“Richard Stallman Speaks About Back Doors After NSA Documents Leak” is the title of a submission that I summarised as follows: “Companies such as Microsoft, Facebook, Apple, and Google are scrambling to restore trust amid fresh litigation over the PRISM surveillance program. Richard Stallman, the founder of the Free Software Foundation and a newly-inducted member of the 2013 Internet Hall of Fame, speaks about not only abandoning the cloud, which he warned about 5 years ago, but also escaping software with back doors. “I don’t think the US government should use operating systems made in China,” he says in this new interview, “for the same reason that most governments shouldn’t use operating systems made in the US and in fact we just got proof since Microsoft is now known to be telling the NSA about bugs in Windows before it fixes them.””

The text for the next part of the interview is not ready yet. I’ve just released this second part (now in the front page of Slashdot), but I need to do some fact-checking on SELinux (NSA-developed) before releasing the next part. Busy times since the NSA leaks…

Identi.ca becomes part of the ‘Internet rot’ problem

“Wait, What? Identi.ca Doesn’t Care About My Data???”

ell, it just had to happen sooner or later. Face the facts. So-called ‘Cloud’ (or Fog as I prefer to call it) Computing is hype and it is dangerous not just in the security sense. Your data, or even your sentimentally-valued information, memories, etc. have no value to others, those who merely provide hosting for self gain or ego or whatever.

Over the years I have seen many so-called ‘clouds’ collapse, whereas with my stuff, almost everything is in tact, even what was online a decade ago (maybe IP addresses changed a little, but it is all still “up there”). This is because to oneself, data matters and data has value. It’s my data. To Fog Computing providers, your data is just mere “content”, something for other users to “consume”, potentially for “monetisation” by the managing party.

Even if the ‘cloud’ is built on Free/Open Source software, and even if that software is made available for download, there is no guarantee that data will be exportable from the database. Identi.ca is a good example of this point (more on that later). What a travesty!

“Why Are You Surprised?”

I previously wrote about my experience losing all my data and work at Digg and Netscape (see the posts “Digg Stabs All Users in the Back, Deletes All Their Content, EVERYTHING!” and “With ‘Cloud Computing’ You Can’t Keep Your Data Under Your Control“).

One could rename and do a rendition of “never fall in love again” as “never fall in cloud again”. Fog Computing is toxic, and the more years go by, the more people (and businesses) will recognise this. From losing access to new binary releases or newly-updated source code people are now losing access to actual data, which they never even retained on their own devices in the first place. What a suicidal decision that would have to be…

“Backup? Export? Where’s the Profit in That?”

Some days ago I made an effort to advise Identi.ca to preserve content and make old URLs accessible, for the sake of preservation. About a week earlier I enquired about the backup feature (experimental) being broken and not allowing me to export my data; this had been broken for years! I never receives a response Right now it says it “provides an incomplete backup” and finally I can actually export some data, but only the past month’s data (I have posted there for over 4 years).

Recently, having suffers another major ‘cloud’ data loss, I made a local copy of all my tweets. Twitter’s archive dump is well-formatted after Twitter announced the feature (months ago) and refined it over time, making it available to all users and not just select few, taking minutes to generate an archive and then sending a notification by E-mail, indicating an archive is ready for download (I have posted nearly 85,000 tweets). This is a commendable move by Twitter, but still, given that Twitter traffic declined 20% in the past 3 months alone (based on Alexa.com), how long will Twitter be hosting the tweets itself and thus keep URLs in tact?

“Sites Go Dark? Never!”

Well, actually it happens all the time, usually financial considerations being a major factor for the operator/s. Consider all the third-party image hosting services and link shorteners such as http://ping.fm/ that I used a lot in 2009-2010. These latter services are a nightmare even in the eyes of the founder of the World Wide Web as lacking any contextual information like link/page, the URLs are worthless; they are utterly dead and useless links, they cannot be recovered even through the Web Archive. Often enough this renders the tweets too rather useless. If people use shorteners in blogs, then they are truly misguided and they too will suffer the consequences.

“Didn’t Identi.ca PR Say it Was Just a Conversion

Identi.ca is trying to call this a “conversion”, but the only thing such sites seem to be planning to convert is user accounts, and not even many of them. Reading “Identi.ca conversion to pump.io” again, it seems clear all user data will be deleted (not left online, thrown away). As manual backup is trimmed/incomplete, this leaves users like myself unable to even pull the raw data while the site and the database are still online. What incompetence; what a betrayal!

“So New Software Platform Means Starting From Scratch?”

It sure seems to be the case. “The Identi.ca social network service will be moving to a new software platform on June 1, 2013,” says the announcement. This is a nice way of saying that the site is reassessing the way it operates and perhaps the business/operating model, neglecting all that was put into it by many users. It started by stating the migration would occur in one of the secondary domains and now this is coming to Identi.ca, as some people feared.

In a month-old post titled “Identi.ca conversion to pump.io” says:

Active accounts will be converted automatically to the new platform. Active users don’t have to do anything to continue using the service.

