Mac OS CU-SeeMe Change History README file 30 Apr 97 by Dick Cogger, revised by Larry Chace and Andy Wyatt This document collects the changes that have occured to CU-SeeMe, through the various releases for the Mac, since the start of the 0.70 series. Most recent changes are at the top of the file. -----------------Changes in 0.87b1 (22 April 1997) This version contains the items previously released for limited test in 0.87a1 and 0.87a2. New features: 1. This version of CU-SeeMe will automagically use Apple's Open Transport, if it is present, instead of MacTCP. 2. The Slide Window facility has been redesigned. a. The overhauled Slide Window now includes: - screen capture - paste PICT from clipboard - save and open slides as PICT files - slide sorter - interface consistent with the rest of CU-SeeMe. b. Known problems: - slide tray concept loses sort order - a few display glitches with remote cursor and scrolling - hard to save a video grab image sometimes because of memory limits converting to PICT - 68K version doesn't draw the beestripe window option well c. Please note that the Slide Window facility is available only in the Cornell version of CU-SeeMe for the Macintosh. d. Slide Window Guide at the following Web site: http://cu-seeme.cornell.edu/slideweb/slidewindow.html Several minor (but obscure) bugs have been fixed. 1. In the "Chat" window, the "home", "end", "page up" and "page down" keys can be used to navigate through the output area at the top of the window. 2. In the "Chat" window, when turn on or off the "filter" facility (to hide or restore selected folks' messages), a "watch" icon will indicate that CU-SeeMe is busy performing the filtration process. (In the past, the program appeared to go to sleep during this time.) 3. For the "Chat" window, messages that you send are kept around somewhat longer, so as to enchance the error recovery process and to thereby help get your messages out to everyone in the conference. We anticipate the release of yet another beta version in the near future that will incorporate additional fixed and enhancements to the Slide Window and to the Chat Window. -----------------Changes in 0.85b1 (22 August 1996) This release is mainly a bug-fix release with some new features. Fixes: 1. Reduce the possibility of a system hang during "Quit". 2. Avoid random memory smashes on PPC machines. 3. Avoid a "push-to-crash" problem with the Audio Window "push to talk" button. 4. Improve the reliability of the Chat Window messages. 5. (Other small items too numerous and obscure to mention.) New features: 1. Include a built-in Chat Window to functionally replace the Talk Window plug-in routine. (See below for more comments on this feature.) 2. Detect color video being sent by White Pine's Enchanced CU-SeeMe and identify it as such in the Participant List by showing the user "name" in red. (Cornell's CU-SeeMe cannot decode color video -- at present -- and so color senders have to be treated as "lurkers", just as if they are sending no video at all.) Details on the Chat Window: 1. This feature is built into CU-SeeMe and so all users of this version will have it. There is no need to download the "TalkFM" program that was used with previous versions, but that program, if available, can still be used. 2. The window can be re-sized in the normal fashion and the division between "output" and "input" can be adjusted by dragging the "black bar" in the "elevator". 3. The "Config" button can be used to change the font and size of the window contents. 4. The "Filter" button can be used to enable or disable message filtering. (Please note that this process can take several seconds -- don't panic!) 5. The "Reset" button can be used to eliminate all message filters. 6. If you want to filter out some user's messages, then just double-click on the user's name in the output portion of the window. When the dialog box comes up, check that you want to filter out these messages and click "OK". You can also double-click the name to turn off the filtering of that user's messages. 7. The messages are retained in memory, even if you have filtered them out, so that you can turn off the filtering and see the messages once more. Known problems: 1. We still cannot use the built-in digitizer (video-in) on a 7500 or 8500 PPC if virtual memory is turned on. You must turn off virtual memory if you want to use that digitizer. (Yes, it *is* a bug!) 2. If you choose to filter the data in your Chat Window, and if there is a lot of data there, then the filtering process may take several seconds when it is turned on or turned off. Do not panic! (We should at least show you a spinning ball cursor to show you that we are hard at work!) -----------------Changes in 0.83b3 1) New version number handling has been implemented, that separates protocol version from application version, and allows more flexibility in the application version display. On the info panel under a remote window, you will see information on the source and date as well as version number. 