motor movement, it is suggested that the wording be changed slightly so that it just requires that there be a complete set of controls or operable parts which would allow access to the full functionality of the product without necessarily requiring that all of the controls meet this specification.
RECOMMENDATION: It is recommended that the language for this item be changed to:
"At least one method for achieving all functions shall be provided which is operable with one hand and shall not require tight grasping, pinching, twisting of the wrist, fine motor control, or simultaneous actions. The force required to activate controls shall be five pounds (22.2 N) maximum."
NOTE: We also note that 5 lbs seems to be a lot for activating controls in E&IT. We do not have another specific number, however.
NPRM: 1194.23 (A) (4) All Actions Available or Required by the Product Shall be operable from the Keyboard or Keypad
COMMENT 1: It is suggested that this requirement be restricted to all "software/firmware operable" actions since there may be a number of operations or actions which are mechanical in nature. It would not be possible to make these operable from a keyboard.
COMMENT 2: The NPRM states that all action should be available from a keyboard or a keypad. This is an expansion from the EITAAC, which recommended simply that all actions be accessible from a keyboard. We would suggest that the language, again, return to just referring to all actions being available from a keyboard. There are many products which have a keypad, but where it would be unreasonable to have all of the functions be available from the keypad. For example, consider a business telephone. It would be quite difficult to design one that would allow for all of the functions of the phone to be operable from the keypad on the phone. It is also not clear how that would increase accessibility over other techniques. It is possible that one mode of access could be provided in that fashion. However, there are other more effective techniques.
On systems equipped with a keyboard, however, complete operation from the keyboard is not only a practical means of providing access, it is usually very effective, especially when coupled with speech.
RECOMMENDATION 1: It is recommended that this item be reworded to apply only to systems with keyboards and reworded slightly as follows:
"All software/firmware operable actions available or required by the product shall be available from the keyboard."
RECOMMENDATION 2; This item also be positioned so that it applies to software as well since it must also support this functionality
NPRM: 1194.23 (A) (5) If keyboard repeat is supported, the keyboard delay before repeat shall be adjustable to at least 2 seconds. Key repeat rate shall be adjustable to 2 seconds per character.
No comment
NPRM: 1194.23 (B) Non-embedded Software Applications and Operation Systems
COMMENT: In general we agree with the provisions in this section. However, some of the provisions need to be applied to devices other than non-embedded software platforms. For example, some of the requirements in this section that deal with keyboard navigation, etc., should apply to any system which has a keyboard and a display, even if it does not fall into the category of non-embedded software.
Also, a number of the requirements in this section deal with compatibility with assistive technology. It is suggested that the items dealing with compatibility with assistive technology be grouped together, perhaps under a section called "compatibility with software assistive technology." We believe that this is an interesting parallel to begin drawing as we move into the future, where software assistive technology is compatible with applications and platforms which are very different from today's personal workstations. It may be possible in the future to access and invoke assistive technology like screen readers from a network, even on computers other than one's personal workstation. In our suggestions for reorganization, we have anticipated such an arrangement. The trick, of course, is to create a set of provisions that make sense and are applicable today with today's technology, and yet won't get us in trouble as the technologies evolve.
It should also be noted that use of the terms "embedded" and "non-embedded" software is very tricky these days. The NPRM defines embedded software as that which is on a chip, and non-embedded as that which is on a storage device and can be updated. An increasing number of products which we thought of as having "embedded" software, now use chips where the software can be "flashed" and updated. In addition, programmable computers which we used to think of as having storage devices like hard drives, etc., now have nothing in them except the same type of flash memory. For example, the NetPliance is an Internet appliance with an operating system (QNX operating system from QSSL), email and an Internet browser. The device has no hard drive or any other type of storage mechanism other than flash memory. When you turn the computer off and turn it back on, it simply runs software which is stored in its non-volatile memory. However, every few days when it connects to the Internet, new firmware is downloaded to give it new capabilities. Hence, the NetPliance is something that looks like a computer, but has no storage medium other than its non-volatile memory chips.
