Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > alt.folklore.computers > #163800 > unrolled thread

CS instead of math?

Started byusenet@only.tnx (Questor)
First post2016-05-21 00:57 +0000
Last post2016-06-02 19:47 -0700
Articles 20 on this page of 136 — 23 participants

Back to article view | Back to alt.folklore.computers

Contents

  CS instead of math? usenet@only.tnx (Questor) - 2016-05-21 00:57 +0000
    Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-20 22:13 -0400
      Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-05-21 11:39 +0000
    Re: CS instead of math? bert <bert.hutchings@btinternet.com> - 2016-05-21 13:23 -0700
      Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-21 22:36 -0400
        Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-05-24 18:00 +0000
          Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-24 20:27 +0200
          Re: CS instead of math? Bob Martin <bob.martin@excite.com> - 2016-05-25 06:56 +0100
            Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-25 07:42 +0100
              Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-25 09:39 +0200
                Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-25 09:37 +0100
                  Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-05-25 13:05 +0000
                  Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-26 07:07 -0400
                    Re: CS instead of math? Morten Reistad <first@Last.name.invalid> - 2016-05-26 13:34 +0200
                    Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-26 14:02 +0100
                      Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-26 19:46 -0400
                      Re: CS instead of math? Dave Garland <dave.garland@wizinfo.com> - 2016-05-26 19:17 -0500
                        Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-05-27 06:56 -0400
                Re: CS instead of math? JimP <solosam90@gmail.com> - 2016-05-25 07:26 -0500
                  Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-25 14:40 +0200
                Re: CS instead of math? Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-05-25 16:19 -0300
                  Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-05-25 16:15 -0400
                  Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-25 22:29 +0200
                  Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-05-25 21:19 +0000
                Re: CS instead of math? sidd@situ.com (sidd) - 2016-05-26 16:34 -0400
            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-26 07:04 -0400
              Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-26 13:28 +0200
                Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-26 14:06 +0100
                  Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-26 18:57 +0200
                    Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-26 18:19 +0100
                    Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-26 18:35 +0100
                    Re: CS instead of math? cb@elaine.df.lth.se (Christian Brunschen) - 2016-05-26 22:05 +0000
                      Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-05-27 15:33 +0000
                        Re: CS instead of math? Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-05-27 15:40 -0300
              Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-05-26 14:43 +0000
                Re: CS instead of math? JimP <solosam90@gmail.com> - 2016-05-26 11:03 -0500
                  Re: CS instead of math? scott@slp53.sl.home (Scott Lurndal) - 2016-05-26 16:27 +0000
                    Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-26 17:52 +0100
                Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-05-26 16:22 -0400
              Re: CS instead of math? usenet@only.tnx (Questor) - 2016-05-31 17:27 +0000
                Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-05-31 20:06 +0000
                  Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-31 21:08 -0400
                    Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 07:17 +0100
                      Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 05:31 -0400
                        Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 11:56 +0100
                          Re: CS instead of math? Andrew Swallow <am.swallow@btopenworld.com> - 2016-06-01 16:00 +0100
                            Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 17:15 +0100
                              Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-06-01 19:28 +0000
                          Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-06-01 17:42 +0000
                            Re: CS instead of math? Gene Wirchenko <genew@telus.net> - 2016-06-01 13:03 -0700
                          Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 17:05 -0400
                      Re: CS instead of math? Gene Wirchenko <genew@telus.net> - 2016-06-01 12:48 -0700
                        Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-06-01 17:34 -0400
                          Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-02 07:50 +0100
                            Re: CS instead of math? Bob Eager <news0006@eager.cx> - 2016-06-02 10:37 +0000
                              Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-02 12:33 +0100
                                Re: CS instead of math? Bob Eager <news0006@eager.cx> - 2016-06-02 12:08 +0000
                          Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-06-02 08:17 +0000
                            Re: CS instead of math? Andy Burns <feb2017-usenet@adslpipe.co.uk> - 2016-06-02 09:49 +0100
                              Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-02 12:09 +0100
                                Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-06-03 13:16 +0000
                              Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-06-02 12:09 +0000
                                Re: CS instead of math? Andy Burns <feb2017-usenet@adslpipe.co.uk> - 2016-06-02 14:09 +0100
                                  Re: CS instead of math? Andy Burns <feb2017-usenet@adslpipe.co.uk> - 2016-06-02 14:12 +0100
                                    Re: CS instead of math? Gene Wirchenko <genew@telus.net> - 2016-06-02 19:44 -0700
                                  Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-06-03 13:16 +0000
                                Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-06-02 15:02 +0200
