I have something like this on 'radiusd.conf': ... modules{ ... pgping { } ... } ... Where 'rlm_pgping' is a module compiled and installed following the manual. My question is why does it not appear to load (i.e. not showing any messages with 'radiusd -X')? _the module code is on the end of the message_ I have other custom modules that works fine. Am I doing something wrong? On the other hand, I've been playing around with 'radiusd.conf' I've discovered that if I add some random module name in the module instantiation section, radiusd doesn't complain at all about it's non-existence! This is a little inconvenient. Thanks in advance, --- Tomás A. Rossi Ministerio de Economía Proyecto de Informática Buenos Aires, Argentina /* * rlm_pgping.c * * Este módulo se encarga de "pinguear" la base de datos primaria e informa si * la misma está caída (FAIL) o funcionando normalmente (OK). */ #include "autoconf.h" #include "libradius.h" #include <stdio.h> #include <stdlib.h> #include "radiusd.h" #include "modules.h" #include "conffile.h" #include "libpq-fe.h" /* * Estructura para la configuración del módulo. */ typedef struct rlm_pgping_t { char *host; char *dbname; char *user; char *password; char *port; int timeout; } rlm_pgping_t; /* * No hay parámetros de configuración. */ static CONF_PARSER module_config[] = { { "string", PW_TYPE_STRING_PTR, offsetof(rlm_pgping_t,host), NULL, "localhost" }, { "string", PW_TYPE_STRING_PTR, offsetof(rlm_pgping_t,dbname), NULL, "" }, { "string", PW_TYPE_STRING_PTR, offsetof(rlm_pgping_t,user), NULL, "" }, { "string", PW_TYPE_STRING_PTR, offsetof(rlm_pgping_t,password), NULL, "" }, { "string", PW_TYPE_STRING_PTR, offsetof(rlm_pgping_t,port), NULL, "" }, { "integer", PW_TYPE_INTEGER, offsetof(rlm_pgping_t,timeout), NULL, "30" }, { NULL, -1, 0, NULL, NULL } /* end the list */ }; static int mandar_pgping(void *inst, REQUEST *req) { char condata[256]; /* Buffer para guardar los datos de conexión. */ PGconn *con; /* Conexión a la base. */ DEBUG("PGPING: Empieza el modulo"); /*req = req;*/ #define INST ((rlm_pgping_t *)inst) snprintf(condata, sizeof(condata)-1, "host=%s port=%s dbname=%s user=%s " "password='%s' connect_timeout=%d", INST->host, INST->port, INST->dbname, INST->user, INST->password, INST->timeout); con = PQconnectdb(condata); DEBUG("Intentando conectar a la base primaria con datos de conexión: '%s'", condata); if (PQstatus(con) == CONNECTION_BAD) { radlog(L_AUTH, "Falló la conexión a la base primaria."); return RLM_MODULE_FAIL; } return RLM_MODULE_OK; } static int pgping_init(void) { return 0; } /* * Hay que leer los parámetros de configuración para la instancia. */ static int pgping_instantiate(CONF_SECTION *conf, void **instance) { rlm_pgping_t *conf_data; /* * Pedir memoria para los parámetros de configuración. */ conf_data = rad_malloc(sizeof(*conf_data)); if (!conf_data) { return RLM_MODULE_FAIL; } memset(conf_data, 0, sizeof(*conf_data)); /* * Si falla el "parseo" del archivo de configuración para el módulo, CHAU!. */ if (cf_section_parse(conf, conf_data, module_config) < 0) { free(conf_data); return RLM_MODULE_FAIL; } *instance = conf_data; return RLM_MODULE_OK; } static int pgping_detach(void *inst) { free(INST->host); free(INST->port); free(INST->dbname); free(INST->user); free(INST->password); free(inst); return 0; } /* * The module name should be the only globally exported symbol. * That is, everything else should be 'static'. * * If the module needs to temporarily modify it's instantiation * data, the type should be changed to RLM_TYPE_THREAD_UNSAFE. * The server will then take care of ensuring that the module * is single-threaded. */ module_t rlm_pgping = { "pgping", RLM_TYPE_THREAD_SAFE, /* type */ pgping_init, /* initialization */ pgping_instantiate, /* instantiation */ { mandar_pgping, /* authentication */ mandar_pgping, /* authorization */ mandar_pgping, /* preaccounting */ mandar_pgping, /* accounting */ NULL, /* checksimul */ NULL, /* pre-proxy */ NULL, /* post-proxy */ NULL /* post-auth */ }, pgping_detach, /* detach */ NULL, /* destroy */ };
Tomás A. Rossi escribió:
I have something like this on 'radiusd.conf':
... modules{ ... pgping { } ... } ...
