Web Service WTF

[boomzilla]

@JBert said:

So I for one can fully understand Blakey's ranting and I would have cut him some slack

You must be new here.

[dkf]

@antiquarian said:

Wouldn't a competent C# programmer be able to learn Java relatively quickly?

The language? Yes. The libraries? Probably not; each ecosystem is now pretty large and they're not nearly as close as the languages are.

[blakeyrat]

@JBert said:

So I for one

Nobody who truly understands me would ever say "I for one". What the fuck is that? Do you normally post as a hivemind of 36 different people, and you put that in there to indicate the rest of the aliens are asleep? Are you trying to create this huge hypothetical audience that disagrees with you, then trying to set yourself aside from it?

I for one!

[flabdablet]

@redwizard said:

All I know is the code was supposed to do the same thing if any of the bits were found to be set.

Then I'm having trouble understanding what your ROL loop was for.

if (flags & (this_bit|that_bit|the_other_bit)) {
    return stuff;
}

or if you want to test every bit that could conceivably be set in flags, just test the entire thing for non-zero-ness:

if (flags) {
        return stuff;
    }

What was the ROL loop intended to achieve?

[tarunik]

@blakeyrat said:

JBert said:
So I for one

Nobody who truly understands me would ever say "I for one". What the fuck is that?

Think of it as a contracted form of "I, for one example" (which is better stated as "I, for instance"). Or, a first-person existential qualifier, if that is how you prefer to think of it.

[blakeyrat]

If the people who used that phrase actually thought about it, they might agree with you.

But I think it's much more likely that it's some kind of textual diarrhea they picked up from someplace (it's endemic on Slashdot, for example), and it just plops out onto the page.

[Keith]

@blakeyrat said:

But I think it's much more likely that it's some kind of textual diarrhea they picked up from someplace (it's endemic on Slashdot, for example), and it just plops out onto the page.

For all intensive purposes, you're probably correct.

[JBert]

@blakeyrat said:

@tarunik said:

Think of it as a contracted form of "I, for one example" (which is better stated as "I, for instance"). Or, a first-person existential qualifier, if that is how you prefer to think of it.

But I think it's much more likely that it's some kind of textual diarrhea they picked up from someplace (it's endemic on Slashdot, for example), and it just plops out onto the page.

Right in one guess; I must have picked it up from Slashdot, that whorific site.

@boomzilla said:

You must be new here.

Oh, only the "everything you post can and will be used against you" thing will be hard to get used to.

[redwizard]

@flabdablet said:

just test the entire thing for non-zero-ness:

That's what my friend did!

Remember, I only had one class in C programming & little experience. I mechanically shortened the 2-page source code to 4 lines with the ROL loop. He checked to see if the variable was non-zero and executed accordingly, hence his (yours) was even shorter (and more to the point).

In effect, I reduced the WTFyness by >90%, you and he eliminated it altogether.

[flabdablet]

@blakeyrat said:

picked up from someplace

folk entomology

[HardwareGeek]

@flabdablet said:

<a href="http://knowyourmeme.com/memes/i-for-one-welcome-our-new-insect-overlords">folk entomology</a>

That is the source for "I, for one, welcome our new X overlords," but "I, for one, ..." itself predates that by decades, if not centuries.

[dkf]

@HardwareGeek said:

That is the source for "I, for one, welcome our new X overlords," but "I, for one, ..." itself predates that by decades, if not centuries.

Slashdot has been around a long time!

[Zecc]

@blakeyrat said:

Nobody who truly understands me would ever say "I for one".

I, personally, would just to annoy you.

[boomzilla]

@Zecc said:

@blakeyrat said:

Nobody who truly understands me would ever say "I for one".

I, personally, would just to annoy you.

Actually, you forgot the "for one" part.

[algorythmics]

@boomzilla said:

Actually, you forgot the "for one" part.

I for one, second this tertiary approach to trolling blakey.

"1 2 3"

teehee

[Onyx]

@algorythmics said:

I for one, second this tertiary approach to trolling blakey.

Approved. Go forth, my friend!

---

Filed under: Punishment for this is already being self-inflicted

[Keith]

@Onyx said:

Approved. Go fo[u]u[/u]rth, my friend!