                                  Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-06-02 16:44 +0000
                                Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-06-02 16:43 +0000
                              Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-06-02 16:42 +0000
                            Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-06-02 16:41 -0400
                              Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-06-02 22:18 +0000
                    Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-06-01 17:42 +0000
                      Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-06-01 22:03 +0200
                  Re: CS instead of math? Gene Wirchenko <genew@telus.net> - 2016-06-01 12:47 -0700
                Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-31 20:58 -0400
                  Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 07:29 +0100
                    Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-06-01 11:59 +0200
                      Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 06:48 -0400
                        Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 12:41 +0100
                          Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 17:04 -0400
                            Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-02 07:59 +0100
                              Re: CS instead of math? Huge <Huge@nowhere.much.invalid> - 2016-06-02 08:18 +0000
                              Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-02 19:53 -0400
                      Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 12:15 +0100
                      Re: CS instead of math? Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2016-06-01 17:42 +0000
                Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-06-01 13:51 +0000
      Re: CS instead of math? "gareth G4SDW GQRP #3339" <no.spam@thank.you.invalid> - 2016-05-22 11:27 +0100
        Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-22 07:21 -0400
          Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-05-22 13:15 +0000
            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-22 09:53 -0400
          Re: CS instead of math? usenet@only.tnx (Questor) - 2016-05-25 07:18 +0000
            Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-25 09:42 +0200
              Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-05-25 13:11 +0000
                Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-05-25 16:15 -0400
                  Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-05-25 22:31 +0200
                    Re: CS instead of math? Andrew Swallow <am.swallow@btopenworld.com> - 2016-05-26 11:23 +0100
                    Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-05-26 11:37 +0100
                Re: CS instead of math? usenet@only.tnx (Questor) - 2016-05-26 17:56 +0000
              Re: CS instead of math? sidd@situ.com (sidd) - 2016-05-26 16:36 -0400
                Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-26 19:51 -0400
            Re: CS instead of math? "Osmium" <r124c4u102@comcast.net> - 2016-05-25 07:43 -0500
              Re: CS instead of math? usenet@only.tnx (Questor) - 2016-05-26 17:55 +0000
            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-26 07:14 -0400
              Re: CS instead of math? usenet@only.tnx (Questor) - 2016-05-27 17:40 +0000
                Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-27 18:46 -0400
                  Re: CS instead of math? usenet@only.tnx (Questor) - 2016-05-31 17:28 +0000
                    Re: CS instead of math? Peter Flass <peter_flass@yahoo.com> - 2016-05-31 16:09 -0400
                    Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-05-31 21:07 -0400
                      Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 08:58 +0100
                        Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 05:48 -0400
                          Re: CS instead of math? Morten Reistad <first@last.name.invalid> - 2016-06-01 12:06 +0200
                            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 06:46 -0400
                            Re: CS instead of math? sidd@situ.com (sidd) - 2016-06-03 16:05 -0400
                              Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-03 20:49 -0400
                                Re: CS instead of math? sidd@situ.com (sidd) - 2016-06-04 22:29 -0400
                          Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-01 12:27 +0100
                            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 17:07 -0400
                              Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-02 08:12 +0100
                                Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-02 19:56 -0400
                                  Re: CS instead of math? Ahem A Rivet's Shot <steveo@eircom.net> - 2016-06-03 09:17 +0100
                          Re: CS instead of math? Gene Wirchenko <genew@telus.net> - 2016-06-01 13:20 -0700
                            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-01 17:10 -0400
                          Re: CS instead of math? usenet@only.tnx (Questor) - 2016-06-02 17:58 +0000
                            Re: CS instead of math? "Osmium" <r124c4u102@comcast.net> - 2016-06-02 13:57 -0500
                              Re: CS instead of math? Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-06-02 17:43 -0300
                            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-02 20:08 -0400
                      Re: CS instead of math? usenet@only.tnx (Questor) - 2016-06-02 17:58 +0000
                        Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-02 20:05 -0400
                          Re: CS instead of math? Bob Eager <news0006@eager.cx> - 2016-06-03 10:38 +0000
                            Re: CS instead of math? "gareth G4SDW GQRP #3339" <no.spam@thank.you.invalid> - 2016-06-03 11:41 +0100
                            Re: CS instead of math? jmfbahciv <See.above@aol.com> - 2016-06-03 13:16 +0000
                          Re: CS instead of math? usenet@only.tnx (Questor) - 2016-06-09 05:49 +0000
                            Re: CS instead of math? "J. Clarke" <j.clarke.873638@gmail.com> - 2016-06-09 19:58 -0400
                              Re: CS instead of math? usenet@only.tnx (Questor) - 2016-06-17 03:35 +0000
                        Re: CS instead of math? Gene Wirchenko <genew@telus.net> - 2016-06-02 19:47 -0700

