1 
  2 09/21/87  tape
  3 Known errors in the current release of tape.
  4 #         Associated TR's
  5 Description
  6 
  7 75  phx20701
  8 tape_ansi_lrec_io [investigating]
  9          Bad padding with ANSI tape in EBCDIC mode.  .
 10 
 11  As time and resources permit this problem will be evaluated and better
 12 defined at a future date.
 13 
 14 74  phx20643
 15 tape_archive [investigating]
 16          "Loads" an incorrect table, which it is then happy to use if the
 17                tape cannot be mounted.  .
 18 
 19  As time and resources permit this problem will be evaluated and better
 20 defined at a future date.
 21 
 22 73  phx18295
 23 Tape Subsytem [investigating]
 24          Since installing release 10.2 we have been receiving excessive
 25                Reready messages when mounting tapes.  .
 26 
 27 As time and resources permit this problem will be evaluated and better
 28 defined at a future date.
 29 
 30 72  phx17261
 31 reorder_archive.pl1 [error] Linked TRs:  phx19390
 32          references the undefined name-
 33                error_table_$record_quota_overflow .
 34 
 35 As time and resources permit this problem will be evaluated and better
 36 defined at a future date.
 37 
 38 69  phx16969
 39 tape_archive (non-mtape_ version) [investigating]
 40          error about a missing argument when trying to write the online
 41                table to tape .
 42 
 43 As time and resources permit this problem will be evaluated and better
 44 defined at a future date.
 45 
 46 68  phx15996
 47 tape_archive [verified]
 48          tape_archive returns the wrong error message following an RCP
 49                failure.  .
 50 
 51 As time and resources permit this problem will be evaluated and better
 52 defined at a future date.
 53 
 54 67  phx14836
 55 tape software [verified]
 56          Asks reready of reel without indication that a write ring is
 57                necessary.  .
 58 
 59 As time and resources permit this problem will be evaluated and better
 60 defined at a future date.
 61 
 62 65  phx20531
 63 mtape_ / ansi_tape_io_ [investigating]
 64           The authentication code written to the tape header by the ANSI             PFM is incorrect in certain cases..
 65 
 66  As time and resources permit this problem will be evaluated
 67  and better defined at a future date.
 68 
 69 62  phx19297
 70 tape utilities [investigating]
 71           should force detach tape after abort.
 72 
 73  As time and resources permit this problem will be evaluated
 74  and better defined at a future date.
 75 
 76 60  phx18913
 77 tape_ibm_, tdcm_ [error]
 78           Bad results when the user must enter a new tape number, and                there is an error in mounting the tape..
 79 
 80  As time and resources permit this problem will be evaluated
 81  and better defined at a future date.
 82 
 83 57  phx18325
 84 tape_in [investigating]
 85           The tape_in command seems to mishandle the "storage_extend"              local statement..
 86 
 87  As time and resources permit this problem will be evaluated
 88  and better defined at a future date.
 89 
 90 55  phx16647
 91 tape_ibm_ [error]
 92           invalid record structure using VBS format for output.
 93 
 94  As time and resources permit this problem will be evaluated
 95  and better defined at a future date.
 96 
 97 53  phx16030
 98 tape_out [verified] Linked TRs:  phx14584 phx17140 phx17480 phx19533
 99           Bad formline_ control string in tape_io_interpret_.pl1 causes              conversion error..
100 
101  As time and resources permit this problem will be evaluated
102  and better defined at a future date.
103 
104 50  phx15547
105 tape_archive [verified]
106           problems using tape_archive when the volume name doesn't match             the external label.
107 
108  As time and resources permit this problem will be evaluated
109  and better defined at a future date.
110 
111 49  phx15545
112 tape_archive [verified]
113           misleading message about no table found when tape could not be             mounted.
114 
115  As time and resources permit this problem will be evaluated
116  and better defined at a future date.
117 
118 48  phx15544
119 tape_archive [verified]
120           can't tell which tape took an error on a compact operation.
121 
122  As time and resources permit this problem will be evaluated
123  and better defined at a future date.
124 
125 47  phx15543
126 tape_archive [verified]
127           poor error message when only one drive is available for a        compaction.
128 
129  As time and resources permit this problem will be evaluated
130  and better defined at a future date.