FTFY

[Onyx]

@Keith said:

FTFY

I hoped that it won't be necessary to point that tidbit out to the audiences here, so I went with the proper spelling and allowed for inference of my intention.

Was I wrong?

[boomzilla]

@Onyx said:

Was I wrong?

Are you saying you want to start Doing It Right?

[Keith]

@Onyx said:

I hoped that it won't be necessary to point that tidbit out to the audiences here, so I went with the proper spelling and allowed for inference of my intention.

Was I wrong?

If you'd done a chubertdev and hidden the 'u' with a tiny font, you would have covered yourself against any confusion. You would also have successfully trolled me.

[Zecc]

@Keith said:

FTFY

I read that as "fifth". Now pretend that was your intention all along.

[Zecc]

@boomzilla said:

@Zecc said:

I, personally, would just to annoy you.

Actually, you forgot the "for one" part.

I figured the "I, personally," would be just as much, if not more, annoying.

[boomzilla]

@Zecc said:

@boomzilla said:

@Zecc said:

@blakeyrat said:

Nobody who truly understands me would ever say "I for one".

I, personally, would just to annoy you.

Actually, you forgot the "for one" part.

I figured the "I, personally," would be just as much, if not more, annoying.

Actually, I, for one applaud you for not dumbing down your scientific trolling.

[antiquarian]

I assumed you were referring to having to program in forth as the self-inflicted punishment and that @Keith missed the joke.

[Keith]

@antiquarian said:

I assumed you were referring to having to program in forth as the self-inflicted punishment and that @Keith missed the joke.

I think a joke has gone awry when it causes this much confusion for so many people.

[boomzilla]

@Keith said:

I think a joke has gone awry when it causes this much confusion for so many people.

I think a joke is awesome when it can have so many ways of being funny.

[Onyx]

@antiquarian's explanation is something I didn't even originally consider, but now I will pretend that it was my sixth sense seeing into the future and picking the spelling with that intention.

[chubertdev]

@Keith said:

If you'd done a chubertdev and hidden the 'u' with a tiny font, you would have covered yourself against any confusion. You would also have successfully trolled me.

Only my wife gets to do a chubertdev.

[Onyx]

@chubertdev said:

Only my wife gets to do a chubertdev.

Must have really good eyesight then.

[chubertdev]

@Onyx said:

Must have really good eyesight then.

She's a sharp one.

[AlexMedia]

I hate you. :)

[abarker]

All right, dragging this back OT with an update:

So the other company's offshore team made some changes to their genericResponse structure. It is now significantly different from the genericRequest in the OP:

<genericResponse>
    <customerid>SOME GUID</customerid>
    <userid>SOME GUID</userid>
    <locationid>SOME GUID</locationid>
    <attributeResponse>
        <attributes>
            <attributeDetail>
                <attributeName>SOME NAME1</attributeName>
                <attributeValue>SOME VALUE1</attributeValue>
            <attributeDetail>
            <attributeDetail>
                <attributeName>SOME NAME2</attributeName>
                <attributeValue>SOME VALUE2</attributeValue>
            <attributeDetail>
        </attributes>
        <attributes>
            <attributeDetail>
                <attributeName>SOME NAME3</attributeName>
                <attributeValue>SOME VALUE3</attributeValue>
            <attributeDetail>
        </attributes>
    </attributeResponse>
    <attributeResponse>
        <attributes>
            <attributeDetail>
                <attributeName>SOME NAME4</attributeName>
                <attributeValue>SOME VALUE4</attributeValue>
            <attributeDetail>
        </attributes>
        <attributes>
            <attributeDetail>
                <attributeName>SOME NAME5</attributeName>
                <attributeValue>SOME VALUE5</attributeValue>
            <attributeDetail>
            <attributeDetail>
                <attributeName>SOME NAME6</attributeName>
                <attributeValue>SOME VALUE6</attributeValue>
            <attributeDetail>
        </attributes>
    </attributeResponse>
</genericResponse>

Now, because this response is used in about 2 dozen service methods, all of those service methods are broken and we have to re-write huge chunks of code on our side. Not only that, but when they made the change, they introduced a structural bug that means we can't even process the WSDL file for the service.