2) A bug was fixed with outgoing audio bandwidth not being counted against the rate cap, which meant that video and other outgoing data would not slow down to make room for audio. It should now be automatic for senders to reduce video (if needed to stay under the cap) whenever sound starts being sent. 3) A bug was fixed with audio slider controls getting stuck whenever moved outside the valid range. 4) A bug causing crashes on abnormal exits was fixed. 5) Audio being sent with 7500/8500 models should now be handled properly. CAVEATS *** Remember, with the 7500/8500,the video digitizer doesn't work (to capture to an offscreen buffer as CU-SeeMe needs) if virtual memory is on. We havn't figured out what we're going to be able to do about it, beyond buying more memory. If you turn VM off, 83b3 (and 83b2) works. RamDoubler also prevents digitizing. *** Also, with the 7500/8500 CU-SeeMe and OpenTransport aren't compatible with each other. We are working on making CU-SeeMe compatible with Open Transport (and hope that Apple is also working on make Open Transport compatible with existing MacTCP-based applications). Meanwhile, do the following: (1) Install OpenTransport COMPLETELY, (2) Remove from the Extensions folder the libraries named "Open Tpt Internet Library" and "OpenTptInternetLib" and remove the TCP/IP control panel, (3) Make sure that you have MacTCP 2.0.6 or later (and for dialup, MacPPP 2.1.1SD or later). *** There is a bug that can cause a crash if you attempt to set a nickname to a domain-name rather than a dot-notation IP address. (Domain-names are OK to type in to the Connect dialog.) -----------------Changes in 0.83b2 1) Talk should work better (no more lost messages), and SlideWindow should be working again. Fixed an Auxdata bug that would prevent completion of large items, and degrade performance for all items has been fixed [Problem was that many aux data packets, including all retries, were being rejected by recipient for having incorrect header.packetLength] 2) Fixed an audio playout bug that would cause audio to cease after 5-10 seconds on machines with a heavily burdened cpu is fixed [Problem was faulty queue management for audio playout buffers that would fail under some combination of routines interrupting each other. Audio playout buffer management has been substantially reorganized]. 3) Apple's internal video for the Performa 630 (and some other models as well) should now work properly [Problem was that it cannot write 8-bit gray scale video to memory, only 16-bit color. Solution was to follow 16 bit frame grabbing with an 16 -> 8 bit translation. BTW, this experience confirms the validity of our "supported list" approach, as the Apple digitizer would crash if we asked it to grab an 8-bit gray scale frame... despite returning no error from the VDPreflightDestination call that is supposed to test the validity of a given configuration. Without our "supported list" logic, CU-SeeMe would have been unuseable even for receive-only on this class of machine]. Probably the 7500 and 8500 PowerMac video is not covered by this fix, but we have not received machines of these models yet. 4) A bug that would sometimes cause remote video to be scrambled when the source resolution was changed with a connection ongoing has been fixed [Problem was the receiver being configured for the wrong size. Interestingly, the problem appeared most frequently with a QuickCam, though it had no direct relation to the QuickCam, and would appear whenever a conference management message with the new configuration arrived before a video packet did] -----------------Changes in 0.83b1 1) some minor interface tweaks. 2) fixed bug causing connection time out followed by immediate resumption, manifested as participant list shrinking to none, then growing back. 3) fixed bug with wrong names appearing on window titles. 4) fixed bug with audio output level not working on 68K machines. 5) fixed bug with "please increase your memory to 2100K...." dialog mysteriously appearing at times in version 83a7. 6) some help with the recently reported problems with people masquerading as others in the Talk window. Now, you are not allowed to change your name while connected. ----------------Changes in 0.83a7 1) The algorithms used for measuring and responding to packet loss have been improved (they will be fully effective only when used with Reflector 4.00b3, available later running at the usual address at Cornell). 2) A new option for the Participant List has senders both hidden and showing combined in the same part of the list. They won't move around as you open and close them. 3) General cleanup of the UI, adjustment of gray shades, etc. has been done. 4) Active squelch pointer is now highlighted in audio window (upper one active when audio incoming from the net, lower otherwise). ----------------Changes in 0.83a5 1) Calibrations in the audio Panel. There are now db scales on the level meters. Zero db on the meters should indicate maximum modulation of the digital envelope (+127, -128 in 8bit). This level of modulation should represent an input signal of 2v RMS or +-2.83v PP input on a PowerMac, if you have the input gain set to 0 db also. These calibrations may not be exact (yet), but we're working on it. The output meter should show the content of the received digital samples and