131 
132 36  phx15087
133 tape_mult_ tends to return obscure error codes and not be very
134 description when diagnosing problems while handling tapes.
135 
136 I am considering this a limitation of the tape_mult_ system.
137 Hopefully, the problem will be corrected when the Multics standard tape
138 format is added to the mtape_ set of per-format-modules.
139 
140 31  phx12545
141 The data_security_erase control order to tape_nstd_ is a no-op.
142 
143 27  phx14374
144 An extra blank record is written by tape_out in format fb and the
145 record is exactly amrl long.
146 
147 26  phx13999
148 Both a Block and Record statment are required when a Format statement
149 is present for labeled tapes, but not stated in the documentation.
150 
151 22  phx17930
152 tape_ioi_ declares certain types of tape errors to be fatal.  Although
153 normally correct, the declaraing of such errors as fatal prevents
154 tape_ioi_'s callers from attempting their own resyncing (by skipping
155 data).  In particular, the reloader's recovery mechanisms are currently
156 broken.
157 
158 20  phx17176
159 When tape_ibm_ extracts record information from a tape, instead of from
160 attachment arguments, it enforces the old 8192 char block size.
161 
162 19  phx01050
163 When a tape is attached using -retain all with tape_ansi_/ibm_,
164 detached and then reattached using -clear, these modules forget that
165 the tape is assigned and screw up when trying to re-assign.
166 
167 18  phx02634
168 When the spooler runs out of initial tapes, it dies after requesting
169 more tapes.
170 
171 17  phx03153
172 tape_ansi_ should allow density change for a volume if, and only if,
173 the attach specifies the creation of a new file set beginning with file
174 1.
175 
176 16  phx03656
177 The "-retain all" option to tape_ansi_/ibm_, when using more than one
178 tape (different attachments) can cause the module to force all tapes
179 onto that one drive.
180 
181 14  phx04667
182 tape_ibm_ refuses to mount for unlabeled operations a tape whose first
183 file contains zero records.
184 
185 13  phx05091
186 After reporting a tape error, tape_ansi_ sometimes throws away the
187 error status, which wasn't explained in the first place.
188 
189 12  phx06346
190 tape_ansi_ should call com_err_ when the com_err switch is on and when
191 the switch to be attached is already attached.
192 
193 11  phx08363
194 The file_status control order to tape_ansi_, from command level,
195 produces a null pointer fault.
196 
197 10  phx08977
198 tape_ansi_ can return a zero error code upon attach when a tape was
199 unreadable even though it printed a message stating it knows the tape
200 is unreadable.
201 
202 9  phx09549
203 Some cleanup handlers in tape_ibm_ and tape_ansi_ have non-local gotos,
204 thus aborting unwinding upon a release.
205 
206 Another manifestation of this is obstinate behavior when hitting QUIT
207 and release while tape_ansi_/ibm_ are waiting for a second tape in a
208 volume set to be mounted.
209 
210 8  phx10157
211 When opened in vb format, tape_ibm_ sometimes makes the last record in
212 a block, upon writing, to be longer than the lrecl.
213 
214 7  phx10167
215 The default lrecl for s type records in tape_ansi_ should be 1044480,
216 not 0.
217 
218 5  phx14863
219 tape_ansi_ does not properly handle tape blocks > 8192 or blocks of
220 length ^= 0 mod 4.
221 
222 4  phx15029
223 A bug exists in tdcm_ workspace management if an error occurs when
224 switching to another tape volume.
225 
226 3  phx15903
227 tape_nstd_ should allow block sizes not 0 mod 4.
228 
229 2  phx15904
230 The write_record operation of tape modules should sense when a tape
231 error occured in nine mode because of a non zero high bit in a byte and
232 produce a corresponding message.
233 
234 1
235 Various tape I/O modules (e.g., ntape_, tape_ansi_, tape_ibm_) do not
236 support the FORTRAN backspace statement.
237 
238 Although this remains a limitation for the referenced I/O modules, the
239 limitation has been partially removed by the introduction of the mtape_
240 tape I/O module in MR10.2, which supports ANSI and IBM tape formats.
241 The mtape_ I/O module fully supports all iox_$positioning modes for
242 sequential_input openings.  If I understand the stated limitation
243 correctly, mtape_ should be able to support the FORTRAN backspace as
244 well as the rewind statements.