Page 4 of 7 — ← Prev page 1 2 3 [4] 5 6 7  Next page →

#164411

Fromjmfbahciv <See.above@aol.com>
Date2016-06-03 13:16 +0000
Message-ID<PM0005345ECE857D5D@aca40481.ipt.aol.com>
In reply to#164306
Ahem A Rivet's Shot wrote:
> On Thu, 2 Jun 2016 09:49:08 +0100
> Andy Burns <feb2017-usenet@adslpipe.co.uk> wrote:
>
>> Huge wrote:
>>
>> > I spent a LOT of time before retiring (last year)
>> > working on restricting the use of live data in testing
>>
>> One system I look after (the o/s and networking parts rather than the
>> software) takes a nightly copy of the live database into the test
>> database, in the process it mangles all the human readable names,
>> addresses, and identifiers. This gives the same size of data with the
>> same relations between rows in the tables to work on.
>
>     Yep distorted real data is often the best test data - at one PPOE
> we had vast amounts of bank data (from many banks) with real names and
> addresses but faked numbers (one set we had to re-randomise because they
> were way out of reasonable range) because we needed to work with real world
> name and address data. Those were sensitive data sets but nothing like as
> sensitive (internally at least) as they would have been with the real
> numbers attached.
>
Of course, junky data also has to be tested.

/BAH

[toc] | [prev] | [next] | [standalone]

#164312

Fromjmfbahciv <See.above@aol.com>
Date2016-06-02 12:09 +0000
Message-ID<PM00053449D55C3826@aca41013.ipt.aol.com>
In reply to#164301
Andy Burns wrote:
> Huge wrote:
>
>> I spent a LOT of time before retiring (last year)
>> working on restricting the use of live data in testing
>
> One system I look after (the o/s and networking parts rather than the
> software) takes a nightly copy of the live database into the test
> database, in the process it mangles all the human readable names,
> addresses, and identifiers. This gives the same size of data with the
> same relations between rows in the tables to work on.
>
> I think it uses a temporary hash to form the mangled version then throws
> it away before committing to the test database, I suppose the mangling
> could be subverted by the developer but he (one left from the original
> team of three) could subvert the system anyway.
>
> It allows testing or demonstrating of problems to the likes of me
> without seeing patient data, and when I need to test printing it ends up
> looking like it's in Klingon, hopefully no pharmacy would ever issue
> such a methadone prescription!
>
How do you test anything which needed to be sorted if all the fields
have junk in them?

/BAH

[toc] | [prev] | [next] | [standalone]

#164319

FromAndy Burns <feb2017-usenet@adslpipe.co.uk>
Date2016-06-02 14:09 +0100
Message-ID<drapg1F89edU1@mid.individual.net>
In reply to#164312
jmfbahciv wrote:

> How do you test anything which needed to be sorted if all the fields
> have junk in them?

The substitution uses English-like syllables, but junk still sorts and 
you humans can tell if it is sorted or not.

[toc] | [prev] | [next] | [standalone]

#164320

FromAndy Burns <feb2017-usenet@adslpipe.co.uk>
Date2016-06-02 14:12 +0100
Message-ID<draplhF89edU2@mid.individual.net>
In reply to#164319
Andy Burns wrote:

> The substitution uses English-like syllables, but junk still sorts and
> you humans can tell if it is sorted or not.

I don't know where that "you" crept in from, it made me smile, I think 
I'm human too ...

[toc] | [prev] | [next] | [standalone]

#164377

FromGene Wirchenko <genew@telus.net>
Date2016-06-02 19:44 -0700
Message-ID<rnr1lbdqmc7ctj5gflsnpl6k07t3hjb9v3@4ax.com>
In reply to#164320
On Thu, 2 Jun 2016 14:12:18 +0100, Andy Burns
<feb2017-usenet@adslpipe.co.uk> wrote:

>Andy Burns wrote:
>
>> The substitution uses English-like syllables, but junk still sorts and
>> you humans can tell if it is sorted or not.
>
>I don't know where that "you" crept in from, it made me smile, I think 
>I'm human too ...

     "People are people, but programmers are something else."

Sincerely,

Gene Wirchenko

[toc] | [prev] | [next] | [standalone]

#164406

Fromjmfbahciv <See.above@aol.com>
Date2016-06-03 13:16 +0000
Message-ID<PM0005345EC343E1C7@aca40481.ipt.aol.com>
In reply to#164319
Andy Burns wrote:
> jmfbahciv wrote:
>
>> How do you test anything which needed to be sorted if all the fields
>> have junk in them?
>
> The substitution uses English-like syllables, but junk still sorts and
> you humans can tell if it is sorted or not.

As someone who worked using the IBM sorter, data doesn't behave well
when it involves proper nouns and numbers.