On the other end of the spectrum, there are simple cell phones which have no Internet capability of any kind and which can only be used to make phone calls. All of the cell phone's programming, however, is contained, again, on a flash memory, so that the software can be upgraded.
Even pagers are available where all of the programming is in non-volatile RAM so that it can be easily upgraded. However, neither the cell phone nor the pager allow you to download and run other software in the background, such as a screen reader.
Thus, you have devices that fit the NPRM's description of embedded systems, but in fact have operating systems and could support assistive technology. We have devices which clearly would fit the definition of non-embedded software, but function as embedded software devices. So there are already technologies on the market which make the definitions of embedded and non-embedded software hopelessly confounded. We therefore suggest that these definitions no longer are valid or meaningful with today's or tomorrow's technologies.
In the suggested reorganization (Part 3), we suggest some slight changes in language and terminology so that the terms "embedded" and "non-embedded" are replaced with functional terminology to avoid the ambiguity that is resulting from the evolving and converging nature of information technologies.
NPRM: 1194.23 (B) (1) Logical Navigation Among Interface Elements Shall Be Provided By Use of Keystrokes
COMMENT 1: This is a good technique requirement. Our only suggestion is that it should apply to any device which has a display with interface elements on it and a keyboard (and not just to "non-embedded" or "OS based" systems). In the past these have all been non-embedded software applications, but this is not true today and will not be true in the future.
COMMENT 2: Also, it should be possible to not only navigate among the items but to be able to operate them as well.
RECOMMENDATION: It is our recommendation that this provision be made to apply to products that have software interface elements on a display and also have a keyboard. It should also support activation of elements, in addition to navigation. Suggested new wording for the provision would be:
"If interface elements are provided on a display, then logical navigation among interface elements and operation shall be provided by use of keystrokes."
NPRM: 1194.23 (B) (2) Software shall not interfere with existing features of other products or operating systems that affect the usability for people with disabilities
No comment
NPRM: 1194.23 (B) (3) Well Defined On-Screen Indication of the Current Focus Shall Be Provided That Moves Among Interactive Interface Elements as The Input Focus Changes. The Focus Shall Be Programmatically Exposed so that Assistive Technology can Track Focus and Focus Changes
COMMENT: This provision actually includes two requirements, both of which are good. One, however, should be more generally applied.
The first requirement: providing a well defined on-screen focus indicator, is something that should be true for embedded and non-embedded software systems. Again, any system that has on-screen elements and a focus should provide a mechanism that is well defined and highly visible. It is recommended that this be generalized beyond just embedded software.
The second requirement deals with the fact that this focus should be programmatically exposed so that it can be tracked by assistive technology. This, of course, should stay with the other requirements that deal with software assistive technology compatibility.
RECOMMENDATION: It is recommended that this requirement be divided into two parts:
NPRM: 1194.23 (4) Sufficient Information About a User Interface Element Including the Identity, Operation and State of the Element Shall be Available to Assistive Technology
COMMENT: This provision is fine as written.
The only issue arises around products where there is no assistive technology available. Creating hooks for AT should not preclude the necessity to either find AT that would work with the platform, provide sufficient incentives that AT would be built for their platform, or build access directly into the product if no AT vendors were interested. Hooks without AT to work with them should never be equated with accessibility.
In the suggested reorganization, we attempt to provide wording to address this problem so that industry is not left with a set of requirements around assistive technology compatibility that it would be unable to address -- nor with the ability to claim accessibility through AT compatibility if there aren't sufficient AT available commercially for their platform to meet the performance guidelines.
NPRM: 1194.23 (5) Where an image represents an interface element or the state of an interface element, there must be a way for assistive technology to associate meaningful text with the image.
No comment beyond those made above.
NPRM: 1194.23 (6) The Use of Images Shall be Consistent Throughout an Application
COMMENT: The discussion mentions that the purpose for this provision is to facilitate use of screen readers by making sure text can be associated with a bitmap image. According to the other requirements in this same section, applications are already required to associate text with any images. It would, therefore, appear that this requirement is redundant with the techniques above and would only be useful on old screen readers. It should be noted that screen readers that are that old would be unable to take advantage of most of the other information discussed in this section.
