HECnet

hecnet@lists.dfupdate.se

5 participants
15580 discussions

Another sillyness. More information in the nodename database on MIM.

by bqt＠softjar.se

On 2013-05-18 22:58, Sampsa Laine wrote: On 18 May 2013, at 18:41, Johnny Billquist <bqt at softjar.se> wrote: Or I could possible scrape files in know locations to manage the updating, if that would make more sense. (The last one would probably be really easy from my point of view...) Johnny Didn't we develop a format for this ages ago, to be stuck at the end of INFO.TXT on the public accessible dir of a machine? I don't think the current INFO.TXT files are that useful, or even easy to scrape. But that might just be me. :-) Not to mention that I have no clue which machines to scrape, if I were to do that. If I were to do it, I'd probably ask for a different format, and have a designated machine per person/area to scrape. Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol

12 years, 3 months

Another sillyness. More information in the nodename database on MIM.

by sampsa＠mac.com

On 18 May 2013, at 18:41, Johnny Billquist <bqt at softjar.se> wrote: Or I could possible scrape files in know locations to manage the updating, if that would make more sense. (The last one would probably be really easy from my point of view...) Johnny Didn't we develop a format for this ages ago, to be stuck at the end of INFO.TXT on the public accessible dir of a machine? sampsa

12 years, 3 months

was Re: [HECnet] TOPS-10 and DECnet Progress Update

by gerry77＠mail.com

On Sat, 18 May 2013 18:05:52 +0000, <Paul_Koning at Dell.com> wrote: Because CTERM was designed, in essence, as a remote procedure call representation of the VMS terminal driver features. The ones in TOPS-10 are entirely different -- just consider the way carriage return echoes. IIRC you once wrote that RTERM is really four (or was it three?) different protocols: one for TOPS-10, one for RSX, one for RSTS, and I'm not sure if the fourth was about TOPS-20 or if that was included in one of the above. Unfortunately, RTPAD.EXE on OpenVMS 8.x has a bug that prevents it to connect to older systems such as TOPS-10 and RSTS, but not to VMS (in RTERM mode). The affected image has ident X-10. Instead ident X-9 (VMS 7.x) works. Now I'm just trying to see if an AXP VMS 7.3 image works on 8.3... G.

12 years, 3 months

Another sillyness. More information in the nodename database on MIM.

by bqt＠softjar.se

On 2013-05-18 20:28, Cory Smelosky wrote: On Sat, 18 May 2013, Ian McLaughlin wrote: (I may be too late answering this, because I'm mobile and there's a billion messages in this thread that I haven't read yet...) I'd really like to see a Hecnet-based solution - like a set host to a captured account on mim or something would be nice. A web page kind of breaks the illusion of an antique network :). Limiting access to hecnet methods sort of self-polices just by itself. A "captured" account? ;) "Captured" accounts under RSX is very natural. However, they call it slaved terminals. :-) Johnny Ian (Trying to keep up with the tread while at my son's birthday party) Sent from my iPhone On 2013-05-18, at 9:58 AM, Johnny Billquist <bqt at softjar.se> wrote: On 2013-05-18 18:52, Cory Smelosky wrote: On Sat, 18 May 2013, Dave McGuire wrote: On 05/18/2013 12:45 PM, Johnny Billquist wrote: I like the idea of giving us all write access. Maybe you can set up regular dumps of the database for restoration in case someone screws up. I can certainly do that. It just becomes a question of people might need to reenter information in case a rollback has to go far. Also trickier to realize if "corruption" actually have happened perhaps. On the other hand, I could atleast have permissions that only allowed random people to modify data, not add new, nor delete. That's a really good idea. I second this. How flexible is Datatrieve's permissions system? You can identify users either by UIC, or by entity-specific passwords. And for each type of thing in there you have Read, Write, Modify, Extend, and Control. And the access is a list, with the first match being the permissions used. So I could either make sure all people who might want to modify this have their own account on MIM, or else I could just create password protected access, and we could have shared or private passwords. Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol --- Filter service subscribers can train this email as spam or not-spam here: http://my.email-as.net/spamham/cgi-bin/learn.pl?messageid=2C48D1D4BFDC11E2A… -- Cory Smelosky http://gewt.net/ Personal stuff http://gimme-sympathy.org Experiments -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol

12 years, 3 months

TOPS-10 and DECnet Progress Update

by bqt＠softjar.se

On 2013-05-18 20:05, Paul_Koning at Dell.com wrote: On May 18, 2013, at 1:28 PM, Cory Smelosky wrote: On Sat, 18 May 2013, G. wrote: ... If you use the CTERM protocol, it will be always slow and sluggish, no matter how much you'll try to tweak it. Try SET HOST /APP=R and appreciate speed :) It's MUCH more usable when using RTERM. Thank you. Why is CTERM so slow on TOPS-10? Because CTERM was designed, in essence, as a remote procedure call representation of the VMS terminal driver features. The ones in TOPS-10 are entirely different -- just consider the way carriage return echoes. So TOPS-10 is forced to do its terminal I/O through a protocol that doesn't come close to matching what it really wants to see. I suspect it simply runs things in single character mode, since that's about the only way to guarantee that you get what you want. Yeah. It's a rather complex abstraction. Come to think of it, TOPS-10 RTERM mode may well be just that, but it's a very lightweight protocol so it's fast. It also has the nice benefit that it exists on systems that don't do CTERM -- like RSTS. I wonder if they are compatible. I know that under RSX, there exists separate implementations to talk to RSTS/E and TOPS-20, in addition to the RSX specific protocol. (Not sure if the T10 protocol is the same as the T20, and I don't know the proper names of any of them. Are they all called RTERM?) Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol

12 years, 3 months

Another sillyness. More information in the nodename database on MIM.

by b4＠gewt.net

On Sat, 18 May 2013, Paul_Koning wrote: On May 18, 2013, at 2:28 PM, Cory Smelosky wrote: On Sat, 18 May 2013, Ian McLaughlin wrote: (I may be too late answering this, because I'm mobile and there's a billion messages in this thread that I haven't read yet...) I'd really like to see a Hecnet-based solution - like a set host to a captured account on mim or something would be nice. A web page kind of breaks the illusion of an antique network :). Limiting access to hecnet methods sort of self-polices just by itself. A "captured" account? ;) Something supported by various DEC operating systems -- an account that runs the login command script but if that exits, logs you right back out. So you can't ever get to an open ended command prompt. I was making a joke about "captured" versus "captive". ;) paul -- Cory Smelosky http://gewt.net/ Personal stuff http://gimme-sympathy.org Experiments

12 years, 3 months

Another sillyness. More information in the nodename database on MIM.

by Paul_Koning＠Dell.com

On May 18, 2013, at 2:28 PM, Cory Smelosky wrote: On Sat, 18 May 2013, Ian McLaughlin wrote: (I may be too late answering this, because I'm mobile and there's a billion messages in this thread that I haven't read yet...) I'd really like to see a Hecnet-based solution - like a set host to a captured account on mim or something would be nice. A web page kind of breaks the illusion of an antique network :). Limiting access to hecnet methods sort of self-polices just by itself. A "captured" account? ;) Something supported by various DEC operating systems -- an account that runs the login command script but if that exits, logs you right back out. So you can't ever get to an open ended command prompt. paul

12 years, 3 months

Another sillyness. More information in the nodename database on MIM.