will not be affected by the setting of the output level control. When we get everything calibrated correctly, an output setting of 0 db (with input also set to 0) should give unity gain through the whole system: that is, the level at the line output of the receiving machine should equal (slightly delayed) that of the input at the sending machine. We are calibrating for PowerMac's since they have line level inputs and outputs. Earlier Mac's with low level (mic) inputs would need an appropriate attenuator. Signals which go above 0 db represent overmodulation or clipping. Ocaisional clippiing of speech is usually acceptable. We're still futzing with the damping for the meters and with access to extra controls, etc. 2) Refinements in Audio Window appearance: When in PTT mode (Push To Talk), the squelch thresholds do not appear. When not in PTT, the labels on the button change and the button is "pushed" automatically when you break squelch. 3) To hear just one participant, Option-Click their speaker icon (instead of Command-click). Option-Click again to restore former conditions. 4) Button bars on windows are constructed offscreen, so the annoying flicker is gone. ---------------Fixes in 0.83a4 1) crash trying to draw audio window icons or drawing wrong audio window icons is fixed. Also, icons "push in" properly again. 2) crash upon startup in non PTT mode is fixed 3) crash upon startup without digitizer is fixed 4) a non-functioning digitizer is treated the same as no digitizer (e.g., instead of error exit on Mac AV with monitor set too deep, the program continues in receive only mode) 5) fixed bug that prevented display of outgoing data rate on digitizer-less machines 6) point-to-point connections no longer require that at least one party have video Regarding 4, there is still a problem with PowerMac AV's which causes them to crash, rather than returning an error code that CU-SeeMe can respond to, when trying to access the digitizer when the monitor is set too deep. This problem may be fixed with an upgrade to the System software, but Tim's 7100/66AV with System 7.1.2 does crash. ----------Version 0.83a3 has the following updates: * Save Window Positions now works properly again. * The button bar icons, background color, recent/last speaker indicators are all in sync with those in the Participants List. Outstanding items/issues at this time (besides various bugs) are: * We're still fiddling with the appearance of icons, etc. * The button bars are updated unnecessarily (you can see them flicker) * VU scales in the Audio window are updated unnecessarily (if you put the cursor on the edge of the scale, you will see it flicker) * The audio input gain control does not operate smoothly. -------------Version 0.83a2 included: * Clear/Enter key works again properly for burnt text-in-video * Participant List no longer condenses long participant names * Participant List doesn't get lost in invisible desktop space anymore (checks for invalid position) * CU-SeeMe» Talk now works... (it was AutoDoubled accidentally by the server) * Color schemes matched * Lots of other small optimizations and features * You can expand the Audio Window by clicking in the Grow Box (upper right corner) to access more options... --------------Version 0.83a1 original notes: Major changes are the following: 1. There is a new algorithm and format for loss reports and more communitcation with the new version of the reflector. You now have parameters to set to limit your overall receive bw. 2. There is a participant list (Show Participant List from the Participants Menu) which will show information about senders and lurkers and provide control of audio, access to statistics, IP address, etc. There are several options in the Prefs dialog to control how the list is resized, and you can twiddle small triangles to show or hide portions of the list, in a fashion similar to finder list views. The Participants list essentially consists of the button bars for various users, giving you the audio controls and the "eye-con" that shows who is watching your video stream. Enhanced indication of who is sending audio, who was the last speaker, who has recently been speaking, etc. is provided as well. You can tell the difference between someone having their audio completely off versus just off for you (big X, little x). Note that details of these UI changes are still being adjusted; there will probably be additional alpha releases over the next week or so. 3. The Disconnect item in the Conference Menu now shows the IP of the other end of your connection, either user or reflector as the case may be. 4. There is a new Audio window with level indicators for both your mic (as before) and for incoming audio. Also there is an input gain control to adjust the sensitivity of the mic and an output volume control to adjust the speaker level. The input level meter indicates the level to be transmitted and is affected by the setting of the input gain as well as by how loud you speak. The output level control shows the level received in the packets (it should show the same level the person sending saw on their meter) and is unaffected by the output volume control--that just controls how loud it sounds at your end. Other features are experimental: a dual squelch control allows you to raise the squelch level when sound is being received from the net. 5. You can use the Control key as a push-to-talk activator instead of having to mouse to the right place in the audio window. CU-SeeMe needs to be in front, however. 