CONTACTS 
0201 - JPS Net                            800 711 5353 fax or 800...
020101 - Mr. Joe NLN; Customer Support Rep; Technical Service  =415 376 0121
0202 - IBM Internet                       800 821 4612 fax or 800...
020201 - Mr. Donald McKenzie; Technical Support

SUBJECTS
Web Site, SDS Data Base, Diary 02 Testing
Intelligence Business
Test Delivery, Access by Web Browser over
Email Access to SDS Record via Links
SDS Attach to Email HTML Netscape, IE

1007 -    ..
1008 - Summary/Objective
1009 -
100901 - Follow up ref SDS 16 4305, ref SDS 13 1380.
100902 -
100903 - Developed planning for test messages to evaluate using the Internet as
100904 - a delivery medium for business intelligence.
100905 -
100906 - Problems occured that prevented any messages from being issued.  The
100907 - experience discovering a solution illustrates the challenge of
100908 - compound problems that lead to false understandings and rework, which
100909 - is related to the objective of providing "business intelligence."
100910 -
100911 - Submitted ref DIT 1 1909 explaining problems, ref SDS 0 3668, and
100912 - asking about accessing our data.
100913 -
100914 -     [On 980318 confirmed solution to accessing files on the Web site.
100915 -     ref SDS 20 2877]
100916 -
100917 -     [On 980318 encountered problem accessing links in Web pages, ref
100918 -     SDS 15 4880.  Got solution from Netscape. ref SDS 20 2190]
100919 -
100920 -     [On 980318 accessing links using Internet Explorer was still
100921 -     failing, ref SDS 20 5888.]
100922 -
100923 -     [On 980320 got solution from Microsoft. ref SDS 22 5182]
100924 -
100925 -     [On 980321 seem to have successful test of system. ref SDS 23
100926 -     4345]
100927 -
100928 - More concerns arise about reliability of JPS as our Internet Service
100929 - Provider (ISP). ref SDS 0 2211 and ref SDS 0 2126.
100930 -
100931 -     [On 980321 JPS server failed again. ref SDS 24 1827]
100932 -
100933 -
100934 -
1010 -
1011 -
1012 - Analysis
1013 -
101301 -  ..
101302 - Compound Mistakes Increase Rework Time Exponentially
101303 - Timely Notice Critical to Avoid Costly Rework False Positives
101304 -
101305 - I am providing a little more detail in this SDS record to illustrate
101306 - how the mind is led down blind alleys.  It omits some stuff and
101307 - probably does not have the sequence exactly correct, but makes a good
101308 - case study.
101309 -
101310 - This is a microcosm of meaning drift that afflicts big projects, as
101311 - reported in the case study of a ship that sank at sea, presented on
101312 - 940611 at Asilomar. ref SDS 1 8473
101313 - ..
101314 - The actual solution time required by Welch to today's problem
101315 - was less than a second to add a dot in a file spec, ref SDS 0 3424,
101316 - but it required 6 hours or so to discover, in part due to lack of
101317 - disclosure (i.e., feedback) mechanisms in other systems. ref SDS 0
101318 - 3078
101319 -
101320 -     [On 980318 hours of rework by multiple organizations discover
101321 -     cause of problem that takes a few minutes to fix re upper
101322 -     case/lower case for elements of SDS record link specification. see
101323 -     ref SDS 20 3389]
101324 -
101325 -     [On 980320 hours of rework on removing a space in a test string.
101326 -     ref SDS 22 5182]
101327 -
101328 -     [On 991001 space probe crashes on Mars because communication is
101329 -     not aligned on small matter. ref SDS 33 0001]
101330 -
101331 - This illustrates the critical role of "notice" in managing events,
101332 - which causes increased costs to recover from rework.
101333 -
101334 -     Rework was identified as a source of extra costs at ref SDS 2
101335 -     4472.
101336 -
101337 - The initial test message had IP address errors which were masked by
101338 - the report from the system that the server could not access requested
101339 - files.  Since the ISP had previously reported its system was down, the
101340 - initial search for a solution was in the direction of the ISP.  Since,
101341 - however, the ISP incorrectly reported its systems were all running
101342 - correctly, ref SDS 0 3874, attention turned to the data entry
101343 - specification, and this disclosed the error, after working past the
101344 - distraction of the IBM email system being down.