Where 'rlm_pgping' is a module compiled and installed following the manual. My question is why does it not appear to load (i.e. not showing any messages with 'radiusd -X')? _the module code is on the end of the message_ I have other custom modules that works fine. Am I doing something wrong?
On the other hand, I've been playing around with 'radiusd.conf' I've discovered that if I add some random module name in the module instantiation section, radiusd doesn't complain at all about it's non-existence! This is a little inconvenient. List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html Sorry, I know what it was happening with the module. It seems that radiusd it doesn't take in account the instantiation of any module unless it is used in some other section (i.e. authorize). Though this behavior could be enhanced.
Thanks --- Tomás A Rossi Ministerio de Economía Proyecto de Informática Buenos Aires, Argentina
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
Sorry, I know what it was happening with the module. It seems that radiusd it doesn't take in account the instantiation of any module unless it is used in some other section (i.e. authorize). Though this behavior could be enhanced.
To do what, exactly? The whol point of loading a module is to use it in a section (e.g. authorize). If you want to load it, but not use it, that doesn't make sense. The server will *not* be changed to load modules it's not going to use. Alan DeKok.
Alan DeKok escribió:
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
Sorry, I know what it was happening with the module. It seems that radiusd it doesn't take in account the instantiation of any module unless it is used in some other section (i.e. authorize). Though this behavior could be enhanced.
To do what, exactly?
The whol point of loading a module is to use it in a section (e.g. authorize). If you want to load it, but not use it, that doesn't make sense.
The server will *not* be changed to load modules it's not going to use.
That's not my point. I'm not trying to say that you should do that but rather to print some kind of message showing that the server has read the module instantiation. Note that I'm not being pretentious: you could warn the user with a message that the module will not be loaded since it's not used and *that* would be being pretentious. FreeRADIUS is a very nice product though I think at least you should concede me that it lacks of good documentation IMHO. I was trying to add a module not knowing of the above mentioned behavior. It results in only adding the entry inside the instantiation section and testing if 'radiusd -X' returned some message to give me a clue that the module instantiation was parsed at least. You'll surely think I'm a jerk (and I'm not saying the opposite :-P ), but I lose about an hour till I figured out what was the whole issue. Thanks for the answers, --- Tomás A Rossi Ministerio de Economía Proyecto de Informática Buenos Aires, Argentina
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
That's not my point. I'm not trying to say that you should do that but rather to print some kind of message showing that the server has read the module instantiation.
Why? It doesn't, in fact, read the module instantiation because it doesn't even look for it, because the module isn't being used.
FreeRADIUS is a very nice product though I think at least you should concede me that it lacks of good documentation IMHO.
You're joking, right? Have you ever tried to use a *commercial* server? Many of those make our documentation look world-leading.
I was trying to add a module not knowing of the above mentioned behavior.
No. You were trying to add a module without telling the server to use it, and you were surprised that the server didn't use it. If the server *had* printed out the message you wanted, odds are that your next question would be "why doesn't the server use the module when I send it a packet? I listed it in the instantiate section!" Yes, we've been through this dicussion before with other people. You're not the first to run into this. And the end result of what you want is an endless series of messages explaining why the server isn't doing what you think it's doing. It's a lot easier to have the server to print out what is *is* doing, and then tell people to read the documentation to see why it isn't doing what they expect. Usually, their expectations are wrong, and the documentation contains the information to correct the expectation.
It results in only adding the entry inside the instantiation section and testing if 'radiusd -X' returned some message to give me a clue that the module instantiation was parsed at least.
And what did you expect the server would do after that? You still haven't answered that question. So you're not explaining what you thought was going on, or what you were trying to do. You're just complaining that the server isn't doing what you want. Alan DeKok.
Alan DeKok escribió:
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
That's not my point. I'm not trying to say that you should do that but rather to print some kind of message showing that the server has read the module instantiation.
Why? It doesn't, in fact, read the module instantiation because it doesn't even look for it, because the module isn't being used.
I mean a message saying something like: *WARNING: module _blabla_ instantiated but not used, radiusd will ignore it*
FreeRADIUS is a very nice product though I think at least you should concede me that it lacks of good documentation IMHO.
You're joking, right? Have you ever tried to use a *commercial* server? Many of those make our documentation look world-leading.