Accounts that have not been used since May 1, 2012 will not be converted. If you have friends or people you like on Identi.ca that you think should keep being users, please let them know. Just posting one notice will mean their account gets converted.

If you’re interested in seeing how pump.io works right now, you can set up an account by going to http://pump.io/ and clicking the “try it” button.

pump.io has a very different API than StatusNet. If you use a desktop or mobile client for Identi.ca, please check with the software developer to see if they’re planning to port to pump.io.

Backups of all public data will be available on archive.org after the switchover. You can also make a manual backup.

pump.io is under active development; some features you’re used to from StatusNet will be unavailable or will be implemented by third parties. There are a lot of things that pump.io does better, though. Social games, sharing pictures, and web-wide social buttons are just part of the new fun.

How hard would it have been to just keep the old CMS in tact, even for the sake of old URLs being accessible? Probably trivial bar space and CPU concerns, right? Identi.ca should reconsider its position on this. Maybe Evan (Identi.ca founder) can ‘kickstart’ a fund-raiser to help sponsor this; I would put my money in to preserve my data. Maybe others would, too.

New Identi.ca means the following: Heaps of broken URLs, disregard for people’s work which was posted online (essentially just like in the case of Digg, Netscape/Propeller, etc.), and elimination of many connections like “Followers”/”Following”. It was bad enough when theme-related information got dumped as part of the previous software upgrade. Not the same is being done with post data. Only user data is preserved (name, E-mail, etc.). Imagine if YouTube did the same thing, throwing people’s videos out with the bathwater… YouTube did throw away people’s theme-related information when it applied some updated, but these are often restorable with some effort. The same goes for Facebook with its layout tweaks. Imagine the outrage resulting from a Facebook announcement that it is dumping all old posts and photos…

“So It’s All Gone in a Few Days?”

Seems so, unless Identi.ca decides to keep the StatusNet setup in tact, as least for legacy purposes (I have thousands of links to Identi.ca URLs out there, and they are needed for context).

I have some mirroring of selected Identi.ca accounts in a IRC channel, which I back up and make available online for good. Alas, that is hardly a substitute.

“What Can I Learn From This Disaster?”

When your online work (including Facebook, Twitter, etc.) will no longer align with someone else‘s business model, say goodbye to it all. Yes, seriously. This is not a charity.

Yesterday I wrote about self-hosting one’s photographs. No guarantee of export options in Flickr, eh? Are you listening, Flickr MicroHoo! users? As my friend Tract put it in “Tracy’s photo album,” this is “better than flickr! I won’t lose all of my pics when flickr disappears.”

Recently, speaking to relatives or mine, I advised them to access Friendster to export or save their accounts’ contents before it’s too late to do so. A stampede to export would cost a site in the process of shutting down a lot in terms of bandwidth, reducing incentive to provide such an option, especially when there is no brand/reputation to protect anymore. And if you think Twitter and Facebook are any different, think again. The only difference is, those sites are probably quite a few years away from shutting down and throwing the content down the drain. Why else would the Library of Congress already amass tweets of everyone? Spying concerns aside (profiling people based on their posts from decades in the past), this shows that the US government too recognises that all Fog Computing ends up the same way — it ends up down. Not up, down. Offline, probably stored on some magnetic tape/disc in some warehouse owned by some company which had nothing to do with the data and not making accessible online, even to those who provided all this data. In due course this storage media too will erode, collect dust, and become inaccessible (incapable or getting salvaged), in essence destroying the data for good and not even giving data contributors a chance to preserve/curate the data themselves.

“What Should Identi.ca Do?”

It’s simple. Keep the data up. Keep the old URLS in tact. Make the privately-owned database accessible one way or another. If the backup feature permits complete download of all data for a given user, then it doesn’t resolve all the issues, but it may resolve some.

Y BT Internet connection has been faulty since the beginning of this year. I have spent no less than about 10 hours speaking to support representatives in an offshore call centre, all of whom go through the script and a list of steps that “test” the connection, never mind if a dozen people before them ran the same tests. I have been polite but assertive, especially after these issues persisted for months. But never ever did they send out an engineer (meaning, a UK-based person paid at UK rates) to address the issue. At one stage they sent out a replacement router, but unsurprisingly this did not resolve the problem.

Imagine having a flaky connection when you work from home (in the employment sense). Calls are dropping, SSH sessions are dropping, IRC logging and conversations are choppy, and even Web browsing is very erratic. Imagine this going on for about 4 months. Imagine having your ISP refusing to just fix the issue by sending an actual person to the site for investigation.

My issue have been escalated internally numerous times and I have just spoke to their manager about it. No compensation can ever recover or make up for the time and work lost due to BT’s systemic incompetence. But wait, it gets worse. Not only is BT too ‘cheap’ (must increase shareholders value!) to send out an engineer; it is unable to even follow up with calls that it promises to make. The automated phone reminder which says they would call works correctly, even phoning me to wake me up at 7 AM on a Sunday. But the actual representative ‘forgets’ to call. Oops. I guess the customer does not matter enough to inform. If the customer stays home for a 2-hour time slot allocated for a call, they can just be left out in the cold, right? Well, that’s BT.

