World: r3wp
[!Uniserve] Creating Uniserve processes
older newer | first last |
[unknown: 5] 17-Mar-2005 [143] | LOL - yeah I understand how that goes. |
Dockimbel 17-Mar-2005 [144] | ;-) |
Anton 18-Mar-2005 [145x6] | My last mail to Graham was Undeliverable. This could be because of the content I put in it :) |
Reporting-MTA: dns; sv7.prth.eftel.com X-Postfix-Queue-ID: E836A143ADA X-Postfix-Sender: rfc822; <anton's email address> Arrival-Date: Thu, 17 Mar 2005 21:00:45 +0800 (WST) Final-Recipient: rfc822; [gchiu-:-compkarori-:-com] Action: failed Status: 5.0.0 Diagnostic-Code: X-Postfix; host pop.compkarori.com[203.79.110.37] said: 552 write failure (in reply to end of DATA command) | |
Doc, just reading your task-master doc linked above. Grammar corrections: provides you an integrate framework -> "provides you with an integrated framework" | |
This service provides you : -> "This service provides :" | |
theses -> "these" needs the following files been installed -> "needs the following files installed" its spawns -> "it spawns" job in background -> "job in the background" transfer to it your data -> "transfers your data to it" additionnal -> "additional" service which dispatch -> "service which dispatches" If there's no free processes -> "If there are no free processes" it become available -> "it becomes available" properties, event and method -> "properties, events and methods" optionnally -> " optionally" Maximum number of background process -> "Maximum number of background processes" controled -> "controlled" performances -> "performance" | |
I checked most of those first error strings above and they are unique in the document. | |
Dockimbel 18-Mar-2005 [151x2] | Thanks Anton, I'm fixing these errors at once. |
Ok, done. | |
Graham 18-Mar-2005 [153x2] | Anton, that was more likely a bug in my save-email routine, which I hope is now fixed. |
I'm ready for the next round of testing after some bug fixing. Anyone care to send an email to me at [gchiu-:-compkarori-:-com] ? Thanks. | |
Ammon 18-Mar-2005 [155] | incoming |
Graham 18-Mar-2005 [156x3] | Got mail from didec and ammon. |
both classed as spam due to smtp challenge timeouts :( | |
clearly got to work on that ... my smtp service checks to see if the sender has a valid mailbox ... | |
Graham 19-Mar-2005 [159x5] | this is my first crude webmail interface to my smtp service. http://www.compkarori.co.nz:8001/ use userid: "test" password: "account" If you send a message to [test-:-compkarori-:-com], it will take 15mins at least to get thru ( which is the default delay I have built in for mail from new users). But after that it should be very quick. |
And then you will be able to see the message in the webmail interface. | |
I have truncated the message display to the first 5000 chars | |
Anyone like to collaborate on building a decent web interface ? | |
BTW, that is a test to the email account: [test-:-compkarori-:-com] ... and not me :) | |
Anton 20-Mar-2005 [164] | Doc, could you have a look in !webmail group. Graham's cheyenne server is not getting through to me properly (hangs on second read-io of 2048 bytes). |
Dockimbel 21-Mar-2005 [165] | I'll look at it. But remember that Cheyenne's RSP are early alpha and they are still a several issue to solve and bugs to fix before having making them reliable. |
Anton 21-Mar-2005 [166] | Sure. :) I am helping Graham to find the bugs - I do not demand any level of service. :) |
Dockimbel 21-Mar-2005 [167] | The main issue with current RSP is that it includes a session variables synchronization system that's not yet finished. So, the freezing issues should be related to that. This system queues the RSP jobs if it detects that they use the same session variables. (It garanties that your session variable values would not be corrupted by the concurrent execution of RSP). It needs more testing and tuning to be fully reliable. |
Dockimbel 17-May-2005 [168x4] | Just a quick update info about Cheyenne: I'm working on a new version of the RSP module that should fix the previously encountered issues. It needs a lot of testing so it will take a few days before I'll publish anything. |
About UniServe, I'd like to include some protocols/services done by Graham in UniServe's distribution (if he agrees). If anyone has done some protocols or services implementations that would be useful to others, please post some info about them here. I'll be also glad to have from UniServe's users some feedback about UniServe's API, missing features that you would like to be included, etc... | |
To answer a question of MikeL, UniServe is not tied in any way to LNS. | |
In fact, from what I know about LNS, they could be implemented above UniServe using UniServe's service API. | |
Graham 17-May-2005 [172] | Sure, include away ... :) |
Pekr 17-May-2005 [173] | what are those services, protocols? |
Graham 17-May-2005 [174x3] | ummm... thinking hard ... |
popd .. which is a pop server | |
and smtpd which is a smtp server ( receives mail only so far ) | |
Pekr 17-May-2005 [177] | what about your sync protocol? Is that separate one, or part of Uniserve architecture? |
Graham 17-May-2005 [178x3] | That's separate ... |
It was a contracted job that is still not finished ... | |
Anyway, the smtp service is currently being used to process my mail, and it is being picked up by mail clients using the pop service | |
Pekr 17-May-2005 [181] | hmm, and how is your sync stuff going? Is it reliable already? :-) |
Graham 17-May-2005 [182x3] | Well, it works for me .. but that is not adequate amount of testing! |
What I want to see with Uniserve is to be able to encrypt binary data as it is sent | |
Your welcome to test the sync service ... if it works, you won't need to use IOS :) | |
Pekr 17-May-2005 [185] | you can encrypt streams yourself, no? |
Graham 17-May-2005 [186] | yes .. and no |
Pekr 17-May-2005 [187] | OK, I can test .... I will write also some proposal for new kind of app, I think some money can be earned that way :-) |
Graham 17-May-2005 [188x5] | the uniserve client detects that a "packet" is sent by knowing how many bytes are coming ( as in http ), or in knowing at unique end sequence |
as in "." + crlf | |
but when you stream a file that is encrypted, the file length is now changed | |
if you encrypt the whole file at the start, it defeats the point of streaming which is to save cpu | |
the replication server and client are at http://www.compkarori.com/coyote/ | |
older newer | first last |