/BAH

[toc] | [prev] | [next] | [standalone]

#164322

FromMorten Reistad <first@last.name.invalid>
Date2016-06-02 15:02 +0200
Message-ID<jbn52d-tff.ln1@sambook.reistad.name>
In reply to#164312
In article <PM00053449D55C3826@aca41013.ipt.aol.com>,
jmfbahciv  <See.above@aol.com> wrote:
>Andy Burns wrote:
>> Huge wrote:
>>
>>> I spent a LOT of time before retiring (last year)
>>> working on restricting the use of live data in testing
>>
>> One system I look after (the o/s and networking parts rather than the
>> software) takes a nightly copy of the live database into the test
>> database, in the process it mangles all the human readable names,
>> addresses, and identifiers. This gives the same size of data with the
>> same relations between rows in the tables to work on.
>>
>> I think it uses a temporary hash to form the mangled version then throws
>> it away before committing to the test database, I suppose the mangling
>> could be subverted by the developer but he (one left from the original
>> team of three) could subvert the system anyway.
>>
>> It allows testing or demonstrating of problems to the likes of me
>> without seeing patient data, and when I need to test printing it ends up
>> looking like it's in Klingon, hopefully no pharmacy would ever issue
>> such a methadone prescription!
>>
>How do you test anything which needed to be sorted if all the fields
>have junk in them?

It is not junk. It is encrypted transformations of the data.

Example from telephone logs.

There we have fields like caller, callee, when, duration etc.
The caller and callee are telephone numbers. These must match the
routing tables, which are prefix, cost based.

So the scambler matches the caller and callee with the cost
prefix, keeps the numbers that are necessary for that, and scrambles
the rest. The duration is kept, but the time is moved a random 
time back or forward with a 12 hour limit.

e.g. 

I call my parents, caller=4722688356, callee=4755280165, when=1.june.2005 12:33, duration=4m38sec
prefix for the caller is 472 and for the callee 475. So the record in the test
version becomes caller=4720123456, callee=4750987654, when=1.june 2005 18:14 duration=4m38sec.

All the billing code should work fine with this. It is also important that the
scrambled numbers are the same for every scrambling in that run.

Doing such stuff requires a bit of insight into the nature of the data, and
the limits of their confidensiality.

-- mrr

[toc] | [prev] | [next] | [standalone]

#164329

FromHuge <Huge@nowhere.much.invalid>
Date2016-06-02 16:44 +0000
Message-ID<drb639FaofbU4@mid.individual.net>
In reply to#164322
On 2016-06-02, Morten Reistad <first@last.name.invalid> wrote:
> In article <PM00053449D55C3826@aca41013.ipt.aol.com>,
> jmfbahciv  <See.above@aol.com> wrote:
>>Andy Burns wrote:
>>> Huge wrote:
>>>
>>>> I spent a LOT of time before retiring (last year)
>>>> working on restricting the use of live data in testing
>>>
>>> One system I look after (the o/s and networking parts rather than the
>>> software) takes a nightly copy of the live database into the test
>>> database, in the process it mangles all the human readable names,
>>> addresses, and identifiers. This gives the same size of data with the
>>> same relations between rows in the tables to work on.
>>>
>>> I think it uses a temporary hash to form the mangled version then throws
>>> it away before committing to the test database, I suppose the mangling
>>> could be subverted by the developer but he (one left from the original
>>> team of three) could subvert the system anyway.
>>>
>>> It allows testing or demonstrating of problems to the likes of me
>>> without seeing patient data, and when I need to test printing it ends up
>>> looking like it's in Klingon, hopefully no pharmacy would ever issue
>>> such a methadone prescription!
>>>
>>How do you test anything which needed to be sorted if all the fields
>>have junk in them?
>
> It is not junk. It is encrypted transformations of the data.

This is a good read;

https://en.wikipedia.org/wiki/Tokenization_%28data_security%29

-- 
Today is Pungenday, the 7th day of Confusion in the YOLD 3182
                  I don't have an attitude problem.
    If you have a problem with my attitude, that's your problem.