6. If you hold down (instead of clicking) on someone's mic icon, it functions as a private-to-them push-to-talk. If you Command-click someones speaker icon, instead of turning them off, it turns everyone else off. 7. Lurker windows can be opened, and you get a local window if you are a lurker (just title bar and button bar) to give access to various controls. In particular, a lurker can access the control panels under the local window for audio and receive bandwidth. 8. The Participant List will show up to 24 participants (up from previous limit of 16 on the menu), so you can keep track of more senders/lurkers on a busy reflector. Be warned, though, that the conference management messages for 24 users can pretty well use up a 14.4 modem link. 9. QuickCam Audio works (yea!), although the sound is not the highest quality. QuickCam provides 5khz sampling, and the present implementation simply interpolates that up to 8khz so it will go thru the encoders as they stand. Puffing up the bandwidth utilization at the sending end by 8/5ths is clearly a sub-optimal idea, but to interoperate with everything it was seen as a necessary mode and a starting point. Larry Chace, Tim Dorcey, Jeff Han, and John Lynn contributed to the code and design for this release. ------- 0.80b1 --------- Slide Window stuff: 1. Slide Window AuxData type has been changed from 2 to 'CUsp' (transparent to user except different versions do not recognize each other-- i.e., 0.80a3 is obsolete.) 2. When you send a slide, the AuxData routines keep it available for receivers to request retrys of missing pieces until one of the following conditions is true: 1. A new slide is sent 2. MAX(30 seconds,twice original transmission time) elapses 3. The slide window is closed Previously, only 1. was controling. 3. Slides will not be received from a participant unless you are also receiving video from that participant. 4. The slide window menu item now alternates between "Open slide window" and "Close slide window" 5. The vertical dimension of the slide window is adjusted to fit on the monitor. Transmitted slides are the same size as is displayed, and will be positioned at the top of larger destination slide windows. Slides arriving from a larger source are truncated at the bottom. 6. The slide window large cursor will not appear and no remote pointing/queuing will be allowed unless the item has been sent at least once. Previously, it appeared to work, but nothing was sent. 7. When you leave Local mode, the slide that would have been queued had you not been in Local mode is queued. In other words, you catch up with the current state of the conference. 8. Fixed an anomaly in the button behavior that would occur if a new slide arrived while you were in Local mode, but yet had no slides in your stack 9. If you try to open SlideWindow without Quicktime installed, it will not be allowed, since SlideWindow uses the QT JPEG compressor. Even if you are receive-only, you need QuickTime in order to receive slides. Miscellaneous: 1. The application beeps if a connection request is received while you are in Waiting state (regardless of whether the acceptance dialog appears) 2. The application clicks whenever a new participant joins the conference, subject to a (new) preference item. 3. Video windows do not automatically come to foreground when they open, solving problem with Talk window typing being misrouted 4. Menu item for audio window brings it to the foreground (instead of just making it visible). 5. Reflector messages terminate the connection immediately (instead of waiting for user OK). 6. The reception hold down interval (during which time packets from a recently closed connection are ignored) has been increased from 30 sec. to 60 sec. Some comment on this is necessary to insure that it doesn't confuse people in point-to-point testing situations. 7. Audio squelch level is saved in preferences 8. Frame-Differencing and square decompression routines in PPC version have been modestly optimized. 9. Packet sequence number wrap-around is now allowed (to account for nv's 2-byte seqnum). 10. The kbps display in the rates bar now includes all packets sent or received on the CU-SeeMe port, with the exception of packets that are received from an IP address for which there is no rate bar to display it. Previously, only video was counted in the outgoing direction, and only _valid_ video, aux data, and OC packets were counted incoming. E.g., out of sequence packets, video packets for a closed window, etc. were not previously counted, nor was audio. 