Consistency in design is also good for individuals with cognitive disabilities and is generally a good human factors principle. It is suggested, however, that rather than including this as a requirement, this item might be changed to an advisory. Another reason for doing this is that often individuals will use an icon to depict different things themselves. This provision might prevent an application from allowing icons to be used differently by a user.
RECOMMENDATION: It is recommended that this provision be an advisory rather than mandatory.
NPRM: 1194.23 (7) Text shall be provided through an application programming interface supporting interaction with assistive technology or use system text writing tools. The minimum information that shall be available to assistive technology is text content, text input caret location, and text attributes.
NPRM: 1194.23 (8) A Minimum of Eight Foreground and Eight Background Color Selections Capable of Producing a Variety of Contrast Levels Shall be Provided
COMMENTS: First, the problem that this requirement is attempting to address has little to do with the number of colors, but actually the number of color settings (which include hue, saturation and darkness). Users need to be able to adjust the color (hue), but also the saturation and the darkness. A system could have a small or large number of colors (hues) and be accessible (including 8). The key to accessibility is in having the ability to set the other dimensions, particularly lightness/darkness. Good contrast is achieved by having some colors set to light and other colors set to dark. If this is done, then regardless of the colors (and color deficiencies) there will be good contrast between figure and ground and the person will be able to see the text or other elements.
We suggest that contrast, rather than the number of colors, be the focus of the requirement. That is, that there be sufficient contrast for visibility and readability. Even if the individual is provided with a wide range of colors, most individuals do not have the ability to go play with the color palettes in such a fashion that it would result in different applications providing them with good contrasts. The best language we have heard is that "color choices representing the full range of brightness, including pure black and pure white, and at least six choices with brightness values approximately evenly spaced in between those two extremes."
RECOMMENDATION 1: That the emphasis be moved from providing a large number of colors to providing applications which provide good contrast.
COMMENT 2: The second component of this provision in the EITAAC recommendation was that large fonts and different font styles, including Sans Serif fonts, be available. In the NPRM discussion, it was stated that screen enlargement software would be required and could be used by people who need increased font size. The problem with screen enlargement software is that it causes most of the text to move out of view, requiring an individual to pan back and forth repeatedly in order to try to read text. This makes it very difficult to track off the end of one line of text and on to the beginning of the next. When fonts can be enlarged, the words will automatically wrap, and the individual can read the text without having to constantly pan the screen left and right. The ability to read text without panning can greatly increase a user's efficiency.
RECOMMENDATION 2: It is recommended that the ability to adjust font size be added back in at least on an advisory basis.
NPRM: 1194.23 (9) An Option Shall be Provided to Ignore Individual Application Display Attributes so System Wide Settings will be Maintained
COMMENT: The problem sometimes manifests itself as a program that will not allow users to override the display attributes of documents as well.
RECOMMENDATION 1: That the words "and document" be added after the word Application.
RECOMMENDATION 2: That keyboard settings be removed from the discussion of this item since this item only deals with display settings and it is confusing.
NPRM: 1194.23 (10) Electronic forms shall allow people using assistive technology to access the information, field elements, and functionality required for completion and submission of the form including all directions and cues. Inaccessible electronic forms may be used, if an alternative accessible electronic form with equivalent information, field elements, and functionality is also provided.
COMMENT:. This topic seems to already be covered in the previous sections. Furthermore this seems to provide an exemption for forms that is not available for anything else. Is this what is intended? If so, shouldn't an exemption be provided for anything where there is an alternate way to achieve it?
Perhaps there should be a general "at least one mode for " phrase attached to some of the items in this group.
NPRM: 1194.23 (11) If Animated Moving Text is Provided, it Should Also be Displayable in at Least One Static Presentation Mode at the Option of the User
COMMENT 1 : In the write-up, this is stated as being a provision to address a problem faced by screen readers. In fact, it is a problem also for individuals with low vision who have trouble reading moving text, individuals who are deaf and who may need more time, individuals with physical disabilities that affect the stability of their head or ocular tracking.