[toc] | [prev] | [next] | [standalone]

#164328

FromHuge <Huge@nowhere.much.invalid>
Date2016-06-02 16:43 +0000
Message-ID<drb60tFaofbU3@mid.individual.net>
In reply to#164312
On 2016-06-02, jmfbahciv <See.above@aol.com> wrote:
> Andy Burns wrote:
>> Huge wrote:
>>
>>> I spent a LOT of time before retiring (last year)
>>> working on restricting the use of live data in testing
>>
>> One system I look after (the o/s and networking parts rather than the
>> software) takes a nightly copy of the live database into the test
>> database, in the process it mangles all the human readable names,
>> addresses, and identifiers. This gives the same size of data with the
>> same relations between rows in the tables to work on.
>>
>> I think it uses a temporary hash to form the mangled version then throws
>> it away before committing to the test database, I suppose the mangling
>> could be subverted by the developer but he (one left from the original
>> team of three) could subvert the system anyway.
>>
>> It allows testing or demonstrating of problems to the likes of me
>> without seeing patient data, and when I need to test printing it ends up
>> looking like it's in Klingon, hopefully no pharmacy would ever issue
>> such a methadone prescription!
>>
> How do you test anything which needed to be sorted if all the fields
> have junk in them?

That's one of the reasons that data anonymisation is very hard ...

:o)

-- 
Today is Pungenday, the 7th day of Confusion in the YOLD 3182
                  I don't have an attitude problem.
    If you have a problem with my attitude, that's your problem.

[toc] | [prev] | [next] | [standalone]

#164327

FromHuge <Huge@nowhere.much.invalid>
Date2016-06-02 16:42 +0000
Message-ID<drb5v1FaofbU2@mid.individual.net>
In reply to#164301
On 2016-06-02, Andy Burns <feb2017-usenet@adslpipe.co.uk> wrote:
> Huge wrote:
>
>> I spent a LOT of time before retiring (last year)
>> working on restricting the use of live data in testing
>
> One system I look after (the o/s and networking parts rather than the 
> software) takes a nightly copy of the live database into the test 
> database, in the process it mangles all the human readable names, 
> addresses, and identifiers. 

Very, very hard to do well. Especially since we had use cases where things
like customer names and address *were* the data to be tested with, so
could not be desensitised without destroying their utility.

> This gives the same size of data with the 
> same relations between rows in the tables to work on.
>
> I think it uses a temporary hash to form the mangled version then throws 
> it away before committing to the test database, I suppose the mangling 
> could be subverted by the developer but he (one left from the original 
> team of three) could subvert the system anyway.
>
> It allows testing or demonstrating of problems to the likes of me 
> without seeing patient data, 

That's the kind of thing we were working on.


-- 
Today is Pungenday, the 7th day of Confusion in the YOLD 3182
                  I don't have an attitude problem.
    If you have a problem with my attitude, that's your problem.

[toc] | [prev] | [next] | [standalone]

#164354

FromPeter Flass <peter_flass@yahoo.com>
Date2016-06-02 16:41 -0400
Message-ID<1468754751.486592510.050241.peter_flass-yahoo.com@news.eternal-september.org>
In reply to#164299
Huge <Huge@nowhere.much.invalid> wrote:
> On 2016-06-01, Peter Flass <peter_flass@yahoo.com> wrote:
>> Gene Wirchenko <genew@telus.net> wrote:
>>> On Wed, 1 Jun 2016 07:17:39 +0100, Ahem A Rivet's Shot
>>> <steveo@eircom.net> wrote:
>>> 
>>> [snip]
>>> 
>>>> Edge cases are another matter, sometimes there's more code for the
>>>> edge cases than there is for the main flow.
>>> 
>>> Sometimes?  I would say it is at least usually.
>>> 
>> 
>> This is often a big shock to new college graduates starting their first
>> real job.  Most assignments the prof just wants to see something that runs
>> using the data provided, and there is limited time to get it done, so
>> usually no attention is paid to detecting and handling errors.  All of a
>> sudden the newbie has to write a program that will be handed any old
>> garbage data, and just ABENDing is not a valid option. (speaking from
>> experience, even if ancient)
> 
> Still goes on today. I spent a LOT of time before retiring (last year)
> working on restricting the use of live data in testing for security
> reasons.  One of the (true) reasons the developers gave for wanting to
> use live data for testing was that the production databases contained
> a lot of instances of pathological combinations of data that no-one had
> tested for because either they hadn't even thought of them, or everyone
> had assumed they were impossible.
> 

I had one customer that didn't want to release live data, and as a result
lots of problems showed up that might have been fixed in test.  Finally
they just scrubbed any identifying info and gave me the complete dataset.