My issues with BT were serious back in 2011 when they were unable to simply set up my connection, incurring weeks in delay. I should have taken the hint and taken my money elsewhere, but BT has a monopoly on the lines. So I stayed with BT, only after their cancellations department was very insistent and successfully persuaded me to give them another chance. They also compensated me which was an admission of guilt more than it was a compensation for all the time lost and the agonising experience lasing weeks.

BT’s issues are not technical. BT’s issues are systemic. The company assumes its customers are dumb. It insists on running simple tests rather than addressing low-level issues that have been ongoing for months. It would rather have you suffer for days and talking to poorly-paid employees than send out a person who — through direct physical contact with the infrastructure — can probably remediate the issue immediately.

BT is not a company that cares about people. It cares only about money to the extent where it forgets what customers actually mean and why bad service will give them bad reputation and discourage new customers from joining,

Today, after months of bad service, BT said they would send an engineer (at long last!) to my house, but only in two days from now (I stood firm on quick action), meaning that I would suffer from faulty connection for a couple more days until I go on vacation (Monday). I also need to wait at home for a five-hour time slot on Saturday. Great, eh? See how much bad service from BT impacts one’s life on a daily basis.

If you never relied on BT for anything, do yourself a favour and never do. BT doesn’t care about people, it will take your money and run up a tree, then tell you that you must be dumb and the fault must not be theirs. You are just a fool with his/her money…

LEXA data is not an accurate measure of site popularity (see my views from 2005 and from 2006), but trends as judged by Alexa can sometimes — especially for large sites — indicate if a site is going mainstream or going away. With statistically-meaningful deviations from the baselines it is now fair to say that Twitter has jumped the shark. The amount of communication I get in that site is definitely not increasing and it seems to be turning more and more into a hub for celebrities, perhaps because 140 characters are enough for them or their PR agents to communicate with. Many former Twitterers seem to be logging in less (some never at all), or reading less, certainly communicating less in the comments — something which is also a growing issue in Facebook and Google Plus, less so in JoinDiaspora, which is my favourite social network these days.

OMETIMES I get asked how much traffic my Web sites are getting. The only honest answer I can offer is that I don’t know. It depends a lot on how it’s measured, what measures it (if anything), when it is measured (peaks taken into account), and how spiders or spam traffic get culled out. Bot traffic is increasingly made more sophisticated, so it is hard to classify one thing as a bot viewer and another as a human viewer. In any event, by far the biggest site that I run is Techrights. It has almost 20,000 pages that I wrote over the past 6.5 years. Schestowitz.com, this one particular site, predates Techrights and has more pages in it than Techrights, but some of the content is not of high quality, e.g. my USENET posts. Then there is the site of my relative Harvey, who lives in Florida. I set up that site for him and have helped him maintain it since 2004. Recently, my friend Mark and I set up Medivasc.com, which also attracts a vast amount of traffic. Those are just 4 of my sites; there are about a dozen in total (an almost complete list of domains is here, but it is not complete). Techrights is believed to be dealing with millions of hits per week, based on Varnish logs. It is hard, however, to dissect those logs because they’re all routed through a cache proxy and therefore have the same IP address for almost all traffic. My second most-accessed site is Schestowitz.com and this month (so far) it is looking as follows:

Tobkes.othellomaster.com (subsite alone)

Medivasc.com looks like this

OR ALMOST a month now BT has been throttling speeds on its exchange, thus crippling/impeding connections, breaking what was previously working without issues. Interlacing and full capacity, they appear to have decided, can just arbitrary be disabled, disconnecting people with all their lives sessions at any time. When phoning BT — as I have done for about 3 hours over the past month (in total) — they acknowledge the issue, although their technical staff seems unable to even recognise the problem or communicate it properly (among peers), which leaves them having to issue pathetic compensation (also requires hours on the line) while continuing to throttle the connections. This is a technical issue at their end — one that they can address and resolve remotely, but why do they knowingly break their systems in the first place? They make a lot of verbal promises — hence not legally binding — only to break those promises and leave people like me wasting a dozen or so hours, incapable of doing proper work or even keep good temper.

BT are failing because their own traffic management systems are defective by design. Calling this inevitable error is like saying that accidents are OK because you feel like committing traffic violations. This was not the case in prior years. Why is BT crippling its own infrastructure at cost to the clients? Why are they abusing their monopoly on landlines? Maybe it’s time to really break this monopoly rather than hide it behind the fake impression of “choice”. They know that people cannot cease reliance on landline (bar mobile networks), so they continue to throttle connections to the point where people get none of what they paid for. They just get a sporadic connection and a headache.

8 years after I last updated the front page of this site I finally got around to modernising it a little. It’s more business-friendly now.