COMMENT 2 : The discussion seems to indicate that any animated element should be addressed, but the guideline only mentions text.
RECOMMENDATION 1 : It is recommended that this provision be categorized as a generally applied provision as it was in the EITACC report rather than being treated only as an assistive technology compatibility issue.
RECOMMENDATION 2 : That the language be changed to any animated element of the user interface.
NPRM: 1194.23 (C) Web Based Information or Applications
COMMENT 1: There is a concern with the promulgation of multiple standards for web accessibility. Although the 508 provisions are based on the Web Accessibility Initiative (WAI) Web Content Accessibility Guidelines, they differ in a number of places sufficiently that it is causing confusion among people who don't know whether to follow the language in one set of guidelines or the other.
In some places, confusion is introduced by the variation in terminology in the NPRM provisions themselves. In other cases, it is in the language that accompanies the provisions, which explains how they are to be applied, that leads to confusion. In several places the NPRM language provides a quite different interpretation in the explanatory text than the WAI Web Content Accessibility Guidelines for the same WAI guideline.
RECOMMENDATION 1: Ideally, the 508 regulations could refer to the Web Content Accessibility Guidelines (WCAG) directly. If desired, a copy could be included as an appendix to the regulation.
RECOMMENDATION 2: If it is not possible to directly refer to the WCAG guidelines, then the language in the NPRM should be harmonized more closely with the WAI's. It is possible that much of the disharmony is accidental rather than intentional changes to the guidelines.
COMMENT 2: The 508 provisions only included priority one guideline from the Web Content Accessibility Guidelines (except for the one Priority 3 item).
This sets a minimum standard for accessibility. There is a concern that this level of accessibility will establish a new legal standard for accessibility that is quite low. This concern stems from the fact that "accessible" is defined as complying with these guidelines, and these guidelines only require Priority 1 items from the WAI guidelines. This is a very conservative level of compliance and will still leave web sites very difficult to use for many people. If individuals need to use the web as part of their daily livelihood (which is becoming increasingly common), this could be a severe hardship.
RECOMMENDATION: It is recommended that level two WAI WCAG guidelines be required, or, if that is not possible, then at least included as an advisory.
COMMENT 3: It is noted that the cognitive aspects of the web guidelines and the EITAAC guidelines have all been dropped. One of the larger groups of individuals in the federal government with disabilities would include people with learning disabilities. It is recommended that the Access Board reconsider including the cognitive disability related guideline. If this guideline cannot be included as a requirement, it should be included as an advisory.
COMMENT 4: The EITAAC recommended that the WAI User Agent Accessibility Guidelines (UAAG) and the WAI Authoring Tool Accessibility Guidelines (ATAG) also be included. Although the general software accessibility provisions in the NPRM will address some of the issues around making these tools accessible, the web tool area (both for generation and for browsing) has a number of unique requirements which are not addressed. This area is becoming important enough that the Access Board may want to re-examine the WAI guidelines to identify key issues that should be incorporated in the NPRM.
RECOMMENDATION: That Priority 1 items of the WAI User Agent Accessibility Guidelines (UAAG) and the WAI Authoring Tool Accessibility Guidelines (ATAG) also be included as required and that Priority 2 items be advisory.
NPRM: 1194.23 (D) Telecommunication Functions
COMMENT: We concur with the Access Board's movement of certain provisions from the general level in the EITAAC report into this narrower application. Although certain provisions could apply to a wider range of technologies than just telecommunications, it is not practical to require them for all E&IT.
In the reorganization, we do suggest that two provisions, (hearing aid compatibility and volume specifications) be applied to communication systems in general, rather than narrowly focused only on telecommunication. This is particularly important for the items related to volume control, as noted below.