101345 -
101346 - After correcting the IP address mistakes, access continued to fail.
101347 -
101348 - Since the prior error was with data entry, the initial focus was to
101349 - test the data entry spec.  When that evaluation did not reveal a cause
101350 - for failure, attention was forced back on the JPS server as the
101351 - problem.  But when they were called their phone messages said all
101352 - systems were running.  Since they were not reporting system problems,
101353 - that moved attention to other possible causes.  I then thought that
101354 - because our Domain name was activated today, after having created a
101355 - data base using the IP address, it could be that we can no longer use
101356 - the IP address, and have to use only the Domain name.  Another
101357 - possibility was that JPS changed access to the data.  It turned out
101358 - that JPS server had a rare failure of the http process, and it did not
101359 - update its telephone message.
101360 -
101361 -
101362 -
101363 -
1014 -
1015 -
1016 - Progress
1017 -
101701 -  ..
101702 - Email Supports HTML for Correspondence in Netscape, IE Not Sure
101703 - Planning Program to Test Viability of Business Intelligence
101704 -
101705 - Want to send email to some people who have a fair amount of history in
101706 - the SDS record, so they can roam around and provide feedback on the
101707 - usefulness of this form of accessing information about matters that
101708 - impact their work.
101709 -
101710 - I think Netscape's email program supports HTML, so can create normal
101711 - SDS correspondence, then use cut and paste to load it into an email
101712 - composition screen.
101713 -
101714 - No, that didn't work. Cut and paste simply enters HTML as text rather
101715 - than commands.
101716 -
101717 - Called Morris.
101718 -
101719 - He suggested trying "Attach."
101720 -
101721 - This worked.  However there is a trick.  We can get thing to display
101722 - in the email by attaching the local version, but then the links will
101723 - not work for people who received the letter, because they all point to
101724 - the local disk.  We have to open the target record on the web, grab
101725 - the address, with operating systems copy function.  Open Netscape
101726 - email, write the letter, and then call "Attach."  Use paste to enter
101727 - the web address from the prior step.
101728 -
101729 - The attached HTML stuff does not display in the email composition
101730 - screen.  It only appears as a listed attachment.  It shows up in the
101731 - sent version.
101732 -
101733 -      [On 980323 some email programs cannot handle HTML in the email
101734 -      composition. ref SDS 25 0001]
101735 -
101736 -      [On 980329 Morris suggests use link in email. ref SDS 27 3343]
101737 -
101738 -      [On 000225 forgot about "Attach." ref SDS 34 0001]
101739 -
101740 - Eventually (over next several days) got this to work, per below.
101741 - ref SDS 0 3739  Letters appears the same in the email composition area
101742 - as in the Netscape web browser with links highlighted and functional,
101743 - and all of the formatting applied.  This provides a simple means to
101744 - accomplish the goal of delivering...
101745 -
101746 -
101747 -                  intelligence anytime, anywhere
101748 -
101749 -
101750 - ...per initial planning on 971021, ref SDS 3 9999, and also on 971105,
101751 - ref SDS 4 1459, and further explained on 971206. ref SDS 5 4862
101752 -
101753 - ..
101754 - Will send test letters to...
101755 -
101756 -      Bill DeHart       PG&E
101757 -
101758 -         [On 980328 sent test to Bill linked to web. ref SDS 26 2904
101759 -
101760 -         [On 980404 test was successful. ref SDS 30 4022]
101761 -
101762 -      Morris Jones      Chips
101763 -
101764 -         [On 980328 sent email to Morris linked to web. ref SDS 26 3402
101765 -
101766 -         [On 980329 Morris' email program didn't handle mail from
101767 -         Netscape in HTML, so switched to entering links only.
101768 -         ref SDS 27 3343]
101769 -
101770 -      Max Wideman       Wideman
101771 -
101772 -         This is the first one, prepared in another record today.