-- 
Pete

[toc] | [prev] | [next] | [standalone]

#164362

FromHuge <Huge@nowhere.much.invalid>
Date2016-06-02 22:18 +0000
Message-ID<drbpmiFeq3jU2@mid.individual.net>
In reply to#164354
On 2016-06-02, Peter Flass <peter_flass@yahoo.com> wrote:
> Huge <Huge@nowhere.much.invalid> wrote:
>> On 2016-06-01, Peter Flass <peter_flass@yahoo.com> wrote:
>>> Gene Wirchenko <genew@telus.net> wrote:
>>>> On Wed, 1 Jun 2016 07:17:39 +0100, Ahem A Rivet's Shot
>>>> <steveo@eircom.net> wrote:
>>>> 
>>>> [snip]
>>>> 
>>>>> Edge cases are another matter, sometimes there's more code for the
>>>>> edge cases than there is for the main flow.
>>>> 
>>>> Sometimes?  I would say it is at least usually.
>>>> 
>>> 
>>> This is often a big shock to new college graduates starting their first
>>> real job.  Most assignments the prof just wants to see something that runs
>>> using the data provided, and there is limited time to get it done, so
>>> usually no attention is paid to detecting and handling errors.  All of a
>>> sudden the newbie has to write a program that will be handed any old
>>> garbage data, and just ABENDing is not a valid option. (speaking from
>>> experience, even if ancient)
>> 
>> Still goes on today. I spent a LOT of time before retiring (last year)
>> working on restricting the use of live data in testing for security
>> reasons.  One of the (true) reasons the developers gave for wanting to
>> use live data for testing was that the production databases contained
>> a lot of instances of pathological combinations of data that no-one had
>> tested for because either they hadn't even thought of them, or everyone
>> had assumed they were impossible.
>> 
>
> I had one customer that didn't want to release live data, 

That's just as dumb as throwing live data about like confetti. If you
don't test with live data at some point, the first time your code will
see live data is the first production run. And you don't want that.


-- 
Today is Pungenday, the 7th day of Confusion in the YOLD 3182
                  I don't have an attitude problem.
    If you have a problem with my attitude, that's your problem.

[toc] | [prev] | [next] | [standalone]

#164263

FromCharlie Gibbs <cgibbs@kltpzyxm.invalid>
Date2016-06-01 17:42 +0000
Message-ID<nin6pu2ipl@news4.newsguy.com>
In reply to#164202
On 2016-06-01, J. Clarke <j.clarke.873638@gmail.com> wrote:

> In article <nikqro01pm6@news4.newsguy.com>, cgibbs@kltpzyxm.invalid 
> says...
> 
>> On 2016-05-31, Questor <usenet@only.tnx> wrote:
>> 
>>> Testing cannot definitively prove that code is correct.
>> 
>> Even if it could, it can't determine whether the specification to
>> which the code was written has any bearing on what the customer needs.
>> Or wants.  Or will want next week.
>
> Well, that depends on the testing.  If the testing involves the customer 
> sitting in front of the machine, testing can very quickly determine 
> whether the code has any bearing on what he wants.

True.  Unfortunately, there are often intermediaries who get in the way,
and who turn the whole exercise into the telephone game.

I remember the first time I actually got to sit with a customer and
talk about what he really needed.  It was amazing how we were able
to clear the air.  But by then, I had already turned in my notice...

-- 
/~\  cgibbs@kltpzyxm.invalid (Charlie Gibbs)
\ /  I'm really at ac.dekanfrus if you read it the right way.
 X   Top-posted messages will probably be ignored.  See RFC1855.
/ \  HTML will DEFINITELY be ignored.  Join the ASCII ribbon campaign!

[toc] | [prev] | [next] | [standalone]

#164274

FromMorten Reistad <first@last.name.invalid>
Date2016-06-01 22:03 +0200
Message-ID<cmr32d-cla.ln1@sambook.reistad.name>
In reply to#164263
In article <nin6pu2ipl@news4.newsguy.com>,
Charlie Gibbs  <cgibbs@kltpzyxm.invalid> wrote:
>On 2016-06-01, J. Clarke <j.clarke.873638@gmail.com> wrote:
>
>> In article <nikqro01pm6@news4.newsguy.com>, cgibbs@kltpzyxm.invalid 
>> says...
>> 
>>> On 2016-05-31, Questor <usenet@only.tnx> wrote:
>>> 
>>>> Testing cannot definitively prove that code is correct.
>>> 
>>> Even if it could, it can't determine whether the specification to
>>> which the code was written has any bearing on what the customer needs.
>>> Or wants.  Or will want next week.
>>
>> Well, that depends on the testing.  If the testing involves the customer 
>> sitting in front of the machine, testing can very quickly determine 
>> whether the code has any bearing on what he wants.
>
>True.  Unfortunately, there are often intermediaries who get in the way,
>and who turn the whole exercise into the telephone game.
>
>I remember the first time I actually got to sit with a customer and
>talk about what he really needed.  It was amazing how we were able
>to clear the air.  But by then, I had already turned in my notice...

I have learned to always do this, as early in the development process
as possible. I usually first present a pure mockup, and get the 
customer to have their users run on the mockup, and make notes.