[Scarlet_Manuka]

Here's a different web service WTF I ran into yesterday:

• call operation "getAllScheduledReportHistory" as defined in the WSDL, and as instructed by vendor support
• get back a fault response saying "No such operation 'getAllScheduledReportHistory'"

OK then...

[JBert]

@abarker said:

All right, dragging this back OT with an update:

So the other company's offshore team made some changes to their genericResponse structure. It is now significantly different from the genericRequest in the OP:

<genericResponse>
    <customerid>SOME GUID</customerid>
    <userid>SOME GUID</userid>
    <locationid>SOME GUID</locationid>
    <attributeResponse>
        <attributes>
...
        </attributes>
        <attributes>
...
        </attributes>
    </attributeResponse>
    <attributeResponse>
        <attributes>
....
        </attributes>
        <attributes>
...
        </attributes>
    </attributeResponse>
</genericResponse>
```</blockquote>
I wonder why they would suddenly be repeating the `attributeResponse` and `attributeDetails` elements. Do they somehow compose this response based on the results of some other of their webservice calls? It looks as if they **really** don't want to merge those attributes into one big list...

---
Filed under: <a>There are lazy programmers and lazy programmers.</a>

[abarker]

They're trying to mimic tables and rows. with this structure. So each attributeResponse is a "table", each attributes is a "row", and each attributeDetail is a "field".

[Onyx]

@abarker said:

They're trying to mimic tables and rows. with this structure. So each attributeResponse is a "table", each attributes is a "row", and each attributeDetail is a "field".

Well then. The obvious question now is: can you JOIN them?

[Arantor]

With a suitable script and an XML parser, sure, why not?

[Onyx]

@Arantor said:

why not?

Oh, sure, it's not a big deal. Unless you think this is an inappropriate way to greet your boss the next morning:

[JBert]

@abarker said:

They're trying to mimic tables and rows. with this structure. So each attributeResponse is a "table", each attributes is a "row", and each attributeDetail is a "field".

With an inconsistent number of fields for each row and table??? For real???

Good luck JOINing them.

---

Filed under: TDEMSYR(s)

[abarker]

@Onyx said:

Well then. The obvious question now is: can you JOIN them?

Yes, they are providing sufficient information that they could be joined. If the damn services weren't broken.

[abarker]

@JBert said:

With an inconsistent number of fields for each row and table??? For real???

Good luck JOINing them.

---

Filed under: <a>TDEMSYR(s)</a>

Yes, this is the shit I am dealing with. Once again. the company producing this is outsourcing the work to India, so it's cheap, imported shit.

[boomzilla]

@Onyx said:

Oh, sure, it's not a big deal. Unless you think this is an inappropriate way to greet your boss the next morning:

What? Smiling and eager looking?

[Onyx]

@boomzilla said:

What? Smiling and eager looking?

I always look eager when carrying an axe as well, yes.

[FrostCat]

@Severity_One said:

Our experience with off-shoring to India has been very positive.

I had this wonderful meeting a couple weeks ago where someone said that some of our offshore programmers keep ignoring our coding conventions. I kept wondering why we didn't reject the code. I guess that's TRWTF.

[dkf]

Reading that, I'm a little surprised that they didn't decide to just write it all in Turtle (one of the many RDF serialization formats). Like that, they'd have at least as much power and generic-ness as they have now, but with the added bonus that there's far fewer standard libraries for working with it. They could also do something nasty with URLs in the process to make it all absolutely incomprehensible too.

And when people complain, they could just switch back to the XML rendering of RDF. It's XML! It's generic! It's a true standard! It's fucking awful!

[abarker]

Another update:

We just got the other company to add a few new services that we needed for the project. When they finished the update, they sent us the updated documentation. According to the documentation, the genericRequest structure was changed to look like the most recent version of the genericResponse structure. Well, all of our integration to this point is broken, but this was expected. So we pull down the updated WSDL and get ready to start updating all of our integration points. But wait! The genericRequest structure didn't change! Instead, they added a new genericRequestWithList structure, which is probably a good thing.

So they actually did something right. But instead of just adding the new stuff to the documentation, they "updated" all the existing documentation so that it is no longer correct.

I really hate this project.