11. The transmission rate cap applies only to video+aux data, so OC packets and audio packets may cause the outgoing kbps to consistently exceed the cap. 12. The pull down stats display and the packet loss statistics that are used to adjust the rate cap are based on video + oc packets, and are only accumulated when video is being received. This approach is required since aux data and audio data have different sequence number spaces. This is essentially the same as before, except that there was a logic error that would sometimes allow a packet to enter the byte count without entering the packet count. Hence, the average packet size may have been over-estimated before. 13. Copyright and Permission notices have been modified: Contained in the application's About and also in the README.0.80.basic... file, it has been changed slightly: permission to modify has been deleted (not very meaningful with binary only, will be spelled out for Plug-In example sources), permissions granted explicitly only for the current version, reserving the right to modify permissions with future releases. ------- 0.80a3 --------- 1. A fix to the pop-up in the local window control panel for selecting digitizers which appeared in 80a2 (and in 70b13). The previous version *may* have introduced a freezing problem (although it appeared to work fine), especially when using System 7.5, especially on PowerMac's (we think). This fix is also in 0.70b15. 2. Please do not connect to "Self" as it will probably hang up your machine. ------- 0.80a2 --------- Note that the Plug-Ins are new and won't work with 80a1 and the old plugins won't work with 80a2. What's Different: 1) QuickCam support--The ability to select vdig (so you can use QuickCam) is brought forward from 0.70b14. However, the QuickCam won't produce a 640x480 image, so the SlideWindow is less useful than with a normal video source. 2) Delta-Mod 16K audio -- Charlie Kline's own 16Kbps audio encoding which first showed up in 70b13 has been brought forward. Remember that vat users will not be able to decode this signal; just Maven and CU-SeeMe for now. 3) The four required Apple Events are now supported, thanks to Aaron Giles. 4) DNS names are now supported in the connect dialog, but not yet in the Nickname file. A further release will have support for url format nicknames and saving bookmark files containing a url which will launch CU-SeeMe and connect to the site pointed to by the url. Apple events will also work for launching/connecting, so it should be possible to set up web brousers to establish CU-SeeMe sessions. But we havn't had time to wring this stuff out for this release. 5) There is a new preference item to open windows automatically when new video sources are discovered, which was previously the standard behavior. The new default is NOT automatic. That means that you would need to check this box to get the old way of working. The purpose of this change is to prevent naive first-time users from hooking to a busy reflector via a modem link, with default settings, and getting a flood of 8 windows most of which can do nothing but jam up their slip server. For the rest of us, it will allow you to get the folks on the screen you want and then not be annoyed by additional windows ceaslessly popping up. The full utility of this mode must await implemntation of the "Participants List" window which will have an entry for each particpant who is not displaying in a window. This option only applies to reflector connections. In 2-party, available video is always opened. 6) There is a new preference item to disable direct screen drawing, and omit black borders if direct screen drawing is disabled. Also, black borders are 2 pixels wide rather than 4. For those doing demo's with powerbooks and projectors, wanting to run in Mirror mode, this choice should allow the desired operation. 7) Default "Change Tolerance" is now 30 instead of 24. The result should be just a little less bandwidth, but little visible impact. On AV Mac's, I find that a value of 34 seems tolerable. 8) Change in IP address no longer wipes out old Preferences. 9) Preference file handling now allows compatibility across versions, so that we can add additional fields without destroying old information. Running an old version will, however, wipe out any new preference fields that it is not aware of. This will require a minor update to the 0.70 series (look for 0.70b15) to allow seamless switching between .80 and .70 series (except that running.70 will erase new preference items 5) and 6) above) 10) The "Local/Remote" button on slide projector behaves differently: it always displays as Local, but when you click it highlights (rev video) meaning you are *in* local mode. When it is normal black on white, it means you could click it to *go into* local mode. 11) The Aux Data facility has been improved in several ways. It can now transmit various items in an interleaved fashion, so that a large item will not delay the transfer of other items. It will repeat very short items a few times to enhance the probability of correct reception. It will avoid sending unnecessary control information. It will more efficiency handle error recover for multiple recipients. The plug-in interface now keeps Aux Data items around for 15 seconds after they have been sent greatly enhancing the probability of the items being received correctly. In our testing, the result is that the Talk window responds quickly and