I've not used any other commercial server though I know what you're saying is very true by have listened from other people. No offense, but I still think FreeRADIUS *lacks* of an homogeneous and well organized documentation. I mean, don't take me as an ungrateful guy. We appreciate this product very much, I'm just trying to make some constructive criticism. If there could be some kind of index of contents for the docs (at least) one wouldn't have to search trying to guess upon the filenames which of those textfiles has the information I need. Maybe some PDF or html documentation some day would make the product more attractive. Just a suggestion.
I was trying to add a module not knowing of the above mentioned behavior.
No. You were trying to add a module without telling the server to use it, and you were surprised that the server didn't use it.
No, I wasn't surprised by the server not using it but by the server not telling me anything about it.
If the server *had* printed out the message you wanted, odds are that your next question would be "why doesn't the server use the module when I send it a packet? I listed it in the instantiate section!"
This isn't true at all. In every moment I was conscious that the module wouldn't be used when some packet arrived, since I haven't included it in some of the other sections. If the right message were printed the odds would have been the opposite.
Yes, we've been through this dicussion before with other people. You're not the first to run into this. And the end result of what you want is an endless series of messages explaining why the server isn't doing what you think it's doing.
I can't see why.
It's a lot easier to have the server to print out what is *is* doing, and then tell people to read the documentation to see why it isn't doing what they expect. Usually, their expectations are wrong, and the documentation contains the information to correct the expectation.
Note that my suggestion only would have added a simple WARNING message saving you to discuss with that other people and about future concerns as well. Don't forget that many many people would try-and-test first and then read the docs (though I'm sure people finds difficult to navigate through README-based docs to find what they want to find). I'm not sure that what you say is in _every situation_ a lot easier.
It results in only adding the entry inside the instantiation section and testing if 'radiusd -X' returned some message to give me a clue that the module instantiation was parsed at least.
And what did you expect the server would do after that? You still haven't answered that question. So you're not explaining what you thought was going on, or what you were trying to do. You're just complaining that the server isn't doing what you want.
Sorry I didn't feel you've asked that question before though I'll answer: "I did NOT expect the server to do ANYTHING after that, just to tell me it PARSED the module instantiation". I thought the server wasn't parsing well 'radiusd.conf' or my module was having some kind of coding error preventing the server to load it. Maybe my complain (I'll call it suggestion) is not that relevant to worth server modifications, it's only that I thought it was better for you to know in which whirlpool FreeRADIUS users get stuck in sometimes. Kind regards, --- Tomás A. Rossi Ministerio de Economía Proyecto de Informática Buenos Aires, Argentina
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
I mean a message saying something like: *WARNING: module _blabla_ instantiated but not used, radiusd will ignore it*
Perhaps you didn't read my response. The module is *not* instantiated, because it's not used.
No offense, but I still think FreeRADIUS *lacks* of an homogeneous and well organized documentation. I mean, don't take me as an ungrateful guy.
Then write some and submit it. Other people have done that. Without complaining, too.
If there could be some kind of index of contents for the docs (at least) one wouldn't have to search trying to guess upon the filenames which of those textfiles has the information I need. Maybe some PDF or html documentation some day would make the product more attractive.
Feel free to do the work and submit it back to the project. If you're not going to do that, your comments sound suspiciously like you're asking other people to do work, for free, that you're unwilling to do.
Yes, we've been through this dicussion before with other people. You're not the first to run into this. And the end result of what you want is an endless series of messages explaining why the server isn't doing what you think it's doing.
I can't see why.
Hmm... That's pretty much my point.
Note that my suggestion only would have added a simple WARNING message saving you to discuss with that other people and about future concerns as well.
In nearly 7 years of working with FreeRADIUS, we've only had 3-4 comments about this particular behavior. So it's simply not worth the effort to fix it, because so few people see it as a problem.
Maybe my complain (I'll call it suggestion) is not that relevant to worth server modifications, it's only that I thought it was better for you to know in which whirlpool FreeRADIUS users get stuck in sometimes.
I understand. My point is that much of the whirlpool people get stuck in is because of misunderstandings how the server works, or unreasonable expectations. I answer messages on the list (in part) to address those issues. But there is a minority of people who simply refuse to believe my explanations, and insist on being stuck in the whirlpool. For those, there isn't anything I can do. Alan DeKok.
Alan DeKok escribió:
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
I mean a message saying something like: *WARNING: module _blabla_ instantiated but not used, radiusd will ignore it*
Perhaps you didn't read my response. The module is *not* instantiated, because it's not used.