101773 -         ref SDS 17 0001, which failed.  On 980321 finally was
101774 -         successful sending an email test to Max. ref SDS 23 4237
101775 -
101776 -         [On 980323 seems to have failed. ref SDS 25 0001]
101777 -
101778 -         [On 980404 solved it using Morris' suggestion. ref SDS 31 0001
101779 -
101780 -      Don Harms         Transamerica
101781 -
101782 -      Tom Keesling      USACE CESPN
101783 -
101784 -         [On 980403 sent letter to Tom; problems using the links as
101785 -         HTML letter in email. ref SDS 29 3920]
101786 -
101787 -         [On 980502 Morris' solution helped USACE. ref SDS 32 1906]
101788 -
101789 -      Max Blodgett      USACE CESPN
101790 -
101791 -         [On 980403 Max feels Com Metrics on Internet is powerful new
101792 -         business system. ref SDS 29 4756]
101793 -
101794 -         [On 980520 Max said Com Metrics on Internet can support goals
101795 -         for "virtual office" concept. ref SDS 32 2204]
101796 -
101797 -
101798 -
101799 - If these tests work, we can send marketing efforts to...
101800 -
101801 -      Dave Vannier      Intel
101802 -      Ross Chittenden   Caltrans
101803 -      Lee Calloway      PG&E
101804 -      MoonJa Kim        USACE CERL-PL-B
101805 -      Mike Capik        USACE CESPK
101806 -
101807 -
101808 - Spent a few hours converting links from the old "line" spec to the new
101809 - method, so that people can click around more freely on our Web site,
101810 - which is our concept of "knowledge space."
101811 -
101812 -
101813 -  ..
101814 - Sent Test Email to Myself to Verify System is Working
101815 -
101816 - Prepared the first letter to Max Wideman, ref DIP 1 0001, in another
101817 - record, ref SDS 17 0001, and sent it to myself to make sure links are
101818 - properly entered so that it can be accessed.
101819 -
101820 - Wasn't able to test Internet Explorer because could not figure out
101821 - how to run their email system.  There seems to be several different
101822 - programs...
101823 -
101824 -                        Outlook
101825 -                        Outlook Express
101826 -                        Mail
101827 -
101828 -
101829 - I also need test the SDS records, linked in the email, to make sure
101830 - they have citations that can be opened, since this is what we are
101831 - demonstrating. We have only put a small number of records on the site,
101832 - ref SDS 15 1378, and so the records I send, may not have a rich enough
101833 - set of connections to convey our point about business intelligence.
101834 -
101835 -
101836 -  ..
101837 - Failed Test Illustrates Compound Errors Mislead
101838 -
101839 - We had a number of technical problems with this first test effort:
101840 -
101841 -      1.  I entered the link incorrectly. ref SDS 0 3424
101842 -
101843 -   [On 980318 additional problems of access showed 2 different coding
101844 -   errors by Welch. ref SDS 19 1281]
101845 -
101846 -      2.  IBM mail program crashed in San Francisco. ref SDS 0 2217
101847 -
101848 -      3.  JPS server crashed for HTTP ops, vendor fails to give notice.
101849 -          ref SDS 0 3366 and ref SDS 0 2126.
101850 -
101851 - Only one of these problems required action by Welch.  Problem one
101852 - required pressing the period key on the keyboard, about half-a-second.
101853 -
101854 - But the background, circumstances and manner of encountering these
101855 - problems led to expenditure of 4 hours or so of effort to discover
101856 - that such efforts were not needed.
101857 -
101858 -     [On 980318 confirmed solution to accessing files on the Web site.
101859 -     ref SDS 20 2877]
101860 -
101861 -     [On 980318 additional problems of access showed 2 different coding
101862 -     errors by Welch. ref SDS 19 1281]
101863 -
101864 -
101865 -
1019 -

SUBJECTS
Cannot Access Data on JPS Server
Testing
SDS Linking System
Linking Address for SDS Web Pages
Home Page
Domain, Login, welchco.com

1609 -
160901 -  ..