NPRM: 1194.23 (D) (1) Telecommunications products which provide a function allowing voice communication and which do not themselves provide a TTY functionality shall provide a standard non-acoustic connection point for TTYs. It shall also be possible for the user to easily turn any microphone on and off to allow the user to intermix speech with TTY use
COMMENT: A new standard for connecting TTYs that has been developed is the use of a 2.5mm headset jack such as those used on cellular phones. This standard would be a very valuable thing for industry to follow, since it would be quite convenient and economical to implement. However, manufacturers not in the cellular technology area may not be aware of this standard.
RECOMMENDATION: It is suggested that a note be added after this item that would read as follows:
NOTE: A 2.5mm headset jack with industry standard signal levels would meet this requirement.
NPRM: 1194.23 (D) (2) Telecommunications products which include voice communication functionality shall support use of all cross-manufacturer non-proprietary standard signals used by TTYs.
No comment
NPRM: 1194.23 (D) (3) Voice Mail, Auto Attendant, and Interactive Voice Telecommunication Response Systems Shall be Usable by TTY Users with Their TTYs
COMMENT: It is recommended that these systems be accessible, not only by people who are accessing these systems directly, but also by those accessing these systems through a telephone relay service operator.
RECOMMENDATION: Recommend that the TRS option be added, so that it would read:
"Voice mail, auto-attendant, and interactive voice response telecommunications systems shall be usable by TTY users directly with their TTYs and via TRS operators including the ability of the TRS operators to repeat and pause messages so that they can keep up with the recording."
NPRM: 1194.23 (D) (4) Voice mail, messaging, auto-attendant, and interactive voice response telecommunications systems shall provide at least one mode which does not require users to respond within a timed interval or allows users to adjust the timing and repetition of those intervals to a minimum of 5 times the default.
COMMENT 1: This item is duplicative with the timing item in the general requirements category. However, it does no harm to put it down here next to the other voice mail items, unless it implies that people following the telecommunication functions do not need to follow the provisions in the general requirements section.
COMMENT 2: In some circumstances, depending upon the system and default setting, a value of five times the default setting may not provide users with sufficient time to respond. For example, if a user is presented with a question or information, he or she may need time to decipher and comprehend the information before choosing a response. Preferably, the system will allow users the maximum of flexibility in setting response time. A timeout extension could be accomplished by asking the user if he or she needs more time. This has the affect of extending the amount of time available to them to carry out an action. Users should also be given the option of repeating messages delivered by the system.
RECOMMENDATION: The wording in the provision could be changed as follows to reflect the special nature of interactive voice response systems:
"Voice mail, messaging, auto-attendant, and interactive voice response telecommunications systems shall provide at least one mode which allows users to repeat messages and does not require users to respond within a timed interval or allows users to adjust the timing and repetition of those intervals to a minimum of 5 times the default or offers the user an option to extend their time and provides at least 10 seconds to respond to the offer. "
NPRM: 1194.23 (D) (5) Where provided, caller identification and similar telecommunications functions shall also be available for users of TTYs, telecommunications relay services, and for users who cannot see displays.
No comment.
NPRM: 1194.23 (D) (6) For transmitted voice signals, telecommunications products shall provide a gain adjustable up to a minimum of 20 dB. For incremental volume control, at least one intermediate step of 12 dB of gain shall be provided.
COMMENT: This should probably be generalized beyond just telecommunication to other communication systems in the work place. A suggestion for how to do that is included in our reorganization.
NPRM: 1194.23 (D) (7) If the telecommunications product allows a user to adjust the receive volume, a function shall be provided to automatically reset the volume to the default level after every use but not before.
COMMENT: For private workstation telephones, having it reset following each use can be very disconcerting and inefficient. It is, therefore, recommended that there be a way of overriding this reset feature.
RECOMMENDATION: We suggest that the following be added to the end of this item so that it reads:
"Telecommunication products which allow a user to adjust the receive volume shall provide a function to automatically reset the volume to the default level after every use, but not before. Unless the volume can exceed 120 dB SPL, the reset function shall be a user option on non-public devices."
NPRM: 1194.23 (D) (8) Where a telecommunications product delivers output by an audio transducer, which is normally held up to the ear, a means for effective magnetic wireless coupling to hearing technologies shall be provided.