Repeat with even intervals as the application gets underway.

This way the handover process is also a lot smoother. 

-- mrr

[toc] | [prev] | [next] | [standalone]

#164270

FromGene Wirchenko <genew@telus.net>
Date2016-06-01 12:47 -0700
Message-ID<ureukbhefapcd0dhp8jk5o45dv79db1jc7@4ax.com>
In reply to#164182
On 31 May 2016 20:06:16 GMT, Charlie Gibbs <cgibbs@kltpzyxm.invalid>
wrote:

>On 2016-05-31, Questor <usenet@only.tnx> wrote:
>
>> Testing cannot definitively prove that code is correct.
>
>Even if it could, it can't determine whether the specification to
>which the code was written has any bearing on what the customer needs.
>Or wants.  Or will want next week.

     Or is required to follow.  Or, or, or.

>> Formal verification is an entirely valid field of study within computer
>> science, and indeed, it is part of the "science" in computer science.
>> The methods and tools are widely used in hardware design, presumably
>> because once something is committed to silicon, correcting errors is
>> expensive.  The spirit of getting it right the first time has long
>> been lost in the software realm, to say nothing of actually shipping
>> a finished product.
>
>"There's never time to do it right, but always time to do it over."

     Oh?

     "No, we can't redo it.  We have so much invested in this way."

>Even if you could get it right (first time or otherwise), the marketing
>department's obsession with constant change will have them demanding
>that you break your perfect solution for the next release.

     If not, sooner?

Sincerely,

Gene Wirchenko

[toc] | [prev] | [next] | [standalone]

#164200

From"J. Clarke" <j.clarke.873638@gmail.com>
Date2016-05-31 20:58 -0400
Message-ID<MPG.31b7fd2a2f1a293298a1ed@news.eternal-september.org>
In reply to#164171
In article <574dc948.9600174@nntp2.rawbw.com>, usenet@only.tnx says...
> 
> On Thu, 26 May 2016 07:04:54 -0400, "J. Clarke" <j.clarke.873638@gmail.com>
> wrote:
> >In article <dqkt3mFqjklU2@mid.individual.net>, bob.martin@excite.com 
> >says...
> >> in 665169 20160524 190052 Charlie Gibbs <cgibbs@kltpzyxm.invalid> wrote:
> >> >On 2016-05-22, J. Clarke <j.clarke.873638@gmail.com> wrote:
> >> >> And anybody who has had to debug someone else's code knows that there's
> >> >> also a strong dose of detective skills.
> >> >
> >> >That's not CS, that's the real world.  Different universes.
> >> 
> >> See how a plumber or carpenter gets on debugging code!
> >
> >And part of the problem with "CS" in academia is that they seem to think 
> >that it doesn't have anything to do with "the real world".  Hence the 
> >sort who think that we should "prove code correct" by some kind of 
> >mathematical analysis instead of by testing.
> 
> Testing cannot definitively prove that code is correct.
> 
> Formal verification is an entirely valid field of study within computer science,
> and indeed, it is part of the "science" in computer science.  The methods and
> tools are widely used in hardware design, presumably because once something
> is committed to silicon, correcting errors is expensive.  The spirit of getting
> it right the first time has long been lost in the software realm, to say nothing
> of actually shipping a finished product.


"Formal verification" depends on a verifiable specification.  You want 
to tell me how to get a verifiable specification out of the Powers That 
Be, go right ahead.

In the real world, it's pie in the sky.

[toc] | [prev] | [next] | [standalone]

#164208

FromAhem A Rivet's Shot <steveo@eircom.net>
Date2016-06-01 07:29 +0100
Message-ID<20160601072943.f8e8d7cf1380df1066575b48@eircom.net>
In reply to#164200
On Tue, 31 May 2016 20:58:07 -0400
"J. Clarke" <j.clarke.873638@gmail.com> wrote:

> "Formal verification" depends on a verifiable specification.  You want 
> to tell me how to get a verifiable specification out of the Powers That 
> Be, go right ahead.

	Specification quality varies with application, most business
application specifications are certainly not verifiable, consistent is all
too often too much to ask for and complete is never going to happen. Things
get better as you get closer to accountants (the specifications may seem
insane but they are generally pretty complete and internally consistent),
and I would like to hope they get much better for things like embedded
controllers in dangerous machinery (cars, planes, factory floor gear ...).

	It would be nice to think that at least some of the last category
have verifiable software specifications, they are after all key components
in what should be very well engineered systems.