160902 - Cannot Access SDS Records on Web Server
160903 -
160904 - Sent a test message to myself with links to 2 SDS records, ref SDS 16
160905 - 1583, per ref SDS 0 3739
160906 -
160907 - I used Internet Explorer (IE) to retrieve the email so I could open
160908 - the links to our Web site SDS records, since it had worked last
160909 - Friday, and Netscape had failed. ref SDS 12 5215
160910 -
160911 - However, this evening, IE would not open the link in the email.
160912 -
160913 -     [On 980320 got solution from MS. ref SDS 22 5182]
160914 -
160915 - The email file showed only a portion of the link specification was
160916 - highlighted and underlined, which, in fact, indicated the entire link
160917 - spec was defective, i.e., I had entered it incorrectly.
160918 -
160919 - I did not, however, initially recognize the format condition indicated
160920 - the link spec had a problem.
160921 -
160922 - I thought that the email may have been corrupted in transmission, so I
160923 - opened Netscape and sent the message again.
160924 -
160925 - This time there was a considerable delay in sending the second test
160926 - message, which was only about 5 minutes after the first one.
160927 -
160928 - I opened IE and found the same problem with the link spec.
160929 -
160930 - For fun (it was not a "hunch," because I knew it would not work based
160931 - experience last Friday), I tried the Netscape mail program, which had
160932 - been failing to open links to records on our Web site last Friday. ref
160933 - SDS 8 3828
160934 -
160935 - I was amazed to find that Netscape opened the link.
160936 -
160937 - This is the reverse of the experience last Friday. ref SDS 12 5215
160938 -
160939 -
160940 -  ..
160941 - Discovered Link Address to Web Site SDS Record Entered Incorrectly
160942 -
160943 - I then realized that I was clicking on the 2nd link in the email.  I
160944 - tried the first one and it failed to open.
160945 - ..
160946 - As a result of this sequence, I looked more closely at the IP
160947 - address entered in the test letter, and noticed that I had left out a
160948 - dot separating elements of the IP address in the first link.
160949 -
160950 - It took a second to put the dot in the IP address to correct that
160951 - problem.
160952 -
160953 -   [On 980318 additional problems of access showed 2 different coding
160954 -   errors by Welch. ref SDS 19 1281]
160955 -
160956 - I wanted to try IE again, so I had to send the message again because
160957 - once a message is opened and downloaded, it cannot be downloaded.
160958 -
160959 - About this time, I checked Internic and found that our Domain name is
160960 - registered. ref SDS 18 3126
160961 -
160962 - I therefore edited the test letter to Max and added alternate forms of
160963 - the link address using the Domain name and the IP address.
160964 -
160965 - When I tried to send this test email, there was another long delay in
160966 - executing the transmission.  Seemed to take about 20 minutes.
160967 -
160968 - This was a distraction.  I wound contacting IBM and finding that their
160969 - server in San Francisco is down. ref SDS 0 2217
160970 -
160971 - I retrieved the mail in Netscape and found that it opened the links
160972 - after a long delay, which I attributed to the IBM mail server problem.
160973 - ref SDS 0 2217
160974 -
160975 - I sent the message again, so I could open it in IE.  This took another
160976 - 20 minute delay.
160977 -
160978 - I opened IE and retrieved the third test email.
160979 - ..
160980 - IE would not open the link as corrected.
160981 -
160982 - I opened Netscape and found again it was opening the links after a
160983 - delay.
160984 -
160985 - Then I noticed that Netscape was actaully not opening the links, but
160986 - was opening a cached copy on my local disk.
160987 -
160988 - This indicated that there was a problem with the JPS server.
160989 -
160990 - I tried logging on to just the welchco home page, but I was entering
160991 - the string incorrectly as welch, per ref SDS 18 3200.
160992 - ..
160993 - When I tried to enter the SDS directory it returned an error,
160994 - saying:
160995 -
160996 -             There is no URL for "sd"
160997 -
160998 - Similarly I tried to log onto the SDS directories at our IP address,
160999 - but the system returned an error, saying...