by b4＠gewt.net

On Sat, 18 May 2013, Ian McLaughlin wrote: (I may be too late answering this, because I'm mobile and there's a billion messages in this thread that I haven't read yet...) I'd really like to see a Hecnet-based solution - like a set host to a captured account on mim or something would be nice. A web page kind of breaks the illusion of an antique network :). Limiting access to hecnet methods sort of self-polices just by itself. A "captured" account? ;) Ian (Trying to keep up with the tread while at my son's birthday party) Sent from my iPhone On 2013-05-18, at 9:58 AM, Johnny Billquist <bqt at softjar.se> wrote: On 2013-05-18 18:52, Cory Smelosky wrote: On Sat, 18 May 2013, Dave McGuire wrote: On 05/18/2013 12:45 PM, Johnny Billquist wrote: I like the idea of giving us all write access. Maybe you can set up regular dumps of the database for restoration in case someone screws up. I can certainly do that. It just becomes a question of people might need to reenter information in case a rollback has to go far. Also trickier to realize if "corruption" actually have happened perhaps. On the other hand, I could atleast have permissions that only allowed random people to modify data, not add new, nor delete. That's a really good idea. I second this. How flexible is Datatrieve's permissions system? You can identify users either by UIC, or by entity-specific passwords. And for each type of thing in there you have Read, Write, Modify, Extend, and Control. And the access is a list, with the first match being the permissions used. So I could either make sure all people who might want to modify this have their own account on MIM, or else I could just create password protected access, and we could have shared or private passwords. Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol --- Filter service subscribers can train this email as spam or not-spam here: http://my.email-as.net/spamham/cgi-bin/learn.pl?messageid=2C48D1D4BFDC11E2A… -- Cory Smelosky http://gewt.net/ Personal stuff http://gimme-sympathy.org Experiments

12 years, 3 months

Another sillyness. More information in the nodename database on MIM.

by ian＠platinum.net

(I may be too late answering this, because I'm mobile and there's a billion messages in this thread that I haven't read yet...) I'd really like to see a Hecnet-based solution - like a set host to a captured account on mim or something would be nice. A web page kind of breaks the illusion of an antique network :). Limiting access to hecnet methods sort of self-polices just by itself. Ian (Trying to keep up with the tread while at my son's birthday party) Sent from my iPhone On 2013-05-18, at 9:58 AM, Johnny Billquist <bqt at softjar.se> wrote: On 2013-05-18 18:52, Cory Smelosky wrote: On Sat, 18 May 2013, Dave McGuire wrote: On 05/18/2013 12:45 PM, Johnny Billquist wrote: I like the idea of giving us all write access. Maybe you can set up regular dumps of the database for restoration in case someone screws up. I can certainly do that. It just becomes a question of people might need to reenter information in case a rollback has to go far. Also trickier to realize if "corruption" actually have happened perhaps. On the other hand, I could atleast have permissions that only allowed random people to modify data, not add new, nor delete. That's a really good idea. I second this. How flexible is Datatrieve's permissions system? You can identify users either by UIC, or by entity-specific passwords. And for each type of thing in there you have Read, Write, Modify, Extend, and Control. And the access is a list, with the first match being the permissions used. So I could either make sure all people who might want to modify this have their own account on MIM, or else I could just create password protected access, and we could have shared or private passwords. Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol --- Filter service subscribers can train this email as spam or not-spam here: http://my.email-as.net/spamham/cgi-bin/learn.pl?messageid=2C48D1D4BFDC11E2A…

12 years, 3 months

TOPS-10 and DECnet Progress Update

by b4＠gewt.net

On Sat, 18 May 2013, Paul_Koning wrote: On May 18, 2013, at 1:28 PM, Cory Smelosky wrote: On Sat, 18 May 2013, G. wrote: ... If you use the CTERM protocol, it will be always slow and sluggish, no matter how much you'll try to tweak it. Try SET HOST /APP=R and appreciate speed :) It's MUCH more usable when using RTERM. Thank you. Why is CTERM so slow on TOPS-10? Because CTERM was designed, in essence, as a remote procedure call representation of the VMS terminal driver features. The ones in TOPS-10 are entirely different -- just consider the way carriage return echoes. So TOPS-10 is forced to do its terminal I/O through a protocol that doesn't come close to matching what it really wants to see. I suspect it simply runs things in single character mode, since that's about the only way to guarantee that you get what you want. Ahhhh. Yeah. That would explain it. Come to think of it, TOPS-10 RTERM mode may well be just that, but it's a very lightweight protocol so it's fast. It also has the nice benefit that it exists on systems that don't do CTERM -- like RSTS. Yup. Very beneficial on systems without CTERM. ;) paul -- Cory Smelosky http://gewt.net/ Personal stuff http://gimme-sympathy.org Experiments

12 years, 3 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

HECnet