Ok, *WARNING: module _blabla_ not instantiated because it's not used*
No offense, but I still think FreeRADIUS *lacks* of an homogeneous and well organized documentation. I mean, don't take me as an ungrateful guy.
Then write some and submit it. Other people have done that. Without complaining, too.
I wish I could do that but I'm not that familiarized with FreeRADIUS yet. No complains at all, just concede me the point and do not deny it by comparing it with other commercial and worst-documented software, please.
If there could be some kind of index of contents for the docs (at least) one wouldn't have to search trying to guess upon the filenames which of those textfiles has the information I need. Maybe some PDF or html documentation some day would make the product more attractive.
Feel free to do the work and submit it back to the project.
If you're not going to do that, your comments sound suspiciously like you're asking other people to do work, for free, that you're unwilling to do.
Again, I will gladly do that if I was more familiarized with FreeRADIUS and its docs. Not necessity to suspect from my intentions because I'm not unwilling to collaborate, it's just I can't because of my limited wisdom about the project. As I said I'm new to FreeRADIUS. Now you're sounding offensive.
Yes, we've been through this dicussion before with other people. You're not the first to run into this. And the end result of what you want is an endless series of messages explaining why the server isn't doing what you think it's doing.
I can't see why.
Hmm... That's pretty much my point.
I'm not taking it as bad as it sounds. Just don't see why an "endless series of messages". You mean if you print a message for every request of the kind?
Note that my suggestion only would have added a simple WARNING message saving you to discuss with that other people and about future concerns as well.
In nearly 7 years of working with FreeRADIUS, we've only had 3-4 comments about this particular behavior. So it's simply not worth the effort to fix it, because so few people see it as a problem.
Ok, that I agree.
Maybe my complain (I'll call it suggestion) is not that relevant to worth server modifications, it's only that I thought it was better for you to know in which whirlpool FreeRADIUS users get stuck in sometimes.
I understand. My point is that much of the whirlpool people get stuck in is because of misunderstandings how the server works,
Because of the not-so-easy-to-explore issue of the docs (sorry I'd to repeat that).
or unreasonable expectations. I answer messages on the list (in part) to address those issues.
I don't think my expectations were so unreasonable though they were a little naive and illusory.
But there is a minority of people who simply refuse to believe my explanations, and insist on being stuck in the whirlpool. For those, there isn't anything I can do.
Ok, I hope you aren't putting me in that bag. I believe 100% in your explanations but not agree/understand 100% of them. I frequently post on other mailing lists of other open source and commercial software and never been treated this way in any of them. Sorry if I have offended you with my questions/concerns. Regards, --- Tomás A. Rossi
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
I wish I could do that but I'm not that familiarized with FreeRADIUS yet. No complains at all, just concede me the point and do not deny it by comparing it with other commercial and worst-documented software, please.
Sorry, I can't concede that the docs are that bad, because many people think they're pretty good. I concede that *you* think they're bad, but that's your opinion, and only one of many.
I'm not taking it as bad as it sounds. Just don't see why an "endless series of messages". You mean if you print a message for every request of the kind?
There is only a small amount of functionality in the server. It prints out messages describing that functionality. There is an *infinite* amount of things the server doesn't do. If we printed out a message about everything the server didn't do, that list would be infinite. That's what you're asking for, and that's why I'm saying "No.". I'm sorry this upset you.
Ok, I hope you aren't putting me in that bag. I believe 100% in your explanations but not agree/understand 100% of them.
That's fine. But recognize that you don't control what goes into the server.
I frequently post on other mailing lists of other open source and commercial software and never been treated this way in any of them.
Treated like what? Having someone disagree with you, and explain why? Alan DeKok.
Alan DeKok wrote:
Treated like what? Having someone disagree with you, and explain why?
No. [Quote] Feel free to do the work and submit it back to the project. If you're not going to do that, your comments sound suspiciously like you're asking other people to do work, for free, that you're unwilling to do.
Yes, we've been through this dicussion before with other people. You're not the first to run into this. And the end result of what you want is an endless series of messages explaining why the server isn't doing what you think it's doing.
I can't see why. Hmm... That's pretty much my point. [/Quote]
Just to make my point, you seem to answer concerned people from a superior stand point with the primitive of not changing anything because there's no way you can be wrong. Maybe I misunderstood but a couple of your replies were a little offending to me. They sounded somewhat like you was mocking me. Maybe I'm being too sensitive I don't know. Please just forget everything and lets start again. Thanks, Tom;
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
Just to make my point, you seem to answer concerned people from a superior stand point with the primitive of not changing anything because there's no way you can be wrong.
