Wikipedia:Reference desk/Archives/Computing/2010 April 29

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

April 29

IRC Aggregator Bot?

Is anyone aware of such a thing? Basically what I would like is to be able to park a bot in 5 IRC rooms plus it's own room (##BOT or something) and have it repeat everything said in the other 5 IRC rooms? ^Avicenna_sis @ 05:51, 29 April 2010 (UTC)

I don't know of one offhand, but the IRC protocol is pretty simple, so depending on how much programming you want to do, you could make one yourself. There are lots of IRC libraries for easy use too. Shadowjams (talk) 05:40, 2 May 2010 (UTC)

Wikipedia use ontologies for manage the knowledge and the relations between them?

Wikipedia use ontologies (as an informatic concept) for manage the knowledge and the relations between concepts? with some special software for include intelligence to the web site? —Preceding unsigned comment added by 157.159.40.82 (talk) 09:52, 29 April 2010 (UTC)

The above question was asked by 157.159.40.82 (talk). 24.189.90.68 (talk) 10:05, 29 April 2010 (UTC)

I'm not sure if I understand the question, but if you are asking, "could the links made between articles and concepts on Wikipedia be made, along with specialized software, into some kind of rich semantic web for an A.I. program?" the answer is, "maybe." I don't see any a priori reason why that couldn't be the case. But writing an A.I. program that could take advantage of such things is non-trivial. There are a number of smaller scripts that mine Wikipedia for semantic data, but Wikipedia's data is not very uniform in most cases, which makes anything but very trivial things hard to do. --Mr.98 (talk) 12:50, 29 April 2010 (UTC)

Firefox SSL warning

Is there a way to disable the warning about a wrong SSL certificate for certain sites? I'm accessing a few servers via SSH's port forwarding, so I'm always using https://127.0.0.1 to access the servers, with the real IP being hidden by the SSH connection. In its default configuration, Firefox allows me to whitelist a certain certificate for a certain IP, it seems, but if the certificate changes (because the server behind the SSH connection changes), it nags me again. Is there an about:config hack that allows me to whitelist 127.0.0.1 or 192.168.0.0/16, so that wrong certificates for this IP or IP range are never reported (or at least not with such an endless array of click-throughs, a simple "Are you sure?" would be acceptable)? As I'm using the same Firefox for regular browsing, I would prefer it if I don't have to disable the warning mechanism entirely. -- 78.43.60.58 (talk) 10:14, 29 April 2010 (UTC)

Why are the certificates changing? If you're making these certificates yourself, make a CA and install it in your Firefox as a trusted root, and sign your certificates with it, instead of having them self-sign if that's what you're doing now. 69.228.170.24 (talk) 23:38, 29 April 2010 (UTC)

They are changing because they're at different locations, and I'm not in the position to install different certificates on them. So I really need to disable the feature in Firefox. -- 78.43.60.58 (talk) 09:11, 30 April 2010 (UTC)

Can't you whitelist multiple certificates for the same IP in firefox? I suppose you could set up a proxy server on your local client that gives Firefox a certificate it likes, but it sounds like you're really trying to stop checking the certificates, in which case why even use ssl? Why are you tunnelling ssl through ssh anyway? If you trust the machines at the other end and you trust the ssh fingerprints and the security of ssh, just run http through ssh. 69.228.170.24 (talk) 05:58, 1 May 2010 (UTC)

No, I can't whitelist multiple certificates for the same IP (at least there is no obvious way to do so), otherwise I wouldn't be asking for help here. As soon as I tell it to save the second certificate for 127.0.0.1, it forgets about the first. If you happen to know how to change that, please do share.

Again, I have no control over the remote endpoints - they speak HTTPS and only HTTPS, so I have to deal with that. I need to get Firefox to comply, not the remote endpoint. -- 78.43.60.58 (talk) 16:00, 3 May 2010 (UTC)

The workaround I'd recommend is using ssh SOCKS forwarding (the -D switch) instead, and telling Firefox to use 127.0.0.1 as a SOCKS proxy. You can even tell FF to only use the proxy for some specific sites by writing your own proxy auto-config script. —Ilmari Karonen (talk) 12:24, 4 May 2010 (UTC)

My pen drive .

My pen drive don't allow to format .How can I format it. —Preceding unsigned comment added by 192.248.87.1 (talk) 16:11, 29 April 2010 (UTC)

Some pen drives have a little switch on the side which prevents the computer modifying data on the drive unless it's set in the "write" mode. Or the drive might have just failed, in which case there is nothing you can do to fix it and you would need to get a new one. .tkqj (talk) 16:21, 29 April 2010 (UTC)

HP has a utility that might work: http://files.extremeoverclocking.com/file.php?f=197 The brand and model of your drive might let use help you more. ---— Gadget850 (Ed) ^talk 00:56, 30 April 2010 (UTC)

jit compilers

the machine language code blocks produced by the jit compilers are deleted when the application is closed? t.i.a. --87.5.125.159 (talk) 20:11, 29 April 2010 (UTC)

Yes. Memory space is allocated and code is emitted into that space; when the application closes the space is recovered and the code destroyed. I can't think of a circumstance where that wouldn't apply - I've never heard of emitting code into shared memory, and mobile code like JRMI is transmitted as an intermediate code like java bytecode. -- Finlay McWalter • Talk 20:26, 29 April 2010 (UTC)

I would hope that the JITted code is deleted when the application is closed, but, of course, there might be some badly-written system that keeps the code around in memory. Comet Tuttle (talk) 20:27, 29 April 2010 (UTC)

It really wouldn't be a JIT compiler if the code persisted. It would just be a regular compiler, like GCJ, that creates machine-code. Nimur (talk) 03:35, 30 April 2010 (UTC)

I thought that the .NET Framework caches its machine code in the Native Image Cache and the GAC inside the C:\WINDOWS\assembly folder. Otherwise, a .NET application would be slow to start up every time. When the native code is cached, it is only slow to start up the first time. The first time it starts, I think it compiles the bytecode JIT into native code and then caches it. The view of the assembly folder inside Windows Explorer is not very accurate, though. If you plan on taking a look to see what I mean, you should use the command prompt.--Best Dog Ever (talk) 07:18, 30 April 2010 (UTC)

Outlook / Outlook Express / Windows Mail / Windows Live Mail

Where do these programs store user and server names (POP, IMAP, SMTP)? The reason I'm asking is that I'd like to prepare a script that pulls the account info (sans password) from the file/registry, so that it can be fed into an installer for an alternate mail client like Thunderbird. -- 78.43.60.58 (talk) 21:11, 29 April 2010 (UTC)

For Outlook Express see here. -- Finlay McWalter • Talk 21:20, 29 April 2010 (UTC)

For Windows Mail see here. -- Finlay McWalter • Talk 21:26, 29 April 2010 (UTC)

Rough info about Outlook 2003 is here. -- Finlay McWalter • Talk 21:31, 29 April 2010 (UTC)