COMMENT: It is suggested that this requirement be applied beyond telecommunication to any situation where an acoustic output device is held up to the ear. This would include kiosks or museum exhibits or even intercoms, where a handset-like device is used for conveying voice signals.
RECOMMENDATION: We suggest that the requirement be reworded slightly and moved up to a general audio requirement as follows:
"Where a product delivers information by an audio output device which is normally held up to the ear, a means for effective magnetic wireless coupling to hearing technologies shall be provided, except where all information is also provided visually."
NPRM: 1194.23 (D) (9) Interference to hearing technologies (including hearing aids, cochlear implants, and assistive listening devices) shall be reduced to the lowest possible level that allows a user of hearing technologies to utilize the telecommunications product.
COMMENT: It is recommended that this item be moved up to the general requirements category, since it is not just telecommunication products that can generate interference with hearing technologies.
NPRM: 1194.23 (E) Video or Multi-Media
COMMENT: We suggest this section be retitled to read "Video, Multi-Media, and Animation," thus representing the full range of activities in new generation digital media which may cause accessibility problems.
NPRM: 1194.23 (E) (1) All television displays 13 inches and larger, and computer equipment that includes television receiver circuitry, shall be equipped with caption decoder circuitry which appropriately receives, decodes, and displays closed captions from broadcast, cable, videotape, and DVD signals.
COMMENT: This provision appears to be written so that it would generally apply to computers and other display devices that would be displaying captioned video. The way it is worded, however, it is not clear that this would cover devices which do not have receivers in them. We suggest that the language be modified slightly to include the fact that all devices which project or display NTSC signals, would need to include a closed caption decoder so that captions would be visible.
RECOMMENDATION: We suggest the following slightly revised wording:
"All television displays and projectors 13" and larger, and computer equipment that includes television receiver circuitry or accepts NTSC input shall be equipped with caption decoder circuitry which appropriately receives, decodes, and displays closed captions from broadcast, cable, video, DVD signals, etc."
NPRM: 1194.23 (E) (2) Television tuners, including tuner cards for use in computers, shall be equipped with secondary audio program playback circuitry.
No comment.
NPRM: 1194.23 (E) (3) All video and multimedia productions, regardless of format, that contain speech or other audio necessary for the comprehension of the content, shall be open or closed captioned if the production is procured or developed for repeated showings to audiences that may include people with hearing impairments.
No comment.
NPRM: 1194.23 (E) (4) All video and multimedia productions, regardless of format, that contain visual information necessary for the comprehension of the content, shall be audio described if the production is procured or developed for repeated showings to audiences that may include people with visual impairments.
No comment.
NPRM: 1194.23 (E) (5) Display or presentation of alternate text presentation or audio descriptions shall be user-selectable unless permanent.
No comment
NPRM: 1194.23 (F) Information kiosks and transaction machines shall be usable by people with disabilities without requiring an end-user to attach assistive technology to the information kiosk or transaction machine.
No comment.
NPRM: 1194.23 (F) (2) Where information kiosks and transaction machines deliver audio output, including speech, a mechanism shall be provided for private listening and user interruptability.
COMMENT: Generally we agree with this requirement, except that it probably should be at the general requirements level. Thus, whenever speech is provided for access on any product (not just on kiosks and ATMs), it should be interruptible. Also, wherever the headphone jack is used, it should be possible for the user to listen privately as well.
RECOMMENDATION: Move interruptability and private listening to the general level, as separate items. Private listening could be attached to the headphone jack requirement, since it is so closely related.
NPRM: 1194.23 (F) (3) Where information kiosks and transaction machines deliver voice output, incremental volume control shall be provided with output amplification up to a level of at least 65 dB. Where the ambient noise level of the environment is above 45 dB, a volume gain of at least 20 dB above the ambient level shall be user selectable.
COMMENT: It is recommended that this requirement be applied more generally. In particular, if there is any product that is providing voice output, it is recommended that there be a volume control that provides sufficient volume. The volumes specified here are actually about the same as someone talking loud and within the range of output of even small pocket recorders.