161000 -
161001 -             There is no URL for "sd"
161002 -
161003 - Since I got the same error message for both, I was not led to see if
161004 - the domain name was entered correctly.  It did not occur to me to
161005 - investigate that possibility, since I was focused on what had happened
161006 - to our data that took 4 or 5 hours to upload last night. ref SDS 15
161007 - 4305, ref SDS 15 5830
161008 -
161009 - I thought possibly that since the Domain name was now registered, that
161010 - this impacted the way our data was stored on the JPS server, or
161011 - possibly the data had been deleted or moved.
161012 -
161013 - As a result, I called JPS Technical Support. ref SDS 0 3874
161014 -
161015 -
161016 -
1611 -

SUBJECTS
Dialer, IBM
Dialer, IBM for W95
299795  Dialer Not Sending Email

1906 - 1920 Called IBM
1907 -
190701 -  ..
190702 - IBM's Server is Down, Caused Delay in Sending Mail
190703 -
190704 - Talked to Janna
190705 -
190706 - I was a bit exasperated that in trying to test the JPS Web site
190707 - system, the IBM mail system was causing delays.
190708 -
190709 - Janna checked and said that San Francisco is reporting their mail
190710 - server has crashed.  They have people working on it but do not give an
190711 - estimate of when it will be back on line.
190712 -
190713 - I asked if there is a backup number we can call so we can procede
190714 - with our work.
190715 -
190716 - Janna said there is no alternate method.  She suggested I contact
190717 - IBM's suggestion box.
190718 -
190719 -
1908 -
1909 -
1910 - 1948 Mail working
1911 -
191101 - The mail operation seems to be working better now.
191102 -
191103 -
1912 -

SUBJECTS
Cannot Access Data on JPS Server
Access Failed to SD
Reliability
Server - Barberra

2407 - 2025 called JPS
2408 -
240801 -  ..
240802 - JPS Reports All Networks are Running Correctly
240803 -
240804 - As a result of getting messages that our SDS directory does not exist,
240805 - ref SDS 0 2855, I called JPS Technical Support.
240806 -
240807 - The JPS technical support phone number initially gives callers a
240808 - standard message about the status of the network.  It says to stay on
240809 - the line for help by an engineer, and also says callers can submit
240810 - questions by email using the JPS Web site or email address.
240811 - ..
240812 - This evening the JPS message says:
240813 -
240814 -            All regional networks are running normally
240815 -
240816 - This turned out to be an error, because the JPS server computer named
240817 - "Barberra" was not working properly.
240818 -
240819 - Since I did not know their system was down, I stayed on line to speak
240820 - with a technician.
240821 -
240822 -
240823 -  ..
240824 - JPS Technical Support Telephone Message Failed
240825 -
240826 - At the conclusion of the JPS generic message, there was a
240827 - continual series of beeps, maybe 10 or 15, and then a standard phone
240828 - company message was given...
240829 -
240830 -          If you would like to make a call, please hang up
240831 -          and try again!
240832 -
240833 - This indicated the JPS technical support phone line was defective.
240834 -
240835 - I tried calling back several times over about 20 minutes and got the
240836 - same result each time.
240837 -
240838 -
240839 -  ..
240840 - JPS Fails to Respond to Email Request for Technical Support
240841 -
240842 - I therefore tried sending an email requesting assistance at...
240843 -
240844 -                        support@jps.net
240845 -
240846 - Submitted ref DIT 1 1909 explaining the following background and
240847 - asking about accessing our data.
240848 -
240849 -
240850 -
240851 -  ..
2409 -
2410 -
2411 - 2311 called JPS
2412 -
241201 -  ..
241202 - Delay of 20 Minutes Waiting for Support Staff
241203 -
241204 - We have not received an answer to our email sent at 2025, ref DIT 1
241205 - 1909, so I called to expedite correction of the problem.
241206 -
241207 - JPS commits to responding to email within 15 minutes at ref SDS 9
241208 - 2236.
241209 -
241210 - The JPS Technical Support line seems to be working now.  The beeps
241211 - that occurred previously are gone.  I got connected to their support
241212 - section and was told to hold that someone would be with us shortly.
241213 -
241214 -
241215 -
241216 -
241217 -  ..
2413 -
2414 -
2415 - 2334 finally talked to support staff
2416 -
241601 -  ..
