lauri@svax.cs.cornell.edu (Georges Lauri) (12/13/88)
My esperience with Acknowledge has been more on the hegative side. I bought the product because I do a lot of telecom with Europe and wanted to automate a lot of it. Acknowledge is really just an interpreter; it comes with a set of demo "Communication Programs" that do things like BBS, terminal emulation, etc... The first problem I discovered were bugs in their programs, which are supposed to be usable as is. Since source code is delivered, it was not too hard to fix them, although their communication language is a monstrosity of ugliness (unless you like basic). The second problem was that their file transfer is limited in terms of protocols (only xmodem) and flaky. The third problem is that modifying the provided applications to say, suit my modem setup was extremely frustrating because all their programs are written in the worst of spaghetti code. So while the idea of Acknowledge is good, I found the implementation to be sorely lacking in power, flexibility, and cleanliness of organization. It does the job, but slowly, buggily, and I am not using it for the purpose I intended. Georges Lauri lauri@cs.cornell.edu
stuartb@microsoft.UUCP (Stuart Burden) (12/13/88)
In article <23387@cornell.UUCP> lauri@svax.cs.cornell.edu (Georges Lauri) writes: | My esperience with Acknowledge has been more on the | hegative side. I bought the product because I do a lot of | telecom with Europe and wanted to automate a lot of it. | Acknowledge is really just an interpreter; it comes with a | set of demo "Communication Programs" that do things like | BBS, terminal emulation, etc... The first problem I | discovered were bugs in their programs, which are supposed | to be usable as is. Since source code is delivered, it was | not too hard to fix them, although their communication | language is a monstrosity of ugliness (unless you like | basic). The language is a little obtuse, but it's great advantage is that you can change it on the fly, to debug (even while online, if it's written correctly), and the whole process is very interactive. I have found that the best approach is not to try to modify someone elses code in Acknowledge. Most of the sample applications supplied seem to be written by different people, this means that the source is commented in a different manner, sometimes well, sometimes not. Sometimes the code is really threaded, and in other examples it is quite structured. It does appear kind of piece-meal when you first look at it. I found that by printing out the samples and looking for neat techniques to acomplish particular tasks I had in mind, was the easiest way to put the samples to use, rather than using them as written. After a short period, I also discovered that modifying any of the sample code was definately not the way to go, cos your just too confined to how someone else invisaged thier program to be, and your forever falling over what someone else wrote. I started from scratch and built up a complete Mac-Xenix e-mail system, and I'm satisfied with it. This is not to say that there are not points in Acknowledge that don't always glow with gold or that I'd never change a thing about the operation or functionality of Acknowledge, if I had the chance. Acknowledge is a telecommunications construction kit, and I think if it is viewed in this light, then it makes sense that a reasonable amount of development effort will have to be put into developing the kinds of applications that are specific to individual needs. I've found only a few bugs in Acknowledge, so I would certainly like to know what kinds of problems you've been having, perhaps you can e-mail me (if they are code problems, I don't want to know.. I've long since stopped looking at the samples for guidence, and prefer to roll my own, but perhaps you might take the time to e-mail SuperMac). I've found that intermittent bus errors on a Mac ][, seemed to only happen when my source reached and hovered around a particular size. I've also had problems with line noise causing Acknowledge to lock up the serial port. There is also a problem with text loss, at uninterupted high speed text sends (9600 baud). Red Ryder and Acknowledge are the only comms apps I have that do this when sending text to my host. Microphone, Versaterm and pcLink (now PacerLink) work 100% with no line or character delay, and I get screaming throughput. Acknowledge also had screaming throughput, however, without substantial line delays, causes a lot of text to be lost, and thus decreases functional throughput. I have the feeling that Acknowledge is munging with one of the port parameters to do this, cos if I run pcLink after Acknowledge, I get text losses from pcLink also, and that is the only time I have ever seen text corruption in pcLink. If anyone can shed some light on this, I'd appreciate it. | The second problem was that their file transfer is | limited in terms of protocols (only xmodem) and flaky. Text, XMODEM and MacBinary are supported. YMODEM and Kermit are not supported, but am told that Kermit will be a plug in module supplied by Apple with the comms manager. YMODEM, I don't know when we are likely to see this. I've had no problems at all with either MacBinary or XMODEM. | The third problem is that modifying the provided | applications to say, suit my modem setup was extremely | frustrating because all their programs are written in the | worst of spaghetti code. So while the idea of Acknowledge | is good, I found the implementation to be sorely lacking | in power, flexibility, and cleanliness of organization. Hmm, I tend to disagree with only one point here, and that is the flexability issue. Appart from a few annoyances, like not being able to create array variables from a lookup table (in fact general inability to use arrays outside of specific dialog functions), flexability is one of it's best points. I guess it depends on how much you want to do and how much you expect Acknowledge to do for you. | It does the job, but slowly, buggily, and I am not using | it for the purpose I intended. Yes there are some areas that are painfully slow, but if managed correctly then they are less obtrusive. When you write anything with Acknowledge you must think "basic".. in fact it helps if you think "Applesoft Basic", and structure your programs accordingly. Lookup Table variables take an age to load in. Double clicking a line is not nearly as responsive as it should be. An additional command enhancment to the LOOK FOR function would be LOOK FOR DoubleClick (look for triple click would be nice also, cos currently the triple click function is just not reliable enough). All in all, Acknowledge, although it doesn't shine brightly all the time, it's still worth a look for anyone considering doing custom telecommunications programming. | Georges Lauri | lauri@cs.cornell.edu Stu. __Paths to my door:_______________________ microsoft!stuartb@beaver.cs.washington.edu - Usual disclaimer, that all microsoft!stuartb@uw-beaver.arpa - the above is pure fantasy microsoft!stuartb@uunet.UU.NET - and Microsoft only [DE01HB]stuartb@DASNET# {from AppleLink} - gave me the Mountain Dew stuartb@microsoft.uucp {well connected} - to dream it all in a D2012 {AppleLink - shared account} - caffeine haze :-) __________________________________________________________________________
jay@mitisft.Convergent.COM (Jay O'Conor) (12/16/88)
In article <61@microsoft.UUCP>, stuartb@microsoft.UUCP (Stuart Burden) writes: > > Text, XMODEM and MacBinary are supported. YMODEM and Kermit are not > supported, but am told that Kermit will be a plug in module supplied by > Apple with the comms manager. YMODEM, I don't know when we are likely to > see this. > Acknowledge will only be truly useful (to me) when it allows the creation of other communication protocols. My immediate need is for the UUCP 'g' protocol. After that, It would be desirable to have it deal with UUCP over Ethernet via AppleTalk connected to a Kinetics Fastpath or equivalent. Other protocols would be nice, too. It would be interesting to attempt to duplicate the AppleLink or MacNET protocols to allow the creation of a custom front end to these services using Acknowledge. At the August MacWorld Expo I asked SuperMac what their plans were for being able to support other protocols, and of course their answer was "wait for Apple to release the comm manager". Anybody know anything about the comm manager? The only knowledge I have is it's ability to handle more communications ports on a board in an SE or II. What will it's relationship be with a program like Acknowledge? Jay O'Conor Convergent Technologies