Changes between Initial Version and Version 20 of Ticket #23


Ignore:
Timestamp:
01/11/14 15:49:24 (10 years ago)
Author:
ossi
Comment:

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #23

    • Property Status changed from new to closed
    • Property Severity changed from to no branch
    • Property Reporter changed from slavazanko to ossi
    • Property Cc zaytsev added
    • Property Type changed from defect to enhancement
    • Property Branch state changed from to no branch
    • Property Milestone changed from to 4.7
    • Property Resolution changed from to wontfix

  • Ticket #23 – Description

    initial v20  
    77||Release:||current (CVS or snapshot)||Operating System:||All|| 
    88 
    9 Discussion: 
     9Original submission: 
    1010{{{ 
    11 Wed 20 Sep 2006 03:41:52 PM UTC, comment #8: 
     11so there is this old_esc_mode option ... 
     12it really needs some more exposure. in fact, it should be enabled  
     13by default, and the timeout should be lowered from one second to at  
     14most 100 ms (that is, virtually instantaneous). 
     15high-latency network links should not cause problems, as the chars  
     16of one keypress should still be in one packet, keeping the sequence  
     17intact. if there really is some problem with this, the patch posted  
     18by denis vlasenko on 2004-10-25 to make the timeout configurable  
     19via an environment variable can be used.  
     20}}} 
    1221 
    13 I'd like to be able to set it to 0 ms, i.e. Escape behaving as  
    14 Escape and nothing else. 
    15         Miguel Pérez <wiseman1024> 
    16 Sat 12 Aug 2006 07:22:02 AM UTC, comment #7: 
     22Comment 1 by Pavel Tsekov <ptsekov> at Wed 01 Mar 2006 05:11:49 PM UTC: 
     23{{{ 
     24Please, test the attached patch. Comment, flames, etc will be much  
     25appreciated. 
     26}}} 
    1727 
    18 Denis Vlasenko's patch has been commited. 
    19         Pavel Tsekov <ptsekov> 
    20 Project AdministratorIn charge of this item. 
    21 Wed 01 Mar 2006 08:09:41 PM UTC, comment #6: 
     28Comment 2 by Oswald Buddenhagen <ossi> at Wed 01 Mar 2006 06:59:42 PM UTC: 
     29{{{ 
     30first a correction on my side: meanwhile i did actually experience  
     31a sequence being torn apart, so the "one packet" theory simply does  
     32not hold. :( 
     33going slightly overboard, one could apply some heuristics to  
     34determine the worst-case line latency automatically. 
    2235 
    23 > [...] I cannot imagine why one would need to adjust it more than  
    24 once. 
     36i'm not sure it is a good idea to make the timeout a proper option  
     37only. it makes sense to have the setting depend on the location, or  
     38rather on the route to the location. an env variable seems to be  
     39the by far the most appropriate means of achieving this. having  
     40separate settings for certain $TERM seems interesting as well, as i  
     41might want to extend the timeout for terms where i have to enter  
     42esc sequences manually - but that can be scripted in the startup  
     43script as well. 
     44other than that the patch seems reasonable (i did not test it, but  
     45i don't see any code paths which you could have skipped if you  
     46tested it at all ;). 
     47}}} 
     48 
     49Comment 3 by Pavel Tsekov <ptsekov> at Wed 01 Mar 2006 07:17:35 PM UTC: 
     50{{{ 
     51Well, one could always adjust it as per his immediate requirements.  
     52Have you noticed that I've added a new configuration dialog ? Why  
     53do you think an environment variable is more appropriate ? 
     54}}} 
     55 
     56Comment 4 by Oswald Buddenhagen <ossi> at Wed 01 Mar 2006 07:25:41 PM UTC: 
     57{{{ 
     58> Well, one could always adjust it as per his immediate requirements. 
    2559> 
    2660 
    27 you can log into the same machine from different sources - locally,  
    28 through serial line, through networks with different latencies. and  
    29 also you can share the same config among several machines if you  
    30 have a networked home. both cases are not exactly exceptional in  
    31 heterogenous environments like unis or even when simply ssh-ing  
    32 into your box at home. 
     61this is slow. and extremely annoying if done several times a day. 
    3362 
    34 > > no, i did not read your patch. ;-) 
    35 > 
    36 > You could try it at least :) [...] 
     63> Have you noticed that I've added a new configuration dialog ? 
    3764> 
    3865 
    39 oil your irony detector, dude. ;) 
     66no, i did not read your patch. ;-) 
    4067 
    41 > In this case Denis Vlasenko's patch should suffice. 
     68> Why do you think an environment variable is more appropriate ? 
    4269> 
    4370 
    44 i think the env var should override the setting, but both make  
    45 sense. ideally, the config dialog would show whether the setting  
    46 will take effect - to avoid surprises. 
    47         Oswald Buddenhagen <ossi> 
    48 Wed 01 Mar 2006 07:47:46 PM UTC, comment #5: 
     71it is scriptable. without sed-ing the config file, that is. 
     72in the scenarios i drafted this option is rather imposed by the  