reliably, even when one or more slides are being transmitted. ------- 0.80a1 ------- OVERVIEW: new in 0.80 1) A "SlideWindow" facility with which you can capture 640x480 stills, transmit to all participients (who have SlideWindow open) and then point with the mouse and have everyone see the pointer. You can also switch among a number of slides and have the remote partipants see the same thing. SlideWindow is an "AuxData Application" which is invoked with the SlideWindow menu item on the Conference menu. 2) An "Auxilliary Data Transport" providing reliable one-to-many delivery of data constituting auxilliary materials for sharing in a conference. Data is transmitted as "Items," with an item being defined as a string of bytes contained in memory. The AuxData transport takes care of packetizing and any neccessary retrys to ensure reliable delivery. This transport is used by the built-in SlideWindow application or by plug-in AuxData applications (see below). 3) A Plug-In interface so that seperate modules can add functionality to CU-SeeMe. There will be (soon but not with the alpha release) a "Function Module Developer's Kit" consisting of documentation and sample plugins and a CodeWarrior project with libraries to facilitate development. If one or more plug-ins is in the same folder as CU-SeeMe at launchtime, a menu will be added (for each). 4) A "talk" plug-in (in source as an example) for sharing typed messages as with IRC. The Talk module implements a "chat" window that allows exchange of text messages. 5) An AuxData tracing tool for examining data flows when debugging or testing AuxData applications. The AD-Trace module is a testing or debugging tool that shows all "AuxData" traffic. 6) Also look for a new version of the reflector which will prevent sending of AuxData to participants who are not running the appropriate AuxData application (or who don't have it). This version is currently running at Cornell at the usual address, 132.236.91.204. You can use this reflector or you can make point-point connections to test the AuxData applications provided with the alpha 0.80. If any of the alpha-test group can run the new reflector, currently 2.50b1, please do and advise this group. With luck, there will also be a further enhanced reflector out soon, with additional facilities for interoperating with vat and nv. That version will also support AuxData transmission. ------- 0.70 b15 changes --------- 1. A fix to the pop-up in the local window control panel for selecting digitizers which appeared in b13. The b13 version *may* have introduced a freezing problem (although it appeared to work fine), especially when using System 7.5, especially on PowerMac's (we think). 2. A fix to a problem with handling Preferences files, especially when changeing back and forth from 0.70 to future (alpha) versions. This fix erroneously announced as being in b14, previously. -------- 0.70 b14 changes --------- 1. There is an additional audio encoding-- Charley Kline's delta-mod 16K bps encoding (thankyou again, Charley). This should work much better with low bandwidth links, such as 64K ISDN links, and possibly with 28.8K modem connections. However, compared to the DVI 32K encoding, assuming no packets lost in either case, the sound quality is noticibly poorer, although speech is entirely intelligible. You will not be able to use this encoding to converse with anyone using an earlier version of CU-SeeMe. You will not be able to interoperate with VAT. You should be able to use it with recent versions of Maven. ------- 0.70 b13 changes --------- 1. Support for the Connectix QuickCam (Yea!), the $100 combined camera and digitizer that plugs into your serial port, including an addition to the control panel you see under your own window that allows you to choose which of available digitizers to use. 2. Correct strange behavior when using a conference ID greater than 255. 3. Correct behavior when your own video is "paused" -- it was sending at much too high a rate. 4. Internal change dealing with packet sequence numbers and interoperation with VAT for periods longer than a couple hours. NOTE: This version will replace existing "CU-SeeMe Preferences" file, causing all settings to revert to default values. Before running this version, you may wish to take note of your current settings (e.g., Compression Tolerance, your name if not easily recalled, etc.). ------- 0.70 b12 changes --------- 1. Fix a bug introduced in 0.70b10 that caused audio connectivity to be flaky. In other words, some participants would receive your audio while others would not. 0.70 b11 changes: 1. Fix a bug in the processing of audio which potentially could be the cause of random crashes. 2. Allow incoming audio packets to continue to be processed even while the mouse is held down in the menus. For example, this allows incoming audio to continue to play out while you are looking through the list of participants in the Participants menu. Audio will continue to play out for 10 seconds at which point your connectiong will time-out in the reflector and it will stop sending you CUSM packets. ------- 0.70 b10 changes ------- 1. Nicknames >30 char no longer cause a crash on startup (but will be truncated to 30 characters). With the previous b9, the workaround for this bug was to avoid long nicknames. 