-- 
Steve O'Hara-Smith                          |   Directable Mirror Arrays
C:>WIN                                      | A better way to focus the sun
The computer obeys and wins.                |    licences available see
You lose and Bill collects.                 |    http://www.sohara.org/

[toc] | [prev] | [next] | [standalone]

#164221

FromMorten Reistad <first@last.name.invalid>
Date2016-06-01 11:59 +0200
Message-ID<89o22d-d28.ln1@sambook.reistad.name>
In reply to#164208
In article <20160601072943.f8e8d7cf1380df1066575b48@eircom.net>,
Ahem A Rivet's Shot  <steveo@eircom.net> wrote:
>On Tue, 31 May 2016 20:58:07 -0400
>"J. Clarke" <j.clarke.873638@gmail.com> wrote:
>
>> "Formal verification" depends on a verifiable specification.  You want 
>> to tell me how to get a verifiable specification out of the Powers That 
>> Be, go right ahead.
>
>	Specification quality varies with application, most business
>application specifications are certainly not verifiable, consistent is all
>too often too much to ask for and complete is never going to happen. Things
>get better as you get closer to accountants (the specifications may seem
>insane but they are generally pretty complete and internally consistent),
>and I would like to hope they get much better for things like embedded
>controllers in dangerous machinery (cars, planes, factory floor gear ...).
>
>	It would be nice to think that at least some of the last category
>have verifiable software specifications, they are after all key components
>in what should be very well engineered systems.

The controllers that are certified to be immediatly connected to
dangerous machinery often have chainable inputs, so they can take
commands from a non-certified controller; and just validate that it
does not do anything fatal.

Been doing a lot of work on such uncertified controllers. And even
if there is a certified one in the chain I still stand back when
I run a new program on a welder/cutter/bender or even on a 
knitting machine/robot.

And then there is payroll. Which consists of a few thousand edge
cases, and no logic.

-- mrr

[toc] | [prev] | [next] | [standalone]

#164227

From"J. Clarke" <j.clarke.873638@gmail.com>
Date2016-06-01 06:48 -0400
Message-ID<MPG.31b8879e86a1312098a1f9@news.eternal-september.org>
In reply to#164221
In article <89o22d-d28.ln1@sambook.reistad.name>, 
first@last.name.invalid says...
> 
> In article <20160601072943.f8e8d7cf1380df1066575b48@eircom.net>,
> Ahem A Rivet's Shot  <steveo@eircom.net> wrote:
> >On Tue, 31 May 2016 20:58:07 -0400
> >"J. Clarke" <j.clarke.873638@gmail.com> wrote:
> >
> >> "Formal verification" depends on a verifiable specification.  You want 
> >> to tell me how to get a verifiable specification out of the Powers That 
> >> Be, go right ahead.
> >
> >	Specification quality varies with application, most business
> >application specifications are certainly not verifiable, consistent is all
> >too often too much to ask for and complete is never going to happen. Things
> >get better as you get closer to accountants (the specifications may seem
> >insane but they are generally pretty complete and internally consistent),
> >and I would like to hope they get much better for things like embedded
> >controllers in dangerous machinery (cars, planes, factory floor gear ...).
> >
> >	It would be nice to think that at least some of the last category
> >have verifiable software specifications, they are after all key components
> >in what should be very well engineered systems.
> 
> The controllers that are certified to be immediatly connected to
> dangerous machinery often have chainable inputs, so they can take
> commands from a non-certified controller; and just validate that it
> does not do anything fatal.
> 
> Been doing a lot of work on such uncertified controllers. And even
> if there is a certified one in the chain I still stand back when
> I run a new program on a welder/cutter/bender or even on a 
> knitting machine/robot.
> 
> And then there is payroll. Which consists of a few thousand edge
> cases, and no logic.

You think that's bad, try insurance.

[toc] | [prev] | [next] | [standalone]

#164234

FromAhem A Rivet's Shot <steveo@eircom.net>
Date2016-06-01 12:41 +0100
Message-ID<20160601124147.ab49801ea98da5e53414d911@eircom.net>
In reply to#164227
On Wed, 1 Jun 2016 06:48:59 -0400
"J. Clarke" <j.clarke.873638@gmail.com> wrote:

> You think that's bad, try insurance.

	I thought the renewal quote algorithm was trivial, add a random
amount between $MIN_INCREASE and $MAX_INCREASE.

-- 
Steve O'Hara-Smith                          |   Directable Mirror Arrays
C:>WIN                                      | A better way to focus the sun
The computer obeys and wins.                |    licences available see
You lose and Bill collects.                 |    http://www.sohara.org/

[toc] | [prev] | [next] | [standalone]

Page 4 of 7 — ← Prev page 1 2 3 [4] 5 6 7  Next page →

Back to top | Article view | alt.folklore.computers

csiph-web