No. In the message you quoted, I very clearly said you could make the changes you wanted, and submit them back. The problem is you made it clear you *won't* make those changes. But you want *us* to make those changes. And you keep arguing, trying to convince us to make those changes, rather than doing the work yourself. It's not about being right or wrong. It's about you asking us to do work that you are unwilling to do. The only possible response then is "Well, we're not going to do it, so it's only going to happen if you do it." At which point you claim to be offended. If asking you to do work is offensive, then there's very little anyone can do to help you. Alan DeKok.
Alan DeKok escribió:
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
Just to make my point, you seem to answer concerned people from a superior stand point with the primitive of not changing anything because there's no way you can be wrong.
No. In the message you quoted, I very clearly said you could make the changes you wanted, and submit them back.
The problem is you made it clear you *won't* make those changes. But you want *us* to make those changes. And you keep arguing, trying to convince us to make those changes, rather than doing the work yourself.
I keep arguing, not to convince you so you make those changes, just to convince you that my concerns are valid. Then if I convince you I would contribute doing it myself. If what you're saying is true PLEASE point me where in my messages did I mention I wanted you to do the changes.
It's not about being right or wrong. It's about you asking us to do work that you are unwilling to do. The only possible response then is "Well, we're not going to do it, so it's only going to happen if you do it." At which point you claim to be offended.
I'm not unwilling to do anything! And your response isn't that one, your response is something like "no, your suggestions makes no sense to us, we think adding that message isn't the right choice, so forget about it". Of course, then it would be futile to submit local changes concerned to this 'cos you do not agree with them. For me, now that I know the quiet behavior of the server in this aspect, it's not important to have that enhance anymore, but for other beginners that would be nice. Saying that I'm unwilling to contribute and that I'm trying to take advantage of your time (terribly bad judging my intentions) *is* indeed offensive, don't you think so?!
If asking you to do work is offensive, then there's very little anyone can do to help you.
No. You say that if I don't like docs then I have to improve them and submit them. I reply to that saying I would gladly help if I had the wisdom required for that, don't you read anything I write? You could simply admit, "you are right, the docs aren't the best they could be, if you can help I'll gladly accept your contributions". I *never* make clear (as you say) that I won't make contributions to the docs. Maybe if I could get sufficient insight of FreeRADIUS in the future (that only can become true with time), I could help. Asking me to do work isn't offensive at all, your bad-ass attitude is offensive. Why should I submit changes if all my suggestions has been shot down and you never admitted neither of my concerns? Please don't wanna argue anymore with you, there's very little I can do to help you. Tomás A. Rossi
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
Why should I submit changes if all my suggestions has been shot down and you never admitted neither of my concerns?
I'm sorry that you feel your suggestions were "shot down" when I gave reasons for not doing what you want. This may be news, but not everyone's contributions to the project are accepted. Alan DeKok.
Alan DeKok escribió:
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
Why should I submit changes if all my suggestions has been shot down and you never admitted neither of my concerns?
I'm sorry that you feel your suggestions were "shot down" when I gave reasons for not doing what you want.
This may be news, but not everyone's contributions to the project are accepted.
Do not be sorry. You're a bipolar guy, "submit instead of complain!!, no no, wait, not everyone's contributions to the project are accepted (specially from guys that we hate) so keep complaining, either way we'll do nothing because we make it for free, so better, please don't complain either!". Tomás A. Rossi
Tomás A. Rossi wrote:
Do not be sorry. You're a bipolar guy, "submit instead of complain!!, no no, wait, not everyone's contributions to the project are accepted (specially from guys that we hate) so keep complaining, either way we'll do nothing because we make it for free, so better, please don't complain either!".
Your comments surely shattered Alan's fragile ego and catapulted your side into the lead. We all agree with you now.... Ad hominem. The last refuge of a weak argument. -- Dennis Skinner Systems Administrator BlueFrog Internet http://www.bluefrog.com
Dennis Skinner escribió:
Tomás A. Rossi wrote:
Do not be sorry. You're a bipolar guy, "submit instead of complain!!, no no, wait, not everyone's contributions to the project are accepted (specially from guys that we hate) so keep complaining, either way we'll do nothing because we make it for free, so better, please don't complain either!".