     73operation environment than by user preference, so it just makes  
     74sense to use something that is naturally part of the environment. 
     75}}} 
    4976 
     77Comment 5 by Pavel Tsekov <ptsekov> at Wed 01 Mar 2006 07:47:46 PM UTC: 
     78{{{ 
    5079> > Well, one could always adjust it as per his immediate  
    5180requirements. 
    … …  
    79108 
    80109In this case Denis Vlasenko's patch should suffice. 
    81         Pavel Tsekov <ptsekov> 
    82 Project AdministratorIn charge of this item. 
    83 Wed 01 Mar 2006 07:25:41 PM UTC, comment #4: 
     110}}} 
    84111 
    85 > Well, one could always adjust it as per his immediate requirements. 
     112Comment 6 by Oswald Buddenhagen <ossi> at Wed 01 Mar 2006 08:09:41 PM UTC: 
     113{{{ 
     114> [...] I cannot imagine why one would need to adjust it more than  
     115once. 
    86116> 
    87117 
    88 this is slow. and extremely annoying if done several times a day. 
     118you can log into the same machine from different sources - locally,  
     119through serial line, through networks with different latencies. and  
     120also you can share the same config among several machines if you  
     121have a networked home. both cases are not exactly exceptional in  
     122heterogenous environments like unis or even when simply ssh-ing  
     123into your box at home. 
    89124 
    90 > Have you noticed that I've added a new configuration dialog ? 
     125> > no, i did not read your patch. ;-) 
     126> 
     127> You could try it at least :) [...] 
    91128> 
    92129 
    93 no, i did not read your patch. ;-) 
     130oil your irony detector, dude. ;) 
    94131 
    95 > Why do you think an environment variable is more appropriate ? 
     132> In this case Denis Vlasenko's patch should suffice. 
    96133> 
    97134 
    98 it is scriptable. without sed-ing the config file, that is. 
    99 in the scenarios i drafted this option is rather imposed by the  
    100 operation environment than by user preference, so it just makes  
    101 sense to use something that is naturally part of the environment. 
    102         Oswald Buddenhagen <ossi> 
    103 Wed 01 Mar 2006 07:17:35 PM UTC, comment #3: 
     135i think the env var should override the setting, but both make  
     136sense. ideally, the config dialog would show whether the setting  
     137will take effect - to avoid surprises. 
     138}}} 
    104139 
    105 Well, one could always adjust it as per his immediate requirements.  
    106 Have you noticed that I've added a new configuration dialog ? Why  
    107 do you think an environment variable is more appropriate ? 
    108         Pavel Tsekov <ptsekov> 
    109 Project AdministratorIn charge of this item. 
    110 Wed 01 Mar 2006 06:59:42 PM UTC, comment #2: 
     140Comment 7 by Pavel Tsekov <ptsekov> at Sat 12 Aug 2006 07:22:02 AM UTC: 
     141{{{ 
     142Denis Vlasenko's patch has been commited. 
     143}}} 
    111144 
    112 first a correction on my side: meanwhile i did actually experience  
    113 a sequence being torn apart, so the "one packet" theory simply does  
    114 not hold. :( 
    115 going slightly overboard, one could apply some heuristics to  
    116 determine the worst-case line latency automatically. 
    117  
    118 i'm not sure it is a good idea to make the timeout a proper option  
    119 only. it makes sense to have the setting depend on the location, or  
    120 rather on the route to the location. an env variable seems to be  
    121 the by far the most appropriate means of achieving this. having  
    122 separate settings for certain $TERM seems interesting as well, as i  
    123 might want to extend the timeout for terms where i have to enter  
    124 esc sequences manually - but that can be scripted in the startup  
    125 script as well. 
    126 other than that the patch seems reasonable (i did not test it, but  
    127 i don't see any code paths which you could have skipped if you  
    128 tested it at all ;). 
    129         Oswald Buddenhagen <ossi> 
    130 Wed 01 Mar 2006 05:11:49 PM UTC, comment #1: 
    131  
    132 Please, test the attached patch. Comment, flames, etc will be much  
    133 appreciated. 
    134         Pavel Tsekov <ptsekov> 
    135 Project AdministratorIn charge of this item. 
    136 Mon 11 Jul 2005 05:27:22 PM UTC, original submission: 
    137  
    138 so there is this old_esc_mode option ... 
    139 it really needs some more exposure. in fact, it should be enabled  
    140 by default, and the timeout should be lowered from one second to at  
    141 most 100 ms (that is, virtually instantaneous). 
    142 high-latency network links should not cause problems, as the chars  
    143 of one keypress should still be in one packet, keeping the sequence  
    144 intact. if there really is some problem with this, the patch posted  
    145 by denis vlasenko on 2004-10-25 to make the timeout configurable  
    146 via an environment variable can be used.  
     145Comment 8 by Miguel Pérez <wiseman1024> at Wed 20 Sep 2006 03:41:52 PM UTC: 
     146{{{ 
     147I'd like to be able to set it to 0 ms, i.e. Escape behaving as  
     148Escape and nothing else. 
    147149}}}