RECOMMENDATION: It is recommended that this requirement be moved up to the general requirements section.
NPRM: 1194.25 (A) All products that act as a transport or conduit for information or communication shall pass through cross-manufacturer, non-proprietary, industry-standard codes, translation protocols, formats or other information necessary to provide the information or communication in a usable format. Technologies which use encoding, signal compression, format transformation, or similar techniques shall not remove information needed for access or shall restore it upon delivery.
No comment.
NPRM: 1194.25 (B) Where Provided, At Least One of Each Type of Expansion Slots, Ports, and Connectors Shall Comply with Publicly Available Industry Standards.
COMMENT: While this is a good recommendation for things like personal computers, there is a problem when you try to extend the concept to all electronic and information technologies. For example, the Merlin phone systems provide a phone with an expansion slot on the side. You can snap additional modules on the side of the phone to add features or call buttons. Under this provision, it appears as if the government would not be able to purchase those phones, since it has a non-standard slot on the side and there are no other expansion slots provided. Even if Merlin were to publish the slot specifications as a standard, it would not provide any particular accessibility advantage, since there are no assistive technologies that would be designed for that particular phone expansion slot. Similarly, there are small handheld devices that have slots on them. Even cell phones have proprietary slots on the bottom for adding peripherals and expanding functionality. For example, the Nokia phone has a connector on the bottom that expansion modules can be connected to. This is a non-standard connector.
However, the existence of standard connectors on products, particularly I/O connectors, is quite valuable. Combined with assistive technology software, it provides an alternate mechanism to easily transport information in and out of the product to assistive technology peripherals. Perhaps, if the word "slot" was dropped and the rest of the provision that deals with standard ports and connectors were retained, it would achieve most of its functionality while avoiding the problem. Most assistive technologies today are software based and/or connect through the external connectors and high speed buses such as the universal serial bus. As a result, slot access is less important than in the past.
RECOMMENDATION: Either drop this item or restrict it to standard ports.
NPRM: 1194.25 (C) Operating system software shall not interfere with assistive technology.
COMMENT: This one is very general and sweeping. Some more specifics are needed here. AT can be anything. Not clear what is desired or required. This one will drive both manufacturers and evaluators crazy without some specific guidance.
NPRM: 1194.25 (D) Products providing auditory output shall provide the auditory signal at a standard signal level through an industry standard connector.
COMMENT 1 : This is a good rule, except there are some products whose auditory output consists of just simple beeps. Where these beeps are also represented in an equivalent visual fashion, it does not seem that the provision of a headphone jack should be required.
COMMENT 2 :This would also be a good place to mention that there should be a mechanism for cutting off the audio (i.e., silencing any speakers) when the headphone jack is used. This is important, since individuals often turn the volume up. It does not have to be required that it would cut off (silence) the speaker, but it should be possible.
COMMENT 3 :Finally, this requirement mentions that the audio should be provided through a standard connector. However, the purpose for putting it on a standard connector is so that assistive technologies can be connected to it. If an arbitrary number of different standard connectors are used, it would mean that people with disabilities would have to carry around adapters for the different connectors. Since almost all audio today is put out through either a 2.5mm or 3.5mm phone jack, it is recommended that these either be required or they be as an advisory. (The manufacturer would have the option of using either connector.) The 2.5mm is typically used as a headset jack on all telephones and very small devices. The 3.5mm phone jack is used on all miniature devices, including miniature radios and recorders.
RECOMMENDATION: It is recommended that the wording would be expanded slightly to read:
"Products providing auditory output shall provide the auditory signal at a standard signal level through an industry standard connector. Auditory cutoff of speaker shall be available to allow the user to listen privately.
"Exception: Products that produce only simple non-speech sounds that are already presented in equivalent visual form shall not be required to have such a connector.
"ADVISORY: A 2.5mm or 3.5mm phone jack is preferred."
NPRM: 1194.27
Functional Performance Criteria
First, we think that simplifying the functional performance criteria by removing the sub-items and bringing them to the bottom of the document as an overall e