Transliterations responses 
Author Message
 Transliterations responses


>> What we need is not ANOTHER way to represent these, but AGREEMENT on
>> which SINGLE ONE to use.
>   I think this has essentially been done in the paper "Porting APL
> Programs via ASCII Transliteration" by J. Mitloehner in the APL92
> Proceedings.

 Mitloehner has come up with ANOTHER way to represent APL symbols, and it
 may be (according to some) a particularly nice way to solve the problem.
 The thrust of my posting was to gain agreement, which Mitloeher CANNOT
 have done in the paper.  This agreement can only take place here.

 Now, perhaps someone has done such a study, and perhaps that person
 is J. Mitloehner. HOWEVER, I see no strong move to adopt his stuff as
 the Right Way To Go.  I will certainly voice no opposition to adopting
 this scheme as a preferred method of posting.

 This leads to....

> But what mechanism should we use to pick one?
> 1. wait for the standards commitee to decide

     No. The standards committee has decided to be descriptive in nature.
     This boils down to meaning that the standards committee will not
     address this.
     (correct me, or make a more exact statement, Lee{*filter*}ey!).

> 2. have a vote here in this newsgroup

    No, I don't want a vote. Too many people know too few of the translations
    that are out there.  I think there is too little experience with more
    than 1 translation to make good judgements individually.

    Once we all have more experience at this, by using (reading, writing)
    a particular translation scheme, I would want to have a vote.

> 3. have a vote in quote quad

    No. Same problems as [2]. In addition, it's hard to get a good sample
    (poor response rates to postal surveys). We should try a standard (after
    study) see if it works, and then discuss it's points.  Once there is
    braod agreement on this, we can make reccomendations, etc.

> 4. pole the vendors

    Not a bad idea... BUT...
    Almost whatever we do will cause most vendors to spend time and money
    working on this issue.  IMHO, there are more pressing things for vendors
    to focus on! Once we have tried and tested a scheme here, we will have
    a stronger case to go on with vendors.  Also, interested vendors (those
    who are watching this newsfeed) can always contribute comments and/or
    If someone wants the vendors polled, I can supply (outside the newsfeed)
    a fairly definitive mailing list.  Canotact me privately if you are

> 5. just start writing code for your favorite and see what happens

     Also not a bad idea, but don't expect your email to be understood
     by many people if you use some not-so-well-known, but "good"

This is why someone has to take the list Lee has collected off into a corner
someplace, look at alternate methods, write some translation code, etc.
and post something.  This activity should be done in a quiet, peaceful place
where you can concentrate.  Then, we can all look at the results.

Now, Mitleonher HAS written some code, and made some nice choices. He has
written a paper about this work.  Perhaps our volunteer needs to read this
paper and look at the code as part of the study.  But I really urge that
the work be done as suggested. This approach will result in progress, first
for participants in this group, and possibly for all APL users in the future.

 |  Robert G. Brown                      |  "We don't believe in wasting    |
 |  Executive Editor, APL Quote-Quad     |   valuable Bandwidth on a lot    |


Tue, 10 Oct 1995 08:44:56 GMT  
 [ 1 post ] 

 Relevant Pages 

1. Function return values revisted (response to response)

2. Python 2.1.1 ASP/Response object does not HONOR Response.End()

3. httplib: vs

4. Slightly Off Topic : restrictions on the response part in a SOAP-response

5. APLASCII Transliteration Summary

6. Updated transliteration workspaces


8. Transliteration

9. ASCII Transliteration

10. APL Transliterations (and parsing)

11. APL Transliteration & WS Transfer

12. Transliteration schemes, draft 1


Powered by phpBB® Forum Software