Your comments surely shattered Alan's fragile ego and catapulted your side into the lead. We all agree with you now....
Ad hominem. The last refuge of a weak argument.
Think what you want. I've finished my argumentation a lot messages ago (and it was useless). This is not to defend my posture anymore, just a criticism that Alan for sure won't take into account because he's perfect, but it isn't a fallacy. If you aren't going to add further useful information to solve the conflict please remain silent, for your ironical message does only contribute to your own ego. Tomás A. Rossi
Tomás A. Rossi wrote:
Think what you want. I've finished my argumentation a lot messages ago (and it was useless). This is not to defend my posture anymore, just a criticism that Alan for sure won't take into account because he's perfect, but it isn't a fallacy. If you aren't going to add further useful information to solve the conflict please remain silent, for your ironical message does only contribute to your own ego.
Finished, yet still talking. Thank you, I hadn't thought about my ego. I guess it is bigger now, ooo and shiny! As to content, I agree with alan (small A). I don't need fluff in the debug output. It is hard enough to get new users to read the whole thing (read the list for a couple weeks). Follow the instructions and it works. If you don't like the instructions, stop complaining and submit patches. I don't know that Alan has rejected any documentation related patches. I like the fact that he rejects software patches that don't make sense or are for a very limited audience. It makes a smaller, more stable, and less confusing product. Show me another product, open or closed, that has an open forum like this and posts from the main dev make up almost 20% of the traffic (as of this morning, 18.5% since Jan 1, 2005 2307/12448). I deal with several commercial vendors, and most of the time the actual devs are walled off behind several layers of CS and TS levels and are all but impossible to talk to. -- Dennis Skinner Systems Administrator BlueFrog Internet http://www.bluefrog.com
Dennis Skinner escribió:
Tomás A. Rossi wrote:
Think what you want. I've finished my argumentation a lot messages ago (and it was useless). This is not to defend my posture anymore, just a criticism that Alan for sure won't take into account because he's perfect, but it isn't a fallacy. If you aren't going to add further useful information to solve the conflict please remain silent, for your ironical message does only contribute to your own ego.
Finished, yet still talking. Thank you, I hadn't thought about my ego. I guess it is bigger now, ooo and shiny!
As to content, I agree with alan (small A). I don't need fluff in the debug output. It is hard enough to get new users to read the whole thing (read the list for a couple weeks). Follow the instructions and it works. If you don't like the instructions, stop complaining and submit patches. I don't know that Alan has rejected any documentation related patches.
I like the fact that he rejects software patches that don't make sense or are for a very limited audience. It makes a smaller, more stable, and less confusing product.
Show me another product, open or closed, that has an open forum like this and posts from the main dev make up almost 20% of the traffic (as of this morning, 18.5% since Jan 1, 2005 2307/12448). I deal with several commercial vendors, and most of the time the actual devs are walled off behind several layers of CS and TS levels and are all but impossible to talk to.
Ok. Lets make peace, I got out of my mind. It's just that Alan doesn't seem to reply very friendly sometimes, I don't wanna fight anymore. Not my intention at all. In the future I wont complain without submitting patches and please don't mock me anymore cause I was at least being respectful (till some point of the thread). Willing to forget aggressions and begging pardon of everyone (specially of Alan), Tomás A. Rossi
hi, I dont want debug output that lists every module not loaded. that sort of thing should be left to a tool such as 'radiuscheck' or somesuch. THAT tool is ideal for printing out 'module blahblah instantiated but never used!' types of messages. theres my 0.01 euros alan
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
On the other hand, I've been playing around with 'radiusd.conf' I've discovered that if I add some random module name in the module instantiation section, radiusd doesn't complain at all about it's non-existence! This is a little inconvenient.
It's a bug. Alan DeKok.
Alan DeKok escribió:
=?ISO-8859-1?Q?=22Tom=E1s_A=2E_Rossi=22?= <tomas@mecon.gov.ar> wrote:
On the other hand, I've been playing around with 'radiusd.conf' I've discovered that if I add some random module name in the module instantiation section, radiusd doesn't complain at all about it's non-existence! This is a little inconvenient.
It's a bug.
Is it fixed in a newer version? (I'm using 1.0.4) Regards, --- Tomás A Rossi Ministerio de Economía Proyecto de Informática Buenos Aires, Argentina
participants (4)
-
"Tomás A. Rossi" -
A.L.M.Buxey＠lboro.ac.uk -
Alan DeKok -
Dennis Skinner