2. Old Prefs files no longer cause crashes when opening a sound channel (as when clicking Push To Talk). With the previous b9, the workaround was to through away the preferences file and be sure to quit cleanly after using audio to save a good prefs file. 3. There is a small change to the OpenContinue message (CU-SeeMe's conference control protocol messages) as to the indication of changes-- result is much less work for the reflector as people come and go on a busy reflector. As before with bug fixes, there should be no user visible changes (apart from lack of crashes for the unwary). Reflector operators may notice many less log messages when folks have converted to this version. ------- 0.70b9 changes -------- 1. One case of "push to crash" related to wrong initialization of a sound channel. Not clear whether this fixes all cases. 2. A bughalt on a Mac IIcx because it was unable to initialize 4 sound channels. CU-SeeMe will now be happy if it can initialize at least one. It still trys to open 4 sound channels at initialization, but if it gets an error return after the first, if will settle for what it has. 3. Bughalts will now not send you to the dumper (or MaxBug) but will give a dialog with a message and then quit cleanly when you click OK. 4. There was a problem where the MinKbps control under Transmission under the Local window worked improperly. In particular, if the MinKbps were set to a value greater than 100, the first digit might be invisible. It could then appear that Min was, say, 10 but the Max could not be set below 110 or 210. ------- 0.70b5 --------- Be sure you use SoundManager 3.0 extension with Audio capable CU-SeeMe (but NOT on the PowerMac). CU-SeeMe0.70b5 is now available on gated.cornell.edu /pub/video for anonymous ftp. ***WARNING*** This is beta software. It is based on the previous 0.70b1 and contains no new features, only bug fixes. Several bugs in the area of audio have been fixed, and we do not expect that we have introduced new problems, but software folks always say that and aren't always correct, so beware. There is at least one crash producer still in this code. ------ New Stuff in 0.70 ------- 0.70b1 The principle enhancement in 0.70 versions is the inclusion of Audio, based on Charlie Kline's Maven code. It works with the new reflector and supports multiparty operation. BUT. ***Now Hear This*** (and tell your friends): AUDIO WILL NOT WORK CONNECTING WITH A MODEM!!! I know it would be great if it did, but it just doesn't. You need at least 32K bits/sec just for the audio (when speaking or listening) and modems only do 14.4K. If you decide to be a tester, watch out for the following: * Don't run CU-SeeMe0.70b1 and FTPd at the same time. At least on a Quadra 840AV it crashes after a while. Don't know if b5 changes this. * Problems have been reported with LC's and Quadra 650's using Spigots. * One particular Powerbook 180 and one PowerMac 8100 had crashes, but others worked fine. * Results trying to use MacRecorder (on older Mac's, like the venerable cx and ci, which don't have built-in audio) have been problematic. We've ordered a MacRecorder and will try to come up with a clear understanding. Any info/experience would be welcome. * Use the new reflector. If you hook to a reflector, and it's running the new code, you will probably get a Message Of The Day, telling you whatever the reflector operator wants you to know. 70bN will probably work with the old reflector code for video, but there may be problems with the audio. Shaking out these model-specific problems may take a while, since it's really hard for us to reproduce all the variations here. If you do experience crashes (and still want to continue), please try without any extra inits, or perhaps without other apps running; we have had problems with the moire screensaver, for example. REFLECTOR There is also a new version of the reflector, which you need to run to have the audio properly handled. (Note: we are one release further to b2 compared with the version released last week. You need this new version to support the new MOTD feature in the reflector/client.) It will also permit the use of stand-alone Maven. It is on the usual anonymous ftp site, gated.cornell.edu /pub/video. AUDIO Audio is controlled from a small window and works just like Maven. You can choose Push-To-Talk mode (PTT) or set a squelch level with a small pointer to the left of the audio-level bar. There are boxes in the audio window to turn receive and send on and off entirely. Or you can turn off audio from a particular particpant by pushing the speaker button under their video window (they will get the mic on your window X'd to indicate you won't hear when they speak). Also, you can choose to turn off receiving from standalone Maven clients who connect to the reflector (a temporary hack until we have a list-window for participants not displaying video). If you push the mic button on someone's window, you establish a private talk channel to them. Push again to go back to general conference audio