241602 - JPS Server for Welch Failing to Perform HTTP Function
241603 -
241604 - Talked to Nick
241605 -
241606 - I asked if he was aware their phone message system was evidently broke
241607 - for awhile this evening? ref SDS 0 3668
241608 -
241609 - Nick said no one has reported a problem with the phone message.
241610 -
241611 -    I did not take time to discuss that if there was a problem, then
241612 -    the Technical Support staff would have had a period when they were
241613 -    not receiving calls.
241614 -
241615 -    If that did not happen, then some investigation is needed on why
241616 -    the Welch calls failed to get through.  Could be a problem with our
241617 -    local phone line.
241618 -
241619 - Nick said he has not seen our email requesting support. ref DIT 1 1909
241620 -
241621 - He does not have the ability to review it for background.
241622 -
241623 - I explained our problem. ref SDS 0 1935 and ref SDS 0 1787.
241624 - ..
241625 - Nick asked for our account information to test access to our Web
241626 - site at his location.
241627 -
241628 - He said that he is getting the same lack of response when he tries to
241629 - access our IP address.
241630 -
241631 - I explained having received notice that our Domain name has been
241632 - registered, and that we seem to have a Web home page, but it does not
241633 - permit access to any of our directories. ref SDS 0 2855
241634 -
241635 - Nick tried the welchco.com and said he is getting the same response
241636 - he got for our IP address that the server is not responding.
241637 -
241638 - Nick said that our account is missing an identification as a
241639 -
241640 -                     Public HTML Directory
241641 -
241642 - He feels this may be causing the lack if response by the JPS server.
241643 -
241644 - He asked me to hold on while he talks to the Network Administrator.
241645 - ..
241646 - Nick came back on line and said the problem is that the computer
241647 - server that stores our data is not performing the http function that
241648 - permits access using http protocals, which is the method used by Web
241649 - browsers.
241650 -
241651 -     [On 980318 confirmed solution to accessing files on the Web site.
241652 -     ref SDS 20 2877]
241653 -
241654 -     [On 980321 server failed again. ref SDS 24 1827]
241655 -
241656 - Nick said this is a new computer which he called
241657 -
241658 -                        Barberra
241659 -
241660 -      [On 980402 received notice from Network Solutions that are
241661 -      servers have a different identification, per ref SDS 28 4326]
241662 -
241663 -
241664 -  ..
241665 - Public HTML Directory Not Needed for New Server "Barberra"
241666 -
241667 - I asked if the problem is related to the matter he mentioned earlier
241668 - of not having assigned to the Welch account a...
241669 -
241670 -                     Public HTML Directory
241671 -
241672 - Nick said that Baberra is a new computer.  He thinks it does not
241673 - require a Public HTML Directory, so it should be working.
241674 -
241675 - He said people are working fixing the problem.
241676 -
241677 - I asked how long he expects it will be before we can transfer
241678 - information and access it on our Web site.
241679 -
241680 - Nick said the FTP operation is runing okay, so we can go ahead and
241681 - upload information at any time.
241682 -
241683 - I asked how long JPS estimates it will take to fix the problem, so we
241684 - can procede with the testing operation to access our data on the Web?
241685 - ref SDS 0 3739
241686 -
241687 - Nick said that the Network Administrator has not given an estimate of
241688 - when the system will be back up.
241689 -
241690 -     [On 980318 confirmed solution to accessing files on the Web site
241691 -     was getting the JPS server fixed, ref SDS 20 2877]
241692 -
241693 - I asked if they have received many calls since we evidently began
241694 - experiencing the problem about 1900?
241695 -
241696 - Nick indicated they have had a lot of calls on it.
241697 -
241698 -    We did not discuss why Nick did not point out when I first called
241699 -    that th Barbara computer was down. ref SDS 0 5947
241700 -
241701 - ..
241702 - I asked why the JPS telephone message says
241703 -
241704 -            All regional networks are running normally
241705 -
241706 -     ...when there seems to have been a problem for several hours?
241707 - ..
241708 - Nick said that sometimes the phone message does not get updated.
241709 -
241710 -
241711 -
241712 -
241713 -