Older blog entries for dwmw2 (starting at number 42)

Happy Happy Joy Joy. With prumpf's fixes, JFFS no longer appears to be reliably falling over on the toy. Now I just need to implement the buffer write command rather than writing 2 bytes at a time, and see if I can eke some vaguely respectable performance out of it.

People are starting to give me MTD code too, which is good - I now have a DiskOnChip Millennium driver, and code to support the "AMD/Fujitsu Standard" CFI command set.

Managed just a little bit of documentation for the MTD code in 2.4.0-test3 - Configure.help entries. Investigated DocBook briefly and went back to coding/testing. I will do it, I promise - just not today.

It seems that Bryce still can't spell properly - it's spelled with a 'u', like this:

Fucking Useless Telco.

Went to the UKUUG's Linux 2000 Linux Developers' Conference at the weekend. Was fun, although the promised fight between sct and Hans Reiser didn't happen - but there was a concerted effort at one point to take the roving microphones off the pair of them during the Q&A session after Luke's talk.

Luke showed me the ETRAX board which Axis have Linux running on - it's cute. Embedded Linux is definitely where the cutest toys are - none of this having to move house to fit them all in, like sct and _A_.

Also I learned how the Bonobo component architecture got its name.

Haven't had to deal with a telco this week yet, thankfully. Although I should perhaps chase up Vodafone, who haven't managed to respond yet to the email I sent them about two months ago.

Although the GSM standards lay out a simple way for them to turn on and off a dedicated voicemail icon on customers' handsets just by sending a specially-formatted SMS message, and although they already have three or four configurable options for customers to select how they're notified of new voice mail, they claim that they can't add the standard method to their list of possibilities.

The best they can do is send you a normal text message saying "Recall has $n new messages". Which you promptly read and delete, then dial up to listen to it, get interrupted and forget all about it. Proper GSM providers will keep the voicemail icon on your screen until you've actually listened to the message. This is one of the few things that Orange actually got right, but it seems that Vodafone can't.

Fucking Useless Telco.

Cool. Linus included the MTD code into 2.4.0-test3-pre4. Best write the documentation at some point, I suppose. And dig myself into a trench to prepare for a wad of silly questions.

Heard from NTL today. They claim that it's not their fault they're displaying the wrong Caller ID - BT are providing them with the wrong number and it's all BT's fault; so I should take it up with BT.

I pointed out that BT are providing the correct number to every other telco in the world, and it seems to be just NTL that get it wrong - and they persisted.

I pointed out that Orange had the same problem, and after briefly blaming it on BT they fixed it and admitted that it was a misconfiguration on their own switch - and still they persisted.

I said that if it's a signalling problem between BT's switch and theirs, then it's not something that I, as a customer, should have to deal with - they should talk to BT directly and get it sorted out. They said they were trying to, and they'd call me back.

I'm not holding my breath.

Fucking Useless Telco.

OK the BT HomeHighway installation is completed. The engineer who was supposed to be there "between 8am and 10.30am" was in fact there between around 10am and 12.30pm - but what does writing off half of a customer's day matter?

Turns out it took him longer than expected because there was a DACS on my old analogue line which needed to be removed - and the old modem which I'd found lying around and had declared to be a 28.8K one after testing can in fact do 33.6K now it's got a decent line.

It seems that I get to keep all three of the numbers on the various letters I received. One for each analogue 'line' and one for the ISDN 'line'.

However, in spite of the fact that they told me last week I'd see all the incoming voice calls on the ISDN S-bus, so the computer has a chance to answer them and do voicemaily-things, that's not the case - only calls to one of the numbers appear on the S-bus, and the computer can't set its Outgoing Caller Identity to either of the other two numbers either.

Perhaps I would have been better off just getting an ISDN2e installation. It appears that BT are charging £88 per quarter for ISDN, but £39.99 per month for HomeHighway. So although I'd have to get a TA to allow me to plug in my old analogue equipment, I'd probably be better off in the long run, and I'd be able to play.

Fucking Useless Telco.

Hmmm. Bryce hasn't been in the US long, and he seems to have forgotten how to spell already. Is it something in the water? The name of the political party governing the UK is the "Labour" party.

Before you ask - No, I don't have anything better to do than sit here and be pedantic. I'm stuck at home waiting for a BT engineer to finally turn up and install my ISDN line. I've been waiting an hour and a half so far.

Fucking Useless Telco.

Happy Happy Joy Joy. MTD code is working nicely on my toy with JFFS as the root filesystem. So I made a new release and also sent it to Linus in the hope that he'll put it into 2.4 as it's "only new drivers." Now I can work on making a CONFIG_BLK_DEV option so that I can turn it off.

Managed to catch the elusive sct during a brief period of connexion and prod him for vmap_iobuf() again. He's still on dialup because BT haven't managed to move his leased line yet - only two or three weeks after they were supposed to have done it.

Fucking Useless Telco.

Tried to call BT last night at 7:45pm, to confirm that they're really not going to try to take my phone number away from me.

Got a recorded message along the lines of "We're sorry the helpdesk is closed at the moment. Our working hours are Monday to Saturday, 8am to 8pm. Please call back during those times."

I put the phone down and called back immediately, as requested, but just got the same message.

Fucking Useless Telco.

Grrr. I changed the structure presented to userspace by the MEMGETINFO ioctl, which is a pain in the arse because I can't currently compile user-space binaries for by test board.

They'll call me back, apparently. I bet that if they ever do, they've not managed to fix the problem, and probably not even managed to understand it either.

They didn't, they hadn't, and when I eventually called them back, they hadn't.

Fucking Useless Telco.

In today's second round of entirely predictable and prediected cluelessness from a telephone company, it turns out that NTL really are supplying my Caller-ID box with the Network number of the incoming call, rather than the Presentation number.

Upon calling to point this fact out to them, I received the normal "Blame it on the other end" response, followed by utter confusion when I explained in more detail.

They'll call me back, apparently. I bet that if they ever do, they've not managed to fix the problem, and probably not even managed to understand it either.

Fucking Useless Telco.

Hmmm. Three letters from British Telecom about the upgrade of my landline to ISDN^H^H^H^H "BT HomeHighway".

The first says "Your new telephone number will be xxxx".
The second says "Your new telephone number will be yyyy".
The third says "Your telephone number is zzzz", which actually matches my existing phone number.

Somewhat confused, I called BT. I thought that HomeHighway only had two phone numbers - your original one and an extra one for the data line. They categorically stated that they're not going to attempt to change my old number, which would cause me a lot of hassle and be an entirely stupid thing to do.

This is bloody annoying, because hearing that from BT, of course, means that in all probability they _are_ going to change my bloody number.

Fucking Useless Telco.

33 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!