Kir Kolyshkin Голосов: 1 Адрес блога: http://k001.livejournal.com/ Добавлен: 2008-02-19 12:45:30 блограйдером Lurk

epic fail

2011-04-14 18:10:06 (читать в оригинале)

OK now I can tell what happened during my last business trip. Scroll to the bottom for words of wisdom that cost me more than $1000. Yeah I am unbelievably generous today -- you get it for free!

We flew from Moscow (SVO) to New York (Kennedy), then from NY to LAX. On the second flight my colleague asked for a headphones, and I was lazy to get it from my backpack, so I asked a flight attendant to bring a pair for me. He brought it but (unlike on the international flights) they are $2/pair and they only accept credit cards. I was shy to reject it (like I can't afford spending $2, guess it's a typical feature of Russians). So I get my backpack (which I tried to avoid in the first place) and pulled my wallet out of it. Paid, then left it in the front seat pocket.

I forgot it there. Yes, my wallet left in the front seat pocket then I left the plane. Not sure why it happened, I was not even drunk (like it happens on a long flights sometimes), only extremely tired and with a headache, trying to not sleep in order to switch to -11 hours time zone change quicker.

In the forgotten wallet there was my personal credit card, my corporate credit card, and $1100 in cash, from which $1000 was because someone asked me to bring the most expensive version of iPad2 in white color. Yeah… another reason to hate Apple, even if they are not guilty in this case.

The wallet was never found (I guess because of the cash). Delta was very unresponsive — every phone line is a robot, and if you can make it to a human it also acts like a robot, asking you to go to delta.com and file a «lost item» report. Then you file such a report, it apparently goes nowhere.

My colleague had a personal credit card with him, but Fox rent-a-car refused it (although we were able to draw some cash from an ATM). Call to the issuing bank (Alfa) was non-productive: they said card is fine, unblocked and they are aware that we are in US not Russia, and they marked it appropriately, but they don't see any requests for a transaction except for the successful cash draw attempt).

I called my friend in LA to ask for help, but he texted me back saying he's in Thailand. He although suggested another friend of him who could help next morning (the time was like 2am).

So, we were unable to rent a car and unable to drive to the motel I reserved. We went to a Travelodge across the street, paid cash.

Next morning a friend came, it was Sunday and he had a bad hangover, and he was 1 hour by car away from us, but he came. He took the car under his name and using his credit card (with my colleague listed as a second driver).

Next scene, Sunday evening, 10pm, hotel lobby. Colleague's card is not working, we can't check into the hotel. Went out to try drawing some cash, didn't work. Fortunately it was already Monday morning in Moscow, so the call to the bank was productive, card became working, we have finally checked in.

To make a long story short -- the card was blocked a few more times, some places were not accepting it, one time we _almost_ ran out of money, fortunately the conference provided free food. :)

Now the morale of the story. Do not be a fool like me, follow these simple rules and you will be, well, not fine but OK.

1. Do not take more than $100-200 in cash, maybe even less. You can lose it all. If someone asks you to buy something, ask them to transfer money to your account, do not get cash!

2. Have more than one credit card, and put them all in different pockets/places. Cash should be in a different place, too.

3. Do not buy anything on board (especially as stupid stuff such as headphones). Well this is actually questionable and optional advise.

---

горло болит

2011-03-31 12:06:24 (читать в оригинале)

Больное горло перед командировкой становится доброй традицией (см. предыдущий и пред-предыдущий). Вот и сейчас, доктор, когда пишу я эти строки, горло моё саднит и чешется.

Выпил вчера пару рюмок водки перед сном -- не особо помогло. Я понял, что надо пить не водку, а виски или коньяк, потому что водку я пью махом, а виски и коньяк по чуть-чуть.

---

Google Android bugs and CM7 RC4

2011-03-31 05:39:33 (читать в оригинале)

This is a quick followup to my previous post on this topic. First bug mentioned by me (#3106: screen fails to refresh or goes black aka Android issue #15135) is now fixed in both Google AOSP repository and in CyanogenMod. Apparently it was the reason to roll out RC4.

/me likes.

---

CyanogenMod 7 RC3 Nexus One major issues

2011-03-29 01:56:07 (читать в оригинале)

These two issues are show-stoppers for "normal" people wishing to use CyanogenMod on their Nexus Ones, and I hope they will be fixed before CM7 goes final. Both bugs looks to be upstream ones, i.e. they are not born in CyanogenMod team but in Google itself.

(1) #3106: screen fails to refresh or goes black; logs say "/dev/pmem: no more pmem available" -> "OutOfResourcesException locking surface".
This one is an upstream bug, see Android issue #15135. The only known workaround is to reboot, I am testing a solution (modified lib) suggested here.

(2) Call wonkiness. I heard that it's an upstream bug as well (a problem in a driver). Workaround: install the dewonkificator app.

On the good side, a kernel bug which made it lost 45-60 seconds of time every day (aka issue #1222) is now fixed. Kernel is doing a good work accurately keeping time.

Generally, I wish Google Android guys could've paid more attention to quality assurance. Yes, I know, nobody is perfect, and QA is rocket science (indeed it is, and a gap between good and very good is huge), but still I see a room for improvement here. I am very happy with their pace of innovation, but the tendency to trade quality for speed can ruin the whole effort easily.

For other (not so major) bugs I faced recently, see my blog post about a bug in browser find dialog and my patch fixing a bad glitch with keyboard layout. These two are not critical, although the first one is misleading and the second one can be very annoying, and they both are examples of bad UI QA.

Wake up, Android, if you want to keep eating apples for breakfast!

---

android 2.3 browser find on page dialog bug

2011-03-26 22:49:13 (читать в оригинале)

Here is a piece of Java code:

private void setMatchesFound(int found) {
    String template = mBrowserActivity.getResources().
        getQuantityString(R.plurals.matches_found, found, found);

    mMatches.setText(template);
}

and the corresponding XML resource:

    <!-- Displayed on the Find dialog to display the number of matches
         found in the current page. -->
    <plurals name="matches_found">
        <!-- Case of no matches -->
        <item quantity="zero">No matches</item>
        <!-- Case of one match -->
        <item quantity="one">1 match</item>
        <!-- Case of "few" (two) matches -->
        <item quantity="few"><xliff:g id="number" example="2">d</xliff:g> matches</item>
        <!-- Case of several matches -->
        <item quantity="other"><xliff:g id="number" example="137">%d</xliff:g> matches</item>
    </plurals>

It works just fine, showing something like "12 matches" in a "Find text on page" dialog. The code was added by this git commit (oh year, with a "very descriptive" log message).

Now, there is a git commit that changes the above Java code, here's the relevant diff:

-                getQuantityString(R.plurals.matches_found, found, found);
+                getQuantityString(R.plurals.matches_found, mNumberOfMatches,
+                mWebView.findIndex() + 1, mNumberOfMatches);

As you can see, the aim is to print something like "Match 3 of 5". The problem is that XML code is not changed, so it shows something like "Matches: 3" instead of what it should. Bad and broken UI. Android 2.3.3.

The good thing is it is open source so anyone (incl. me) can fix it. I an not an expert in the field but I'll try.