sending. CONFERENCE MENU The Conections menu has been renamed to "Conference" since it now includes an item to open the Audio Window (if you've closed it). In future, there will be more conference-management related items on this menu. Of course, as time goes by, we may need to re-organize the menus to make sense of future developments which aren't even gleams in the eye today. THE EYE Also on remote window button bars is an eye-con (sorry, couldn't resist). If open, it means that the person in the window is looking at you-- or actually that they have a window open on their desktop showing your video stream. If they close the window, the eye at your end closes. When the eye is open, it looks a little sinistre, some folks think. Icons are still being edited. MOTD With the new reflector, you can set up a message which will be displayed to folks as they connect. (If they use 0.70 or later.) If you operate a reflector, you might want to use this facility to explain policies, etc. that apply to use of your reflector. SAVE WINDOW POSITIONS New on the File menu is an item you can use to save current window positions. Until you use it, they open tiling left to right. You probably want to move them around while you remember the order they came up and then save. The Participants menu lists the order that windows open as you initially connect to a reflector. You can go back to defaults by chucking your preferences file. This facility will be improved to make it easier to get the results you want. THE BLACK BORDER If part of the video portion of a CU-SeeMe window is covered, the program uses QuickDraw to move video into the window. Otherwise, optimized routines draw directly to the screen. Particularly on a slower Mac, drawing with QuickDraw is very slow. So you will know when this is happening, a black border shows on such a window. PREFERENCES LIMIT ON NUMBER OF WINDOWS You can set a preference for the maximum number of windows you want to show at one time (default 8, set between 2 and 8), including your local window. This feature will help if you have limited screen real-estate. CONFERENCE ID New on the Connection dialog is a box for conference ID, default of zero. The new reflector will reject all participants with the wrong number, and it can be configured with a conference ID other than zero. If both reflector and participants go with the defaults (zero), it works the way it always has. -------------Basic CU-SeeMe Info--------------- Get the latest file labeled CU-SeeMe.README.txt. for a consolidated collection of information on CU-SeeMe. CU-SeeMe was initially written for the Macintosh by Tim Dorcey with design assistance and sponsorship by Richard Cogger of the Advanced Technology group in the Network Resources division of Cornell University's Information Technology department (CIT). Important early contributions came from: Cornell University Medical Colleges (CUMC), Scott Brim, and John Lynn. Since Oct. 1, 1993, the CU-SeeMe Project receives funding from the National Science Foundation. A very significant collaborative effort at Cornell University Medical Colleges (CUMC) is contributing substantial expertise and code. This material is partially based on work sponsored by the National Science Foundation under Cooperative Agreement No. NCR-9318337. The Government has certain rights in this material. *****The fine print***** Copyright 1993, 1994, 1995, 1996, 1997 Cornell University Cornell hereby grants permission to use, copy, modify, and distribute this program for any purpose and without fee, provided that these copyright and permission notices, and those of Cornell's collaborators included below, appear on all copies and supporting documentation, the name of Cornell not be used in advertising or publicity pertaining to distribution of the program without specific prior permission, notice be given in supporting documentation that copying and distribution is by permission of Cornell. CORNELL MAKES NO REPRESENTATIONS OR WARRANTEES, EXPRESS OR IMPLIED. By way of example, but not limitation, CORNELL MAKES NO REPRESENTATIONS OR WARRANTIES OF MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OR THAT THE USE OF THIS SOFTWARE OR DOCUMENTATION WILL NOT INFRINGE ANY PATENTS, COPYRIGHTS, TRADEMARKS, OR OTHER RIGHTS. Cornell shall not be held liable for any liability with respect to any claim by the user or any other party arising from use of the program. The audio portion of CU-SeeMe was provided by Charley Kline's Maven. Charley Kline, cvk@uiuc.edu University of Illinois Computing and Communication Services 28 Feb 1994 Copyright (c) 1994 by Charley Kline and the University of Illinois Board of Trustees. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. 3. All advertising materials mentioning features or use of this software must display the following acknowledgement: This product includes software developed by the University of Illinois, Urbana and its contributors. 4. Neither the name of the University nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE TRUSTEES AND CONTRIBUTORS ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE TRUSTEES OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.